University of Toronto Press
  • Literal and LiteraryLanguage and the Representation of Chinese Poetry


The evaluation of translation is beset with all sorts of subjectivities, a bewildering array of often conflicting theories, and it is invariably determined by very personal tastes. Too many evaluations of translation follow what might be appropriate for food—de gustibus non disputandum est—but not for literary constructs. This paper is an attempt to provide at least a template for the evaluation of translations that will establish some form of objective measure for judgment. We will look at translations through two prisms: the literal and the literary. These two foci will provide four permutations: (1) false to the letter; false to the spirit; (2) true to the letter; false to the spirit; (3) false to the letter; true to the spirit; (4) true to the letter; true to the spirit. Most readers would dismiss at least (1) and (2) as unsatisfactory; some would also dismiss (3). Few readers would dismiss (4).

In establishing these criteria, I am not trying to establish an absolute metric of measurement, I am merely trying to narrow the parameters of evaluation, so that one’s own presuppositions and assumptions can be clearly identified. Before we begin, let me confess to a preliminary bias of my own. I believe a literary construct is ambiguous and ambivalent, not a discrete or discriminate code to be decoded into one meaning. To be successful, a translation must convey the nuances as well as the feelings that the original text conveys to, or elicits from, the source audience.

With any criteria, it’s useful to explore exemplars, so that one can understand in practical terms how the criteria may be applied. Presumably, as we proceed from (1) to (4), we can approach increasingly more satisfactory translations. [End Page 18]

I. False to the letter, False to the spirit

We begin with (1) translations that are false to the letter, and false to the spirit. Since we are only interested in examples and not practitioners, I will not identify the translators. Here is Li Bai’s (李白) famous farewell to Meng Haoran (孟浩然)




My old friend, going west, bids farewell at Yellow Crane Terrace. Among misty blossoms of the third month, goes down to Yang-chou. His lone sail’s far shadow vanishes into the azure void.2 Now, only the Long River flowing to the sky’s end.

There are at least two mistakes in this translation that stem from over-literalness. First, 西辭 means “leaving from the west,” not “going west” or “going to the west”; second the phrase 煙花 is a compound constructed of the words for ‘mist’ and ‘flowers;’ the phrase does not mean ‘misty blossoms,’ but refers to the lush spring scene in Yangzhou (like ‘cherry blossom time in Washington, D. C.’); in the words of one commentator, this translation strays far from the poet’s intention “to describe the prosperity and liveliness of the spring-time Yangzhou.”3

A second example involves the poetry of Wang Wei, the so-called “Buddhist poet.” Although Wang Wei, like most if not all traditional Chinese poets, speaks from personal experience, he is far from being an ‘autobiographical poet’ bent on exploring the self, as in the case of William Wordsworth, the most familiar autobiographical nature poet. Furthermore, Wang Wei’s language is known for its simplicity and its accessibility, even when it is at its most suggestive and profound. Let us take his most translated poem, sometimes translated as “Deer Enclosure”:






How much of this almost miraculously limpid poem can be gleaned from the following:

The empty mountain: to see no men, Barely earminded of men talking - countertones, And antistrophic lights-and-shadows incoming deeper the deep-treed grove Once more to glowlight the blue-green mosses—going up

(The empty mountain...)

The first disparity one notices are the odd words in this version: ‘earminded,’ ‘countertones,’ ‘antistrophic,’ ‘deep-treed,’ ‘glowlight.’ Not a single word in Wang Wei is so bizarre. Clearly, the translator is struggling to capture something foreign, and what he produces is something so cumbersome and odd that it scarcely sounds like a poem: it certainly will not convey to the English reader illiterate in Chinese why the Chinese admire Wang Wei as a poet. This is what can happen if [End Page 19] one follows a policy of ‘foreignizing’ the translation. The result is not literature, but the dead carcass of a living work of art, with nothing of its power to inform, to touch, and to inspire readers twelve centuries later, and in another language.

In this case, the translator is not entirely unaware of the shortcomings of his attempt: he writes, “a still inadequate, yet philologically correct, rendition ... (with due attention to grapho-syntactic overtones and enjambment).”4 One might sympathize with the translator on the difficulty of the poem, but one must take issue with his claim to philological correctness, unless philology is intended merely to obfuscate the original. Surely philology does not require the erasure of the literariness of the poem? Or its simplicity? And one is nonplussed to guess what the “grapho-syntactic overtones” in the poem might be (it’s a recurrent tendency of some Western translators who rely on the text, and neglect the sound of the original, to put more emphasis on the grapheme than most Chinese would.) The ‘going up’ for 上 in the last line misconstrues a simple post-position that is equivalent to the English preposition ‘on.’ And there is, so far as one can see, no enjambment in these lines, which occurs rarely, if ever, in jueju poetry.

