SusanSontagFoundation prizeForTranslation SusanSontag about events donate
blank Winners 2008Winners committee about SontagOnTranslation
The World as India

The St. Jerome Lecture on Literary Translation

In Memoriam W. G. Sebald

To translate means many things, among them: to circulate, to transport, to disseminate, to explain, to make (more) accessible. I’ll start with the proposition – the exaggeration, if you will – that by literary translation we mean, we could mean, the translation of the small percentage of published books actually worth reading: that is to say, worth rereading. I shall argue that a proper consideration of the art of literary translation is essentially a claim for the value of literature itself. Beyond the obvious need for the translator’s facilitations in creating stock for literature as a small, prestigious import-export business, beyond the indispensable role that translation has in the construction of literature as a competitive sport, played both nationally and internationally (with rivalries, teams, and lucrative prizes) – beyond the mercantile and the agonistic and the ludic incentives for doing translation lies an older, frankly evangelical incentive, more difficult to avow in these self-consciously impious times.

In what I call the evangelical incentive, the purpose of translation is to enlarge the readership of a book deemed to be important. It assumes that some books are discernibly better than other books, that literary merit exists in a pyramidal shape, and it is imperative for the works near the top to become available to as many as possible, which means to be widely translated and as frequently retranslated as is feasible. Clearly, such a view of literature assumes that a rough consensus can be reached on which works are essential. It does not entail thinking the consensus – or canon – is fixed for all time and cannot be modified.

At the top of the pyramid are the books regarded as scripture: indispensable or essential exoteric knowledge that, by definition, invites translation. (Probably the most linguistically influential translations have been translations of the Bible: Saint Jerome, Luther, Tyndale, the Authorized Version.) Translation is then first of all making better known what deserves to be better known – because it is improving, deepening, exalting; because it is an indispensable legacy from the past; because it is a contribution to knowledge, sacred or other. In a more secular register, translation was also thought to bring a benefit to the translator: translating was a valuable cognitive – and ethical – workout.

In an era when it is proposed that computers – “translating machines” – will soon be able to perform most translating tasks, what we call literary translation perpetuates the traditional sense of what translation entails. The new view is that translation is the finding of equivalents; or to vary the metaphor, that a translation is a problem, for which solutions can be devised. In contrast, the old understanding is that translation is the making of choices, conscious choices, choices not simply between the stark dichotomies of good and bad, correct and incorrect, but among a more complex dispersion of alternatives, such as “good” versus “better” and “better” versus “best,” not to mention such impure alternatives as “old-fashioned” versus “trendy,” “vulgar” versus “pretentious,” and “abbreviated” versus “wordy.”

For such choices to be good – or better – was assumed to imply knowledge, both wide and deep, on the part of the translator. Translating, which is here seen as an activity of choosing in the larger sense, was a profession of individuals who were the bearers of a certain inward culture. To translate thoughtfully, painstakingly, ingeniously, respectfully was a precise measure of the translator’s fealty to the enterprise of literature itself.

Choices that might be thought of as merely linguistic always imply ethical standards as well, which has made the activity of translating itself the vehicle of such values as integrity, responsibility, fidelity, boldness, humility. The ethical understanding of the task of the translator originated in the awareness that translation is basically an impossible task, if what is meant is that the translator is able to take up the text of an author written in one language and deliver it, intact, without loss, into another language. Obviously, this is not what is being stressed by those who await impatiently the supersession of the dilemmas of the translator by the equivalencings of better, more ingenious translating machines.

Literary translation is a branch of literature – anything but a mechanical task. But what makes translation so complex an undertaking is that it responds to a variety of aims. There are demands that arise from the nature of literature as a form of communication. There is the mandate, with a work regarded as essential, to make it known to the widest possible audience. There is the general difficulty of passing from one language to another, and the special intransigence of certain texts, which points to something inherent in the work quite outside the intentions or awareness of its author that emerges as the cycle of translations begins – a quality that, for want of a better word, we call translatability.

This nest of complex questions is often reduced to the perennial debate among translators – the debate about literalness – that dates back at least to ancient Rome, when Greek literature was translated into Latin, and continues to exercise translators in every country (and with respect to which there is a variety of national traditions and biases). The oldest theme of the discussion of translations is the role of accuracy and fidelity. Surely there must have been translators in the ancient world whose standard was strict literal fidelity (and damn euphony!), a position defended with dazzling obstinacy by Vladimir Nabokov in his Englishing of Eugene Onegin. How else to explain the bold insistence of Saint Jerome himself (ca. 331-420) – the intellectual in the ancient world who (adapting arguments first broached by Cicero) reflected most extensively, in prefaces and in letters, on the task of translation – that the inevitable result of aiming at a faithful reproduction of the author’s words and images is the sacrifice of meaning and of grace?

