Haim Watzman

220px-Vladimir_NabokovIn the preface to the English translation of his novel Invitation to a Beheading— translated by the author’s son under the direction of the author himself — Vladimir Nabokov states unambiguously that a translator’s faithfulness to his author is a supreme obligation even if the product is ludicrous:

My son proved to be a marvelously congenial translator, and it was settled between us that fidelity to one’s author comes first, no matter how bizarre the result. Vive le pedant , and down with the simpletons who think that all is well if the “spirit” is rendered. . . .

But in that same preface Nabokov contradicts himself when he explains his English rendering of the book’s title. He acknowledges that the English name he has chosen is not a literal translation of the Russian one. His excuse is a strange one for a writer who insists on fidelity to the text, for he says that the Russian title is not really the one he wanted, but was in fact imposed on him by the phonetics of his native language:

The Russian original of this novel is entitled Priglashenie na kazn’. Notwithstanding the unpleasant duplication of the suffix, I would have suggested rendering it as Invitation to an Execution; but, on the other hand, Priglashenie na otsechenie golov (Invitation to a Decapitation) was what I really would have said in my mother tongue, had I not been stopped by a similar stutter.

So Nabokov is admitting that there are situations in which fidelity to the source text is in fact an error. A translator cannot assume, simply because a writer chose a certain string of words in his native language, that he would have chosen the literal equivalent of those words had he been writing in some other language.

Hence my fundamental claim that a translator cannot escape being an editor of the text he is translating. There is, of course, an obligation to the source text and the author, but the translator must take a critical approach to both in order to produce a good translation.

As a translator who works mostly on non-fiction, I will not express an opinion here about the right, or responsibility, of a translator confronting a work of fiction to alter the original. But in non-fiction — in particular, scholarly and academic non-fiction — such alterations are not only inevitable but are in fact a sacred duty. This does not imply any disparagement of non-fiction as a genre. Quite the opposite — my opinion is that the quality of the writing in non-fiction today is in many cases much higher than that in what is termed “literature” by those for whom that term is synonymous with fiction. No writer in any field sits down to write a book or article without investing time and thought in his writing. Most of my clients consider their books to be literary works even if they are not novels or stories.

There are, however, fundamental differences between a novel and a book of non-fiction, say a work of history. A writer of fiction can choose a complex style, make use of symbols, of a variety of techniques; in short, he can impose upon his reader in all sorts of ways, and that is considered art, whether successful or unsuccessful. But a person who writes a history that no one understands misses his target. So a fundamental axiom of the translation of non-fiction is that the translator has to get the writer’s message across clearly, and sometimes that demands certain changes in the text — in its style, in its choice of words, sometimes even in its contents. So the translator of non-fiction is actually doing the work that a good editor at a respectable publishing house does — bridging between the writer and his audience, negotiating between the need of the author to express himself as he sees fit and the right of the reader to receive a comprehensible manuscript.

Translating non-fiction into English is generally, by nature, work with living, available authors who know the target language. If the translator has a question, wants to point out an unclear passage, or offer a suggestion for improvement, he can pick up the telephone or meet with the author, explain the problem, justify his proposed solution, and make a decision together with the author. When the author knows English well, he can choose between alternative wordings suggested by the translator and can also criticize the draft translation, pointing out places where the translator has not properly understood the original or has not succeeded in getting across some subtlety that the author thinks important. When the author is available and able to criticize his work, the translator can afford to be freer with the text without fear of misrepresenting the original to the reader.

A non-fiction work can present any of a number of specific problems that the translator must solve. The first and worst case is a text that may have a lot to say but is simply not well-written in the original. I have no experience with other countries, but in Israel this situation happens more frequently than it ought to because the status of an editor at an Israeli publishing house is much weaker than that of his American counterpart. For example, some Israeli publishers will take a Ph.D. thesis and publish it, almost unchanged, in book form. Such books would benefit from significant cutting and rewriting because they were originally written for a very small group of scholars engaged in the specific discourse engaged by the thesis. For them to be readable and enlightening for interested scholars outside that specific field, not to mention for educated laymen, they require considerable reworking. If that has not been done in the production of the Hebrew book, then it must be done as part of the process of producing an English version.

