Translation: Its Genealogy in the West

Andr¨¦ Lefevere

From Susan Bassnett and Andr¨¦ Lefevere (eds.) Transaltion, History And Culture. London and New York: Pinter. 1990

The history of translation in the West may be said to begin with the production of the Septuagint. Like all early 'historical facts', this one, too, is conveniently shrouded in legend. Conveniently, because the legend will allow us to isolate the basic constraints that have influenced, and continue to influence the history of translation in the West and the other parts of the world it came into contact with.
The Septuagint is the first translation of the Hebrew Old Testament into Greek. It was made by seventy (or seventy-two) translators, all working in separate cells. They all translated the whole text, and all translations turned out to be identical. The translators were sent to Alexandria by Eleazar, High Priest of Jerusalem, at the request of Ptolemy II, Philadelphus, ruler of Egypt. The translation was made for the benefit of those Jewish communities in Egypt who could no longer read the original. It became the basis for later translations into Old Latin, Coptic, Armenian, Georgian and Slavonic.
So far the story. Now for the moral. Translation involves expertise: the seventy translators all produce the same version. They must know their trade. Their knowledge is guaranteed and probably checked by some event beyond their group. A supernatural event most likely, in legend - an all too natural event most likely, in actual fact. Translation also involves commission: a person in authority orders the translation to be made. There are, of course, many instances in which the translator 'auto-commissions' his or her own translation, simply because s/he 'falls' for a text. In this case the problem of 'commission' or at least 'acceptance' of the translation by a publisher is only deferred to the next stage in the process. Translation fills a need: the audience will now be able to read the text again, and the person in authority will have enabled the audience to do so. Translation involves trust: the audience, which does not know the original, trusts that the translation is a fair representation of it. The audience trusts the experts, and, by implication, those who check on the experts. As it happened in the case of the Septuagint, this trust was misplaced. Various versions were found to differ greatly among themselves, and later versions became so 'Christianized' that the Jewish communities stopped using the translation altogether. Texts that start their career as translations do not always remain so, in other words, but they can remain a central text in the history of a culture. The King James Bible comes to mind. But the fact that the Septuagint was, in reality, a 'bad' translation did nothing to undermine its image - on the contrary, it still is the translation used by the Greek Church to this day, and it served as the basis for translation into many other languages of the Ancient Mediterranean world.
The legend of the Septuagint has given us the basic categories of the history of translation. These categories are: authority (the authority of the person or institution commissioning or, later, publishing the translation: the patron; the authority of the text to be translated, in this case a central text in the source culture; the authority of the writer of the original, in this case the most absolute authority one can imagine, and the authority of the culture that receives the translation), expertise, which is guaranteed and checked, trust, which survives bad translations, and image, the image a translation creates of an original, its author, its literature, its culture.
Now take the other possibility: a case in which translation is neither commissioned nor encouraged, but resisted and even forbidden. The central text in this case is the Koran. No translations of it were allowed to be made by the faithful. Yet, the original can be said to have had a pervasive influence on world history, and not just in the area of its own historical dominance. If the central text is not translated, the faithful simply have to learn the language of the central text. If they do not, there will always be experts telling them what is in it, paraphrasing or interpreting it without actually translating it - but still creating an image of it. Translations, then, are only one type of text that makes an 'image' of another text. Other types would be criticism, historiography, commentary and anthologizing. They will be left out of consideration here. They should not be left out of consideration in studies of translation. The trust readers will have to give to those experts will have to be greater than the trust they will have to give to translators, since the possibilities for checking are more limited.