Another example of false-to-the-letter being also false-to-the-spirit involves a translation of the Daodejing (Tao Te Ching), which when confronted by the conventional omission in classical Chinese of a subject-agent supplies them in English, and in obedience to contemporary political correctness offers optional third personal singular references between ‘he/him’ and ‘she/her.’ Reasonable as the practice adopted might be in ‘politically correct’ times, they are totally inappropriate to an ancient Chinese text. The translator’s defense of this practice is nothing if not trendy:

The reader will notice that in the many passages where Lao-tzu describes the Master, I have used the pronoun ‘she’ at least as often as ‘he.’ The Chinese language doesn’t make this kind of distinction;5 in English we have to choose. But since we are all, potentially, the Master (since the Master is, essentially, us), I felt it would be untrue to present a male archetype. . . .

This practice misunderstands the omission of the subject-agent in Chinese as making the subject-agent optional, but in fact, the omission is usually made because the subject-agent is assumed as obvious. It is not true that “the Chinese language doesn’t make this kind of distinction”; it does. What traditional Chinese texts allow is to leave the reference to the subject-agent unsaid: it is implied. The fact that it is implicit does not mean that it is optional. In traditional China, since virtually the entire literate population was male, the implied unstated subject-agent was obviously male. This male chauvinism might offend modern sensibilities, but that does not justify the conversion of the male chauvinism of traditional China into gender sensitive liberalism. And while we moderns might resonate with the egalitarian vision of wisdom—a wisdom of, in, and for all—implicit in the nicely inclusive “But since we are all, potentially, the Master (since the Master is, essentially, us)”, this proposition would be, for the Chinese, preposterous. There was no sense in ancient China (nor, I suspect, in modern China) of egalitarian, communal wisdom, no belief that the source of all wisdom is ‘us.’ And finally, the implication in the first line that the subject can be equally ‘she’ as well as ‘he’ [End Page 20] is simply wrong. There is no warrant to believe that the Master could have been anything but male.

The problem of making explicit (and arbitrary) what was implicit and intuitive is that the unstated background against which emphases are highlighted disappear. In a culture so male-dominated, the stress that Taoism in general and the Tao Te Ching in particular makes on the female is particularly salient. It also diminishes the significance of Daoist cosmology, which holds, in contrast to Aristotle, that something emerged from nothing, and that nothing, regarded metaphorically as ‘the mother,’ is, therefore, the source of all creation. In a sense, the practice of alternating male and female pronouns runs counter to a basic Taoist yin-and-yang theory, which presupposes that nothing is exclusively male or female, and that each individual is a composite, in varying degrees, of both male and female.

This artificial political correctness imposed on a classic Chinese text is most disastrous in the case of Chapter 52, which, though it does not explicitly include either a male or female subject-agent, nevertheless implies definitively a subjectagent that is female. The first part of Chapter 52 in the original text reads as follows:

天 下 有 始 , 以 為 天 下 母

既 得 其 母 , 以 知 其 子 ,

既 知 其 子 , 復 守 其 母 , 沒 身 不 殆

The translation follows:

In the beginning was the Tao. All things issue from it; all things return to it. To find the origin, trace back the manifestations. When you recognize the children and find the mother, you will be free of sorrow.

Despite the lip service the translator pays in the foreword to the emphasis on the female in the Tao Te Ching (“of all the great world religions the teaching of Lao-tzu is by far the most female”), he scarcely does justice to the female in this instance. The word ‘mother’ occurs but once in the translation; it occurs three times in the original. A more faithful translation of these lines might be:

All creation has its beginning, and this beginning can be seen as the Mother of all creation. When we know the Mother, we can understand her offspring. By understanding her offspring, and abiding with the Mother, our lives will be free from peril.

No text in the Tao Te Ching is as emphatic about identifying ‘nothingness,’ exemplified by the void of the womb, as the origin of all things. This is a text in which the gender is indisputably female. The word 其 occurs four times; even though as a possessive, 其 is unlike English not gender specific, the phrase 其 子 clearly means ‘her offspring’—referring to 母, ‘mother.’ So, the translator misses completely the emphatic feminine bias of this excerpt by avoiding the female possessive. [End Page 21] And the repetition of ‘you’ and ‘your’ in the translation has no warrant, though it doubtless makes the translation more attractive as a modern self-help guide. There is no word in the original text that can be translated as ‘you’ or ‘your.’

