Skip to content
Give readers a window on the world. Click to donate.

Translating the Classics: An Interview with Lydia Davis

By Ilan Stavans and Regina Galasso

Image: Lydia Davis at the University of Georgia Chapel (with George Cookes 1847 painting, "Interior of St. Peter's"), Wikimedia Commons.

In the spring of 2016, as co-teachers of the course Translating the Classics, we and our eight students engaged in an electronic conversation with Lydia Davis, acclaimed author of short stories, collected in volumes like Can’t and Won’t (2014), Varieties of Disturbance (2007), and Break It Down (1986); and translator of Proust’s Swann’s Way (2003) and Flaubert’s Madame Bovary (2010), among other classics in French literature. The course focused on translating classics like Cervantes’s Don Quixote, Dostoyevsky’s Crime and Punishment, Kafka’s The Metamorphosis, and even, intralingually, Shakespeare’s Macbeth. The dialogue with Davis was about her strategies. What follows is an edited version.

Is it a sin to be bored when translating a classic?

Lydia Davis (LD): I can admit to being bored at moments while translating Proust—although maybe I should say not exactly bored but feeling suddenly, Not again, not more, not the same thing today as yesterday! But that is inevitable, with a long job. Swann’s Way is a long book, and I did not feel this more than very occasionally, I’m happy to say. I think that if one is bored a lot of the time when translating, whether working on a classic or not, then one shouldn’t be doing it. It’s not a sin, just a sign—that one is in the wrong business.

Would you correct errors that you find in the original?

LD: Some translators do this. There are errors in Proust—I forget the specifics now, but he refers in one spot to four friends on a trip to Italy together and in another spot specifies three. But I believe it is very important not to tamper with the content of the original in that way, much as one might be tempted. One of the obligations of a translator is to try to reproduce something like the way the text is experienced by a native reader. Mistakes and all . . . I would, though, want to say something about the mistake in an endnote.

Are there specific moments in Madame Bovary where you felt that your translation achieved meaning or musicality that did not exist in the original text?

LD: I don’t know about introducing meaning, but no doubt there are moments in my translation that introduce musicality that was not in the original. This is not deliberate on my part, or I should say rarely is it deliberate. But I believe in a sort of equivalence in a translation, whereby what it lacks in musicality compared to the original in one spot can be made up for by musicality in another that may be missing in the original. But not all texts are meant to be musical anyway.

The reason I say “I don’t know about meaning” is partly that meaning is a very elusive thing and in part dependent on a reader’s interpretation; and also that, as I translate I don’t always, or maybe even often, think about meaning per se. I trust—and it seems to work—that if I am faithful to what I understand as I read the original, the “meaning” and everything else—humor, irony, pathos, etc.—will come through. I try to keep the act of translating as straightforward as possible. (This, by the way, applies to translating prose—the translation of poetry is entirely different.)

Does a translator need to dominate the culture of both the language she is translating to and the culture of the language she is translating from?

LD: By “dominate,” do you mean “master”? Or, even better: “have a deep and thorough understanding of it”? I want to clarify, because the attitude of a writer, including a translator, toward his or her own culture, as well as the culture of the original text, should be that of a seeker rather than a dominator. One is always seeking to understand. One gains some understanding, but one never understands completely—true of any culture in which one is working or living.

But to answer more simply: let’s assume that the translator has a good, deep understanding of her own culture. Then the question is how deep does her understanding of the other culture need to be? I found, in translating Madame Bovary, that a good deal of the text was understandable, and translatable, without that deeper knowledge of nineteenth-century French culture in a provincial town. Certain human behavior seems to be fairly universal, or at least common, to Western civilizations of the last couple of centuries. (I should beware of generalizations—there are always exceptions!) Other habits, customs, expressions are not as familiar to us in the twenty-first century. Still, translating the way I do, staying close to the original—even when it comes to expressions such as “to put straw in one’s boots” or “other dogs to beat” (yes? is that what Homais says to the beggar?)—rather than seeking equivalent expressions in English, the customs, habits, even modes of thinking of Flaubert’s time come through quite well. But I may translate accurately what is on Emma’s mantelpiece without knowing what her taste in decor “means”—and it would be good to know, even though that wouldn’t change my translation, in this case. For Flaubert, of course, what she had on her mantelpiece indicated her slavish following of current fashion, her striving for bourgeois gentility. His readers at the time would have known that. I use many reference books, learn what I can, write endnotes to help readers of the translation, but I do not feel I have to become a scholar of the culture Flaubert was writing about, or within. (Long answer! Third cup of coffee!)

We notice in Madame Bovary that you tend to preserve words related to social rank and position in the original French (e.g. marquise, vicomte, curé) but find English equivalents for food, though it can also be seen as culturally specific. Is this kind of distinction a system that you would apply to any prose translation, or is it contingent on the content of the text? 

LD: I haven’t thought about whether I have a general rule for this sort of choice, probably because I don’t have a general rule. Or the general rule is: when to translate the French term would give a false equivalent, as in the case of the titles you mention, since a duke is not the same as a duc, or at least has different associations, in English, then I don’t translate it. (There was disagreement, by the way, among the translators of the Penguin Proust, about whether or not to translate curé—but my feeling is that the associations with the figure of the French curé are quite specific and I wouldn’t want to lose them by translating the term.) As for the foods, I don’t at this point remember the actual instances, but the same argument might well be applied to the translation of foods. Yet you do have to be careful not to have a translation that is too filled with terms left in the original.

Was there a defining moment for you as a young, aspiring translator? What was the first translation you did?