This passage is from the preface Jerome wrote to his translation into Latin of the Chronicle of Eusebius. (He translated it in the years A.D. 381-82, while he was living in Constantinople in order to take part in the Council – six years before he settled in Bethlehem, to improve his knowledge of Hebrew, and almost a decade before he began the epochal task of translating the Hebrew Bible into Latin.) Of this early translation from Greek, Jerome wrote:

It has long been the practice of learned men to exercise their minds by rendering into Latin the works of Greek writers, and, what is more difficult, to translate the poems of illustrious authors though trammeled by the farther requirements of verse. It was thus that our Tully literally translated whole books of Plato…[and later] amused himself with the economics of Xenophon. In this latter work the golden river of eloquence again and again meets with obstacles, around which its waters break and foam to such an extent that persons unacquainted with the original would not believe they were reading Cicero’s words. And no wonder! It is hard to follow another man’s lines…It is an arduous task to preserve felicity and grace unimpaired in a translation. Some word has forcibly expressed a given thought; I have no word of my own to convey the meaning; and while I am seeking to satisfy the sense I may go a long way round and accomplish but a small distance of my journey. Then we must take into account the ins and outs of transposition, the variations in cases, the diversity of figures, and, lastly, the peculiar, and, so to speak, the native idiom of the language. A literal translation sounds absurd; if, on the other hand, I am obliged to change either the order or the words themselves, I shall appear to have forsaken the duty of a translator. (tr. W.H. Fremantle, 1892)

What is striking about this self-justifying passage is Jerome’s concern that his readers understand just how daunting a task literary translation is. What we read in translation, he declares later in the same preface, is necessarily an impoverishment of the original.

If any one thinks that the grace of language does not suffer through translation, let him render Homer word for word into Latin. I will go farther and say that, if he will translate this author into the prose of his own language, the order of the words will seem ridiculous, and the most eloquent of poets almost dumb.

What is the best way to deal with this inherent impossibility of translation? For Jerome there can be no doubt how to proceed, as he explains over and over in the prefaces he wrote to his various translations. In a letter to Pammachius, written in A.D. 396, he quotes Cicero to affirm that the only proper way to translate is

…keeping the sense but altering the form by adapting both the metaphors and the words to suit our own language. I have not deemed it necessary to render word for word but I have reproduced the general style and emphasis.
Later in the same letter, quoting Evagrius this time – one must assume that there were many critics and cavillers – he declares defiantly: “A literal translation from one language into another obscures the sense.” If this makes the translator a coauthor of the book, so be it. “The truth is,” Jerome writes in his preface to Eusebius, “that I have partly discharged the office of a translator and partly that of a writer.”

The matter could hardly be put with greater boldness or relevance to contemporary reflections. How far is the translator empowered to adapt – that is, re-create – the text in the language into which the work is being translated? If word-by-word fidelity and literary excellence in the new language are incompatible, how “free” can a responsible translation be? Is it the first task of the translator to efface the foreignness of a text, and to recast it according to the norms of the new language? There is no serious translator who does not fret about such problems: like classical ballet, literary translation is an activity with unrealistic standards, that is, standards so exacting that they are bound to generate dissatisfaction, a sense of being rarely up to the mark, among ambitious practitioners. And like classical ballet, literary translation is an art of repertory. Works deemed major are regularly redone – because the adaptation now seems too free, not accurate enough; or the translation is thought to contain too many errors; or the idiom, which seemed transparent to the contemporaries of a translation, now seems dated.

Dancers are trained to strive for the not entirely chimerical goal of perfection: exemplary, error-free expressiveness. In a literary translation, given the multiple imperatives to which a literary translation has to respond, there can only be a superior, never a perfect, performance. Translation, by definition, always entails some loss of the original substance. All translations are sooner or later revealed as imperfect and eventually, even in the case of the most exemplary performances, come to be regarded as provisional.


Saint Jerome was doing translations – from Hebrew and from Greek – into Latin. The language into which he was translating was, and was to remain for many centuries, an international language.