II. True to the letter, False to the spirit

In the initial encounter between two cultures, there is an assumption that a word-for-word parallelism exists. This leads to an overemphasis on words rather than meanings, and reflects an overweening dependence on the dictionary. But language is not a dictionary, nor is a dictionary a language; it is a compilation of words, but not always a key to meaning. Fledgling translators tend to rely excessively on dictionary meanings rather than on semantic contexts.6 In the early stages of comprehending China and Chinese, nineteenth century sinologists reflected this addiction to the dictionary. In addition to saddling the study of Chinese with an absurd and non-intuitive transliteration system (the Wade-Giles), they also managed to render a few key Chinese cultural terms with mindless literalism and a tin ear for the English language: they translated the words, but not the meaning. My favorite example is their rendering of 唱歌女 as ‘sing-song girl.’ It is true that 唱 means ‘to sing’ and 歌 means ‘song,’ but the 唱歌女 was a professional entertainer with many talents, not restricted to singing, who had many ways to amuse and enchant a client, not unlike the Japanese geisha; a few of them even composed poetry. Whatever they were, they were lively, talented, and diverting, and anything but ‘sing-song.’7

Then there are the mistranslations of ‘duplicatives’ for Chinese phrases that repeat the same word. The dictionary monger sees the same word occurring twice, and dutifully renders the definition twice, missing the salient fact that most duplicatives are compounds with a single meaning, not a meaning invoked twice. If we think of duplicatives in English, we would realize how silly it is, when rendering a duplicative, to translate the word repeated twice. A ‘boo-boo’ is a mistake: it is not ‘Boo!’ said twice to scare people; ‘hush hush’ means ‘confidential’: it does not mean ‘Be quiet!’ twice. Consider the idiocy of translating ‘no-no’ into French as pas pas instead of interdit, or into German as Nein Nein instead of tabu sein or into Chinese as 不不 instead of 絕不. This is what occurs when some translators slavishly render the seven duplicatives that open Li Qingzhao’s famous, and famously difficult, ci to the tune 聲聲慢:


Seek, seek! Search, search! Cold, cold! Bare, bare! Grief, grief! Cruel, cruel! Sorrow, sorrow!

What in Chinese is a series of brilliant attempts at an onomatopoeia of sorrow, enacting (rather than describing) desolation and sadness, this rendering becomes a comical litany of increasingly meaningless and mindless repetitions. The translator misconstrues the text: the duplicatives repeat the sound, not the meaning. The English reader, by contrast to the Chinese reader, will find it hard to reach the end of these lines without laughing. The translator overlooks the fact that classical Chinese is predominantly monosyllabic, and can repeat words easily without [End Page 22] sounding verbose. For example, 濛 means ‘mist’ or ‘misty,’ but 濛 濛 also means ‘misty;’ it does not mean ‘mist, mist’ or ‘misty, misty;’ and 鬱, ‘luxuriant’ or ‘sad,’ can be doubled in Chinese to 鬱 鬱 to fine effect, but they both mean ‘luxuriant’ or ‘sad;’ 鬱 鬱 cannot mean ‘luxuriant, luxuriant’ or ‘sad, sad.’ In the case of Li Qingzhao’s poem, the last line merely reinforces what the opening implies, that the emotions experienced are beyond the power of words to designate or describe, which is why the seven duplicatives give the impression of someone stuttering, on the verge of speech, but failing to achieve articulation. The ending says it all by pointing out that any attempt to express the feeling in one word will fail:

怎一个 愁字了得

How can one word—’sorrow’—capture it?

On a more complex level of style and characterization, let us consider David Hawkes’ rendering of a particularly spirited passage from the Hongloumeng, from Chapter 59:

Swallow’s mother was still smarting from her unsuccessful quarrel with Parfumée and was angry with Swallow for not having taken her side.

“Little strumpet!” she shouted, bearing down on her wrathfully and slapping her across the head. “How long now have you been working with those young madams?—it hasn’t taken you very long to pick up their airs and graces! But don’t you go thinking I can’t lay my hands on you any more. A foster-daughter’s one thing, but you are my own flesh and blood. I can still look after you when I feel like it. Little painted whores, telling me I can’t go inside where you can go! I wish you’d go inside and stay there: perhaps if you stayed inside long enough, you might find a customer!”

In the original, I think it would be fair to say that the language is vivid, and vigorously vulgar, reflecting the earthiness of the emotions, and the need of a mother to put her high-falutin’ daughter down a peg or two by reminding her of her low-class origins.