Sometimes a writer produces a vague sentence. This may be intentional or unintentional. It could be that she is simply unsure of what she wants to say, or that she knows but refrains from saying what she wants to say. Perhaps she is shy, perhaps she fears the reactions of her colleagues, or it could be that she is burning to make a statement that the documentary evidence she has at hand is simply not sufficient to support. Since I work as a journalist as well, I have a preference for frank, outspoken sentences, but certainly academic work sometimes requires more modesty than newspaper reporters normally have. Yet even a justifiably vague sentence is a problem for the translator of scholarly work, where it is important to know the author’s intention in order to translate accurately. As I once told a client who complained that I had made her intentionally ambiguous sentences too unequivocal, I cannot translate the specific ambivalence you intend unless you can define me that ambivalence precisely. Ambiguity can be a virtue, wrote Robert Graves, pointing to the Balfour Declaration as an example: “[A writer] may not always wish to be so understood; a good deal of play is made in English with deliberate looseness of phrase.” Scholars, however, are not diplomats, and in general ought to be more explicit.

Hence my tendency to nag my clients with endless questions about what they mean, and why they are saying it in that specific way. A historian I have done several books with once told me about the anxiety he feels when he gives me a manuscript. A scholar, he said, stands naked before his translator. Everything he has tried to fudge, to get around, every place he was a little too lazy or hesitant to think a thought out clearly, gets put under the spotlight by the translator, forcing him to explain exactly what it is he means.

Then there are the specific problems that are always involved in moving from one language to another. Sometimes a certain string of words or metaphor cannot be translated in any way approaching the literal because the same comparison or turn of phrase just does not exist in the target language. The translator must find a substitute with a similar meaning, but inevitably this adds new connotations and loses old ones. Sometimes this can be a more general problem. Hebrew prose, for example, has a high tolerance for certain things that are considered bad style in English. One of these is passive voice, a problem because writers often take advantage of this to hide something they don’t know. The translator, who must make some of this passivity active, must either guess or ask who is doing the action. In Hebrew, repeating a certain word, in different declensions or forms, adds emphasis and rhythm; doing the same in English often sounds repetitious and boring.

Hebrew also tolerates sentences much longer than those considered acceptable in good English prose. A translator must respect an author’s style and endeavor to convey it into the target language. But this cannot be done by photocopying the style from one language to the other. Instead, an equivalent style must be found. The long, flowing, multi-commaed sentences of a Hebrew writer cannot be transferred as is, with the same punctuation, into English, because a long, flowing English sentence is seldom as long as a Hebrew one. The Hebrew sentences have to be cut up to a certain extent to achieve the same effect in English.

Another problem is the one Nabokov raises. Sometimes the translator can sense with a great deal of certainty that a writer would not have chosen a certain turn of phrase had she been writing in English. One example is the title of a recent work of history published in Israel. The author, writing about the Holocaust refugees who were brought to Israel from Europe in the late 1940s, chose the book’s title from a poem by one of Israel’s classic poets, Natan Alterman. In Hebrew it echoes with humanity. But were a translator to try to render anything like it in English, he would be doing the author a grave injustice, because it would come out as some variation on “The Gold of the Jews” and sound like the title of an antisemitic tract.

There are other times, I admit, when I encounter an idea or narrative in a text I am working on and am overcome with the certainty that I can express the idea or tell the story better. This takes a certain amount of nerve, but if the relations between the author and the translator are good, I see no problem, and in fact a positive obligation, to draft an alternative and present it to the author as a possibility. In my experience, these suggestions are accepted in good spirit. After all, both author and translator have an interest in producing a good manuscript. I do not accept the excuse given by so many translators when a reader points out a badly-written exposition or a monotonous paragraph — “But that’s the way it was in the original!” To take that attitude is, in my book, to be disloyal to your client.