I am giving this talk in the new international language, estimated to be the mother tongue of more than 350 million people, and spoken as a second language by tens of millions of people throughout the world.

I am here in England, where the language I am speaking and in which I write was born. I shall take the simple view that we are not divided by a common language, as the old quip has it. And in my country, we don’t call the language most of us speak “American” (for all that the title page of the French translations of my books say “traduit de l’américain”). Apparently, however, there are people in the United States who don’t know why they call it English.

Some years ago as English friend of mine, a writer with a rich Oxbridge accent, visited America for the first time with his wife and teenage daughters and decided the best way to have the whole American experience was to rent a car and drive across the country, from New York to California. Stopping at a filling station somewhere in Iowa on a sweltering summer day, my friend was asked by the lone attendant servicing his car, following a few moments of chat, “Where do you folks come from?” “England,” replied my friend, wondering what that might provoke. “No kidding,” exclaimed the filling station attendant. “You really speak English very well for a foreigner.”

Of course, most of us do know why it’s called English. And it is the glory of the literature of my country, which is not much more than two hundred years old, that it gets to be written in your thousand-year-old language.

Each day I sit down to write, I marvel at the richness of the language I am privileged to use. But my pride in English is somewhat at odds with my awareness of another kind of linguistic privilege: to write in a language that everyone, in principle, is obliged to – desires to – understand.

Though seemingly identical now with the world dominance of the colossal and unique superpower of which I am a citizen, the initial ascendancy to international lingua franca of the tongue in which Shakespeare wrote was something of a fluke. One of the key moments was the adoption in the 1920s (I believe) of English as the international language of civilian aviation. For planes to circulate with safety, those who flew them and those who directed their flight had to have a language in common. An Italian pilot landing in Vienna speaks to the tower in English. An Austrian pilot landing in Naples speaks to the tower in English. More, it produces the oddity that an Italian plane going from Naples to Palermo, a Swedish pilot going from Stockholm to Malmö, a Brazilian pilot going from São Paulo to Rio – each should be communicating with the tower in English. We take this for granted now.

More powerfully, and I think decisively, the ubiquity of computers – the vehicle of another form of transport: mental transport – has required a dominant language. While the instructions on your interface are likely to be in your native language, going online and using search engines – that is, circulating internationally on the computer – requires a knowledge of English.

English has become the common language that unifies linguistic disparities. India has sixteen “official languages” (actually, many more vernacular languages are spoken), and there is no way that India, given its present composition and diversity, which includes 180 million Muslims, is ever going to agree to, say, the principal language, Hindi, becoming the national language. The language that could be a national language would precisely not be a native one but the language of the conqueror, of the colonial era. Just because it is alien, foreign, it can become the unifying language of a permanently diverse people: the only language that all Indians might have in common not only is, it has to be, English.

The computer has only reinforced the preeminence of English in our global India. Surely the most interesting linguistic phenomena of our era are, on the one hand, the disappearance of many lesser languages – that is, languages spoken by very small, isolated, impoverished peoples – and the unique success of English, which now has a status unlike any other language used on the planet. English is now advancing in every part of the world, through the dominance of English-speaking media – which means media in which English is spoken with an American accent – and the need for business people and scientists to communicate in a common tongue.

We live in a world that is, in several important respects, both mired in the most banal nationalisms and radically postnational. The primary feature of the trade barriers may fall, money may become multinational (like the dollar, which is the currency in several Latin American countries, and of course the euro). But there is one intractable feature of our lives that roots us in the old boundaries that advanced capitalism, advanced science and technology, and advanced imperial dominance (American style) find so encumbering. That is the fact that we speak so many different languages.

Hence the necessity of an international language. And what more plausible candidate than English?

This globalization of English has had an already-perceptible effect on the fortunes of literature, that is, of translation. I suspect that fewer literary works of foreign literature, especially from the languages felt to be less important, are being translated into English than, say, twenty or thirty years ago. But many more books written in English are being translated into foreign languages. It is now extremely rare for foreign novels to appear on the New York Times bestseller list, as they did twenty, thirty, fifty years ago. Celebrated novels by Kundera and García Márquez and Lampedusa and Pasternak and Grass were bestsellers in the United States. A little over a half-century ago Thomas Mann’s Doctor Faustus was for a time number one on the bestseller list – inconceivable today.


It is often taken for granted that the aim of a translation is to make the work “sound” as if it were written in the language into which it is being translated.