In this connection, the locutions of an Oxford don which one finds in Hawkes’ version hardly seem apposite: ‘Little strumpet’ might be an epithet hurled at an Eliza Doolittle by a Professor Henry Higgins, but it is not a likely outburst from a serving maid to her uppity offspring. How many maids do you know use the word ‘strumpet’?

One of the challenging texts for the translator of Chinese is the famous haoliao ge from Chapter 1 of the Hongloumeng [The Dream of the Red Chamber / The Story of the Stone]. Sung by an itinerant Daoist monk, the song sounds like nonsense, but in some measure it turns out to be a microcosm of the entire novel and its theme: the vanity of human wishes. Key to the song are the two rhymes, 好 and 了, two of the most familiar words in the Chinese language, the first meaning ‘good,’ the second a final particle, indicating past tense, and meaning ‘finished’ (but which may also mean ‘to understand,’ as in the phrase 了解). The song is referred to by its rhymes: the Haoliao ge. Although one can normally dispense with rhyme in the translation of traditional Chinese poetry (the cost in semantic loss being usually too high to justify rhyme), in this case, the rhymes are essential, not only because they are crucial to the meaning of the poem, but also because they lend [End Page 23] a certain antic, singsong quality to the song, making it appear daft and eccentric.

Because the rhymes are essential to the meaning and to the character of the song, any translation that omits the rhyme loses a crucial element of the text. That is why the following version, by C. C. Wang, one of the earliest attempts, can be considered false to the letter as well as false to the spirit. Furthermore, his title, “Forget and Be Free” doesn’t quite convey the sense of hao and liao.

世人都曉神仙好, 惟有功名忘不了! 古今將相在何方, 荒塚一堆草沒了.

世人都曉神仙好, 只有金銀忘不了! 終朝只恨聚無多, 及到多時眼閉了.

世人都曉神仙好, 只有姣妻忘不了!77 君生日日說恩情, 君死又隨人去了.

世人都曉神仙好, 只有兒孫忘不了! 痴心父母古來多, 孝順兒孫誰見了.

Forget and Be Free

We all envy the immortals because they are free, But fame and fortune we cannot forget. Where are the ministers and generals of the past and present? Under neglected graves overgrown with weeds.

We all envy the immortals because they are free, But gold and silver we cannot forget All our lives we save and hoard and wish for more, When suddenly our eyes are forever closed.

We all envy the immortals because they are free, But our precious wives we cannot forget, They speak of love and constancy while we live, But marry again soon enough after we are dead.

We all envy the immortals because they are free, But our sons and grandsons we cannot forget, Many there are, of doting parents, from ancient times, But how few of the sons who are filial and obedient!8

This paraphrases with fair accuracy the meaning of the poem, but it does not enact the gestural character of the song, sung on the streets with no pretensions to philosophical profundity, and hauntingly memorable. It conveys the message of the song, but not the song itself. And it captures nothing of the mock-ironic cynicism of a street-urchin Daoist monk. It sounds more like a sermon from a pulpit—or a psalm from scripture.

David Hawkes attempted to capture the ‘signature’ rhymes in the Haoliao ge and came up with the following: [End Page 24]

Won-Done Song

Men all know that salvation should be won, But with ambition won’t have done, have done. Where are the famous ones of days gone by? In grassy graves they lie now, every one.

Men all know that salvation should be won, But with their riches won’t have done, have done. Each day they grumble they’ve not made enough. When they’ve enough, it’s goodnight everyone!

Men all know that salvation should be won, But with their loving wives they won’t have done. The darlings every day protest their love: But once you’re dead, they’re off with another one.

Men all know that salvation should be won, But with their children won’t have done, have done. Yet though of parents fond there is no lack, Of grateful children saw I ne’er a one.

This is, in many ways, an admirable attempt. It tries to capture the key rhymes in “won” and “done”; it is ingenious in the variation of the rhymes; and it is eminently readable. Hawkes even replicates in English the verse ends with the same rhyme-word, as in Chinese. (Hawkes offers: ‘everyone;’ ‘everyone;’ ‘another one;’ ‘another one;’ and ‘ne’er a one;’ in Chinese, it’s 功名忘不了; 草沒了; 忘不了;眼閉了; 忘不了; 人去了; 忘不了; 誰見了). But, if anything, Hawkes’s version is overly inventive. The rhyming scheme in the original is less varied. As one can see, 忘不了 occurs four times.

The virtues of Hawkes’s rendering also implies its weaknesses. It is, as I’ve said, eminently readable, but it is definitively not singable. The words do not look like, or more important, they do not sound like the lyrics of a song. It is true that Hawkes tries to emulate the refrain of a song by repeating “have done, have done” in the first two verses, but he abandons it in the third and fourth verse, only to resume it in the last verse.

