Poets on Translation: Otherwise the Same

Poets on Translation is a series of short essays in which poets examine the intersections of poetry and translation in relation to questions of language, identity, authorship, and more.

Who’s to blame for what Damion Searls, in The Philosophy of Translation, calls “the annoying claim that translation is impossible”? Despite all evidence to the contrary, the notion that some things (particularly poetry) are untranslatable is so prevalent that it’s often taken as a truism. Back in 2012, in my preface to The FSG Book of Twentieth-Century Italian Poetry, I wrote:

The most notorious and thoughtlessly repeated remark in English about translation is the chestnut attributed to Robert Frost: “Poetry is what is lost in translation.” Though Frost’s authority on the subject is dubious, his remark—like the Italian phrase traduttore traditore (“translator betrayer”)—lends epigrammatic zing to the old notion that the translation of poetry is an impossible task. Arthur Schopenhauer, ever the pessimist, declared that “Poetry cannot be translated.” The great Peruvian poet Cesar Vallejo said simply that “everyone knows that poetry is untranslatable.” And Roman Jakobson, the Russian linguist, argued that “poetry is by definition untranslatable.”

Such claims—many other philosophers, poets, and linguists have made similar ones—are so prevalent in part because they contain a kernel of truth. To the degree that the essence of poetry is embodied in its actual words and their particular sounds, poetry certainly does get lost in translation: all the original words and their sounds disappear. As the noted linguist Steve Martin remarked after a visit to Paris: “Chapeau means hat. Oeuf means egg. It’s like those French have a different word for everything.” Martin’s joke gets at the serious difference between originals and translations that Vallejo, for example, had in mind: poetry is untranslatable, he explained, because “translated into other synonymous but never identical words, it’s no longer the same.”

In his translation diary Catching Fire, Daniel Hahn, describing a book he is preparing to translate, comes to a somewhat different conclusion: “The new novel won’t have any of the same words as the old one, but I’m hoping it will be otherwise the same book.” Hahn’s “otherwise” does a lot of work in that sentence and might remind us of the old “Ship of Theseus” paradox, which philosophers have used for more than 2,000 years to wrestle with questions of identity and difference: as the individual planks of the Argo wear out, they are replaced one by one with new planks, until none of the original ship remains. Can we say the new version is (“otherwise”?) the same ship—it’s still the Argo, still looks the same, and still carries sailors across the sea in the same way—while also admitting that it’s not identical to its original version?

While what’s lost in translation is admittedly enormous, to conclude that poetry is therefore untranslatable is to fundamentally misrepresent both what poetry is and what translation is, and Jakobson’s pronouncement that “poetry by definition is untranslatable” in fact hinges on definitions of poetry and of translation that are both unreasonably narrow. Echoing Jakobson (and Vallejo), Italo Calvino wrote: “We all know that poetry is by definition untranslatable.” But neither Jakobson nor Calvino can be understood without the context of their following sentences, in which they both reveal their statements to be largely rhetorical. Jakobson adds that “Only creative transposition is possible,” and Calvino adds that “Literary translators are those who put their whole selves on the line to translate the untranslatable.” Jakobson, that is, suggests that something called “creative transposition,” as opposed to translation, is possible (as if literary translation has ever really meant anything other than creative transposition), and Calvino, leaning into the paradox, makes clear that the impossible task—translating the untranslatable—is, somehow, possible.

Wikipedia, to its credit, currently specifies in the opening paragraph of its page on “Untranslatability” that “the term arises when describing the difficulty of achieving the so-called perfect translation” (emphasis added). All notions of untranslatability depend in fact on the fallacy that “perfection” (or “sameness,” to use Vallejo’s term) is an ideal that translations should aspire to and be judged by. Only if we conceive of the relation between a poem and its translation in this way, as some kind of mathematical equation meant to result in perfect sameness, can poetry (and pretty much everything else) be said to be untranslatable.

