On the Art of Translation

馬悅然論：翻譯的藝術

There are four kinds of translators:

Firstly, there are the scholar/translators, to whom the translation of a text serves as the final argument in a piece of philological research. The structure of such translations normally strictly follows that of the original text, with the addition of square brackets, indicating words which the target language forces the translator to include. My venerated master professor Bernhard Karlgren’s translation of the ancient Chinese anthology Shijing (The Book of Odes) which goes back to the first half of the first millennium before our era, is literal, and is meant to be literal, rather than literary. Had Karlgren chosen to produce a literary translation, he would most certainly have been able to do so. Already in lower middle school, Bernhard Karlgren proved himself a master of translating Latin and Greek poetry into an exquisite Swedish, which closely mirrored the metrical structure of the original poems. His renderings of Classical poetry are far superior to those produced, at about the same time, by venerated professors of Latin and Greek. In his verse-drama The Golden Fleece, produced when Karlgren was still in high-school, he used a variety of classical metres. The drama contains a long passage in faultless hexameter, with many highly effective enjambments. (Bernhard Karlgren’s father was a high-school master in Classical languages and insisted on sometimes speaking Latin to his sons at home. He gave up that idea after having heard a seven-year old son produce the following sentence: “Här kommer (Here come) puellae vicinitatis!”)
The Shijing has in its entirety been translated into English also by the eminent British translator Arthur Waley and by Ezra Pound. Waley had a good knowledge of both Chinese and Japanese. Ezra Pound did not know Chinese. In his translation of the Shijing he based himself on others’ translations and on his poetic intuition. Sometimes he trusted in the strange notions of Ernest Fenollosa (1853-1908), an American orientalist and poet who spent many years in Japan. Much thanks to Ezra Pound, Fenollosa’s essay “The Chinese Characters as a Poetic Medium” was posthumously published in 1920. The essay had a great impact on translators of Chinese poetry in the Western world, especially on Florence Ayschough. Fenollosa and Pound both asserted that the graphic form of the Chinese characters contains rich semantic information. Fenollosa based his theory on the mistaken notion that Chinese characters are pictograms and ideograms, depicting things and ideas. True, the Chinese script does contain pictograms and ideograms, but they amount to far less than one percent of all characters. The Chinese characters are best described as logograms, each denoting a monosyllabic unit, a morpheme.
The 113th poem in the Shijing, in which poor farmers bitterly complain about cruel tax-collectors, has the title Shishu. Shu means “rat”. What shi means in this context nobody knows. Waley, following an early gloss, translates shishu as “big rats”. Karlgren suggests that shi was a kind of rodent and refrains from translating it. The character for shi 碩 is made up of two graphs: the graph for “stone” 石 on the left and the graph for “head”頁 on the right. Now, listen to Ezra Pound’s magnificent translation (if that is what it is):

Rats,
Stone-head rats lay off our grain,
Three years pain,
Enough, enough, plus enough again.
More than enough from you, deaf you,
We´re about thru and ready to go
Where something will grow
Untaxed,
Good earth, good sown,
And come into our own.

There are many truths in this world. The poetic truth which Ezra Pound arrived at in his translation of this poem is no less true than that which Bernhard Karlgren arrived at through his meticulous philological methods.
Secondly, there are the professional translators who earn their living, and often a rather meagre living at that, from their craft. Whether they are employed or free-lancing, they cannot always themselves choose works to translate.
Thirdly, there are the amateur translators, to whom translation is a work of love. (The term “amateur” is here used in the sense of “One who cultivates anything as a pastime, as distinguished from one who prosecutes it professionally”). I consider myself lucky to belong to this category. I have never yet translated a single piece of literature that I did not appreciate and consider worthy of translation.
There is a fourth kind of translator, a writer/poet cum translator, a rather dangerous species, to whom I shall return later on.