III. False to the letter, true to the spirit

A more imaginative effort is required for the translator who takes liberties with the letter, but who wishes to capture the spirit of what s/he is translating. These efforts—false to the letter, true to the spirit—take a bit of intuitive daring, since there is no textual authority to corroborate the ‘correctness’ of the rendering. We may examine some successful examples.

Nursery rhymes offer a particularly promising genre in which to find successful adaptations:

London Bridge is falling down, falling down, falling down, London Bridge is falling down, my fair lady! [End Page 25]

In Cantonese, this emerges as:

有隻雀仔跌落水,跌落水, 跌落水 yau jaet juek tsai dit lok shui, dit lok shui, dit lok shui 有隻雀仔跌落水,跌落水,被水沖去 yau jaet juek tsai dit lok shui, bei shui chung hui.

which translates into English as:

There is a little sparrow who fell in the water, fell in the water, There is a little sparrow who fell in the water, which flushed him away.

Is there anyone who would protest the disappearance of London Bridge from this rhyme? Or who laments the disappearance of ‘my fair lady’? Is not the spirit of the rhyme, especially if sung to the proper tune, preserved even if the words deviate from the original?

One of the most subtle and difficult poems in Chinese is Liu Zongyuan’s poem “River Snow” 江雪. The last line is especially evocative and ambivalent.






The three words 寒江雪 defy subordination to ‘cold, river, snow,’ and are virtually impossible to render syntactically. The old man is fishing alone in the cold, in the river, and in the snow. Despite the countless vain attempts to render this poem, I tried another:

On a number of mountains, birds fly off, and are gone, Over numerous footpaths, there is no one. On a solitary boat, an old man in straw hat and slicker, Fishes in the cold river snow . . . alone.

I was sufficiently encouraged that, although I hadn’t reproduced the remarkable ru-sheng rhymes in Chinese: 絕, 滅, 雪, I had managed three ‘eye-rhymes,’ i.e., words that didn’t rhyme to the ear but that looked similar to the eye: ‘gone,’ ‘one,’ and ‘alone.’ Emboldened by this unexpected success and filled with hubris, I tried an ear-rhyme. It struck me that the second line indicated that the mountain was deserted and unfrequented, and I thought of the common Tang dynasty trope of citing the overgrown grasses to indicate a place where there was no traffic. This gave me the rhyme for the second line:

From one mountain to another, the birds have flown, Along one footpath to another, the grasses have grown. On a solitary boat, an old man in straw hat and slicker, Fishes in the cold river snow . . . alone.

Before I could savor this triumph, an interlocutor punctured my balloon, and reminded me that this was winter and that there could not be any question of grasses [End Page 26] grown or ungrown. So, another vain attempt was added to the pile of valiant failures. The effort to prove true to the letter as well as true to the spirit fell short.

While I am citing my failures, let me share another. Because I felt Hawkes’ rhymes won/done didn’t quite fit with the hao/liao theme of the poem, I lent my hand some years ago at what I called the “Crave/Grave Song”:

People all know being an immortal is fine, But it’s success we seem to crave. Where are the great men of the past? Buried in a grassy plot, a barren grave.

People all know being an immortal is fine, But it’s gold and silver we seem to crave. All day we gripe: there’s never enough, But when there is—we’re in the grave.

People all know being an immortal is fine, But it’s a pretty wife we seem to crave. From day to day, we express our love; But she’s off with another, once we’re in the grave.

People all know being an immortal is fine, Filial children are what we crave. There’s never an end to doting parents, But how many offspring will sweep our grave?

For years, I included this along with three other translations in my translation seminar, with the identity of each translator disguised, and asked the students in the class to express their preferences. I wish I could report that they preferred my version in every case to the three others, but—alas!—that hasn’t been the case. I then shared this with a colleague, an aficionado of the Hongloumeng who has studied the text for years. I offered her the same four texts, again with the identity of the translator disguised, and, gratifyingly, she preferred mine: perhaps she recognized the others, figured out the remaining version was mine, and was merely being polite.

Whatever the virtues of this version may be, its failings are manifest: it doesn’t have the colloquial vitality of the original; the crave-grave rhyme, while thematically apposite, lacks the familiarity of the hao-liao rhyme in Chinese; and I would have to admit that my version is no more singable than Hawkes’.

