Poetry and the perils of translation

Tim Parks in his essay Translating in the Dark rightly sounds a cautionary note to poets who would be translators (or perhaps even translators who would be poets) at the same time. I can find Parks's observation relevant even to my own casual attempts at poetry translations. Since he himself is a translation instructor and his concern at the moment is with Swedish poet Thomas Tranströmer, & the most recent translation of his writings by Robert Robertson, the whole essay has the look of infinite regression: a translator commenting on a translator commenting on a translator, etc. until we're left with the impression that 'good translation' is a receding ideal. The discussion is decidedly winding & in the end it's not clear that any reliable guidelines to poetry translation have been established.

“We must believe in poetry translation, if we want to believe in World Literature.” Thus Thomas Tranströmer, the Swedish poet and winner of this year’s Nobel Prize in Literature, quoted in a recent essay by Robert Robertson, one of his translators. Robertson goes on to describe the difficulties of capturing Tranströmer’s spare voice and masterful evocation of Swedish landscape in English, particularly if you don’t know Swedish well. Robert Lowell, Robertson tells us, translated Tranströmer with only a “passing knowledge” of the language. Robertson himself describes a process wherein his Swedish girlfriend gives him a literal line-by-line translation into English, then reads the Swedish to him to give him “the cadences,” after which he created “relatively free” versions in English. 

This approach to translation is not uncommon among poets (W.H. Auden gave us his versions of Icelandic sagas in much the same way). Nevertheless, Robertson feels the need to call on various authorities to sanction a translation process that assumes that poetry is made up of a literal semantic sense, which can easily be transmitted separately from the verse, and a tone, or music, which only a poet is sufficiently sensitive to reconstruct.
Thus:

In his introduction to Imitations (1962), Robert Lowell writes that “Boris Pasternak has said that the usual reliable translator gets the literal meaning but misses the tone, and that in poetry tone is of course everything.”

Here the “of course” skates over the fact that tone is always in relation to content: if the content were altered while diction and register remained the same, the tone would inevitably shift. One notes in passing the disparagement of the “usual reliable translator”—the fellow knows his foreign language, but doesn’t understand poetry.

Good translation seems, according to this, to have been out of the reach of even the most accomplished versifiers. Robert Lowell, for example, attempted a translation of Thomas Tranströmer with only a "passing knowledge" of the original & his results were about as good or bad as Auden's translation of the Icelandic sagas. If we take literalness and the cadences of the original language as yardsticks, translations seem to rise or fall by the degree to which translator is a poet, or not, or is intimately acquainted, or not, with a foreign language. Presumably the ideal translator is both.

A translator who both knows her foreign language and understands poetry, acutely sensitive to both literal meaning and tone, is Rosmarie Waldrop. In "Irreducible Strangeness" (from Dissonance) the translator of Edmond Jabès makes (to me the much more interesting) case for the irreducibility of translation to its original source in terms of a hermeneutical 'space' of culture, language & tradition through which the translator must pass at their peril. Presumably, like Waldrop herself, the translator must possess a thorough (almost scholarly) familiarity with the French language & its contemporary literary forms but in the end there's only irreducibility to reckon with. American poet Frank Samperi, on the other hand, with no academic training both read and produced a very competent translation of Dante's Paradiso. Samperi's was the more spiritual 'space' of Catholicism and a Thomistic viewpoint he always brought to bear on his own poetry.

T.S. Eliot is then cited as having warned Lowell not to present his ‘imitations’ of Tranströmer and others as “translations”: 

If you use the word translation in the subtitle it will attract all those meticulous little critics who delight in finding what seem to them mis-translations. You will remember all the fuss about Ezra Pound’s Propertius. 

Here collocating “meticulous” with “little” does the job that Lowell/ Pasternak achieved with “usual reliable”: there are people who always interfere, but don’t understand.
Robertson also calls on the British poet Jamie McKendrick who, he feels, is “surely right” when he says “The translator’s knowledge of language is more important than their knowledge of languages.” How vague this remark is! Does it mean that the translator has one kind of knowledge of how language in general achieves its effects, and another of the nuts and bolts of the different languages he knows, the first kind being “more important” than the second? If that is the case, then to what degree more important? Wouldn’t the two, rather, be interdependent and mutually sustaining?

These perplexities apart, the thrust of McKendrick’s argument is clear enough: we are sweeping aside the objection that a profound knowledge of a foreign language might be required to translate its poetry, or prose for that matter, thus clearing the path for a translation by someone who is an expert in the area that counts: our own language.

