Inverse translation is “a term used to describe a translation, either written or spoken, which is done from the translator's native language, or language of habitual use” (Shuttleworth & Cowie 1997: 90). It is the opposite of direct translation, which refers to the translation done into the translator's native language, or language of habitual use (Shuttleworth & Cowie 1997: 41). Inverse translation is also named “service translation” (Newmark 1988: 52). Though it is a common phenomenon in many parts of the world, especially in Russia, China, Germany, and some East- European countries, inverse translation has long been in the periphery in modern translation studies. The names it's given suggest something about its “nature” and status: “inverse” could mean that it goes against common sense and moves in an uncharted direction, and “service” brings to our minds words like “servant”, “servitude”, implying it can only play a supporting role. Whether declared openly or assumed tacitly, inverse translation is something sensible translators should avoid doing if they do not want to get humiliated. And most of the translation theories that have come out so far are based on and provide guidance only for translation practice that is conducted from a foreign language into one's mother tongue—if they take into account translation practice at all. This biased stance concerning inverse translation is not conducive to the development of translation studies, and runs contrary to translation reality, where translation from one's mother tongue into a foreign language is possible, permissible and some times even desirable.

The reasoning behind those who see inverse translation in the negative light is that a translator's mastery of a foreign language is seldom, if not never, sound enough to ensure an easy maneuver of it, so the readability of the translation will be in question. This mind-set puts too much emphasis on expressiveness and reception of translation, without taking into account the translators' comprehension of the source text, an integrated part as important as expression in any translation process. The argument against this view could be: if the translators translating out of their mother tongues into foreign languages do lose points at the expression end (this is subject to dispute), what they gain at the comprehension end may well offset the losses, leaving this controversy open for further probe.

But has inverse translation always been in such a pitiful state, hence something close to plague in the history of translation?

Contrary to common belief, the principle that translation should always be done into one's mother tongue does not have a long history. On the contrary, translation into a nonmother tongue can also be found at the dawn of Western history” (Pokorn 2005:34).

Inverse translation was also practiced in early periods of China. According to some research the first translations of Buddhist texts from Sanskrit into Chinese in the second century AD were done by foreign missionaries, not by Chinese native speakers (Baker 2009:85; Chu 2000:43-53). Many other records in the history of translation also show that inverse translation has been a common practice, not something that is abnormal and will inevitably result in deformed or inferior product. The relegation of inverse translation to its present peripheral state probably started with Martin Luther (1483—1546), who assumed that the best translations were always into the mother tongue, and the translation out of the mother tongue could be regarded only as a pedagogical exercise (Baker 2009: 85).

The questions to be asked concerning inverse translation after this introduction are understandably whether its product is indeed inferior to that of direct translation, and whether inverse translation shows some distinctive features of its own which are incompatible to direct translation. To find immediate and universally accepted answers to these questions is by no means easy, but some researchers have ventured into this field and come up with thoughtprovoking results. Stuart Campbell, for example, by looking into the status quo of the field of translation in Australia, provides the first comprehensive discussion of translation into one's second language. It's suggested that “second language translation output be seen as a development system rather than a substandard version of some ideal target” (Campbell 1998:175). And in May 1997 an international conference was held at the University of Ljubljana “on a subject which had long been a taboo for translation theory: Translation into Non-Mother Tongues”, and the collected papers, most of them questioning the fixed ideas and prejudices, were published in 2000 (Grosman et al 2000). The most recent and indepth probe into inverse translation is conducted by Nike K. Pokorn. Beginning with cross-examining concepts like “native speaker”, “mother tongue”, and “bilingualism”, she then makes a well-designed comparative study of the English translations of Slovene prose works by Ivan Cankar done respectively by native speakers of English or Slovene, nonnative speakers of English or Slovene, and pairs of translator, and arrives at the conclusion:

The quality of the translation, its fluency and acceptability in the target language environment depend primarily on the yet undetermined individual abilities of a particular translator, on his/her translation strategy, on his/her knowledge of the source and target cultures, and not on his/her mother tongue and the direction into which he/she is translating” (Pokorn 2005:XII).

We can also see from her research that inverse translation does not show obviously exclusive traits which set it entirely apart from direct translation.