For a more successful example of being false to the letter and true to the spirit, I need cite no other than Ezra Pound’s rendering of Shijing #23, which I’ve had occasion to admire. Pound’s version captures the ambivalent innocence and desire implicit in the original ballad, where the innocence is reinforced by the simplicity of form and diction:

    野有死麕 野有死麕、白茅包之。有女懷春、吉士誘之。 [End Page 27] 林有樸樕、野有死鹿。白茅純束、有女如玉。 舒而脫脫兮、無感我帨兮、無使尨也吠。

Ezra Pound’s version is a miracle of re-creation. Though he departs from the literal here and there, he captures the dramatic scene, as well as the lyric mood, perfectly.

Lies a dead deer on yonder plain whom white grass covers, A melancholy maid in spring   is luck   for   lovers.

Where the scrub elm skirts the wood, be it not in white mat bound, or a jewel flawless found,   dead as doe is maidenhood.

Hark! Unhand my girdle-knot,   stay, stay, stay   or the dog   may   bark.

The departures from the literal are, however, not arbitrary. For 有女如玉, substitutes value for color (“a jewel flawless found”), because he wishes to avoid the connotations of jade as green that are prevalent in the West (in China, jade will more often connote white, or flesh color, rather than green); Pound finesses the visual image, and invokes the trope of perfection (“flawless”). Pound also leaves out 吉士 which Waley renders as “A fair knight;” Pound approximates the seduction scene with the phrase “A melancholy maid in spring / is luck / for lovers,” rendering the wonderfully suggestive phrase in Chinese, 懷春, which literally means ‘to harbor spring in one’s heart,’ a discreet metonymy for the natural sexual lusts that come with spring. For shui 帨, which, faithful to the dictionary, Waley improbably renders as “handkerchief,” not bothering to question why anyone would be concerned about having her handkerchief touched. Pound’s rendering of 帨 as “girdle-knot” makes the situation both more urgent and provocative. Symbolized by the dead doe in the first stanza, it’s the death of virginity that is being described here, and the conflicted plea of the maiden begging her love both to stop, and yet not to stop, is perfectly captured in: “stay, stay, stay,” enacting the desire and prudence that complicate virginity.

IV. True to the letter; true to the spirit

In selecting examples in this category, I am mindful of a certain subjectivity in judgment. In my defense, I have chosen a few examples where I am not the only one who considers the translations true to both the letter and the spirit. Let me begin with an unexceptionable example, the Cantonese version of “Twinkle, twinkle, little star”: [End Page 28]

一 閃 一 閃 小 星 星,一 顆 一 顆 亮 晶 晶 yat sim yat sim siu sing sing, yat fo yat fo leung jing jing 仿 似 許 多 小 寶 石,掛 在 天 空 放 光 明 fong qi hui doh siu bo shek, gwa joi tin hung fong gwong ming 一 閃 一 閃 小 星 星, 一 顆 一 顆 亮 晶 晶 yat sim yat sim siu sing sing, yat fo yat fo leung jing jing

Translated back into English, this would come out as:

A flash, a flash, little star, a drop, a drop of bright crystal, Just like many little jewels, suspended in the sky radiating brightness, A flash, a flash, little star, a drop, a drop of bright crystal.

But for the final distich, where the translation opts for a repetition of a previous phrase instead of attempting a version of “How I wonder where you are?” the version is a pretty close approximation of the original, and is eminently singable.

An obvious example of a translation that is both true to the letter and to the spirit would be Ezra Pound’s famous “A River Merchant’s Wife: A Letter,” his version of Li Bai’s 長干行, despite the obvious mistakes (“the river Kiang”, for example). There is no need to revisit this remarkable rendering. What I offer is Pound’s simple, literal, and literary version of the word 君 in the line 十四為君婦, “At fourteen, I married my Lord, you.” Pound catches what many linguists miss, which is that many Chinese pronouns are not just pronouns, they are also proper nouns: they are not merely terms of address, they are also terms of reference, identifying the addressor or the speaker, and indicating a relationship to the addressee. For example, the word 妾 means ‘I,’ first-person singular, but it can only be used if the speaker is a women addressing her husband; and the term 君 means ‘you’ as an honorific intended to show deference. To translate 君 as, simply, “My Lord,” as Amy Lowell does, loses the intimacy of the locution, making it appear formal and pompous; to translate it as, simply, ‘you,’ as most translators do, loses the sense of deference and submissiveness in the line: after all, she married when she was at a very young age. Pound’s inspired, “At fourteen, I married my Lord, you,” captures both the intimacy and the deference. That is why I offer it as an example of a version that is both true to the letter and true to the spirit.