And then there are the in-between situations. As we know from history, the Romans translated, but they did not really have to. Educated Romans could just as well have gone on reading Greek literature and philosophy in the original, since they were bilingual anyway. Moreover, the percentage of educated Romans was relatively small when compared to the total population of the empire, or even the city of Rome. A similar situation prevailed in the Middle Ages: the learned did not need translation, and they did relatively little of it. In fact, they often did not write in their own language, but translated their thoughts directly into Latin, simply because the conventions of the time demanded this 'reverse translation': one could not be taken seriously as a scholar if one did not write in Latin.
Translation, then, is encouraged and commissioned, resisted and rejected. Obviously the reasons behind these two polar attitudes have little to do with expertise. There must have been Muslims perfectly capable of translating the Koran into other langauges. Trust is a factor, obviously: the central text of a culture should not be tampered with - no graven image should be made of it - precisely because the text guarantees, to a great extent, the very authority of those in authority. Linguistics, therefore, is by no means the overriding consideration in translation history. Translators do not get burnt at the stake because they do not know Greek when translating the Bible. They got burnt at the stake because the way they translated the Bible could be said to be a threat to those in authority.
Before we go on, let us call to mind - and firmly anchor there - the fact that European culture from, say, AD 500 to, say, 1800, was in essence bilingual, or even multilingual. There was a generally respected 'language of authority', first Latin, then French, which would be known by all those professing to be scholars, ecclesiastics or literati. They would know their mother tongues as well, of course, and, in many cases, one or two additional languages. Again, as with the Romans, they would not be all that large in number. European literate culture between 500 and 1800 can therefore be said to have been a bi(multi)lingual coterie culture - a fact so brilliantly repressed by Romantic historians who had to stress the importance of national languages and cultures that it is only now beginning to re-establish itself in the general consciousness of the West.
Obviously, in such a culture, translations were not primarily read for information or the mediation of the foreign text. They were produced and read as exercises, first pedagogical exercises, and, later on, as exercises in cultural appropriation - in the conscious and controlled usurpation of authority. That this usurpation was resented and resisted by those in authority is obvious from remarks like the following, found in the introduction to a translation of Hippocrates' Aphorisms: 'even though he foresees that his labour may incur the anger and the mockery of many who seem to be eager to keep the sciences hidden from the people' Jean Br®®che de Tours, in Horguelin, 1981). Members of the coteries who betray the coterie by making its knowledge available to those outside must be prepared to take the consequences of their actions. Jean Br che de Tours' observation already points forward to the break-up of the coterie culture. That break-up occurs some time around 1800. After the break-up writers on the subject begin to identify different potential audiences for translations, and different ways of translating emerge to matchdifferent audiences. Those who do not know the language of the tl original, and who are increasingly able to read their own ianguage, I will read the translation for information and mediation. Those who t still know the language of the original, at least in theory, will read the translation as a short-cut, a crib, or, still, an intellectual and aesthetic challenge, or even game. By 1900, with English increasingly filling the position of 'language of authority' reluctantly given up by French, the trend towards monolingualization of the audience increases, as does the corresponding trend towards producing translations for information. By 1900 the West has also come into contact with languages and cultures for which it has very few experts available. Trust becomes an important factor again, and images can be produced without being subject to rigorous checking. Fitzgerald's appropriation of Omar Khayyam comes to mind.
After 1800, Goethe can write in Dichtung und Wahrheit: 'If you want to influence the masses, a simple translation is always best. Critical qi translations vying with the original really are of use only for conversations the learned conduct among themselves' (in Lefevere, 1977: 38). But the masses do not always want to be influenced. Translations can be, and are still seen as a threat to the identity of a culture, as Victor Hugo observes in his introduction to the Shakespeare translations made by his son, Francois-Victor:

to translate a foreign poet is to add to one's own poetry; yet this addition does not please those who profit from it. At least not in the beginning; the first reaction is one of revolt. A language into which another idiom is transfused does what it can to resist. (Hugo, 1865: xv)

Not always, though. It does after 1800, and if it feels that the foreign text is a threat to its own authority.
Before 1800, languages were not supposed to resist, nor was translation felt to be an impossible task. On the contrary: Batteux affirms that 'a translator will be forgiven all metamorphoses, on condition that he makes sure that the thought emerges with the same body, the same life' (Batteux, 1824: Vol. II, 242). Language was considered a 3 vehicle for the exchange of thought. Or, in other words, the same thoughts could be conveniently 'dressed' in different languages. The old Latin word for translating: translatare can be taken to mean simply: 'an exchange of signifieds' (Berman, 1988: 25), without overmuch regard for the connotations, cultural and otherwise, carried by the actual signifiers. Translatio, then, can be seen as epitomizing the ideal of 'faithful translation', so dear to the heart of those in authority, who are intent on purveying the 'right' image of the source text in a different language. Translatio is vital for the 'authoritative texts' of a culture:
I insist on treating Holy Writ with such diligence and care because I do not want the oracles of the Holy Ghost to be adulterated by human and earth-bound elements. For it is not without divine counsel that they have been expressed in certain selected words, selected from a certain sphere and arranged in a certain order, for there are as many mysteries hidden in them as there are dots in the text. And did not Christ himself say that not one dot should be erased from the Law until heaven and earth are destroyed? (Huetius, 1683: 23)
But translation is impossible. An exchange of signifieds in a kind of intellectual and emotional vacuum, ignoring the cultural, ideological and poetological overtones of the actual signifiers, is doomed to failure, except in texts in which the 'flavour' of the signifiers is not all that important: scholarly texts, or non-literary texts in general. The historical analogy to the Septuagint in this case would be the translational activities of the Spanish school of Toledo, which translated many Arabic scientific and scholarly works into Latin after the city with its magnificent library fell to the Christians. Translatio tries to regularize the linguistic components of the translation process, without giving much thought to anything else. If it does, it will short- circuit as a result of the inbuilt tension between the linguistic and the cultural components of that process.
Its polar opposite can be designated by a Latin word that never really existed: traductio. As Berman pointed out: 'Leonardo Bruni is said to have translated the past participle traductum used by a Latin author, Aulus Gellius, by the Toscan tradotto. But for Aulus Gellius traductum did not mean "translated" but rather "transported"' (Berman, 1988: 30). Traductio is the more creative counterpart to the more conservative translatio. Traductio is prepared and allowed to give at least equal weight to the linguistic and the cultural/ideological components of the translation process. It will come to the fore in a culture when that culture considers itself'authoritative', central with regard to other cultures. But precisely because it usurps that role, that culture will treat the cultural side of the translation process in its traductio the way translatio treats the linguistic side of the translation process: it will try to regularize it. As Herder puts it in the Fragmente:the French, who are overproud of their natural taste, adapt all things to it, rather than try to adapt themselves to the taste of another time. Homer must enter France a captive, and dress according to fashion, so as not to offend their eyes. He has to allow them to take his venerable beard and his old simple clothes away from him. He has to conform to the French customs, and where his peasant coarseness still shows he is ridiculed as a barbarian. But we, poor Germans, who still are almost an audience without a fatherland, who are still without tyrants in the field of national taste, we want to see him the way he is. (in Lefevere, 1977: 48)
Almost a hundred years later, Fitzgerald writes to his friend E. B. Cowell: 'It is an amusement for me to take what Liberties I like with these Persians who (as I think) are not Poets enough to frighten one from such excursions, and who really do want a little Art to shape them' (Fitzgerald, 1972, VI: xvi). Traductio is a matter of the relative weight two cultures carry in the mind of the translator: obviously, Fitzgerald would never have taken the same liberties with a Greek or Roman author, also because there were too many experts around. But since Victorian England considers itself central, and since he happens to be translating from a culture that is by no means central to it, he takes what liberties he pleases. As we shall see later, traductio can also be used by translators as individual members of a culture,. who are dissatisfied with certain features of it, and want to usurp the authority of texts belonging to another, 'authoritative' culture, to attack those features, defying both experts and those in authority with a certain degree of impunity. In fact, traductio, as described by Nicholas Perrot d'Ablancourt in 1709, sounds suspiciously like Eugene A. Nida's 'equivalence of effect': 'I do not always stick to the author's words, nor even to his thoughts. I keep the effect he wanted to reach in mind, and then I arrange matters according to the fashion of our time.' (Perrot d'Ablancourt, 1709: 23).
A view of language, like Schleiermacher's, which no longer sees the signifiers as essentially neutral vehicles for conveying signifieds, but rather as inextricably bound up with different languages, will have to raise the problem of the very possibility of translation. If, as Schleiermacher holds, 'every man is in the power of the language he speaks and all his thinking is a product thereof' (in Lefevere, 1977: 71), translation appears to be an impossible task. Or rather, what appears to be impossible is translatio, and all translation will have to be transposition, traductio. In his persona of translator, Schleiermacher himself shied away from the consequences of this insight, which makes the second part of his famous maxim, 'move the author towards the reader' the only viable one. But if translation was to remain possible after 1800, it would have to be traductio. Possible or not, though, translations continued to be produced, and their production was to keep increasing.
Both translatio and traductio involve authority, expertise and trust. Authority draws the ideological parameters of the acceptable. It influences the selection of texts for translating, as well as the ways in which texts are translated. In John of Trevisa's 'Dialogue between a Lord and a Clerk upon Translation' (1903: 23) the Lord makes it quite clear that he is paying the piper, and therefore expects to call the tune. The Lord says: 'I desire not translation of these the best that might be, for that were an idle desire for any man that is now alive, but I would have a skillful translation, that might be known and understood.' In other words, something that works - and, in later words: something that sells. The Clerk just wants to make sure: 'Whether is you liefer have, a translation of these chronicles in rhyme or in prose?' Again, the answer is refreshingly blunt: 'In prose, for commonly prose is more clear than rhyme, more easy and more plain to know and understand.' Translators know who pays the piper, and give advice to other translators accordingly. In a little quoted passage from his best-known work, Du Bellay ends his admonitions to translators with: 'what I say is not meant for those who, at the command of princes and great lords, translate the most famous Greek and Latin writers, since the obedience one owes to those persons admits of no excuse in these matters' (1948: 52). Again about a hundred years later, the Earl of Roscommon refers to those in authority, but they are now of a different kind:

I pity from my Soul unhappy Men
Compelled by Want to prostitute their Pen
Who must, like Lawyers, either starve or plead
And follow, right or wrong, where Guineas lead.
(in Steiner, 1975: 82)

Around 1700, with the increasing speed of literacy and the gradual spread of a more open type of society, the authorities are no longer just 'princes and great lords'; they are joined by publishers. If the role of the publisher as the authority who decides what is going to be translated increases, the ideological parameters widen, since the ultimate criterion for deciding is, primarily, money. The publishers of Roscommon's time would publish only a traductio of Homer, which would be acceptable/saleable to their readers. Roscommon advises translators to leave out what they deem unacceptable:

For who, without a Qualm, hath ever lookt
On Holy Garbage, tho by Homer cookt?
(in Steiner, 1975: 78)

Similarly, the Abb Pr vost writes in the introduction to his translation of Richardson's Pamela:
I have suppressed English customs where they may appear shocking to other nations, or else made them conform to customs prevalent in the rest of Europe. It seemed to me that those remainders of the old and uncouth British ways, which only habit prevents the British themselves from noticing, would dishonor a book in which manners should be noble and virtuous. To give the reader an accurate idea of my work, let me just say, in conclusion, that the seven volumes of the English edition, which would amount to fourteen volumes in my own, have been reduced to four. (in Horguelin, 1981).
It would appear that the French reader will be given a rather different 'image' of Pamela than his English counterpart.
As we move closer to the present, the excesses of traductio are more limited. But the case of the translations into English of the Irish national epic, the Ta'in, are a good example of the influence of authority on translation. A scholarly translation of the Ta'in existed in German as early as 1905 (Ernst Windisch's Die altirische Heldensage 'Tain Bo' Cu'ailnge', published in Leipzig). The first comparable translation in English was published only in 1967: Cecile O'Rahilly's Ta'in Bo' Cu'ailnge, Recension 1, published in Dublin. The first English traductio of the complete Ta'in, by the poet Thomas Kinsella, was published in Dublin in 1969. It had been preceded by many partial traductiones, among them Lady Gregory's 1902 version Cuchulain of Muirthemne. There were obviously more than enough qualified translators around, but the intellectual and, primarily, emotional climate in Ireland between 1905 and 1969 was such that nobody would translate in its entirety a national epic that alternates descriptions of noble behavior with descriptions of the Irish as a merrily barbaric bunch, killing, looting, raping and defecating all over the place - precisely the image the intellectuals associated with the 'Irish Renaissance' tried to counteract with all their might.
The experts are employed by those in authority to check each other's expertise. This checking process takes place most obviously in the pedagogical situation. As late as the mid-seventeenth century, Gottsched states in his Ausfuhrliche Redekunst that translation is 'precisely what the copying of a given model is to a beginner in the art of painting. We know that the works of great masters are copied with pleasure and diligence by mediocre artists or by beginners who would like to make their way' (in Lefevere, 1977: 44). The experts also delimit the poetological parameters of translation: will the finished product be acceptable as literature in the target culture? Will it confirm to the poetics currently dominating that culture? Again, traductio appears to be the answer, and some translators go to great lengths to make the source text fit the target culture poetics. De la Motte, for example, states in the introduction to his translation of the lliad:
I have reduced the twenty four books of the lliad to twelve, which are even shorter than Homer's. At first sight you might think that this could only be done at the expense of many important features. But if you pause to reflect that repetitions make up more than one sixth of the lliad, and that the anatomical details of wounds and the long speeches of the fighters make up a lot more, you will be right in thinking that it has been easy for me to shorten the poem without losing any important features of the plot. I flatter myself with the thought that I have done just that, and I even think I have brought together the essential parts of the action in such a way that they are shaped into a whole better proportioned and more sensible in my abbreviated version than in the original. (1714: 17)
Small wonder that Perrot d'Ablancourt, faced with the twin constraints of authority and expertise, began his apology for his translation of Lucian with the diplomatic statement:
Two things can be held against me where this translation is concerned. One has to do with the selection of the work, the other with the way in which I translated it. One group of people will say that I should not have translated this particular author, and another group that I should have translated him differently. (1709: 24)
With the split in the audience after l800, and the rise of philology as a university discipline, the worst excesses of traductio came to an end. The experts could reserve a part of the market for themselves, and produce translations aimed primarily at other experts, in effect recreating the coterie culture, but this time in isolation from the general culture they were part of, even if they would produce the odd traductio for its benefit.
The experts are supposed to guarantee that the trust the audience places in various translations is not misplaced. But they are not always successful. The problem is that the audience places less trust in the experts' stamp of approval of a fida interpretatio than in the reputation of a translator as a fidus interpres. Glyn P. Norton has shown that the well-known Horatian phrase was used earlier by Sallust in the Iugurtha, and that the 'qualifier fidus ... characterizes the personal reliability of the go-between - his mutual trustworthiness in the eyes of both parties - rather than a quality inherent in his translation' (Norton, 1981: 184). This explains why 'bad' translations continue to enjoy great popularity among the general public, even when technically superseded by translations of better quality - a state of affairs which most definitely predates the introduction of copyright. A certain translation achieves a somewhat 'canonized status' and can hardly be dislodged from it. Maybe one of the earliest examples of this state of affairs can be found in St Augustine's seventy-first letter, addressed to SaintJerome. In the letter, Augustine tells the story of a bishop who introduced the use ofJerome's translation 'in the church of which he is the pastor. They hit upon a passage in the prophet Jonah which you translated very differently from the way in which it has established itself in the mind and memory of all, and the way it had been sung for such a long time.' The result is unrest, foul play is suspected, and after consultation with the localJews, who are no help either, the bishop 'was forced to correct himself, as if he had made a mistake, since he didn't want to lose all the people in his church' (Augustine, 1909: no. 71).
Translation, then, is sanctioned by authority, but it can also try to subvert authority by usurping the authority of an authoritative text alien to the target culture. The early translations of the Communist Manifesto into Russian come to mind, as do the translations of the English philosophers into French in the seventeenth and eighteenth centuries. Translation can play an important part in the struggle between rival ideologies, witness Luther's lament in the Tischgesprche:

We are aware of the scribbler in Dresden who stole my New Testament. He admitted that my German is good and sweet and he realized that he could not do better and yet he wanted to discredit it, so he took my New Testament as I wrote it, almost word for word, and he took my preface, my glosses and my name away and wrote his name, his preface and his glosses in their place. He is now selling my New Testament under his name. Oh, dear children, how hurt I was when his prince, in a terrible preface, forbade the reading of Luther's New Testament but ordered the scribbler's New Testament read, which is exactly the same as the one Luther wrote. (in Lefevere, 1977: 22).

Translations also play an important part in the struggle between rival poetics. The case of Pound's Cathay (1915) is too well known to warrant extensive discussion here. Faced with Victorian/Edwardian poetics, Pound manufactured the Chinese T'ang dynasty poets as an 'authoritative' countertext, one that did, as if by miracle, fit all the requirements of the new poetry he, Pound, was trying to create. Since translation awards some kind of limited immunity to those who write it (after all, they are not responsible for what others wrote), attacks on the dominant poetics of a literature often pass themselves off as translations. Horace Walpole's Castle of Otranto (1764), the first of the othic novels', is a case in point. In the preface to the first edition, Walpole tells the reader that the novel he is about to read is a translation of an Italian manuscript, and promises to 'reprint the original Italian' if the novel 'should meet with success' (in Fairclough, ,1964: 43). In the preface to the second edition, the author apologizes: 'it is fit that he should ask pardon of his readers for having offered his work to them under the borrowed personage of a translation' ( 1964: 48). The whole strategem was necessary because of the 'novelty of the attempt ... to blend two kinds of romance, the ancient and the modern' ( 1964: 48) .
Translation usurps authority, but translation also bestows authority. It bestows authority on a language. In Cicero's words:

by giving a Latin form to the text I had read, I could not only make use of the best expressiOns in common usage with us, but I could also coin new expressions analogous to those used in Greek, and they were no less well received by our people, as long as they seemed appropriate. (Cicero, Oratore, book I: 35)