In the discussion of inverse translation, as is pointed out in the works of Stuart Campbell and Nike K. Pokorn, the following factors have to be taken into account: the discrepancy between theory and practice and the gap between ideal and reality. Theoretically and ideally speaking direct translation may be desirable, but in the real world inverse translation is sometimes indispensable. In countries and communities where “minor” languages are spoken it is often the case that there are not enough qualified translators who can translate from their foreign languages (the minor languages like Vietnamese and Slovene) into their mother tongues (the major languages like English), so they have to depend on their own translators who are called upon to translate out of their mother tongues if they are to get involved into the broader world stage culturally, economically and politically. With variations this is also true of China. The Chinese language is certainly not a minor language: in terms of the number of people speaking the language, it might be considered to be one of the most important languages around the world. In terms of its influence and permeation, however, it is still a language in the periphery even though more and more people are now beginning to get interested in Chinese culture and to learn Chinese. Compared to “hegemonic” English, in Venuti's term, Chinese is still one of the least translated languages. In 1987, for example, the global translation output was approximately 65,000 volumes, more than 32,000 of which were from English, and only 216 were from Chinese, other least translated languages including Arabic (479), Bengli (89), Korean (14), and Indonesian (8) (Venuti 1998:160). If we take into account China's economic status and long history this figure may bear more significance. The causes behind this imbalance are many, and one of them is that there are not enough English translators who can work from Chinese into English. This is also true of translation of Chinese into other major languages.

The dispute over the legitimacy of inverse translation in the West is not at all uncommon in China, especially in the latter half of the last century. The opponents of inverse translation share the similar reasoning with its Western counterparts, focusing on the expression end of the translation process while attaching little importance to if not ignoring completely the step of comprehension, and thus believing that only sinologists are qualified translators of Chinese literature. They can find support from the sayings of some famous sinologists themselves, who do not see much value in Chinese translators' translating out of Chinese. One of the most often quoted sinologist is Angus Charles Graham, who said in Poems of the Late T'ang: “ … we can hardly leave translation to the Chinese, since there are few exceptions to the rule that translation is done into, not out of one's own language” (Graham 1965: 37). This sentence has become a golden principle, as it were, which the opponents of inverse translation often turn to, and a long time curse to the Chinese translators who dare to translate out of Chinese.

The proponents of inverse translation also have their own points to make, and sometimes can even gain some ground in this prolonged debate. Some of them point out the mistakes the sinologists made in translating Chinese classic works, suggesting that what has been introduced to the West by the sinologists is often an incomplete or fake portrait of China and Chinese literature. Having given some examples of inadequate translation done by sinologists, Yu Guangzhong says:

It has been almost a century now² since the publication of A History of Chinese Literature by Herbert Giles, but the English translation of classical Chinese literature is still far from satisfying whether in terms of quality or quantity. …. With all that has been achieved by the British and American sinologists in translating Chinese literature to the West, it's time now that native Chinese translators played their parts” ( Yu 2002: 66-81).

There are also those advocates likes Pan Wenguo, a scholar in East China Normal University, who conducts an in-depth analysis of the pros and cons of inverse translation and comes to the conclusion that it is not only feasible but also desirable for native Chinese translators to translate Chinese literary classics into foreign languages (Pan 2004:40-43).

The controversy over inverse translation has been gaining more momentum in recent years as China sets to introduce to the outside world its traditional culture and social and economic changes that have taken place after its initiation of reform and opening up since the end of 1970s. Eager to hasten the introduction process and above all to increase the acceptability of the translated Chinese literary works among foreign readership, some scholars suggest that more emphasis should be given to direct translation as illustrated by the following statement:

The sinologists are undoubtedly the ideal translators in translating Chinese literary works into other languages, and domestication strategy aimed at accuracy, readability, and acceptability should be taken as the common pursuit for all translators (Hu 2010:15).

This suggestion, with all its warm embrace of direct translation and domestication, is typical of the views of the opponents of inverse translation in China. It sounds plausible and goes well with the present-day overriding pursuit of profit in the publishing industry. But so sweeping a statement should be subject to further probe and modification considering the complexity in translation. Firstly, the quality and value of translations of Chinese literary works, or in fact of any translation for that matter, should not be based solely on their reception by the readership. Wholesale domestication, while catering to target readers, often fails to show due respect for the source culture, and though widely practiced in the US it is not at all the only means of translation. In fact domestication has been severely criticized in favor of foreignization by some scholars for its erasing the cultural and linguistic differences (Venuti 1998). Secondly, sinologists do not necessarily pursue domestication when translating Chinese literary works: the translation strategy employed is the result more of personal choice than of language affiliation. As is shown in the following case studies, sinologists may resort to foreignization and attempt to maintain the otherness, while native Chinese translators may, contrary to “common belief”, give priority to smoothness and readability in their translations.