Can anyone, after the literalist manner of Eliot, assume that critics of translated works tend not to get it? Is translation work always doomed to fail in this sense? McKendrick's more radical claim of the primacy of knowledge of 'language' over knowledge of 'languages' seems to be a way to avoid Eliot's own denigration of readers who aren't as language-proficient as they ought. I'm rather inclined to agree with McKendrick. The "earish" poetry translations of Robert Kelly, for example, in which sounds count as the only viable translation items, certainly sweep aside this hard competency requirement as do any of its other forms, reserving for the poet-translator more wildly experimental varieties. To call it translation is no more problematical than calling Mac Low's "chance-determined" methods of composition a right way to write a poem.

I really do not wish to nitpick. I enjoy Lowell’s and Robertson’s translations of Tranströmer, and Pound’s Propertius. I am glad these people did the work they did, giving us many fine poems along the way. As a writer myself who has also done a number of translations I might be expected to have a vested interest in the idea that what skill I have in English sets me apart from the “usual reliable” translator. However, and quite regardless of whether we want to call such work translation or imitation, it does seem that a serious issue is being dispatched with indecent haste here. 

Let us remember our most intense experiences of poetry in our mother tongue, reading Eliot and Pound as adolescents perhaps, Frost and Wallace Stevens, Auden and Geoffrey Hill, then coming back to them after many years, discovering how much more was there than we had imagined, picking up echoes of other literature we have read since, seeing how the poet shifted the sense of this or that word slightly, and how this alters the tone and feeling of the whole. And then let’s also recall some of the finest poetry criticism we have read—by William Empson, Christopher Ricks, or Eliot himself—the ability of these men to fill in linguistic and literary contexts in such a way that the text takes on a deeper meaning, or to tease out relations inside a poem that had been obscure, but once mentioned are suddenly obvious and enrich our experience of the work. 

Now imagine that, having a poet friend who wishes to translate these authors, you offer a literal translation of their poems in your second language. Maybe you read The Four Quartets out loud, line by line, to give him the cadence. But does our translator friend, who doesn’t know our language well, hear what we hear when we read aloud? The onomatopoeia, perhaps. But a dying fall in one tongue may not be the same in another, not to mention the echoes of other texts, or simply of voices in the air in our language. Over my thirty years in Italy I have often been told by uninitiated English friends what a beautiful and harmonious language Italian is; but that is Italian as heard by an ear accustomed to English sound patterns. To the Italian ear, and to mine these days, much of what is said in Italian grates. One hears the language differently when one knows it.

And yet for all of a translated work's 'irreducibility' and fitness for experimental variations, Parks repeats Eliot's foursquare resolve to give the utmost importance to language competency, opting for the more enlightened 'literalist' position. Only "when one knows it" can the language of poetry reveal, usually over the lifetime of its reader, its own delicious nuances & tones and rhythms. Eliot might be saying that being a poet and possessing language mastery are each individually necessary for successful translation. And presumably sufficient as well. As if he were addressing Kelly directly, Parks makes even the way sound patterns are registered a function of language mastery: the better it's known the more easily even contemporary Italian can grate on the ears.

Why do those “usual reliable translators” often give us work that we feel is wooden or lackluster, thus inviting the poets to get involved? Teaching translation, I frequently deal with students who write well in their mother tongue, but whose translations into that tongue lack fluency. This brings us to a paradox at the heart of translation: the text we take as inspiration is also the greatest obstacle to expression. Our own language prompts us in one direction, but the text we are trying to respect says something else, or says the same thing in a way that feels very different. We have come to what Paul Celan meant when, despairing of translating Baudelaire, he remarked that “poetry is the fatal uniqueness of language.” All the same, what often frees the student to offer better translations is a deeper knowledge of the language he is working from: a better grasp of the original allows the translator to detach from formal structures and find a new expression for the tone he is learning to feel: in this case, however, every departure from strict transposition is inspired by an intimate and direct experience of the original.

All this to arrive at the obvious conclusion that while expression and creativity in one’s own language is crucial, a long experience in the language we are working from can only improve the translations we make. But having hit that rather easy nail on the head, we can now ask the really interesting question: why are such intelligent writers as Eliot, Lowell, Pasternak, Robertson, and McKendrick unwilling to consider the question more carefully. Is it because, to return to Tranströmer, “We must believe in poetry translation, if we want to believe in World Literature.” There is no point, that is, in examining what we do too closely if we’ve already decided what we want our conclusion to be.

