MPT FEATURE

One good translation deserves another

Series 3 No. 5 - Transgressions

A line sailed back into mind recently, after several years’ absence: ‘Now autumn mars the green of hills, my gentle beasts.’

I couldn’t at first recall where it came from. But in a moment or so I found it in my Penguin Modern European Poets Quasimodo, translated by Jack Bevan in 1965, bought not long after. Other haunting lines in the same book were close to the surface of the memory: ‘The wind is still there that I remember / kindling the manes of horses…’ and ‘In the spaces of the hills / all winter, the silence / of the light of sailing ships.’

These translated lines bedded down long ago in the same way as did memorable lines from Wordsworth, George Herbert, Elizabeth Bishop and all the others. That they were translated was a feature that the memory overlooked in the pleasure of their possession as lines of poetry in English, and irrespective of their ‘quality as translation’. Only later did I turn to the original, wondering if the Edward Hicks-ish resonances of the line were authentic. I’m still not sure.

When Lowell’s Imitations came out, I read it as poetry in English - without checking, for instance, whether the lines of Valéry’s original Helen were as haunting as: ‘I come from the lower world/ to hear the serene erosion of the surf; / once more I see the galleys bleed with dawn / and shark with muffled rowlocks into Troy.’ That incurious habit was part of the general excitement of finding so much that was newly, or freshly brought into the English.

For whole readerships too, when particular translations become integral parts of our experience as readers, either the original need not be present to the mind at all, or the need for it may fade. The literary success of a translation may mask the very existence of the original. We may forget that Yeats’ ‘When I am old and grey…’ is – or was once - a translation. The King James Bible, Pound’s Rihaku, Miguel Leon-Portilla’s Aztec songs, Milosz’s Herbert, Milner’s and Theiner’s Holub, Ulli Beier’s Yoruba hunter-poems, and so on have in greater or lesser measure transcended their translatedness. For most readers, they have become originals themselves.

The experience of subsequently encountering alternative and plural translations may then, however, disturb such unquestioning acceptance. Reading the Akhmatova versions in MPT 3.3 – including two of the Dante poem – sent me back to Richard McKane’s durable renderings, in the Bloodaxe Collected Akhmatova and in Weissbort’s Penguin Post-War Russian Poetry. Comparing them, however, was a merely ‘literary’ activity, since I know no Russian.

Which is not to undervalue such essential activity. Having two or three versions to compare is usually very illuminating, can even be potentially subversive, not least when the cultural success of one version carries definitiveness or the apparent finality of implying No translation beyond this point. It is also valuable, on the other hand, when such comparison results in this seeming-finality being endorsed or celebrated afresh – however provisionally. Benson Bobrick’s The Making of the English Bible (Phoenix, 2001) has some fascinating examples of those creative leaps in the process of translation which eventually, precariously, gave us the Authorised Version. Interestingly, the successes of that version were often, Bobrick makes clear, the outcomes of decades of tinkering. Here are two examples:

‘How are the mighty overthrown’, Tyndale 1530.
‘How are the mighty fallen’, King James Authorised Version 1611.

‘Hast thou given the horse strength? Or covered his neck with neighing?’, Geneva Bible 1560.
‘Hast thou given the horse strength? hast thou clothed his neck with thunder?’ AV

Some, though, were last-minute inspirations:

‘Faith is a most sure warrant of things, is a being of things hoped for, a discovery, a demonstration that things are not seen.’
‘Now faith is the substance of things hoped for, the evidence of things not seen.’
A V – two final options.

All three ‘simple’-seeming changes – made by collaborative committee endeavour, fascinatingly enough – are revelatory, the first two in their directness and metaphoric clarity, the third for its newly-found rhythmic assurance. In these ways they – and the illumination they give to the moments of inspiration in writing – can be compared with the extraordinary, clinching revisions Hardy made to the last lines of stanzas of ‘During Wind And Rain’.

Early version Final version

The sickened leaves drop down in throngs How the sick leaves reel down in throngs!

And the wind-whipt creeper lets go the wall. And the rotten rose is ript from the wall.

On their chiselled names the lichen grows. Down their carved names the rain-drop ploughs.

Such examples are testimony to the precariousness of all writing, but more pertinently here, underscore the truth that translating poems is no less or more than writing poetry. The poet’s choices, as between this and that word, rhythmic pulse, turn of phrase, syntactical gesture, sentence length, and so on, are those the translator makes. Translations can be compared not just with drafts, but as drafts. Herein lies, then, the indispensable value of a literary scrutiny of differing versions. Take these two versions of a well-known poem by Praxilla, from fifth-century BC Greece.