Now a few words about the requirements of a translator: he or she must have an excellent command of his mother-tongue (into which he or she normally translates) and a very good command of the source language from which he/she translates. He/she must have a keen ear for the euphonic subtleties of the two languages, and especially for the paramount role that rhythm plays in both poetry and prose. 
The translator must also be keenly aware of his/her double responsibility, to the author of the original text, and to his own readers. He must not manipulate the text which he translates, nor must he add anything to or distract anything from the original text. The translator should aim at imitating the author of the original text, and his translation should be a likeness of the original work. Even though the literary qualities of a translation at times, and for various reasons, may appear superior to those of the original work, the translator must never consciously strive to excel the author。
Literary appreciation is of course highly subjective. What one reader considers excellent, another reader may consider as trash. In the first quarter of the 20th century, Lin Shu (1852-1924), who himself did not master any foreign language, translated some 130 works of Western literature into classical Chinese, with the aid of a competent assistant. His translation of Lamb’s Tales from Shakespeare was instrumental in creating a great interest for the study of English literature. Dr. Hu Shi (1891-1962), the initiator of the Literary Revolution in China, has the following to say about Lin Shu’s translations: “It was a tremendous task and exceedingly amusing to read the comic figures in the novels of Charles Dickens talking in the dead language of two thousand years ago.” Arthur Waley voices a diametrically opposed view in his assessment of Lin Shu’s translations of works by Dickens: “Dickens inevitably becomes a rather different and to my mind a better writer. All the overelaborations, the overstatements and uncurbed garrulity disappear.” The famous scholar-writer Qian Zhongshu (1910-1998), who achieved a complete mastery of the major European languages, has somewhere stated that in his youth he preferred to read Rider Haggard’s (1856-1925) novels in Lin Shu’s translations.
American friends of mine, who engage in translation, have told me that their publishers not seldom interfere with their work, demanding that certain passages be deleted, lest they offend the taste of the reading public. The American translation of the Chinese writer Lao She’s novel Luotuo Xiangzi (1936) is a case in point. The novel describes the total degeneration of a once healthy and happy country lad, who settles in Peking in the hope of one day being able to buy a rickshaw of his own. In spite of all his attempts at securing a better life, he loses everything and ends up as a miserable wreck. The American translator substitutes a happy end à la Hollywood for the utterly tragic final chapter of the book. Lao She, whom I knew well in the mid 1950s, was very much upset by this manipulation of his work.

The Foreign Languages Press in Peking has produced a great many translations of Chinese works of literature into European languages, undertaken by Chinese translators. The results are as a whole far inferior to works translated from European languages into Chinese, also by Chinese translators. I do not suggest that translators should abstain from translating literature into what is not his or her mother tongue. Lin Yutang (1895-1976), whose command of English was superior to that of even the most highly educated Englishman, proved that it can be done. 

