I first read Madame Bovary in my teens or early twenties. Although even in high school I was aware of translators and translations, it never, ever occurred to me that the reason I did not like the novel might have been not only its unsympathetic characters (whom Flaubert himself did not like), or the weak and relatively thoughtless heroine (I craved a strong, thoughtful model), but most of all the inadequate translation. There is great trust in translations on the part of many people who don’t know any better and even many who do. Now that I’m aware of how many previous translations of Madame Bovary there are, and how inadequate most of them are, I suspect I read a bad one.

The quality and nature of a translation (let’s say from the French) depends on three things, the first fairly obvious and the second two not quite as obvious: 1) the translator’s knowledge of French language, history, and culture; 2) his or her conception of the task of the translator; and 3) his or her ability to write well in English. These three variables have infinite subsets that recombine infinitely to produce the many different kinds and qualities of translations that we have. Publishers selecting a translator seem to proceed on the assumption that the most important qualification is the first. “Let’s ask Prof. X, head of the French Department at Y!” Often they completely ignore the second factor—how will Professor X approach the task of translating?—and certainly the third—what is Professor X’s writing style like? All three factors are vital, but in many instances, if one has to rank them, the third—how well the translator writes—may be the most important qualification, followed closely or equaled by the second—how he or she approaches the task of translating—and it is the first that comes in last place, since minor lapses in a knowledge of the language, history, and culture may result in mistakes that are, in a beautifully written, generally faithful version, fairly easily corrected, whereas a misconception of the task of the translator and, worse, an inability to write well will doom the entire book through its every sentence.

Eleanor Marx Aveling, daughter of Karl Marx, produced the first translation of Madame Bovary in 1888. The Paul de Man revision of the Marx Aveling translation (Norton, 2005, 1965) retains some of her old-fashioned or inappropriate vocabulary, such as “heretofore” and “conjure” for “beg” or “plead.” It integrates explanations or identifications into the text (“the Chaumière” becomes “the Chaumière dance hall”)—undoubtedly helpful to the reader, but a betrayal of the original—and the writing style is poor, the revision making a rather poor style even poorer. For a while it seemed to me the very worst translation out of the eleven. It isn’t. Maybe it’s the second worst. But then, such a thing is hard to judge, because in certain specific passages, it is the worst. Although Marx Aveling was not a brilliant writer, she was a better writer in English than de Man, so where he corrects a mistake of hers, the correction is often not as well written as the original mistake. (This occurred, also, in the Kilmartin and the Enright revisions of Scott Moncrieff’s Proust, where the original impeccable grammar of the earliest version is replaced by an “improvement” that introduces a grammatical mistake.)

The book exists for a couple of wrong reasons, and people buy it for another wrong reason. Wrong reason #1: Norton chose to use the Aveling translation because it was in the public domain and wouldn’t cost anything (I’m assuming, or I was told—can’t remember which). Wrong reason #2 (I’m guessing here): They asked Paul de Man to revise and edit it not because he was conscientious and an excellent writer in English but because he had prestige, a reputation, and scholarly intelligence. He then apparently asked his wife to do much of the work (this is rumor, but from a good source—I’d be happy to have it proved either right or wrong) and did not acknowledge her. Wrong reason #3: People buy the book not because it is an excellent translation of this important novel, but because it has a useful apparatus of essays, etc.—handy for a teacher, for instance. So readers have a collection of useful material to read about the novel, but are reading one of the most important novels in the history of the novel, and one of the most famous novels, in a poor translation.

Lydia Davis’s translation of Madame Bovary debuted last week. On October 4, she will be speaking at the 92nd Street Y.

See Also: “Group Think”

See Also: Lydia Davis in Feed Magazine, from 2000