Loveliest of what I leave behind is the sunlight,
And loveliest after that the shining stars, and the moon’s face,
But also cucumbers that are ripe, and pears, and apples.
Trans. Richmond Lattimore

Most beautiful of things I leave is sunlight;
Then come glazing stars and the moon’s face;
Then ripe cucumbers and apples and pears. Trans. Willis Barnstone

Lattimore’s musically greater expansiveness is, arguably, truer to the pace of the emotion of relishing things. He risks ‘the’s and ‘and’s to create space between each loved thing; reference becomes momentary contemplation. In the second line he repeats ‘loveliest’ and only arrives at ‘stars’ on the fifth stress of seven. He draws out the third line in an adjectival clause longer than sense requires, marking it with a couple of commas like road-humps. Barnstone’s version by contrast is brisk. He takes musically and syntactically briefer options each time – as in the impatient ‘then’s . His version hurries, I’d say, and sounds cramped.

This, though, is to look at two versions as only, or primarily, poems in English. I don’t know Ancient Greek, and if for some reason which I can’t imagine, Praxilla’s poem actually sounded brisk in Ancient Greek, I suppose Barnstone would have done the better rhythmic job. But insofar as an unawareness of the original language is most readers’ reason for reading a translation, they can only go on their own sense of the poetic rightness of the translation as a would-be-poem in itself. They have nothing else to go by. So I prefer Lattimore.

In the same way, William Soutar’s version of the frist stanza of an old Irish poem seems more elegiacally end-of-summer than Flann O’Brien’s:

Summer is by;
There is nae mair to tell.
Stark on the brae the stags bell:
The drift blaws oot o’ the sky:
Summer is by. From the Irish – William Soutar

Here’s a song
stags give tongue
winter snows
summer goes. Flann O’ Brien

And two versions, both in Penguin, of Osip Mandelstam’s ‘The Stalin Epigram’ are so different they can’t help raise questions of preference. Here are a few lines side by side.

Penguin - trans. Clarence Brown and W S Merwin Penguin - trans. James Greene

Our lives no longer feel ground under them. We exist, without sensing our country beneath us,

At ten paces you can’t hear our words…. Ten steps away our words evaporate….


the ten thick worms his fingers, His fat fingers slimy as worms….

the huge laughing cockroaches on his top lip… His cockroach moustache chuckles…

He pokes out a finger and he alone goes boom. He alone prods and probes…

The concrete immediacy of Brown and Merwin dictates my preference from the first line. Subsequent lines confirm it, as in the punching rhythm of ‘ten thick worms his fingers’ compared with the flaccid ‘fat fingers slimy as worms’; and the finger that doesn’t just prod and probe but ‘goes boom’.

There must surely be some legitimacy in reading like this, with an exclusive regard, as it frequently seems to be, to the literary quality of the translated poem-as-poem in English. How else can one engage with a poem from a language one doesn’t know? And yet, if we are to ‘return to the original’, there needs also to be available the kind of translation that keeps the original language in mind and at hand. Seamus Heaney remarks about Thomas Kinsella, that his translations of The Poems of the Dispossessed ‘are not asking to be taken as alternatives to the originals, but are offered as paths to lead our eyes left across the page, back to the Irish’; hence ‘the refusal of rhyme and the disdain of a charming tune’.

So though the idea of the ‘better’ translation most often points – in most published works – towards the pole of the literary version, it also turns, Janus-faced, to the literal pole, the more ‘faithful’ rendering which guides or goads the reader towards the original. Readers for whom this pragmatic rationale for translation is more pressing, who want to be sent to original texts, will value, as the dominant criterion of translation quality, ‘fidelity to the original’. Not knowing Ancient Greek or Russian, I can’t be sure, according to this criterion, that Barnstone’s version of Praxilla’s poem isn’t better than Lattimore’s, James Greene’s handling of the ‘Stalin Epigram’ more ‘Mandelstam’ than Brown’s and Merwin’s.

There’s clearly a tension between these divergent aims of translation. Is it reconcilable? Towards the end of his Introduction to The Penguin Book of Modern Verse Translation (1966), George Steiner suggests that the ‘only completely honest format’ for his book would have been to have had the original on the facing page, with a prose paraphrase bracketing the principal difficulties, in the margin. He regretfully adduces reasons ( like ‘size, economy’ ) why this ideal didn’t seem possible, but suggests nonetheless that his anthology would ‘defeat itself if it did not…return the reader to the original’.