Translation being an activity not only practiced in every nation but subject to national traditions, there are greater pressures in some countries than in others to efface as much as possible the evidence of foreignness. France has a particularly strong tradition of translation as adaptation, at the expense of strict fidelity to the text. I have often been told by my French publishers, when I pointed out flagrant inaccuracies in a translation of one of my books: “Yes, true…but it reads very well in French.” When I hear that my book or someone else’s, thanks to the translator’s efforts, now reads very well in French, I know that the book has been reshaped according to existing conventions (usually not the most fastidious ones) of contemporary French prose. But since my prose in English is not always conventional in its rhythms or its lexical choices, I can be sure that this is not being transmitted into French. Only the sense – and only a part of that (because the sense seems to me connected essentially with whatever is odd about my prose) – is being transmitted.

The first, and still perhaps definitive, criticism of the idea – so powerfully expounded by Jerome – that it is the job of the translator to completely recast the work to suit the spirit of the new language was made by the German Protestant theologian Friedrich Schleiermacher (1768-1834) in his great essay “On the Different Methods of Translation,” written in 1813.

In arguing that “reading well” is not the primary standard of merit in translation, Schleiermacher does not, of course, mean all translations but only literary translations – those that involve what he calls, appealingly, “the sacred seriousness of language.” As for the rest, he writes:

…as nations appear to mix in our time to a greater extent than they did before, the marketplace is everywhere and these are conversations of the marketplace, whether they are social or literary or political, and really do not belong in the translator’s domain but rather in that of the interpreter. (tr. André Lefevere)

For Schleiermacher, translation – which is far more than a service to commerce, to the marketplace – is a complex necessity. There is the intrinsic value of making known, across a linguistic border, an essential text. There is also a value in connecting with something that is different from what we know, with foreignness itself.

For Schleiermacher, a literary text is not just its sense. It is, first of all, the language in which it is written. And as each person has a core identity, each person has, essentially, only one language.

Just as a man must decide to belong to one country, he must adhere to one language, or he will float without any bearings above an unpleasant middle ground. It is right that even now Latin is being written among us as the language of officialdom, to keep alive the consciousness that this was the scholarly and sacred mother tongue of our ancestors; it is good that this should also happen in the field of the common European economy, to make commerce easier; but in that case, too, it will succeed only to the extent that the object is everything for such a representation and that one’s own opinion and the way in which one combines objects counts for very little.

Substitute English for Latin in Schleiermacher’s extremely reserved encomium for a pan-European (read: global) language required to facilitate pan-European (read: global) technical and scientific exchanges, and you will see how little he expects of this language as a medium of subjective, that is, literary expression.

In the matter of concrete practice, Schleiermacher takes up the exact opposite of Jerome’s position, arguing that the translator’s primary duty is to stay as close as possible to the original text, with the understanding that the result will, precisely, read as a translation. To naturalize a foreign book is to lose what is most valuable about it: the spirit of the language, the mental ethos out of which the text emerges. Therefore if a translation from, say, French or Russian into German sounds as if it were originally written in German, the German-speaking reader will be deprived of the knowledge of otherness that comes from reading something that actually does sound foreign.

The difference between Jerome’s and Schleiermacher’s position is the difference made by the interposition of the idea of national identity as the framework around which linguistic separateness coheres. For Jerome, to speak another language was not to be another kind of person. Jerome lived in a world that was, in ways not unlike our own, significantly transnational or international. For Schleiermacher, to speak another language was to become, in the deepest sense, inauthentic. He writes:

…the aim of translating in a way such as the author would have originally written in the language of the translation is not only out of reach, but also null and void in itself, for whoever acknowledges the shaping power of language, as it is one with the peculiar character of a nation, must concede that every most excellent human being has acquired his knowledge, as well as the possibility of expressing it, in and through language, and that no one therefore adheres to his language mechanically as if he were strapped into it…and that no one could change languages in his thinking as he pleases the way one can easily change a span of horses and replace it with another; rather everyone produces original work in his mother tongue only, so that the question cannot even be raised of how he would have written his works in another language.