Earlier, I cited my attempt at the Haoliao ge, deeming it a valiant failure because it did not capture the vivid vernacular in the original. One reader, as I mentioned, liked the translation, but added: “If you can add a humorous tone to it, it will be perfect because the original song is sung by a Taoist priest who looks shabby and crazy.” Spurred by this remark, I ventured to ‘translate’ my translation into contemporary American vernacular. With what results, I can only leave others to judge:

Y’all know this: it’s dandy to be good, But a good time is what we crave. Where are the big shots of the past? In a grassy plot, a piss-poor grave.

Y’all know this: it’s dandy to be good, [End Page 29] But filthy lucre is what we crave. All day we gripe: there ain’t enough! But when there is — we’re in the grave.

Y’all know this: it’s dandy to be good, But a pretty wench is what we crave. She tells us she loves us every day; But finds someone else who ain’t so grave.

Y’all know this: it’s dandy to be good, Our kiddies’ love is what we crave. You’ll find lots of dotty parents, over the years; How many young’uns are a-sweepin’ our grave?

Whatever the shortcomings this version might have, it cannot be accused of pretention. One could imagine it being “sung by a Taoist priest who looks shabby and crazy.”

With the reader’s indulgence, I offer as the final two examples of attempts that are true both to the letter and to the spirit, two of my own translations. I am emboldened to do this because the translations have been approved by the author. For some time, I have been translating the contemporary Chinese poems written in traditional shi and ci forms by a friend, colleague, and former student, Wann Ai-jen, who has been teaching Chinese at Purdue University for a number of years. The first poem is titled “An Unusual Willow by the Roadside”:

<道旁異柳> 一身憨態不尋常,千綹長鬚及地量。 頭角崢嶸非弱質,中心隱晦是諧裝。 迷離不識晚風烈,解悟何嫌新夢涼? 堪笑古來傷落絮,請觀自在此垂楊。

An Unusual Willow by the Roadside

A really silly sight not often found: Its thousand tendrils reach to the ground. Its tops, like antic antlers, show no weakness, Its heart hidden behind a comic flair. Unaware that fierce night winds blow still, Unfazed at new dreams upset by autumn chill. Funny how, for years, people thought willows shed tears But look at this one; poised, and free of fears.

This poem takes the anthropomorphism of the conceit of a ‘weeping willow’ to its logical extreme, and imagines this particular willow by the roadside as an exception to the rule: a willow that does not weep. The difficulty of the translation is that, in terms of semantic density, there is more weight in the seven words per line in the original Chinese than can be accommodated by a line of ten syllables [End Page 30] in English, and every effort has been made to pare down the syllables so that the English does not appear more verbose than the Chinese.

My last example is a satire, about pumpkins, and empty-headed jack-olanterns:

<外放南瓜燈> 鬼臉堪雕琢, 炎黃耀此洲。 瓜瓜皮相怪, 空腹亮肥頭。     萬愛珍作於癸未年季秋初十

Outside: Jack-o-lanterns

Goblin masks, cunningly carved, Yellow-skin faces posturing in this country. Pumpkin people looking outlandish: Empty inside, they’re all lit up: bright fat heads. Ai-jen Wann, written on October 5, 2003.

The poem is purportedly an apt description of the jack-o-lanterns that populate the North American landscape around Halloween, but the phrase ‘Yellow-skin faces’ suggest an ethnic target. Here, one must assume a certain latitude in color perception: not everyone would see the distinction between orange and yellow, both of which are subsumed in Chinese by the word 黃; and 炎黃 has a double meaning here: ‘burning yellow’ and ‘descendants of the Emperors Yan and Huang,’ i.e., the Chinese. The translation tries to capture the antic satire in the original. Whether it succeeds is for the reader to judge, though the author appears to be initially satisfied with it.

Conclusion: A New Metric for Measuring Translation

Our permutational analysis of four types of translation—(1) false to the letter, false to the spirit; (2) true to the letter, false to the spirit; (3) false to letter, true to the spirit; and (4) true to the letter, true to the spirit—enables us to develop a more precise calibration with which to evaluate translation. Clearly, there would be no dispute about categories (1) false to the letter, false to the spirit and (4) true to the letter, true to the spirit: the first produces translations that are, in a sense that can be generally accepted, wrong; the second produces translations that are, in a sense that can be generally accepted, right. But categories (3) and (4) are more problematic: they cannot be so unequivocally judged ‘right’ or ‘wrong.’ Category (2), true to the letter, false to the spirit, is literally ‘right,’ but misses the sense and spirit of the original. In other words, there may be translations that are ‘correct’ literally, but which fail to capture the essence of the text. So, these translations would be ‘correct’ without being satisfactory. Category (3), false to the letter, but true to the spirit, has the opposite valence: there may be translations that are [End Page 31] literally ‘incorrect’ but which nevertheless manage to convey the essence of the original. In that sense, these translations are ‘correct’ in capturing the meaning, but ‘incorrect’ in misrendering the literal text.