The misunderstanding and prejudice against inverse translation in China, i.e., translation done by Chinese translators, may be summed up in the following points:

• Translations done by Chinese translators often sound foreign and are less readable;
• Translations done by Chinese translators put more emphasis on transferring Chinese culture into the target language, sometimes without considering the needs and wants of the target culture, and thus may result in resistance from the readership;
• Sinologists often stumble in comprehending the Chinese original while Chinese translators often show an inadequacy in the use of the target language.

To demonstrate that the above points on inverse translation are biased and based more on speculation than on facts, we will have a brief comparative study of the English translations of Wenxin Diaolong, a classical work of Chinese literary criticism written by Liu Xie³ . We will look at some examples taken from the three versions⁴ done respectively by American sinologist Stephen Owen and Chinese translators Shi Youzhong (Shih Vincent Yu-chung )and Yang Guobin.

• This, in short, is the first step in the art of writing, and the main principle employed in the planning of a literary piece. (Liu 1959:155)
• This is the foremost technique in directing the course of wen, the major point for planning a piece (Owen 1992:209)
• This is the foremost art of writing and a main feature of composition. (Liu 2003:377)

The sentence is taken from Chapter 26 (神思), a chapter mainly concerned with the importance of imagination, as well as the relationships between imagination, construction and diction of literary works. According to Liu Xie knowledge in this field is essential to successful writing. In the original sentence, 驭literally means “drive”, 文means “writing”, 术 means “art” or “technique”, 谋means “plan” or “design”, 端 means “end”. Compared with many other sentences in the book, this sentence is among the easiest to understand and to interpret. All the three versions successfully transfer the general meaning of the original, but are quite different in structure and diction—Shih and Owen try to re-present the details of meaning in the original by translating every word of it, while Yang's version just keeps to the essence, neglecting words like驭and谋 for the sake of keeping the version transparent and fluent.

• Now to be thin in ideas and fat in words, or confused and disorganized without unity, are sure signs of lack of this kind of bone. (Liu 1959:163)
• If the truths are emaciated but the phrasing fat—a profusion indiscriminately mixed and lacking all governing coherence—then we see no evidence of bone. (Owen 1992:227)
• Shallow thought, profuse language, and poor organization betray the lack of “bone”.(Liu 2003:401)

This example is taken from Chapter 28 (风骨), which is an illustration on two important elements indispensable to good literary writings: 风 and 骨 (the sentence we choose covers only 骨). Like many other Chinese literary terms, 风 骨, which literally means “wind and bone”, defies definition in other languages. Even in Chinese context there is controversy over its exact meaning. What may be accepted about its interpretation is that 风 is chiefly about style while 骨 is concerned more with diction and structure of a literary work.

The three versions, like example 1, again show no distinctive features that can tell the translation by Owen apart from that by the two Chinese translators, though they do have their own characteristics that result from the respective translation strategies employed. The versions by Shih and 5 Owen keep all the three organic metaphors⁵ in the original: 肥—fat; 瘦—thin/emaciated; 骨—bone, and both attempt to transfer every nuance of meaning of the original. There are differences between them as well: the latter is more formal in the choice of words and more complicated in structure. Yang's version, with as few as 12 words (compare that with 26 words of Shih and 25 of Owen), is very compact; it uses three parallel structures, keeps only one organic metaphor and explains the meaning of the other two.

• But one can be considered a good musician only after one has played a thousand tunes, and a collector of arms can be considered a connoisseur only after he has seen a thousand swords; so broad experience and learning are the sine qua non of true wisdom. (Liu 1959:261)
• You can understand sound only after playing ten thousand tunes; you can recognize the capabilities of a sword only after examining a thousand. You must first endeavor to observe widely in order to have the impression (xiang) that comes from comprehensive understanding. (Owen 1992:303)
• An understanding of music comes from playing a thousand tunes. The ability to judge fine weapons comes after seeing a thousand swords. Insight and perception are based on broad observations. (Liu 2003:693)

The two sentences are taken from Chapter 48 (知音) of Wenxin Dialong, a chapter about evaluation and appreciation of literary writing. 知音 literally means “knowing the sound”. The differences between the three versions are obvious, but we find it hard to group them neatly into the two camps of translators mentioned above. Departing from what is often expected of sinologists in translating Chinese, Owen tries every means to mirror the verbal meaning of the original: to translate 声 literally as “sound”, 象 as “impression”, and 博观 as “observe widely”, for example. So his version bears a touch of foreignness, if not “clumsiness” . The versions of the two Chinese translators are not only different from that of Owen but also from one another. Shih's version is more detailed, with such interpretive words and expressions like “a good musician”, “connoisseur”, “sine qua non”, and is more complicated in structure. Yang's version, composed of three simple sentences, is concise, smooth, and easier to understand. It is a free rendering of the original but successfully represents its gist. The overall impression we have of the three versions is that they do not fit into the stereotyping of sinologists and Chinese translators.