The gaps in the translations of Chinese literature into Swedish are so many and so wide that they can hardly be detected! Very little of Chinese literature prior to the May Fourth Movement of 1919 has been translated into Swedish. Bernhard Karlgren was throughout his life too heavily engaged in research to find the time to indulge in translation of works of Chinese literature into Swedish. Karlgren’s translations are limited to one essay by Tao Yuanming (365-427), one essay by Ouyang Xiu (1007-72), a collection of stories from the Ming period (1921) and extracts from the works of ancient philosophers (Confucius, Lao Zi, Zhuang Zi, Lie Zi and Mo Zi, presented in a monograph on Ancient Chinese thought (1929).
Only two Swedish sinologists are rather heavily engaged in the translation of Chinese literature, myself and Anna Gustafsson Chen, married to Chen Maiping, scholar and writer. In the relatively short period of nineteen years she has translated nearly twenty works of contemporary Chinese literature, including three long novels by Mo Yan (The Red Fields and The Garlic Ballades, Life and Death are Wearing me out), one novel by Su Tong (The Red Lantern), two novels by Yu Hua (To Live and Xu Sanguan sells his blood) and three novels by the women writer Hong Ying (The Generation of Naked Dancers, Starving Daughter and K, a Love Story), two novels by Ma Jian, one novel by Han Shaogong and one work by Chen Ren. She has also translated important documentary literature, such as Wei Jingshen’s The Courage to Stand Alone). Her achievements are all the more impressive as she works full-time as librarian in the Stockholm City Library.
My own translations cover fairly wide fields, such as Ancient poetry (selections from the anthologies Shijing (The Book of Odes) and the Chuci (Songs of the South), from the third century B.C., popular poetry from the Han periods (206 BC-220 A.D.), lyrical poetry of the Tang period (618-906) and polymetric poetry of the Song period (960-1279). I have translated two of the traditional Chinese novels, the Shuihuzhuan , the earliest parts of which go back to the Yuan period (1280-1367). (The novel has been paraphrased into English by Pearl Buck, under the title All Men are Brothers, and translated by J. H. Jackson, under the title Water Margin ) and the 16th century Xiyouji (The Journey to the West), excellently translated into English by Anthony Yu. A few years ago I violated a pledge made in my youth never to translate the Daoist Classic Dao De Jing into Swedish.
As far as modern and contemporary literature is concerned, I have chosen to translate authorships rather than separate works. Of modern writers of fiction I have concentrated on Shen Congwen (1902-1988), to my mind the greatest novelist of his time. My translations of his works include the novel Biancheng (Border Town), his autobiography (Congwen zizhuan) and a great many short stories.
I have translated a great deal of the works by two modern poets, Wen Yiduo (1899-1946) , who started writing in the 1920s, and Ai Qing (1910-1996), to my mind the greatest poet of the 1930s. An anthology of mine, published in 1971, includes works by 42 poets of the 20th Century.
In the last two decennia I have concentrated on translating works by contemporary writers. I have translated about 95 % of the works by the Nobel Laureate Gao Xingjian (his two novels Lingshan (The spiritual Mountain), Yige ren de shengjing (One Man’s Bible), all his short stories and fourteen of his 18 dramas. I have also translated four major works by the Shanxi writer Li Rui (Houtu (Sovereign Earth); Wu feng zhi shu (Windless Trees); Wanli wuyun (Clear Skies), and Jiuzhi (The Old Place). In recent years I have translated the Shanxi writer Cao Naiqian’s magnificent novel Dao heiye xiang ni mei banfa (When night falls I can’t help longing for you), a gruesome tale of the life of destitute farmers in a poor mountain village in northern Shanxi during the cultural revolution, and his collection of stories Zuihoude cunzhuang (The last homestead). 
As to contemporary poetry I have translated almost the entire oeuvre by Bei Dao and much of the poetry written by the so-called “misty poets” (menglong shiren), who started to publish their works towards the end of the 1970s. In recent years I have taken a great interest in the works by Taiwanese poets. My Swedish anthology of translations include works by nine Taiwanese poets (Chi Hsien, Lo Fu, Yü Kuang-chung, Shang Ch’in, Ya Hsien, Lo Men, Yang Mu, Lo Ch’ing and Hsia Yü). I have also collaborated with my friend Michelle Yeh of the University of California at Davis in the editing of a large anthology of Taiwanese poetry, translated into English (Frontier Taiwan. An Anthology of Modern Chinese Poetry. New York: Columbia University Press, 2001, 490 pages). In 2011, I published my translation into Swedish of works by the Taiwanese poet Yang Mu, to my mind the greatest poet of our time, writing in Chinese (Den gröne riddaren, Dikter av Yang Mu/The Green Knight. Poems by Yang Mu. Stockholm: Bokförlaget Tranan, 500 pages.).

Each translator approaches work in his or her own way. To me, translation is a work of love. (I should add that my amateur status affords me the privilege of myself choosing what I want to translate.) Before starting to translate, be it a long novel or a poem, I read the work several times in succession, in order to get a feeling for the structure and the flow of the text. While reading and re-reading the text I make mental notes of passages which I know will present a challenge, and ponder over how they might be best translated. I always articulate the text silently when I read, which gives me a sore throat at the end of a long day’s work. The repeated readings make me feel the presence of the author, and the author’s voice. When I eventually arrive at the point when I feel that my own voice, and breathing, are in harmony with the voice and breathing of the author, then the work is almost done. I am aware of the fact that my notions of the author’s voice and breathing may sound like hocus-pocus to you. I am at a loss to explain how it works, but I do know that it does. Once I feel that I have reached this stage, I am ready to devise a language and a style to match those of the original text.
I once discussed this method of translation with a Swedish colleague, who happens to be an excellent translator. He objected that this method would deprive him of the pleasure of surprise and unexpected encounters: to him, turning pages in the book he translated was like following a meandering mountain path, not knowing what view might unfold itself beyond the next bend.