Many poets, particularly, will want to know what the translated poem sounds like in the original, knowing that its music is, in part, the poem. Over the last year or two, reading modern Greek poems, particularly Seferis in Keeley and Sherrard’s translation and Ritsos in both Keeley’s and Pilitsis’, I more and more found I needed the Greek on the opposite page, so as to hear the poems in Greek, or try to, to hear the poet’s music. That way, with perhaps a dictionary and a grammar for company, one can be alerted to slippages like – here for instance – a smoothing out of syntax and rhythm:

Gerontissa, tifli, kaqotan sto katwfli, mpros sto dromo.
A blind old woman was sitting on her doorstep facing the street.
Yannis Ritsos, trans. George Pilitsis

The aspiration to ‘return’ to the original may be frustrated, by the poem’s origin in a script or syllabary inaccessible to the reader, the sheer effort needed to begin to read any language in the original, and so on. Too often, also, there is an original only for the translator, and even then only if the translator is as much at home in the language being translated as a native speaker. If the translator is working on a collaborator’s ‘literal’ or ‘trot’, as Graves called it, then the (second) translator is as reliant on the primary literal as any other reader. With what consequences the reader cannot know. Janos Csokits says of Hughes’ translations of Pilinsky, that ‘without the softening effect of the original metre and rhyme-scheme they sound harsher’ than in Hungarian, so in Hughes’ versions ‘Pilinsky’s view of the world appears grimmer’. Which is the opposite effect from that noted by Heaney in overly mellifluous translations from Irish by writers like Lord Longford and Frank O’Connor. Such insights raise doubts: can Waley be trusted after all, however much ( literary ) pleasure he gives?

Suppose, then, that we wish to ‘return to the original’, how do we do it? By learning a language and getting on with it? Then how much returning can we do? I’d ask also whether publishing has a role to perform. Following Steiner’s hint, is there not much to be said for having all three manifestations of the poem – original, literal and literary versions – seen together? I value the moments in MPT, as elsewhere, where original and translation appear side by side. I also like the format of the old Penguin Italian, French, Spanish etc. collections, which have a prose paraphrase at the foot of the page, with the originals filling the top three-quarters. Much published poetry in translation now offers less than this, offers frequently only the one literary version. It seems there might even be a shift towards English-only translations and away from dual-language editions. The Keeley-Sherrard dual-language Seferis, for instance, appears to have given way to an English-only edition.

An argument for a plurality not just of translations but of the contexts of their appearance in publications amounts to an aggressive form of defence of the idea of translation. Scepticism as to the possibilities of translation is a constant, and perhaps easily underemphasised. It may well lie behind the incuriousness still, in schools for instance, of our cultural relations with other literatures, including those which inhabit these islands, and some of the Englishes beyond them.

This may be true even though readers whose imaginative world has been largely constructed by translated works – the readers of MPT for example – would endorse without a second’s hesitation the positives of translation, facing down the doubt that cripplingly stresses the imaginative or cultural or linguistic distance between original and translation. Not so much the ‘possibility’ of total translation, then, but the total impossibility of not having translations. How could one’s mind have done without – in my case – the Bible in the Authorised Version of 1611, Lattimore’s Homer, Cohen’s Cervantes, Constance Garnett’s Turgenev, David Magarshark’s Dostoevsky, Waley’s Chinese poems, and so on? Or without The Penguin Modern European Poets series? Or without Jerome Rothenberg’s Technicians of the Sacred, Ruth Finnegan’s Oral Poetry, Judith Gleeson’s Leaf and Bone?

But in the world of the poem that lies beyond the confines of dedicated magazines, there is perhaps a need for renewed commitment to the foreign-language poem-in-translation, and the poem in English that, because it belongs to earlier times or a non-English English, needs a different kind of ‘translation’, an interpretative, creative reading without formal translation. One argument which could be made more often ‘in favour’ of translation (George Steiner makes it, in the Introduction referred to above) is to suggest that in the normal activity of reading our own literature, we continuously ‘translate’. Reading the English of then, from say the fourteenth to the twentieth centuries, into the English of now inside our heads, it’s obvious that there are words, rhythms, stresses, that we have to translate:

I that in heill wes, and gladnes,
Am trublit now wiyth gret seiknes,
And feblit with infermite.
Timor mortis conturbat me.
William Dunbar, late 15th-16th century

I’m not sure if the ‘e’ of ‘feblit’ is long or short – short I guess. Are ‘gladnes’ and ‘seiknes’ two long syllables each or a short and a long? I find it a tremendous verse without being sure of my ‘translation’ as I read. Then I ‘hear’ the stressed ‘me’ at the end and the second reading persuades the ear that ‘gladnes’ and ‘seikness’ are two long syllables, and that the ‘ei’ of ‘heill’ and ‘seik’ is heard as a diphthong. And in this act of reading-translating what seems crucial is that the sound and texture of the original, the English of then, is not overlaid by the English of now. One might think ‘sickness’ but hear, or strive to hear, ‘seikness’.

Chaucer is perhaps easier to read-translate, but hearing his unmistakable five-stress line involves de-remembering, or not letting rise to the surface of the auditory memory, other five stress lines – Wallace Stevens’, Shelley’s, Browning’s, Shakespeare’s and so on. It’s hardly straining the term to suggest that the act of hearing these lines involves a kind of ‘translation’ – and of course not just of nouns like ‘pye’ and ‘jargon’.

He was al coltissh, full of ragerye,
And ful of jargon as a flekked pye.
The slakke skin aboute his nekke shaketh,
While that he sang, so chaunteth he and craketh.
Chaucer, Merchant’s Tale

The same consideration works in geographical or cultural terms. The readers of various Englishes need to subjugate themselves to and imaginatively create, starting from their own English, the tunes, the rhythms and syntactic habits of other Englishes as they read. Because William Carlos Williams’ English isn’t Derek Walcott’s and neither is Les Murray’s.

This kind of interpretative reading, this imaginative quasi-translation, can only take place in the absence of full translation in the formal sense – a different printed version, ‘simplifying’ things. Formal translation offered when none is needed becomes an attempt to supplant – the dialect with the standard for instance, the authentic with the smoothed out. We have become familiar with this lethal form of supplantive translation in the editorial erosions of Emily Dickinson. Fine poet though he was, Richard Wilbur published, as late as 1960, an edition (Signet) of her selected poems which reproduces earlier interferences with her texts, even though the authentic or original versions were by then available. Side by side are the original first line of Poem 258 in Johnson’s 1951 Faber edition and Wilbur’s.

There’s a certain Slant of light, There’s a certain slant of light,
Winter Afternoons –   On winter afternoons,
That oppresses, like the Heft That oppresses, like the weight
Of Cathedral Tunes –   Of cathedral tunes.

The process of smoothing down Emily Dickinson produced remarkable alterations of meaning. In Poem 303, ‘The Soul selects her own Society – / Then – shuts the Door –’ , those first lines are without dashes in Wilbur, then he alters the ‘To’ of line 3 to ‘On’ and introduces an ‘Obtrude’ in line 4 to replace ‘Present’, making an imperative clause out of an adjective phrase.

To her divine Majority – On her divine majority
Present no more – Obtrude no more.

In the second verse the Wilbur edition recasts the participle ‘pausing’ as a glib gerund, and in line 3 makes the emperor ‘unmoved’, not the Soul.

Unmoved – she notes the Chariots – Unmoved, she notes the chariot’s pausing
pausing –
At her low Gate –   At her low gate;
Unmoved – an Emperor be kneeling – Unmoved, an emperor is kneeling
Upon her Mat –      Upon her mat.

Instead of assuming that the reader has the skill to perform any necessary reworking – interpretative translating – of difficult verses in his or her head, the editor nervously replaces the true music of the poem with an alternative simpler version, a musak, one might say, feeling a need to translate formally as if from another language, when the only ‘translation’ that’s needed takes place in the head. Why not then edit-translate Huckleberry Finn, for instance?

Doubt about the young or mature reader’s capacity inferentially to construct from what he or she already knows the whole of the partly audible, partly known poem becomes a denial of both reader’s and writer’s skills. One reason for the relative neglect of dialect poetry (not least in anthologies for schools) may be the spurious idea that it needs replacing – in formal printed translation – so as to be understood by non-dialect speakers, who just can’t hear it right. My own dialect was broad Lancashire, but I ‘hear’ William Barnes with enormous pleasure, hear indeed one of the greatest lyricists:

‘Ithin the woodlands, flowr’y gleaded,
   By the woak tree’s mossy moot,
The sheenen grass-bleades, timber-sheaded,
   Now do quiver under voot;
An’ birds do whissle over head,
An’ water’s bubblen in its bed,
An’ there for me the apple tree
Do lean down low in Linden Lea.