(The popular fondness for “untranslatable” words—of which the internet contains many entertaining lists, lists that invariably include translations of the so-called untranslatable words—reflects this tendency to think of translation in terms of equivalence. To say that hygge or abbiocco or schadenfreude or mamihlapinatapai are untranslatable typically means only that no single English word offers a perfect equivalent. It means only, that is, that we must sometimes use more than one word to translate another word—a rather banal observation.)

Much gets lost in our eagerness to acknowledge that much gets lost. Though the original sounds of a poem do indeed vanish in translation, poetry doesn’t live solely in its sounds. There is also something like “content” or “meaning,” which, though often overrated and overemphasized in classrooms, is not nothing. There are also other, often more important qualities: tone or voice, imagery or metaphor or even story; the logic or illogic of a poem that can disrupt or reshape our existing ideas about ourselves or our world; and so on. Such elements can often be conveyed, substantially if still imperfectly, in translation. And here it’s worth remembering an obvious but often overlooked truth: translations are entirely additive. They are not, that is, erasures of but rather addendums to their originals—which continue to exist, having lost nothing, and having perhaps even gained new readers.

Another thing that gets lost if we allow Frost’s dictum to be a mic drop is the fact that, in any good translation, poetry is also what’s found. This opposing truth suggests a corrective corollary—and unsurprisingly poets as different as Octavio Paz, Seamus Heaney, and Charles Bernstein have all been quoted as saying that “poetry is what is found in translation.” Indeed, finding new poetry in a new language to stand in for what’s lost may be considered one of the chief tasks of the translator—and good literary translations generally aim to do exactly that. Only if we understand that both Frost’s dictum and its opposing corollary are simultaneously true do we approach an understanding of what translation is.

One can go further. Bernstein, for example, adds that “there is nothing ‘outside’ translation.” Every time we read a poem we translate it—our reading is our translation—which means, to quote Bernstein again, that “there can be no experiencing the poem, even in your own language, without translating” (emphasis added). Analogously, whenever you listen to, say, Bach, you are listening to a translation, even if you are performing it yourself. Without a reader or a performer, the words of a poem or the notes of a score are just black stones of ink on the white stone of the page. When Yo-Yo Ma plays a Bach cello suite, that performance is not Bach’s original but Ma’s path through it. (Searls calls translation the translator’s “path through” a text.) And when I read Emily Wilson’s translation of the Odyssey, or Clare Cavanagh’s translations of Szymborska, or any translation of any poem, I’m reading a performance of a text, following their path, just as surely as I’m performing my own translation when I read an Emily Dickinson poem—and my path will never perfectly match yours, or Dickinson’s own. If we understand translation as the performance of a reading, would we ever think to speak of its impossibility?

Translation-as-the-Argo and translation-as-musical-performance are two of the countless useful if imperfect metaphors that many translators like to collect—in part, I think, as correctives to the implicitly mathematical metaphor that declares its impossibility. Another is translation-as-metaphor—a meta-metaphor, if you will. It’s helpful to recall that metaphor (from the Greek) and translation (from the Latin) mean roughly the same thing, etymologically: both are a carrying across. Metaphors carry some essence across the gap between the literal and the figurative; translations carry some essence across the gap between two languages. If we understand a translation as a metaphor for its original, would we ever think to speak of its impossibility?

Finally, another useful if imperfect metaphor comes from the Indian polymath A. K. Ramanujan, who ends his brilliant essay on translating Tamil poetry with this parable about translation:

A Chinese emperor ordered a tunnel to be bored through a great mountain. The engineers decided that the best and quickest way to do it would be to begin work on both sides of the mountain, after precise measurements. If the measurements were precise enough, the two tunnels would meet in the middle, making a single one. “But what happens if they don’t meet?” asked the emperor. The counselors, in their wisdom, answered, “If they don’t meet, we will have two tunnels instead of one.”

In practice, of course, the paths never coincide perfectly, and unlike the Argo example, in which only one ship remains, translations always leave us with two tunnels. But if the measurements are good enough, both can carry you across—or make a path through—the same mountain.