When I read a book, or translate a book, I acquaint myself thoroughly with the milieu and with the people who inhabit it: some may become my bosom friends and others my enemies. I would dearly like to share a few pints of wine with Lu Zhishen, one of the heroes of the Shuihuzhuan, but I would hate to spend the night at a Shandong inn together with Song Jiang, the leader of the robbers’ gang. I refuse to see a movie based on a book that I have enjoyed reading, fearing lest the milieu, and the people portrayed in the film, would differ too greatly from those in my own imagination.
The eminent British scholar and translator Arthur Waley, who translated a great many works of Chinese and Japanese literature into his mother-tongue, refused to visit China and Japan. I guess that he wanted to avoid a collision between his imagined world and that of reality.
I have already referred to my translations of works by the Shanxi writer Li Rui, depicting life in a poor mountain village in the Lüliang Range. In the Spring of 2004, Li Rui wrote me a letter, saying that deposits of coal had been found in the close vicinity of the village, which meant that it was doomed to be annihilated. I immediately made up my mind to visit the village, which I did, together with my wife and my friend Li Rui in August of the same year. The village comprised 22 households and altogether about one hundred people. Among them were several of the men and women which Li Rui had portrayed in his books. The landscape and the village were more or less what I had expected. I was told that the villagers had been persuaded by their local Party secretaries to sell their land to a mining enterprise. According to the contracts, the farmers were to receive the annual sum of 500 RMB for each mu of land, equal to one sixth of an acre, for the period of ten years. In addition, they were offered jobs in the mines and on the construction of roads which had to be built to transport the coal. The villagers did not receive a single cent for their land, and none of them were offered jobs in the mines or on the roads. I learnt more on the trip to that village than hundred volumes dealing with rural economy and the plight of Chinese peasants could have taught me. 

I would now like to discuss some of the problems which I have encountered during my fairly long career as translator. In the mid 1970s I translated the medieval picaresque novel Shuihuzhuan (All Men are Brothers). The novel has grown out of the repertoires of traditional story-tellers and it is written in a plain colloquial language which must have been very close to the spoken idiom of northern China in the 16th century. In a review of my translation a critic takes up the following question for discussion: should a translator allow the idiom of the original text to re-emerge in his translation or should he do whatever he can to create the impression that his translation is an original work in his own language? In other words, should the Swedish translation of a medieval Chinese novel sound as if it were written in Sweden in the mid 1970s? In order to answer that question I shall have to make an excursion into the literary history of China. In the development of Chinese literature we can clearly distinguish two parallel currents: polite literature, which was written in the classical language and which carried an enormous prestige, and popular literature, which developed from oral traditions and which was written down in an idiom close to the spoken language. Polite literature and popular literature as a rule utilized different genres.
The grammar and diction of Buddhist legends and sermons of the 9th century do not differ considerably from the grammar and diction of present-day Chinese. Towards the mid 13th century, the phonological structure of Northern Chinese was more or less identical with that of Putonghua, or Common Language of today. I feel rather embarrassed to admit that I recently found out that I have greater difficulty in reading a Swedish verse chronicle from the early 14th century than I have in reading a 14th century Chinese text.
When a listener or a reader enjoys a literary text, a relationship, or a dialogue, is established between him and the author of the text. In my opinion, the task of the translator is to facilitate the establishment of a similar dialogue between his translation and the readers. A translator of works of Chinese literature who allows the distance in time and space to characterize the style of his translation is guilty of sacrificing literary values on the altar of chinoiserie. Naturally, a translator should be free to use archaisms in both diction and grammar, but only to the extent that they are used in the original text.
I am sometimes asked whether it is not terribly difficult to translate Chinese literature into Swedish. My answer is that the difficulties vary from text to text. Translating the Shuihuzhuan, I had great difficulties in making out the shapes of head-gear and shoes. In that regard, I was greatly assisted by wood-cuts from the 16th century, accompanying a Ming edition of the novel. Translating the novel Lingshan (The Spiritual Mountain) by Gao Xingjian, I had great difficulties in identifying his many references to flowers. With the aid of learned friends in the Swedish Museum of Natural History I was able to solve most of the problems. But when five names of flowers remained unidentified, I wrote to Gao Xingjian and asked him to draw the plants (he is after all an excellent painter!). His answer was: “Hah, I made up those names myself!”