LD: I translated some poems by Blaise Cendrars for a college literary magazine, at the request of a friend who was putting together a mini-anthology of French poetry. But even in high school the thought went through my head that I might become a translator (as well as writer—that thought was there from very early on). In working on my first book-length translation, Maurice Blanchot’s Death Sentence, which I began sometime in my twenties, I did not have any aspirations to be a great translator, for instance, or to launch a career in translation. I simply enjoyed doing it: I admired and liked the book (a love story, of sorts, about a dying woman who is called back to life for a moment by the narrator’s love); I liked the kind of activity of writing which translation is. I wanted my translation to be as good as it could be.

Are you ever satisfied with a translation job? Do you believe your translation of Madame Bovary is finished? Is the moment the publisher reads a translation the moment it might indeed be finished?

LD: I can feel that there is no more I can realistically do, within the time frame of the demands of a due date and publication. I won’t send a translation to the publisher until I feel it is good, or as good as I can make it. There will always be moments in a translation, however, that I’m not satisfied with even though I’ve struggled for a long time over them. But translation is often about compromise. In the case of the Proust translation, I made changes for the American edition (the first publication was in the UK) and then again for the American paperback—hundreds of little changes over hundreds of pages. After that publication, I found still more small things I wanted to change. But there has to be a stopping point sometime, unless the “perfect” translation of the book is to be a lifetime’s work. (Which it can and should be, in some cases.)

Classic authors have “something,” a quality of their style that sets them apart. Do translators have a similar “signature”?

LD: In theory, no—or, in theory, a translator should not have a recognizable, “signature” style. Otherwise, he or she may be imposing that style on the original, whatever the style of the original. A translator should enter the skin of the original author (sounds grotesque . . .) and become that author, to a certain extent, become transparent, lose herself, and leave her distinctive style behind. But, inevitably, I’ve come to realize that each writer—and again, a translator is a writer—has her own approach to writing, not only a preferred style but also a preferred vocabulary. The style of the translation will, one hopes, be determined by the style of the original, but since we have a choice of words for the English equivalent of the French, we may choose “our” kind of word rather than another.

Do you think leaving words in the original language is overused? And when left in the original language, should a word be italicized?

LD: Some foreign words, of course, have become entirely familiar in English, like “apropos” and “prix fixe” and so on. Other words, like vicomte, can remain in French without italics. Same for curé. And then others, when they are unusual, unfamiliar, should be italicized. There are rules about this, in fact, in publishers’ style sheets. There should not be too many of those italicized terms—they shouldn’t pepper the text—but I think Anglophone readers don’t mind encountering and figuring out a foreign term now and then. And I favor extensive endnotes (no marks on the page of the text, though!).

Do you think translators should be invisible?

LD: I’m taking these questions in order without reading ahead (the way I also translate—one page at a time, without much looking ahead), so I’ve touched on this earlier, and, yes, I do think the translator should be more or less invisible, except for that opportunity at the end of the book to hold forth in the endnotes, or to speak directly to the reader in an afterword or translator’s note about the translation and the text. I even object to the practice of the translator including a personal dedication to someone or other on one of the first pages—I feel that is an unwanted and unwarranted intrusion of the translator between the reader and the original work.

Would you agree to someone “improving” your translation in, say, a couple of decades? Not doing a new one but updating yours?

LD: Yes, it would be hypocritical of me not to agree to that, in theory, since I’ve often wanted to “improve” someone else’s translation—and not an older one but a contemporary one! In other words, the original translator has presumably done a lot of the legwork, and, let’s say, done a pretty good job, but there are some remaining inaccuracies or infelicities that could be corrected. Now this becomes a little tricky, of course—inaccuracies is one thing, infelicities may be a matter of opinion, and it could be that the “improver” only thinks he has a better sense of style than the original translator, and in fact doesn’t. Actually, that reminds me that the two revisers of Scott Moncrieff’s Proust did not always improve the translation. Some clumsy writing and at least one actual grammatical error were introduced by the “improvers,” which made me quite indignant on Scott Moncrieff’s behalf. So, any improver would have to provide a careful note about what sort of changes had been made. And the original, unimproved translation, one would hope, would still be available to readers.

Is there a difference between a translation of a classic and a classic translation, e.g., a translation so good it stands the test of time? 

LD: Oh, yes. There are over twenty translations of Madame Bovary, but I don’t think any of those is a classic translation. Many are not good. (I looked at thirteen of them when I was working on mine.) Others are good in one way and not in another. The Francis Steegmuller translation has been treated as a classic translation, because it is a good and lively piece of writing, but it sometimes strays rather far from the original—should it be called a classic? The Scott Moncrieff translation of Proust is certainly called, and considered, a classic, and I suppose it has earned this to some degree, having been such a monumental labor of love and quite masterful, if one accepts a certain type of style and a certain approach to translation (freer than I would allow myself, for instance). It is a classic, but I would recommend it only with reservations. Keats’s poem “On First Looking into Chapman’s Homer” is an homage to a classic translation.


Suggested Reading:

“Have We Lost The Lofty? Virgil’s Aeneid and the History of English Poetry” by Aaron Poochigian from WWB’s Dispatches blog, January 2016.

“Keep It Growing: A Translator’s Take on Ezra Pound’s ‘Make It New’” by Olga Nikolova from WWB’s Dispatches blog, February 2016.

Published Jun 22, 2016   Copyright 2016 Ilan Stavans and Regina Galasso

Leave Your Comment

comments powered by Disqus
Like what you read? Help WWB bring you the best new writing from around the world.