Has this attitude cost me work? Here and there, certainly. I’m not embarrassed to say that I’ve had clients who didn’t like my work and some who were positively indignant at what I wanted to do to their manuscript. But when I look back at my career so far, my assessment is that I did my worst work when I was least bold.

You can’t work with just a theory — you need some rules of practice as well. My habits are connected in one way or another with my view of translation, and so might be of interest to others.

When I am first contacted by a prospective client, I explain my conception of translation. I emphasize that there is no such thing as a transparent translator. The translator’s own style and own way of working are inevitably evident in the final product, and therefore it is important that there be compatibility between the translator’s style and that of the author.

I generally offer to produce a small sample of two or three pages, free of charge. This both gives the client the opportunity to see my work on his manuscript, and me an opportunity to try my hand at the text. I find it nearly impossible to estimate how difficult a job will be until I actually try my hand at it.

This sample is not meant to show a final product, but rather a draft of the type that I will later submit to the author for her criticism and comments. Therefore, it is likely to contain questions and comments of my own.

While this is somewhat heretical, I admit that I seldom read a manuscript before translating it. The reasons are both practical and principled. On the practical side — well, to be frank, I simply don’t have the time. Pay is low in this field, hardly enough to justify investing the time needed for a serious pre-reading of a manuscript at work, and as the father of four children I have very little time to read at home.

Still, if I felt it was essential I would do it. Yet when I have invested the time to read a manuscript in advance, I have not found that it has improved the final result. On the contrary, I think it is sometimes harmful. When I am translating something I am reading for the first time, there is tension, excitement, and interest, all of which would be absent if I were to have read the piece before. When I am interested, I produce better work than when I am bored.

Furthermore, translating as part of a first reading puts the translator in the place of the average reader, who will generally be reading the book only once. If something is unclear in this first reading to the translator, it may well be unclear to the reader as well. If the translator is already reading a second time, he is in a preferred position to the reader and that is not necessarily a good place to be.

What I do on occasion read before beginning a project is related literature that will help me understand the background against which the author is writing. This is often more helpful than reading the manuscript itself. For example, before working on a book on Adolf Eichmann, I read Hannah Arendt’s Eichmann in Jerusalem, the book against which my client was arguing.

Generally, I translate a rough draft and edit it twice, once on the computer screen and once on paper, before submitting it to the author, together with my questions and comments. I ask the author to read it through, address my questions, and mark any points of his own on the printout. If necessary, we meet to discuss specific points and make decisions. Then I do one more edit and produce a final copy. In some cases a further round of editing is necessary. Preferably, some time, at least a few days, should elapse between each round of editing, but this is not always possible. When I feel I am having particular trouble, I sometimes have an independent editor go over my work.

If I had my way, I would translate the entire book, edit it, put it aside for a while, come back to it, and edit it again, all before showing the author anything. But scheduling and publication deadlines usually make this impracticable and in most cases I hand the author my draft a chapter or two at a time. This often means giving him material that is really much rougher than I would like it to be, and sometimes my clients have to be reassured that the material will be polished and reworked before it reaches its final form.

To return to Nabokov. In his novel Bend Sinister, he has a metaphor for translation that represents an ideal — to me, at least. From the context, I take it that Nabokov might not think it is a metaphor for a good translation. But one is allowed to disagree even with a master.

Nabokov tells of a sculptor who goes out for a walk in the woods. There he sees a tree, and the tree casts a shadow that fills the sculptor with wonder. The sculptor returns to his studio and creates a sculpture that looks nothing like the tree — but it casts the same shadow.

That, in short, is the job of the translator.

June 1998


  1. In a tiny town in coastal Oregon USA, my husband and I have so enjoyed reading Sapiens, A Brief History, that we have hunted you up. Thank you for a thrilling story we feel lucky to read in English. We’re pleased to learn how it’s done. Thank you.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.