Schleiermacher is not, of course, denying that there is such a thing as the ability to speak and to write in more than one language. But he is assuming that everyone has a “mother tongue,” and the relation to other languages in which one might speak, or even write poetry and philosophize, would somehow be not “organic” – a favorite metaphor of the period. This is, clearly, an ideological position: many peoples have been duoglot, if not polyglot. Italy, for example, where one might speak not only Tuscan (so-called Italian) but also Neapolitan or Romagnolo. Québec, where educated people speak both French and English. In the old Austro-Hungarian Empire, most educated people in the countries now called Austria, the Czech Republic, Romania, and Hungary spoke at least two, sometimes three languages every day. Clearly, Schleiermacher’s position is not merely a descriptive one. (His deep agenda has to do with his notion of nationhood and peoplehood.) In Schleiermacher’s view, it is not that one cannot, but that one should not deploy two languages as equals. The epitome of inauthenticity would be to assume that one could inhabit another language in the same spirit that one could one’s own.


But can one authentically speak more than one language?

Schleiermacher’s question continues to reverberate. What does mastery of a second language mean?

I have been told by American and English friends who are longtime residents of Japan (most with Japanese spouses) that the Japanese typically regard with great suspicion, and even a touch of hostility, a foreigner who speaks their language without making mistakes. But probably this prejudice will fade, as Japan continues to accept that the existence of foreigners in its midst is not an oddity or a misfortune or an adulteration of the national essence.

At the other extreme, a more recent example of what is involved in attaining perfect mastery in a second language – which happens to be English – gives us a perfect Schleiermachian scenario of inauthenticity. I am thinking about one flourishing enterprise in the multibillion-dollar software industry now so important to the Indian economy. These are the call centers, employing many thousands of young women and men who give technical help or take reservations made by dialing 1-800 (that is, toll-free numbers all over the United States). The young people, all of whom already speak English, who compete successfully for these coveted jobs in the call centers, and have completed the arduous course designed to erase all traces of their Indian accent in English (many fail), are being paid what is a munificent salary for office work in India, though of course far less than what IBM, American Express, GE, Delta Airlines, and chains of hotels and restaurants would have to pay to Americans to do the job – reason enough for more and more such tasks to be “outsourced.” It also seems to be the case that the Indians perform the tasks better, with fewer errors, which is not surprising, since virtually all of them have college degrees.

From large floors of office buildings in Bangalore or Bombay or New Delhi, call after call is answered by young Indians seated in rows of small booths (“Hi, this is Nancy. How may I help you?”), each equipped with a computer that allows them to summon with a few clicks the relevant information to make a reservation, maps to give information about the best highway route, weather forecasts, and so forth.

Nancy, or Mary Lou, Betty, Sally Jane, Megan, Bill, Jim, Wally, Frank – these cheerful voices had first to be trained for months, by instructors and by tapes, to acquire a pleasant middle-American (not an educated American) accent, and to learn basic American slang, informal idioms (including regional ones), and elementary mass-culture references (TV personalities and the plots and protagonists of the main sitcoms, the latest blockbuster in the multiplex, fresh baseball and basketball scores, and such), so that if the exchange with the client in the United States becomes prolonged, they will not falter with the small talk and will have the means to continue to pass for Americans.

To pull this off, they have to be plausibly American to themselves. They have been assigned American names and little biographies of their American identities: place and date of birth, parents’ occupation, number of siblings, religious denomination (almost always Protestant), high school, favorite sport, favorite kind of music, marital status, and the like. If asked where they are, they have a reply. For example, if the client is calling from Savannah, Georgia, to make a reservation in a hotel in Macon, Georgia, and is asking directions for the quickest way to drive from Savannah to Macon, the operator might say she or he is in Atlanta. Letting on that they are in Bangalore, India, would get pretend-Nancy or pretend-Bill instantly fired. (All the calls are routinely, and undetectably, monitored by supervisors.) And of course virtually none of these young people has ever left home.

Would “Nancy” and “Bill” prefer to be a real Nancy and a real Bill? Almost all say – there have been interviews – that they would. Would they want to come to America, where it would be normal to speak English all the time with an American accent? Of course they would.


Our ideas about literature (and therefore about translation) are necessarily reactive. In the early nineteenth century it seemed progressive to champion national literatures, and the distinctiveness (the special “genius”) of the national languages. The prestige of the nation-state in the nineteenth century was fueled by the consciousness of having produced great “national” writers – in countries such as Poland and Hungary, these were usually poets. Indeed, the national idea had a particularly libertarian inflection in the smaller European countries, still existing within the confines of an imperial system, which were moving toward the identity of nation-states.