We now see how fuzzy and inconvenient the distinctions of ‘right’ and ‘wrong’ or of ‘correct’ and ‘incorrect’ are when applied to translations. They do not apply unequivocally to two of our four categories, because these labels presume a determinate right answer, as if there were only one ‘right’ and ‘correct’ answer possible. But the translation of literary texts involves the rendering of polysemous meaning, which suggests that while there may be many wrong or incorrect renderings, there may also be more than one (finite and few) right and correct versions. To that end, I propose that we adopt the practice of determining not whether a translation is right or wrong, correct or incorrect, but whether it is apt, or inapt (subsuming translations that are ‘inept’ as well). The definition of ‘apt’ would include (1) appropriateness to the register of the original, and (2) fidelity to the thrust of the meaning. It should be clear that, in some instances, more than one translation, in varying degrees, could satisfy both of these conditions. The determination of the aptness of a translation spares us from the pitfalls of absolute relativism (“a translation is good if the reader likes it”); and of absolute dogmatism (“a translation is good if it is ‘correct’”). The first pitfall undermines any rigor in the evaluation of translation; the second pitfall privileges an inevitably subjective perspective on what is correct.

In the foregoing analyses, there has been no way to disguise a certain assumption: I have assumed that, in translating works of literature, the literary should not be erased. In the translation of literature that is worth translating, dictionary equivalents are almost always unavailing. A translation that is ‘faultless’ merely because it provides the dictionary equivalents for words in the original betrays a shallow view of language and of poetry. First, language is not a mechanical conglomeration of words, whatever deep structuralists might claim; the semantic equivalents between languages, particularly between languages that are not cognate, are few and far between. Second, poetic language is precisely what will not be captured in dictionary definitions, for poetry stretches normal linguistic usage, and is non-normative, even abnormal. Dictionaries are normative; poetry is exceptional. A literal translation merely shows the translator’s ability to look the word up in a dictionary, but it does not necessarily reflect an understanding or appreciation of the original, which is literary rather than literal. Attempts to capture the spirit of the poem even if they are unavailing should be preferred over literal renditions that add nothing to one’s insight into the original. What is worth preserving is not the text, but the spirit of the text. Translators owe the original at least that much: the recognition that what they are doing is not reproducing a text, but recreating an art. [End Page 32]

Eugene Eoyang
Indiana University
Eugene Eoyang

Eugene Chen Eoyang is a practicing translator—some of his translations appear in the May 2010 issue of Renditions: A Chinese-English Magazine—as well as a theorist of translation, having published The Transparent Eye (1993) and Borrowed Plumage (2003). He is Emeritus Professor at Lingnan University and Indiana University.


1. This paper was prepared for the special issue on Translation in the Yearbook of Comparative and General Literature, and was presented as a Confucius Institute lecture at Rutgers University, March 31, 2009. I am indebted to Ms. Wann Aijen and Mr. Lie Jianxi for reading a draft, and for making many helpful suggestions and corrections.

2. This translation renders 孤帆遠影碧空盡; but, as Wann Ai-jen pointed out to me, the original should read 孤帆遠影碧山盡.

3. Wang Ming (汪铭), “On Style Reproduction in Poetry Translation”, M. A. Thesis, 武汉理工大学硕士论文, 2006. 55–56: (56); accessed Dec. 25, 2008. Lest one be accused of indulging in schadenfreude, honesty compels me to confess that I made the same mistake in my initial attempts to render the poem.

4. Peter Boodberg, Cedules from a Berkeley Workshop on Asiatic Philosophy, Berkeley, 1954.

5. The implications of this statement, that the text of the Daodejing does not differentiate between third-person male and female, is incorrect. Weijong Wayne Lu, a personal friend, has calculated that ‘He’/’him’/’himself’ occurs a total 122 times, in 23 chapters: whereas ‘she’/’her’/’herself’ occurs a total of 66 times, in 19 chapters.

6. Borges remarks somewhere that those who rely too much on a dictionary make the false assumption that there are semantic equivalents in different languages.

7. Lie Jianxi, who has been translating Beckett into Chinese, offers a reverse and entirely apposite example: in Waiting for Godot, there is a reference to the English “music-hall”; Lie had initially thought to translate this, literally, as yinyue ting, but then thought better of it, because yinyue ting connotes only the music, but not the low-life bawdiness of the English music-hall.