And yet translations continue to be measured according to a Utopian dream of exact equivalences, and are often dismissed on the basis of a single word, usually by members of foreign language departments, known in the trade as the "translation police." They are the ones who write– to take an actual example– that a certain immensely prolific translator from the German "simply does not know German" because somewhere in the vastness of Buddenbrooks , he had translated a "chesterfield" as a "greatcoat." Such examples, as any translator can tell you, are more the rule than the exception. One can only imagine if writers were reviewed in the same way: "the use of the word ‘incarnadine' on page 349 proves the utter mediocrity of this book."
This is the old bugbear of "fidelity," which turns reviewers into television evangelists. Now obviously a translation that is replete with semantical errors is probably a bad translation, but fidelity may be the most overrated of a translation's qualities. I once witnessed an interesting experiment: average 9-year-old students at a public school in Rochester, New York, were given a text by Rimbaud and a bilingual dictionary, and asked to translate the poem. Neither they nor their teacher knew a word of French. What they produced were not masterpieces, but they were generally as accurate as, and occasionally wittier than, any of the existing scholarly versions. In short, up to a point, anyone can translate anything faithfully.
But the point at which they cannot translate is the point where real translations begin to be made. The purpose of, say, a poetry translation is not, as it is usually said, to give the foreign poet a voice in the translation-language. It is to allow the poem to be heard in the translation-language, ideally in many of the same ways it is heard in the original language. This means that a translation is a whole work; it is not a series of matching en face lines and shouldn't be read as such. It means that the primary task of a translator is not merely to get the dictionary meanings right– which is the easiest part– but rather to invent a new music for the text in the translation-language, one that is mandated by the original. A music that is not a technical replication of the original. (There is nothing worse than translations, for example, that attempt to recreate a foreign meter or rhyme scheme. They're sort of like the way hamburgers look and taste in Bolivia.) A music that is perfectly viable in English, but which– because it is a translation, because it will be read as a translation– is able to evoke another music, and perhaps reproduce some of its effects.
But to do so requires a thorough knowledge of the literature into which one is translating. Before modernism, poems, no matter from where, were translated into the prevailing styles and forms: the assumed perfection of the heroic couplet could equally serve Homer, Kalidasa, or the Chinese folk songs of the Book of Odes. The great lesson of modernism– first taught by Ezra Pound, but learned, even now, only by a few– was that the unique form and style of the original must in some manner determine the form and the style of the translation; the poem was not merely to be poured into the familiar molds. Thus, in Pound's famous example, a fragment of Sappho was turned into an English fragment, ellipses and all, and not "restored" or transformed into rhyming pentameters.
This was based on a twofold, and somewhat contradictory, belief: First, that the dead author and his literature were exotic, and therefore the translation should preserve this exoticism and not domesticate it. Second, that the dead author was our contemporary, and his poems– if they were worth reading– were as alive and fresh as anything written yesterday. An unrestored Sappho was "one of us" precisely because she was not one of us: a foreign (in the largest sense) poet pointing to a way that our poems could be written today.
Modernism– at least in English– created extraordinary works in translation because they were written for modernism: written to be read in the context of modernist poetry. The clich é that the only good poetry translators are themselves poets is not necessarily true: the only good translators are avid readers of contemporary poetry in the translation-language. All the worst translations are done by experts in the foreign- language who know little or nothing about the poetry alongside which their translations will be read. Foreign-language academics are largely concerned with semantical accuracy, rendering supposedly exact meanings into a frequently colorless or awkward version of the translation-language. They often write as though the entire twentieth century had not occurred. (This is especially true in the Asian and Middle Eastern languages.) They champion the best-loved poet of Ruthenia, but never realize that he sounds in English like bad Tennyson. Poets (or poetry readers) may be sometimes sloppy in their dictionary-use, but they are preoccupied with what is different in the foreign author, that which is not already available among writers in the translation-language, how that difference may be demonstrated, and how the borders of the possible may be expanded. Bad translations provide examples for historical surveys; good translations are always a form of advocacy criticism: here is a writer one ought to be reading and here is the proof.
[ page 4 of 7 ]