I don’t read much poetry in translation; in fact, I tend to actively avoid it. As Robert Frost famously said, “Poetry is what is lost in translation.”?

What distinguishes a poem from a story or an essay or a letter or an op-ed column isn’t its subject matter (any number of these things can deal with the same subject) or its ideas (lots of poems, good ones even, have very little in the way of ideas), its imagery (after all, there aren’t any actual images in poems, or in stories or novels: just words, that don’t depict or resemble things) or even its emotional force or poignancy (I have much more often been brought to tears by pop songs that resonated with me than by the most moving poems). It’s the particularity of its language, the way the words interact with one another to produce rhythms and sonic patterns, what Ezra Pound called logopoeia. This is exactly what goes missing when a poem is translated. So I have always agreed with the Italian epigram “Traduttore traditore": “The translator is a traitor,” either to the original text (which he or she replaces with an unconvincing simulacrum) or to his or her own language (which gets mangled or homogenized into a synthetic and flavorless “translatorese”) or, too often, to both.
But, because I also believe that a foolish consistency is the hobgoblin of little minds (and because I wouldn’t want to be too predictable), there are a few translations that have seemed to me not just cribs or reading guides, but poems in English. They aren't always the most literally accurate translations (Clarence Brown and W.S. Merwin’s Mandelstam translations take liberties, and James Greene’s take even more), but they convey, to my monolingual ear and eye, not just an idea of the original poems (what Celan translators Nikolai Popov and Heather McHugh call “a working version of mere meaning”), but also a sense of why we should care, why we should read them at all, in any language. Too many translations just convey the sense of the poem (the least important aspect, certainly the least distinctive or unique), without conveying what makes it poetry. I’m willing to accept a degree of straying from the letter if it results in greater faithfulness to the spirit. As Greene notes, what’s important is to carry across the effect of the poem. In this sense, Joseph Brodsky was right when he said, no doubt responding to Frost, that “Poetry is what is gained in translation”: if, that is, one is exceptionally lucky.
As Clarence Brown writes in his introduction to his and Merwin’s edition of Mandelstam’s selected poems, “We have tried to translate Mandelstam into the English that works as an instrument of poetry in our own time, and we have accepted the responsibility entailed in the fact that to translate is to change.” This is what I ask of a translator, as opposed to those Brown calls “Those of my colleagues in the academy who are sent up the wall by ‘mistakes’ in the translation of poetry, those who are happy to maintain that poetry is untranslatable here on earth, and the arbiters of their own brand of literalism everywhere.”

Originally Published: January 8th, 2008

Poet and editor Reginald Shepherd was born in New York City in 1963 and grew up in the Bronx. He earned a BA from Bennington College and studied at Brown University and the Iowa Writers’ Workshop. His first collection, Some Are Drowning (1994), won the Associated Writing Program’s Award in...

  1. January 8, 2008
     Simon DeDeo

    One of the most influential translation projects I know of -- that has brought new things to the sound as well as the sense of English-language poets -- has been that of Inger Christensen, who I think I'm right in saying could be brought into English without having "local" notions determine the structure and sound. Partly it's because a lot of what she does musically is on the level of the word, as opposed to the syllable.
    I should also mention Eugene Ostashevsky's translations from the Russian; actually, there's a tremendous bilingual community in St. Petersburg doing a lot of great translation work, including Eireene Nealand. It seems, strangely enough, that a lot of what is being done in Russian right now is connected -- accidentally, not through direct correspondence -- in really interesting ways to what I think is the most interesting in the United States.

  2. January 9, 2008
     Reginald Shepherd

    Dear Simon,
    Thanks for your comment. My next post will briefly discuss a few poetry translations that I have found powerful and effective, though it makes no attempt to be a comprehensive or representative list.

  3. January 9, 2008
     Kwame Dawes

    I remain puzzled by monolingual, non-translating commentators who seem to speak with authority about the failure of translation and the "liberties" taken by some translators. What? How do you know? So I will no longer take seriously commentary about the quality of translation from people who only have one language. Talk about the quality of the poem, its music, its language--treat it as a poem not a translation. but if you are going to tackle translation for its efficacy and "accuracy" then please, get the other language or get help from someone who has both...
    One love

  4. January 9, 2008
     Vivek Narayanan

    Hi Reginald,
    Frankly, we have no choice but to read poetry in translation. It's as simple as that.
    And I agree with Kwame. And Simon too. The "faithful" translation, that is, the translation with fidelity and reflections, is often the closest reading. We miss the importance of that, sometimes, in the rush to sign one's own name below, to appropriate everything into familiar idioms and rhythms. I'm not saying I don't like to hear that kind of stuff from the mouth of a good poet. I'm saying that too often that's all that turning the translation into a "poem" ends up meaning. Check out a book of dreadful "free" translations of Ghalib by top American poets, edited by Aijaz Ahmed, to see what I mean.

