Is That a Fish in Your Ear?

Translators working in many languages in widely separated cultural fields—manga, subtitles, political jingles, experimental fiction, poetry, and popular verse—confront and overcome stringent formal constraints. Moreover, the forms themselves are often transported across historical, linguistic, and cultural space. These facts make it seem unwise to claim that anything is impossible. The only impossible things in translation are those that haven’t been done.

A less prejudiced way of understanding the work that translators do is to look more closely at the effects of successful matchings of strict form. Has Gilbert Adair improved Edgar Allan Poe? How come that the very diluted version of the Onegin stanza in Adriana Jacobs’s translation of Maya Arad’s imitation of Vikram Seth’s imitation of Charles Johnson’s verse translation of Pushkin resurrects something of the lightness and joy of Onegin’s youth? How has Anthea Bell made Astérix even funnier in English than in French? And why did anyone ever think that translating verse by verse was a dead end? The truth is quite the opposite. When you have to pay attention to more than one dimension of an utterance—when your mind is engaged in multilevel pattern-matching pursuits—you find resources in your language you never knew were there.

Of course there’s never a match that is 100 percent, because that’s not the way of the world. Just as it would be silly to claim that high-quality tailoring is “mathematically impossible” because we’ve never had a suit that was an absolutely perfect fit, it would be unwise to deny the possibility of translating form just because we’ve not yet done so in a way that is utterly impeccable in every respect.

THIRTEEN What Can’t Be Said Can’t Be Translated: The Axiom of Effability

When the baggage carousel comes to a halt and his suitcase isn’t there, the weary traveler goes to the airline service desk and complains that his suitcase has been lost. The desk clerk quite reasonably asks for evidence—a baggage stub, for instance—and a detailed description of what has gone missing, so that it may more easily be found.

People who claim that poetry is what gets lost in translation could be asked to follow a like routine. Granted, there’s no check-in desk for poetic effects, so the missing ticket stub can be excused. But it’s not unreasonable to request a description of the missing goods. If you can’t provide one, claiming that something called “poetry” has been lost is like telling an airline it has mislaid an item that has no identifiable characteristics at all. It doesn’t cut a lot of ice.

A reader who says that poetry is what has been lost in translation is also claiming to be simultaneously in full possession of the original (which is poetry) and of the translation (which is not). Otherwise there would be no knowing if anything has been lost, let alone knowing that it was poetry.

A good knowledge of the two languages involved isn’t sufficient to justify the claim that what has been lost in translation is poetry. You could make a convincing case only if you knew both languages and their poetic traditions sufficiently well to be able to experience the full scope of poetic effects in both of them. Not many people meet the standard, but there’s nothing unreasonable about the test.

You would have to meet this entrance requirement to declare a loss of poetry in either direction—in a translation from a foreign language into your own (say, on reading George Chapman’s version of Homer) or from your own language into a foreign one (if, for example, you wanted to say that the French or Spanish or Japanese version of John Ashbery’s poem “Rivers and Mountains” just doesn’t move you as the English one does). Only if you have these skills in language and in poetry can you make a credible claim that something has been lost; but even if you do have them, you will find it hard to tell the desk clerk just what it is.

It would not be relevant to your complaint to say that the relationship between sound and meaning is not the same in the translation as in the original. With the sounds changed because the language is different and the meaning preserved broadly if never precisely, the relationship between the two—a relationship all linguists since Ferdinand de Saussure insist is an arbitrary one—must perforce be other.