Issues of Re-Translation

Too often, we automatically assume a new translation is superior to an existing translation. It might be, but it also might not be; to decide, the prudent critic must first read the new translation against the original text and then against the earlier translation, or translations. But even then, it will very often come down to subjective responses to slippery questions of taste, stylistic preferences, convention, and, above all, reader response. Affected by a spate of forces, including those of race, age, gender identification, sexual orientation, class, religious inclinations (or the absence thereof), or political persuasion, one person’s reading of a text will often differ radically from another’s. What is richly ambiguous and evocative for one person may be merely tedious for another, and the translation—the product of a particular reading and rewriting—will reflect this diversity of responses. An alleged “improvement” in one area may be offset by a slip or misstep in another. Then again, a perceived “improvement” may, for still other readers, be little more than another possibility, one semantic or stylistic choice among many. It is, therefore, a mistake to simply declare, grosso modo, a new translation “better” than existing ones.

It is, I think, more accurate to begin our consideration of the question of re-translation with what we know to be true: A new translation is different, in whole (rare) or in part (much more common). For those interested in comparative translation analysis, which I find both fascinating and hard proof of the importance of interpretation to the act of translation (the variances are revealing), the methodology required is quite rigorous. And, in the end, it is subjective, a matter of the careful, painstaking reading of a text and a judicious consideration of how and why it works as it does. But always translation and re-translation are functions of both reading and writing. The great translator Gregory Rabassa liked to say that as he was reading his texts in Spanish or Portuguese he was mentally re-writing them in the kinds and levels of English they required. For Greg, the ideal translator was both a meticulous, engaged reader and a creative writer, with the two being so intertwined that they became indivisible.

Sometimes, unfortunately, the already tricky waters of the re-translation question are further, and unnecessarily, roiled by novice translators who feel they have in their readings so completely “taken possession” of a writer that they are led to declare publicly that their versions of this writer are superior to earlier ones. It is one thing to recognize the brilliance of a given writer and to wish to demonstrate it, via your translation, to others, but it is quite another to claim that your understanding and recreation of said author is the “best” one and that it puts all others to shame. To do so does not cover one in glory. This situation is exacerbated when, sadly, these same translators seek to promote their own efforts by demeaning the efforts of those who came before them. There is no need to do this, and to proceed in this fashion is unseemly and unprofessional. And, as anyone who studies the history of a particular author in translation knows, it is also unwise. In time, the same new translations that are so highly regarded now (at least by the people who do them and their friends) will themselves become subject to critical scrutiny (by future readers) and, quite likely, be deemed wanting themselves. Unless, of course, they are truly perfect, which, given the fluid nature of language, interpretation, and writing (the basic ingredients of translation), seems highly unlikely. One can reasonably conclude, then, that a bit of judiciousness, professionalism, and respect are in order when touting the virtues of a re-translation.
Then, too, there is the role an editor plays in the translation process. Very often, it is the editor, and not the author of it, who makes the final decision about what the final version of a translation, or a re-translation, will be like. In my experience, this is particularly so in the realm of stylistic and structural issues, such as spacing and paragraphing, syntax, and punctuation. The stylistic idiosyncrasies that characterize one writer may not be acceptable to the editor who wants to publish that same writer but in a different language system. By its nature, an editor’s job is difficult and conflicted. On the one hand, she or he seeks to present the author of the original text in the best possible light; on the other hand, he or she must also sell books. An editor who backs books that do not make money for the company will not be long employed by that company. For the editor, the final decisions about what a new translation is like may well turn more on questions of profitability than on questions of aesthetics or fidelity to the original text. While the name of the author will be featured on the cover of the translated version, the language in which the public receives this author will be English, and, if the book is to be marketed in the United States, likely modern American English. This means that if the translation, or the re-translation, comes to life in an English that will put off English language readers, then they will not buy it. Editors are keenly aware of this, and they are charged not to allow it to happen.

Of this very point, Greg Rabassa liked to tell the story of one of his books, Paradiso, by the Cuban novelist José Lezama Lima. Lima’s original text, a brilliant example of neo-Baroque writing, was composed of long, compound complex sentences that could run on for pages and involve a multitude of people, things, allusions, and impressions. Recognizing this, Greg’s goal was to re-produce this same dazzling Baroque style in his English translation. He would match the original, word for word, clause for clause, and long, intricate line for long, intricate line. Aware as only a scholar and translator can be that what Lima had achieved in the original was a kind of richly creative “non-Spanish” that needed to be replicated in the translation, Greg invented a parallel “non-English,” one that would match Lima’s surging and swelling Spanish. Rising to the challenge, he succeeded. After much writing and rewriting (which simply proves how close translation is to creative writing itself; in the end, they are much the same thing), he felt he had given Lima, and his English-language publisher, a Paradiso that was as close, in style, structure, and tone, to the original Spanish as one could get. His editor rejected it, arguing that the translation was written in a kind of English that no one read or spoke and that no one would appreciate. He was concerned that, as it was, no one would buy the book, even though its Spanish-language author was a celebrated poet and a major figure among the Spanish American “Boom” authors, whose works were selling well in the U.S. of that period. Distressed that his efforts to reproduce Paradiso so faithfully had been for naught, Greg felt that the English version of the novel that eventually did appear had been “dumbed down” for the English-speaking audience.

When we look at the long history of translation, we can see that there are two major categories that involve the question of re-translation: The first relates to what I have outlined above, where essentially good, serviceable translations are replaced by newer versions. This is the case of writers like Homer and Cervantes but also of sacred texts like the Bible and the Qur’an, the translations of which are as much a matter of politics and marketing as they are religion.
The second category, and the one that is easier to deal with, or decide about, is the one where an earlier translation has left out (sometimes intentionally, sometimes not) information present in the original text. This was the case with A Brazilian Tenement (1926, translated by Harry W. Brown), the first English version of the 1890 Brazilian novel O Cortiço, in which the sexual references so crucial to the original’s main thematic line were bowdlerized out, either by the translator, the editor, or the publisher. Only in David H. Rosenthal’s 2000 re-translation (re-titled as The Slum) were they put back in. But it is also the case of the Colombian classic María (1867), in which all references to Jews and Judaism were carefully excised from the original English version of the novel, which appeared in 1890 (translated by Rollo Ogden). My son, Ezra Fitz, is currently doing a new translation of this seminal and deeply influential text that restores it to its original form. In the case of the Robert Scott-Buccleuch translation of Machado de Assis’ Dom Casmurro, whole chapters are eliminated while others are cut up and spliced together. Not rare, these are cases where, effectively speaking, there are clear-cut errors that can, and should, be corrected in re-translation. On the other hand, there is the position held by Borges and a few others that a translator is always free to “fix up” or “improve” an existing text. As Ezra Fitz shows, we can see this attitude in action in Borges’ masterful if inventive Spanish translation of Faulkner’s The Wild Palms.

But these cases are clearly different from the issues involved in our first category, in which the argument for undertaking a new translation of the Iliad, the Odyssey, the Quixote, or the Qur’an is more a matter of reading and writing strategies than is the one for undertaking a new O Cortiço or a new María.

Riven with all manner of complications and tensions, the question of re-translation ranks among the most fraught of our literary questions. It is time-, culture-, and language-bound in ways that do not affect the original text, which, in contrast to existing translations, lives on forever just as it is. The original does not need to change; translations of it do. The reality of re-translation is therefore inescapable, and so it behooves us to understand it better and to treat it with the care it deserves.


Earl Fitz

Image: Courtesy of Dalkey Archive Press