I have already referred to the need for the translator faithfully to mirror the rhythm of the original text. The problem is that the difference between two languages sometimes is so great that any attempt to transpose poetic forms from the one language into another is doomed to fail. Some 40 years ago I experimented with a kind of a-syntactic translation of Classical Chinese poetry into Swedish. I simply transposed the words of the original into Swedish, with an utter disregard for the syntactical demands of the Swedish language. Here is an example, a short lyrical poem entitled “River Snow”, by the Tang poet Liu Zongyuan (773-819), transposed into a-syntactic English:

Qian shan, niao fei jue,
Wan jing, ren zong mie.
Gu zhou, suo li weng,
Du diao, han jiang xue.

Thousand mountain, bird fly sever,
Ten thousand path, man footprint extinguish.
Solitary boat, rain-cape, bamboo-hat, old man
Alone fish, cold river snow.

When you translate classical Chinese poetry into a Western language such as Swedish or English, the target language forces you to specify what is not specified in the original. You have to decide for yourself whether the nouns should have definite or indefinite reference, and whether they should be given singular of plural form. Tense is not formally expressed in classical Chinese. But the translator’s Western language forces him to decide whether the action or the state referred to pertains to the past, the present or the future. The universality and the timelessness which characterize the original Chinese poem are lost in translation.
These typological differences between classical Chinese and Western languages are not the only obstacles facing a translator of Chinese poetry. Chinese is a tonal language. The four tones of the National Language of today are either falling, or rising, or both falling and rising, and neither falling nor rising, that is level. For metrical purposes these tones are divided into two categories, level and non-level. The rigid rules applying to certain types of Classical Chinese poetry require a fixed alternation between level and non-level tones within the verse and within the stanza. This tonal variation cannot, of course, be echoed in a non-tonal language.
Rhyming plays a very important role in all types of pre-modern Chinese poetry. Since Classical Chinese is a monosyllabic language, in the sense that all morphemes consist of only one syllable, only male rhymes are employed. All rhyme words must belong to the same tonal category. Now, rhyming comes very easily to the Chinese poet. The National Language of today possesses only 420 different syllables, discounting tonal distinctions. If we take the tonal distinctions into account, the distinctive syllables would number about 1300. This means that homophones and potential rhymes are exceedingly common in the Chinese language. Any attempt at rendering long sequences of rhymes in a translation is bound to fail.
While Classical Chinese is monosyllabic, in modern Chinese most words comprise two or more syllables. When translating modern Chinese poetry, it is therefore easier to achieve a closer correspondence between the original verse and the translation. Wen Yiduo (1899-1946) asserted that the form of the poem must satisfy both the eye of the reader and the ear of the listener: the musical features of the spoken language must serve as building stones in a structure of architectonic beauty. The title poem of the collection Sishui (Deadwater) was by Wen Yiduo himself considered to be his most successful experiment in the field of metrical architecture. Each verse in the poem consists of nine syllables, which are grouped into three two-syllabic units and one three-syllabic unit. The three-syllabic unit changes position from verse to verse, without ever occupying a verse-final position. This feature creates a rhythmical tension in the regular structure. The first verse in the original reads as follows in the original:

這是一溝絕望的死水
Zheshi yigou juewangde sishui

Here’s a ditch of hopeless deadwater.

One English translation of this poem amounts to nothing less than cold-blooded murder of one of the finest poems in modern Chinese literature:

“This is a ditch of hopelessly dead water”.

This translation of the first line utterly destroys the rhythm of the original verse. It also treats the compound noun sishui (deadwater) as a phrase, “dead water”. Finally the adjective juewangde, “hopeless”, has been converted into an adverb. This one example goes a long way to show that a translator must have a firm grasp of the linguistic aspects of the text which he is translating.
At the beginning of my talk I referred to a fourth category of translators, the writer/poet. We sometimes find that some translators who are themselves writers or poets tend to take liberties with the original text, which other translators would not dare to do. I shall give you an example of this. A volume of Robert Bly’s translations of the Swedish poet Tomas Tranströmer’s poems entitled Tranströmer, 20 poems (1970) contains the poem “Balakirev’s Dream”. The seventh stanza of this reads as follows in the original:

Det var ett fält där plogen låg
Och plogen var en fågel som störtat.

(There was a field where the plough lay
and the plough was a bird that had crashed)

Robert Bly translates as follows: 

A field appeared in which a plow stood,
And the plow was a bird just leaving the ground.