Concern for the authenticity of the linguistic embodiment of literature was one response to these new ideas and gave rise to intense support for writing in dialects or in so-called regional languages. Another altogether different response to the idea of national identity was that of Goethe, who was perhaps the first to broach – and at a time, the early nineteenth century, when the idea of national identity was most progressive – the project of world literature (Weltliteratur).

It may seem surprising that Goethe could have fielded a notion so far ahead of its time. It seems less odd if one thinks of Goethe as not only Napoleon’s contemporary but as Napoleonic himself in more than a few projects and ideas that could be the intellectual equivalent of the Napoleonic imperium. His idea of world literature recalls Napoleon’s idea of a United States of Europe, since by “the world” Goethe meant Europe and the neo-European countries, where there was already much literary traffic over borders. In Goethe’s perspective, the dignity and specificity of national languages (intimately tied to the affirmations of nationalism) are entirely compatible with the idea of a world literature, which is a notion of a world readership: reading books in translation.

Later in the nineteenth century internationalism or cosmopolitanism in literature became, in powerful countries, the more progressive notion, the one with the libertarian inflection. Progress would be the natural development of literature from “provincial” to “national” to “international.” A notion of Weltliteratur flourished through most of the twentieth century, with its recurrent dream of an international parliament in which all nation-states would sit as equals. Literature would be such an international system, which creates an even greater role for translations, so we could all be reading each other’s books. The global spread of English could even be regarded as an essential move toward transforming literature into a truly worldwide system of production and exchange.

But, as many have observed, globalization is a process that brings quite uneven benefits to the various peoples that make up the human population, and the globalization of English has not altered the history of prejudices about national identities, one result of which is that some languages – and the literature produced in them – have always been considered more important than others. An example. Surely Machado de Assis’s The Posthumous Memories of Brás Cubas and Dom Casmurro and Aluísio Azevedo’s The Slum, three of the best novels written anywhere in the last part of the nineteenth century, would be as famous as a late-nineteenth-century literary masterpiece can be now had they been written not in Portuguese by Brazilians but in German or French or Russian. Or English. (It is not a question of big versus small languages. Brazil hardly lacks for inhabitants, and Portuguese is the sixth most widely spoken language in the world.) I hasten to add that these wonderful books are translated, excellently, into English. The problem is that they don’t get mentioned. It has not – at least not yet – been deemed necessary for someone cultivated, someone looking for the ecstasy that only fiction can bring, to read them.

The ancient biblical image suggests that we live in our differences, emblematically linguistic, on top of one another – like Frank Lloyd Wright’s dream of a mile-high apartment building. But common sense tells us our linguistic dispersion cannot be a tower. The geography of our dispersal into many languages is much more horizontal than vertical (or so it seems), with rivers and mountains and valleys, and oceans that lap around the land mass. To translate is to ferry, to bring across.

But maybe there is some truth in the image. A tower has many levels, and the many tenants of this tower are stacked one on top of the other. If Babel is anything like other towers, the higher floors are the more coveted. Maybe certain languages occupy whole sections of the upper floors, the great rooms and commanding terraces. And other languages and their literary products are confined to lower floors, low ceilings, blocked views.


Some sixteen centuries after Saint Jerome, but barely more than a century after Schleiermacher’s landmark essay on translation, came the third of what are for me the exemplary reflections on the project and duties of the translator. It is the essay entitled “The Task of the Translator” that Walter Benjamin wrote in 1923 as a preface to his translation of Baudelaire’s Tableaux parisiens.

In bringing Baudelaire’s French into German, he tells us, he is not obliged to make Baudelaire sound as if he had written in German. On the contrary, his obligation is to maintain the sense that the German reader might have of something different. He writes:

All translation is only a somewhat provisional way of coming to terms with the foreignness of languages…It is not the highest praise of a translation, particularly in the age of its origin, to say that it reads as if it had originally been written in that language. (tr. Harry Zohn)

The opportunity offered by translation is not a defensive one: to preserve, to embalm, the current state of the translator’s own language. Rather, he argues, it is an opportunity to allow a foreign tongue to influence and modify the language into which a work is being translated. Benjamin’s reason for preferring a translation that reveals its foreignness is quite different from Schleiermacher’s. It is not because he wishes to promote the autonomy and integrity of individual languages. Benjamin’s thinking is at the opposite pole to any nationalist agenda. It is a metaphysical consideration, arising from his idea of the very nature of language, according to which language itself demands the translator’s exertions.