Translation forces a language to expand, and that expansion may be welcome as long as it is checked by the linguistic community at large. Translation can also bestow the authority inherent in a 'language of authority' (Latin, French, English/Russian) on a text originally written in another language, which lacks that authority. Many works written in 'minor' languages, such as Strindberg's dramas, would not belong to 'world literature' if they had not been launched in a language of authority, in this case French. Similarly, Ibsen's dramas were introduced to Europe not in his native Norwegian, but in German by the Volksbuhne in Berlin. The pervasive influence of translation is so great that these works cease, after a while, to be thought of as 'foreign' to the 'language of authority'. English departments now routinely teach both Ibsen and Strindberg, and students tend to find the Scandinavian names a bit of a nuisance, at times. As a cumulative effort, translation eventually builds up a translinguistic and transcultural canon (the 'Penguin Classics' in our day and age) which is, in its turn, invested with authority.
Translation also allows writers in the target culture to 'proceed on the authority' of writers alien to the target culture and introduced into it by trauslators. In other words, translation introduces new devices in the literatures by which it is received. The sonnet, for example, was introduced into Chinese in the 1920s, via Feng Chi's translations. The ode became the major genre of the poets of the Pl®¶iade after it had been translated extensively from Greek and Latin. Translation, under the moralizing aegis of the Jesuits, transformed the picaresque novel into the Bildungsroman in Germany. The alternation of masculine and feminine rhymes in French goes back to Octavien de St Gelais' translations of Ovid. The hexameter was introduced into German by the Homer translations of Johann Heinrich Voss. John Hookham Frere's translations of Pulci reintroduced ottava rima into English, where it was soon to be used by Byron in his Don Juan. Yet Goethe's pious 'hope that literary history will plainly state who was the first to take this road in spite of so many obstacles' (in Lefevere, 1977: 39), tends to remain exactly that. Literary histories, as they have been written until recently, have had little time for translations, since for the literary historian translation has had to do with 'language' only, not with literature - another pernicious outgrowth of the 'monolingualization' of literary history by Romantic historiographers intent on creating 'national' literatures preferably as uncontaminated as possible by foreign influences. Yet on every level of the translation process it can be shown that if linguistic considerations conflict with considerations of an ideological and/or poetiological nature, the latter considerations tend to win. A. W. Schlegel's fateful pronouncement that 'one of the first principles of the art of translation is that, for as far as e nature of a language allows, a poem should be recreated in the same meter' (in Lefevere, 1977: 52), which has been responsible for all inds of metrical contortions in translations made roughly between 1830 and 1930, was obviously not made on linguistic grounds. Browning's insistence on the 'use of certain allowable constructions which, happening to be out of daily favour, are all the more appropriate to archaic workmanship (Browning, 1937: 1095) is responsible for the fact that most Victorian translations of the classics read so monotonously alike. It was not inspired by linguistic necessity, but by the desire to acquire the timeless through use of the archaic. The result did not meet with the translators' expectations.
The creation of the Latin word sacramentum is also revealing in this respect. When the early Christians needed to translate the Greek word musterion, they did not want simply to Latinize it, because it was too close to the vocabulary used by the 'mystery cults' which were Christianity's main competition at the time. For the same reason they rejected words like sacra, arcana, initia, which would have been semantically acceptable equivalents. They hit on sacramentum as a term both neutral and close to the original. But when St Jerome prepared the Vulgate, Christianity had won the battle against the mystery religions, and he felt free to simply Latinize musterion. (cf. Klopsch, 1983: 37-8) Similarly, the AramaicJesus Christ is supposed to have spoken did not have a copula. He can therefore never have said: 'This is my body' when pointing at a loaf of bread. The copula was put in by translators for both linguistic and ideological reasons.
That different types of text need to be translated in different stylistic (not linguistic) ways, was recognized by Gaspard de Tende as early as 1660. 'It would not be advisable,' he says, 'to translate orations that need to be treated with some leeway into a precise style, cut and dry, nor should you translate parables, that need to be short and precise, into a style that would allow them more leeway' (de Tende, 1665: 5).
Finally, untranslatability seems to have a lot more to do with the absence of poetologial equivalents than with the absence of semantic or morphosyntactic equivalents. The qasida, the canonized genre of Arabic poetry, has never been satisfactorily translated in the West, because it has no obvious generic equivalent. This is how Ibn Qutaiba, the Arab poet and critic, describes the genre:

the composer . . . began by mentioning the deserted dwelling places and the relics and traces of habitation. Then he wept and complained and addressed the desolate encampment and begged his companions to make a halt, in order that he might have occasion to speak of those who had once lived there and afterwards departed.... Then to this he linked the erotic prelude and bewailed the violence of his love and the anguish of separation from his mistress and the extremity of his passion and desire, so as to win the hearts of his hearers and divert their eyes towards him and invite their hearts to listen to him.... He followed up his advantage and set forth his claim: thus he went on to complain of fatigue and want of sleep and travelling by night and of the noonday heat, and how his camel had been reduced to leanness. And after representing all the discomfort and danger of his journey, he knew that he had finally justified his hope and expectation of receiving his due meed from the person to whom the poem was addressed, he entered upon the panegyric and incited him to reward, and kindled his generosity by exalting him above his peers and pronouncing the greatest dignity, in comparison with his, to be little. (in Arberry, 1957: 15-16)

It is easy to discover both ideological and poetological elements in this description that would be most unfamiliar to the Western reader. Lyall states that the qasida 'is not epic, nor even narrative . . . still less is it dramatic ... the Greek idyll is perhaps the type which comes nearest to it in classical poetry' (1930: xviii). Nicholson calls it an 'ode' (1922: 76) andJones refers to 'casseidas or eclogues' (1007: X: 341).
Language is not the problem. Ideology and poetics are, as are cultural elements that are not immediately clear, or seen as completely 'misplaced' in what would be the target culture version of the text to be translated. One such element is the camel dung mentioned in Labid's qasida, which can hardly be expected to make a 'poetic' impression on Western readers. Carlyle, the English Victorian translator, leaves it out altogether; to him 'this was simply incomprehensible' (Polk, 1974: xxviii). German translators, on the other hand, try to find a cultural analogy, but with little success: the solution is worse than the problem: 'German scholars, familiar with the peasants of their own land, where the size of the dung heap is some indication of the prosperity of the farmer, merely transported to the desert the social values of Bavaria' (Polk, 1974: xxviii). Most attempts at surveying translation history begin with a pious platitude. This attempted survey ends with one, borrowed from Mme de Stael: 'the most eminent service one can render to a literature, is to transport the masterpieces of the human spirit from one language to another' (de Stael: 328) Pious platitudes invariably surface in discussions of translation, not just because the subject is so complex, but also because it is potentially disturbing: it keeps questioning expertise and it is always potentially subversive of authority. This does not make it exactly a popular topic for research, but it does make it a potentially very rewarding one.
Translation is one of the most obvious forms of image making, of manipulation, that we have. It makes its images together with other media. For many people in the English-speaking world, for example, Crime and Punishment will always be an amalgam of some of the following: the Constance Garnett or the Magarshack translation, histories of Russian literature, articles in mass circulation magazines, television classical drama series. Translation is responsible to a large extent for the image of a work, a writer, a culture. Together with historiography, anthologizing and criticism it prepares works for inclusion in the canon of world literature. It introduces innovations into a literature. It is the main medium through which one literature influences another. It can be potentially subversive and it can be potentially conservative. It can tell us about the self-image of a culture at a given time, and the changes that self-image undergoes. It can tell us about the strength of a poetics and/or an ideology at a certain time, simply by showing us the extent to which they were interiorized by people writing translations at that time (poor vilified De la Motte was not trying to demolish Homer; he was simply trying to be a 'good' writer as best he knew how). Translation can tell us a lot about the power of images and the ways in which images are made, about the ways in which authority manipulates images and employs experts to sanction that manipulation and to justify the trust of an audience - which is why the study of translation can teach us a few things not just about the world of literature, but also about the world we live in.