  5. January 9, 2008
     Simon DeDeo

    Let me partially relieve your puzzlement, at least as someone who reads plenty of translations from languages he doesn't know and presumes to judge it as something other than an English-language text.
    For any translation beyond the literal or interlinear, I think one call that can always be made is about the "strangeness" of the translated text. If Sappho comes out musically like Emily Dickinson, there's something going on. It's a bare problem of causality and the infinite number of possible sounds.
    That said, one of the most famous "strange" translations from C20 are those from the Chinese, done by the Modernists under the influence of Fenollosa. Which are bizarre fabrications of a complete misunderstanding of how ideograms function. So I suppose strangeness is a necessary, but not sufficient, criterion for a "true" translation.

  6. January 9, 2008
     Don Share

    Simon may be referring to the work you'll find collected in the handy (for this purpose) The New Directions Anthology of Classical Chinese Poetry, edited by Eliot Weinberger, which features work from Ezra Pound's Cathay and versions of Chinese poems from such New Directions / Modernist luminaries as William Carlos Williams, Kenneth Rexroth, and Gary Snyder, rounded out with work by David Hinton.

  7. January 10, 2008
     Alicia (AE)

    Thanks for bringing translation to the fore, Reginald. This is a topic that obsesses me--partly because my life is lived here in a sort of translation (there is a constant negotiation between two languages), and partly because that is much of what I do professionally and much of what I deal with.
    I've increasingly come to the conclusion that translation must be bold to succeed; but that boldness is paradoxically a result of intimacy with the original. You cannot be "bold" or "free" with a text when you do not really know what you are being bold with or free about. Sometimes boldness might mean hewing as literally to the text as possible. Sometimes it might mean making an intuitive leap. But you can't leap if you don't know where you are standing.
    And I certainly agree that translation is a kind of close reading. The closest reading perhaps.

  8. January 10, 2008
     Reginald Shepherd

    With regard to Kwame Dawes' comment, I would appreciate it if people would respond to (or at least take note of) what I actually wrote rather than what they project into that writing. Certainly if they're going to write about, and criticize, what I write, it would be helpful if they had an accurate idea of what that actually is.
    Dawes writes that I should "Talk about the quality of the poem, its music, its language--treat it as a poem not a translation." That is exactly what I did in this post--I wrote about the way in which most translations I have read fail as English language poems. I have read many poetic translations (that's why I tend not to read them any more), and very few have conveyed any sense of why one would actually read the translated poet, or given any sense of the qualities of the work that, say, the introduction lays out. One reads about a poet writing in some other language and then reads a translation and there's a disconnect; one gets no sense of the work as poetry.
    Dawes claims that I judge translation for its efficacy (I'm not sure what he means by that word in this context) and its "accuracy" (the quotation marks are his), but I don't. I think that there is an inevitable element of inaccuracy in translation (as I quote Clarence Brown, to translate is to change), and I write that as a reader I am willing to trade literal accuracy, faithfulness to the particulars of the original text, for a text that works as English poetry. My last paragraph, quoting translator and scholar Clarence Brown, was quite explicitly a defense of "taking liberties" in the service of producing a poem that works in English.
    To adapt Dawes' wording, I will no longer take seriously the commentary of people who willl not read with care or simple accuracy in their own language, let alone any other. So we can agree not to take one another seriously.

  9. January 10, 2008
     Reginald Shepherd

    Thanks for your comment. I agree that translation is both dependent upon and is a variety of close reading. To quote John Felstiner, who's done some amazing and richly informed translations of Paul Celan, "in translating, as in parody, critical and creative activity converge. The fullest reading of a poem gets realized moment by moment in the writing of a poem. So translation presents not merely a paradigm but the utmost case of engaged literary interpretation."
    Take care.

  10. January 19, 2008
     Emily Warn

    I'm curious. When you write that the translations of the Modernists under the influence of Fenollosa became "bizarre fabrications of a complete misunderstanding of how ideograms function." what do you mean? In what way did Fenollosa and they misunderstand ideograms? How should ideograms function, especially in a poem? Thanks, Emily

  11. January 25, 2008
     Reginald Shepherd

    With regard to Emily's question, I don't presume to know what Simon meant by his comment, but from my studies of language in general and of writing systems in particular, I know that it's incorrect to speak of Chinese writing as ideograms. While the Chinese writing system, like every other early writing system, began on a pictographic basis, Chinese characters are neither pictograms or ideograms; they are logograms. That is to say, they don't represent either images (pictograms) or ideas (ideograms), but words.
    As linguist Steven Roger Fischer writes in A History of Writing (Reaktion Books, 2001), "The logographic, or 'word-writing,' nature of Chinese writing dominates the system, reproducing units of spoken Chinese. British philosopher Bertrand Russell [along with Ernest Fenollosa and Ezra Pound] once thought Chinese characters were 'ideogaphic,' believing each 'represents an idea.' This is incorrect. Chinese characters, as whole units or blocks of components, are words--single monosyllabic morphemes--in the Chinese language, and nothing else" (172).