Again, even in the face of the 'mystery' of the real text (post-text?), its space and irreducibility, he acknowledges that "what often frees the student to offer better translations is a deeper knowledge of the language he is working from". Parks seems to have caught himself in a circle: the full rich text offered in translation requires a direct knowledge of the original language but, however competent the translator (and Parks has placed translators on a sort of competency continuum, ranging from "passing knowledge" to "intimate and direct experience of the original": from Merwin's  to John Sinclair's Dante ), it's always likely the finished work's going to be a foregone conclusion. Pound's translation of the "Seafarer" is unmistakably Poundian, for instance. What else could it be?

So why is it imperative that we believe in World Literature? It seems we must imagine that no literary expression or experience is ultimately unavailable to us; the single individual is not so conditioned by his own language, culture and literature as not to be able to experience all other literatures; and the individual author likewise can be appreciated all over the globe. It is on this premise that all international literary prizes, of which there are now so many, depend. The zeitgeist demands that we gloss over everything that makes a local or national culture rich and deep, in order to believe in global transmission. There must be no limitation.

I have no quarrel with the aspiration, or all the intriguing translation/imitation processes it encourages. My sole objection would be that it is unwise to lose sight of the reality that cultures are immensely complex and different and that this belief in World Literature could actually create a situation where we become more parochial and bound in our own culture, bringing other work into it in a process of mere assimilation and deluding ourselves that, because it sounds attractive in our own language, we are close to the foreign experience. Tranströmer remarks:

I perceived, during the first enthusiastic poetry years, all poetry as Swedish. Eliot, Trakl, Éluard—they were all Swedish writers, as they appeared in priceless, imperfect, translations…

Try this experiment: pick up a copy of a book mis-titled Dante’s Inferno. It offers 20 celebrated poets, few of whom had more than a passing knowledge of Italian, each translating a canto of The Inferno. The result is inevitably extremely uneven as in each case we feel the Italian poet’s voice being dragged this way and that according to each translator’s assumptions of what he might or might not have sounded like. Sometimes it is Heaney’s Inferno, sometimes it is Carolyn Forche’s, sometimes it is W.S. Merwin’s but it is never Dante’s.

Then dip into the 1939 prose translation by the scholar John Sinclair. There is immediately a homogeneity and fluency here, a lack of showiness and a semantic cohesion over scores of pages that give quite a different experience. To wind up, look at Robert and Jean Hollander’s 2002 reworking of Sinclair. Robert Hollander is a Dante scholar and has cleared up Sinclair’s few errors. His wife Jean is a poet who, while respecting to a very large degree Sinclair’s phrasing, has made some adjustments, under her husband’s meticulous eye, allowing the translation to fit into unrhymed verse. It is still a long way from reading Dante in the original, but now we do feel that we have a very serious approximation and a fine read.

Why the privileging of Sinclair's more scholar-friendly translation over that of Heaney? To whom does Dante really belong? I suggest the act of translation, especially in regards to the strongest poets, must amount to a sort of literary piety: to be felt in the difference, say, between the way a tourist proceeds through the Propylaea to the sacred precincts of Athena Polias and the way Pericles or Phidias did. There is a cultural divide (Waldrop's 'space') across which the historian, poet, traveller must venture in order to be 'there'. I've walked up those crumbling steps myself & felt the ancient language with a real sense of awe and reverence. Were I to translate that sense into poetry or translation of an ancient Greek poet, who's to say the experience wouldn't be as legitimate as anyone else's, provided only the heart of the devotee be pure & true? Do verses under the amateur-translator's hand necessarily turn into crumbling steps?

I've tried my hand at ancient Latin verses and so has Zukofsky: though based on only a "passing familiarity" (and a great deal of dictionary work!), Zukofsky's homophonic Catullus is an admirable artistic achievement. And anyone who's read even parts of it sees it's very loosely based on the original. Kelly's 'earish' model is a direct descendant of the Catullus paradigm. Bly's translation of Antonio Machado is another example of superb translation work by a great contemporary poet (though his is a more literal reproduction of the original), and I can cite many more. In reading these translations & my own I always detect, much more than linguistic mastery, a reverence for 'language' in McKendrick's sense & the forms of adoration it takes.