References:

Arberry, Arthur J. (1957) The Se1Jen Odes. London and New York: Allen and Unwin and Macmillan.
Augustine, ( 1909) 'Letter to Jerome', in The Letters of St Augustine, ed. W. J. Sparrow-Simpson. London: Society for Promoting Christian Knowledge.
Batteux, Charles (1824) Principes de litt®¶rature, 5 vols. Paris.
Berman, Antoine (1988) 'De la translation a la traduction', Etudes sur le texte et ses ransformations 1: 23-40.
Browning, Robert ( 193 7) 'The Agamemnon of Aeschylus' ( 187 7), The Poetical Works of Robert Browning. New York: Macmillan.
De la Motte, Houdar (1714) L'lliade. Amsterdam.
De Tende, Gaspard (1665) R®®gles de la traduction. Paris.
Du Bellay,Joachim (1948) D®¶fense et illustration de la langue francaise. Paris: Didier.
Fairclough, William (ed.) (1964) Three Gothic Novels. Harmondsworth: Penguin Books.
Fitzgerald, Edward (1972) The Variorum and Denitive Edition of the Poetical and Prose Writings, 7 vols. New York: Doubleday.
Horguelin, Antoine (ed.) (1981) Anthologie de la mani®®ire de traduire. Montreal: Linguatech.
Huetius, Petrus Danielus (1683) De optimo genere interpretandi libri duo. The Hague.
Hugo, Victor (1865) 'Pr®¶face de la nouvelle traduction de Shakespeare', in Oeuvres compl®®tes de William Shakespeare, trans. Francois Victor Hugo, 15 vols. Paris: Pagnerre.
Jones, William (1980) The Works of Sir William Jones. London, 1807, 13 vols; repr. Delhi: Agam Prakashan Reprint.
Klopsch, Paul (1983) 'Die mitellateinische Lyrik', in Lyrik des Mitktalters, ed. Heinz Bergner, 2 vols. Stuttgart: Reclam.
Lefevere, Andr®¶(ed. and trans.) (1977) Translating Literature: TheçäCerman Tradition. Assen: Van Gorcum.
Lyall, Charles J. (ed. and trans.) (1930) Translations of Ancint Arabian Poety. New York: Columbia University Press.
Nicholson, Reynold A. (ed. and trans.) (1922) Translations of Eastern Poety and Prose. Cambridge: Cambridge University Press.
Norton, Glynn P. (1981) 'Humanist Foundations of Translation Theory', Canadian Reuicw of Comparatiue Literatur 8: 173-91.
Perrot d'Ablancourt, Nicholas (1709) Luccn. De la traduction. Amsterdam.
Polk, William R. (ed. and trans.) ( 1974) The Coldcn Ode by Labid Ibn Rabiah. Chicago: University of Chicago Press.
de Stael, Mme. (1821) Mclange, Oeuvres compl tes, Paris.
Steiner, T. R. (ed.) (1975) English Translation Theory 1650-1800. Assen: Van Gorcum.
Trevisa, John of (1903) 'Dialogue between a Lord and a Clerk upon Translation', in Fifteenth Centuy Prose and Verse, ed. A. W. Pollard. Westminster: Constable. _1975) English Translation Thcory 1650-1800. Assen: Van Gorcum.
Trevisa,John of (1903) 'Dialogue between a Lord and a Clerk upon Translation', in Fifeenth Centuy Prose and Verse, ed. A. W. Pollard. Westminster: Constable.
Palma Zlateva Let me note, at the outset, that the terms 'adequacy' and 'acceptability' as used in this paper do not allow for a statement like the following: 'A translated text can be located on an axis between the two hypothetical poles of adequacy (source text oriented) or acceptability (target language oriented)' (Toury, 1980: 34). Such a statement seems to exclude the possibility that a translated text could ever be both adequate to the original and acceptable in the target language, since a +adequate translation will always be a¡ª·cceptable one. This logically leads to the conclusion that the only translated text to satisfy the conditions for being a metatext will be a translated text situated right in the middle of this axis, where we can only have zero adequacy and zero acceptability.
I shall try to supply arguments in favour of another option, namely that the acceptability of a translated text in the target language should be considered part of the adequacy of its translation. Any adequately translated literary text becomes a materialfact not only in the target language, but in the target literature as well: it exists in both. The fact of its existence and acceptability in the target language, howez)er, does not necessarily imply that it is, or will be, immediately accepted in the target literature and culture. This is a different matter altogether. It has to do with the translator's choice of a particular work at a particular time, with the core and periphery of the target cultural and literary tradition at that particular moment, and with several other factors. In short, it pertains to considerations located both on the level of pre-text and post-text.