I remember being quite upset when I read that translation. It seemed to me that the translator had perverted the exquisite imagery of the original: the deserted plough resting on one of its shafts, with the other shaft raised at an angle of 45 degrees, exactly like a bird with one broken wing and the other wing ready for flight. Bly’s translation forces the reader to accept a plough not lying, but standing in the field with the plough-share firmly embedded in the ground and stretching its shafts like a bird ready to rise in the air. When I discussed this translation with my friend Tomas Tranströmer, he assured me that he had given his assent to the change and that he thought that Robert Bly’s translation was quite successful.

It does happen that a translator gets stuck in his text. He simply doesn’t understand what the text says, and no dictionary can help him to elucidate the incomprehensible passage. If he cannot solve the problem he will have to give up his intention to translate the text.
In some cases the text may on the surface appear perfectly comprehensible and yet the translator has difficulty in grasping and conveying its meaning. In the title poem of the collection Heihe (Blackbox, 1990), the Chinese poet Bei Dao speaks of a “hujiao huangdi”, a “pepper emperor”. The poem reads as follows in my translation into English:

who is it that waits
for the preordained sunrise

I shut the door
the interior of the poem darkens

in the middle of the table
the pepper king rages

a piece of music memorizes me
and puts down its burden

the inner parts of the watch are scattered
over the royal horizon

event linked to event
passes through the tunnel

When I first read the poem I was quite bewildered. What is a “hujiao huangdi”? The emperor of all the peppers in the world? An emperor made of pepper? And why does he stand in the middle of the table? Bei Dao wrote this poem in Sweden. In his kitchen there was a round table, providing the horizon of which the poet speaks. In Sweden pepper and all kinds of ground spices are normally bought in small glass jars with screw-lids which resemble a royal crown. To Bei Dao the pepper jar resembled the king in chess.

The story of the translation of Swedish works of literature into Chinese is easily told. In the 1920s, works by August Strindberg (1849-1912) and Selma Lagerlöf (1858-1940) were translated second-hand, from translations into English. The following story may illustrate the difficulty in translating second-hand. In an English translation of one of Strindberg’s works we find the following passage: “When we were enjoying ourselves, the uncle of the baron joined the company.” This seemingly very simple sentence presents unsolvable difficulties for the Chinese translator. Was the uncle the brother of the baron’s father or of his mother? Had the translator had access to the Swedish original, the problem had been easily solved, since the Swedish language distinguishes between “farbror” (father’s brother) and “morbror” (mother’s brother). But having ascertained that the baron’s uncle was the brother of his father, the translator is faced with yet another difficulty: was the uncle older or younger than the baron’s father? In the Chinese highly clan-centred society it is necessary explicitly to state the relative ages of members of the clan.
It is only in recent years that Swedish literature has been directly translated into Chinese. The Nestor among Chinese translators from Swedish is Li Zhiyi, who has translated many major works by August Strindberg, Selma Lagerlöf and Astrid Lindgren (1907-2002), a famous writer of children’s books. He has also translated essays and short stories by the Swedish Nobel Laureates Verner von Heidenstam (1916), Erik Axel Karlfeldt (1931), Pär Lagerkvist (1951), Eyvind Johnson and Harry Martinson (both 1974). Chen Maiping has in the last two years translated Harry Martinssons Aniara, the volume Air Mail, containing the correspondence between Tomas Tranströmer and his American translator Robert Bly, and Kjell Espmark’s novel Glömskan (Oblivion).
Wang Ye, a woman translator who lives in Sweden, last year translated Hjalmar Söderberg’s novel Doktor Glas into Chinese and is now completing her translation of Kjell Espmark’s novel  Béla Bartók mot Tredje riket (Béla Bartók against The third Reich). My own translations into Chinese of Tomas Tranströmer’s two latest poetry collections, Sorgegondolen (The Sad Gondola) and Den stora gåtan (The Great Enigma) together with his autobiographical sketch Minnena ser mig (The Memories see me) were published in 2012.

Translators are assiduous workers in the literary vineyards and they play an exceedingly important role as builders of bridges between cultures. Without translators, the world would be a much poorer place to live in. Someone has said that World Literature is Translation. It is highly unfortunate, and greatly unfair, that their unique contributions do not, as a rule, receive due recognition. They are often underpaid, and unlike writers and poets they more rarely figure as recipients of prizes and awards. It is my sincere hope that this unfortunate situation be remedied.