Simplifying and smoothing down (and out), creating replacements, are respectable cultural activities. Chaucer done into rap for schoolchildren, or a BBC Shakespeare to be mainly prose with some patches of verse in it, and other such phenomena, as long as they are ends-in-themselves intended to permanently by-pass the original, deny translation in denying the plurality and provisionality of it, effacing both the original and the invitation to go back to it.

These thoughts about the indispensability to writer and reader of immersion in the processes of producing literal and literary translations, and in so doing going back to originals where possible, were prompted by a recent attempt to ‘sell’ Seferis and Ritsos to members of a writers’ group. I provided the group with a transliteration, a word-for-word ‘literal’ translation, and two drafts of my own literary attempt, and above them the original Greek text of this short poem by Ritsos, one of ‘15 short songs for the bitter motherland’.

LAOS - Populace

Mikros laos kai polema dicws spaqia kai bolia
gia olou tou kosmou to ywmi, to jws kai to tragoudi.

Katw ap ti glwssa tou kratei tous boggous kai ta zhtw
ki an kanei pws ta tragoudei ragizoun ta liqaria.

Mikros laos kai polema dicws spaqia kai bolia
meecros laos kay polema theekos spathia kai volia
small people and it-fights without swords and bullets

gia olou tou kosmou to ywmi, to jws kai to tragoudi..
yia olloo too kosmoo to psomi to fos keh to tragouthee
for all of-the of-world the bread the light and the song

Katw ap ti glwssa tou kratei tous boggous kai ta zhtw
catto ap tee glossa too kratei toos vongous key ta zeeto
underneath the tongue of-it it-holds the howls and the hurrahs

ki an kanei pws ta tragoudei ragizoun ta liqaria.
ke an kanei pos ta tragouthi ragizoon ta litharia
and if it-makes that them it-sings crack the boulders.

draft 1

A small populace / people – it fights/ and they fight without swords or bullets
For the bread of all the world, for its light, its song.

Beneath the tongue it holds the howls and hurrahs
And if it decides / should it be moved to sing them – boulders break apart/ will crack open.

draft 2

A small people, fighting without swords or bullets
for the all the world’s bread, and light, and song.

Under its tongue it keeps screams and rejoicing
and once it decides to sing them, rocks will split open..

The translated outcome was less the point for me than the illuminating group process of moving between the original text with crudish transliteration, and the literal rendering and draft translation. Clearly, if only draft 2 had been available to those new – apart from one - readers of Ritsos, the power of that short poem might not have been as deeply felt and appreciated. Add in the consideration that the sequence as a whole, which is not the most self-evidently lyrical, was set to music by Theodorakis and become part of a popular sung tradition, and one’s sense of the accessibility of the poem-in-translation, and its worth or value, are challenged by a newly discovered strangeness.

Csokits remarks that Ted Hughes had the knack of seeing, as in X-ray, through the literal version he was provided with to the bone-structure of the original Pilinksy. Most of us may have less of that seeing gift, and find another metaphor more appropriate. For instance the Maya myth that describes how the gods jealously punished newly created humankind for the presumption of its all-seeing intelligence:

And the Creator and Maker took flints of obsidian
and chipped at the clear surface of their creatures' gaze,
and blew mist into their eyes and clouded them over
so they saw the world as if in a misted-up mirror.

So at the beginning of time, the first clear wisdom
of men and women was taken from them,
and their understanding was dimmed for ever.

This ‘imitation’ of a translation of a translation might seem a pessimistic metaphor for the encounter with the translated poem; it might be thought quite bracing, though, as well.

Browse our features about translation

Essential reading, MPT, with its sustained intelligence about how poetries work across cultures, has transformed the British landscape since its inception in 1966.Fiona Sampson

Go Digital

Subscribe to the digital edition of MPT for access to all back issues and to the Exactly app.» View free trial issue

Next issue…

Spring 2014

Spring 2015

No 4 / 2014

MPT (Modern Poetry in Translation) appears three times a year – in Spring, Summer and Autumn. The issues are no longer th... » Read more » Submit to MPT

Back to top
Supported by Arts Council England

Copyright © Modern Poetry in Translation and contributors
Website design ashbydesign
Developed by Code Frontiers
Powered by Storemill