The above three examples, taken randomly from the English translations of Wenxin Diaolong, present a concise illustration of the overall translation strategies used by the three translators. A thorough analysis of the translations of the whole book will yield the same result as what we may get from the three examples, a result that does not fit into the assumptions about inverse translation. We may safely say that the translator's language affiliation or the direction of translation should not be given top priority in evaluating the quality of a translation, and what decides to a great extent the looks the translation will take is translation strategy (foreignization or domestication) other than the direction of translation. Both Shih and Yang are inverse translation practitioners, but there are wide and sharp differences in their translations of Wenxin Diaolong. Shih's version moves along slowly and gracefully, keeping looking back to the original and providing many explanatory expressions and annotations for the purpose of explicitation; Yang's version, with far fewer explanations and annotations, is from the beginning to the end very succinct and smooth, thus causing little interruption in the reading and inviting little resistance from the readership. Owen's version is contrary to what is often expected of the translation of sinologists. At a first reading it seems unnatural and clumsy as if always hindered by what is in the original. The reason behind this is not that Owen cannot write fluent English, or that he is unskillful as a translator, but that he deliberately uses this strategy for some special purpose:

In most cases the author has have decided in favor of a literal awkwardness in translation that will permit the English reader to see something of how the original Chinese text works. This relative literalness is not attractive; but in texts of thought, especially from the Chinese, grace in translation is usually a mark of vast concessions to the conceptual habits of the translation's audience” (Owen 1992:15-16).

This use of foreignization strategy in translating a Chinese literary work, at the cost of fluency and readability, is all the more valuable if we take into account the overriding pursuit of profit-making in many publishing houses nowadays.

The very fact that inverse translation is common in many parts of the world calls for more research into this usually neglected field: theoretical bases are to be established, and systematic comparative study between inverse and direct translation is to be conducted. This is especially true of China, where inverse translation has been playing an important role in introducing Chinese culture and Chinese political and economic policies to the outside world as there are not enough sinologists to do the task. The existing scepticism and criticism among some translation theorists and translators over inverse translation in China, by irrationally echoing the long-held prejudice against it, is not only detrimental to the development of translation studies, a young discipline that has to be receptive to remain its momentum, but also incompatible to the reality in China where a large amount of translation is done by Chinese translators. This paper is just a brief demonstration of the feasibility of inverse translation in China, and more followup study is needed to explore the characteristics and potentials of inverse translation, to provide guidance for this endeavour, and in the long run to facilitate the introduction of Chinese culture to the outside world. It's high time that translation studies gave inverse translation the attention it deserves. What we should bear in mind, though, is that to prove the feasibility of inverse translation is not to advocate it, nor to belittle direct translation. In the case of translation of Chinese literature into other languages, sinologists have so far made a marvelous contribution and will continue to play a critical role. What we do suggest is that more research be conducted into inverse translation and more tolerance be shown to this “unpopular” undertaking. Inverse translation in China is a necessary choice, but not a necessary evil.

Note:

¹Due to different interpretations of the title of Wenxin Diaolong （ 文心雕龙） , there are various English translations for it: The Literary Mind and the Carving of Dragons ( Liu 1959), Dragon-Carving and the Literary Mind (Liu 2003), The Book of Literary Design (Liu 1999), Carving a Dragon at the Core of Literature (Yang et al. 1962), etc.. For the sake of convenience this paper will use Chinese pinyin: Wenxin Diaolong, which is also used by Stephen Owen in his translation (Owen 1992).

²It refers to the year 1974 when the article was written, not the year 2002 when the anthology of Yu Guangzhong was published.

³Liu Xie ( 465—520) is a distinguished literary critic in ancient China, whose only literary work Wenxin Diaolong （ 文心雕龙） is considered to be the most important and most comprehensive writing in traditional Chinese literary criticism. Most of the writings on literary theory in ancient China, different from the systematic approach employed in Wenxin Diaolong, are just very brief comments on particular works or genres.

⁴There are so far three complete English versions of Wenxin Diaolong, done chronologically by Vincent Yu-chung Shih (Liu 1959), Wong et al. (Liu 1999), and Yang Guobin (Liu 2003). Stephen Owen translates 18 of the 50 chapters (Owen 1992). The three versions we use for comparison in the paper are those by Shis, Yang and Owen.

⁵The reason that organic metaphor can often be translated literally is because this popular figure of speech in traditional Chinese literary writing is also used in many Western literary works—though to a lesser degree. For details see Qian Zhongshu's study (Qian 2002: 116-34).