Every language is part of language, which is larger than any single language. Every individual literary work is a part of literature, which is larger than the literature of any single language.

It is something like this view – which would place translation at the center of the literary enterprise – that I have tried to support with these remarks.

It is the nature of literature as we now understand it – understand it rightly, I believe – to circulate, for diverse and necessarily impure motives. Translation is the circulatory system of the world’s literatures. Literary translation, I think, is preeminently an ethical task, and one that mirrors and duplicates the role of literature itself, which is to extend our sympathies; to educate the heart and mind; to create inwardness; to secure and deepen the awareness (with all its consequences) that other people, people different from us, really do exist.


I am old enough to have grown up, in the American Southwest, thinking there was something called literature in English, of which American literature was a branch. The writer who mattered most to me as a child was Shakespeare – Shakespeare as a reading experience (actually, a reading-aloud experience), which started with my being given a prettily illustrated edition of Charles Lamb’s Tales of Shakespeare when I was eight; my reading of Lamb and then of many of the plays predated by four years my actually seeing Shakespeare on the stage or in a film adaptation. And besides Shakespeare, retold or straight, there was Winnie-the-Pooh and The Secret Garden and Gulliver’s Travels and the Brontës (first Jane Eyre, then Wuthering Heights) and The Cloister and the Hearth and Dickens (the first were David Copperfield and A Christmas Carol and A Tale of Two Cities) and lots of Stevenson (Kidnapped, Treasure Island, Dr. Jekyll and Mr. Hyde), and Oscar Wilde’s The Happy Prince…Of course, there were American books, too, like the tales of Poe and Little Women and novels of Jack London and Ramona. But in that distant, still reflexively genteel, culturally Anglophile era, it seemed perfectly normal that most of the books I read should come from somewhere else, somewhere older, such as faraway, thrillingly exotic England.

When the “somewhere else” grew larger, when my reading – always in English, of course – came to include wonderful books that had not originally been written in English, when I moved on to world literature, the transition was almost imperceptible. Dumas, Hugo, and on from there…I knew I was now reading “foreign” writers. It didn’t occur to me to ponder over the mediation that brought these evermore-awesome books to me. (Question: Who is the greatest Russian writer of the nineteenth century? Answer: Constance Garnett.) Had I recognized an awkward sentence in a novel I was reading by Mann or Balzac or Tolstoy, it would not have occurred to me to wonder if the sentence read as awkwardly in the original German or French or Russian, or to suspect that the sentence might have been “badly” translated. To my young, beginning reader’s mind, there was no such thing as a bad translation. There were only translations – which decoded books to which I would otherwise not have had access, and put them into my hands and heart. As far as I was concerned, the original text and the translation were as one.

The very first time I raised to myself the problem of a poor translation was when I started going to the opera, in Chicago, when I was sixteen. There I held in my hands for the first time an en face translation – the original language on the left (by this time I had some French and Italian) and the English on the right – and I was stunned and mystified by the blatant inaccuracy of the translations. (It was to be many years before I understood why the words in a libretto cannot be translated literally.) Opera excepted, I never asked myself, in those early years of reading literature in translation, about what I was missing. It was as if I felt it were my job, as a passionate reader, to see through the faults or limitations of a translation – as one sees through (or looks past) the scratches on a bad print of a beloved old film one is seeing once again. Translations were a gift, for which I would always be grateful. What – rather, who – would I be without Dostoyevsky and Tolstoy and Chekhov?

My sense of what literature can be, my reverence for the practice of literature as a vocation, and my identification of the vocation of the writer with the exercise of freedom – all these constituent elements of my sensibility are inconceivable without the books I read in translation from an early age. Literature was mental travel: travel into the past…and to other countries. (Literature was the vehicle that could take you anywhere.) And literature was criticism of one’s own reality, in the light of a better standard.

A writer is first of all a reader. It is from reading that I derive the standards by which I measure my own work and according to which I fall lamentably short. It is from reading, even before writing, that I became part of a community – the community of literature – which includes more dead than living writers. Reading, and having standards, are then relations with the past and with what is other. Reading and having standards for literature are, indispensably in my view, relations with literature in translation.

Published by FSG in At the Same Time: Essays and Speeches.
Copyright © 2007 by the Estate of Susan Sontag.


© 2010 Estate of Susan Sontag. All Rights Reserved.