Poetry Translation

[Mirabell]

From an exchange which starts with Irma Brandeis defending Mandelbaum's Dante translation (which is, incidentally, the only Dante translation I have ever personally read). I have a few bones to pick with Carne-Ross's reply but the last two paragraphs contain much truth, especially about poetry and translation. Here it is (link below)

In D.S. Carne-Ross's rather venomous review [NYR, December 20, 1984] of Allen Mandelbaum's translation of Dante's Paradiso (venomous in the approach gentle followed by the pounce deadly; venomous in pointing out many faults and skipping all virtues in the huge work), there are a number of points one would like to challenge. Given the limitations of space and study-time, I shall confine myself to one: the curious view that John Ciardi is the superior translator. It all depends, I suppose, on what lapses one can overlook and which ones make one's blood boil.
Mr. Carne-Ross begins his comparison of the two by citing Paradiso II, 31–42 in both versions. Let me repeat them together with Dante's text for checking:

Pareva a me che nube ne coprisse Lucida, spessa, solida e polita, Quasi adamante che lo sol ferisse.
Per entro sè l'eterna margarita Ne recepette, com'acqua recepe Raggio di luce, permanendo unita.
S'io ero corpo, e qui non si concepe Com'una dimension altra patio, Ch'esser convien se corpo in corpo repe,
Accender ne dovrìa più il disio Di veder quella Essenza in che si vede Come nostra natura e Dio s'unìo.

It seemed to me that we were covered by
a brilliant, solid, dense, and stainless cloud,
much like a diamond that the sun has struck. Into itself, the everlasting pearl
received us, just as water will accept
a ray of light and yet remain intact. If I was body (and on earth we can
not see how things material can share
one space—the case, when body enters body), then should the longing be still more inflamed
to see that Essence in which we discern
how God and human nature were made one.
(Mandelbaum)
It seemed to me a cloud as luminous and dense and smoothly polished as a diamond struck by a ray of sun enveloped us.
We were received into the elements of the eternal pearl as water takes light to itself, with no change in its substance.
If I was body (nor need we in this case conceive how one dimension can bear another, which must be if two bodies fill one space)
the more should my desire burn like the sun
to see that essence in which one may see how human nature and God blend into one.
(Ciardi)

Of Mandelbaum's lines Mr. Carne-Ross says only "a shade humdrum…here and there rather awkward." Of Ciardi's version he has nothing to say. But I do; and some comparisons. I note that in line 36 the intent of Dante's permanendo unita (remaining unified) is missed entirely. The phrase has nothing to say about "substance." The image of water unshattered by light is there to suggest the inviolability of the heaven, which foreshadows the inviolability of God as we are to see it in Canto XXXIII. If liberties are to be taken in such lines as this, they must convey their meaning. Mandelbaum conveys it. In line 37 Ciardi makes a garble of qui non si concepe (here one does not conceive). The sense is simple and essential and again Mandelbaum knows and conveys it. In line 40, where the text says disio (desire), Ciardi says "my desire." His "like the sun" is not present in Dante: it fills out a short line. His final line appears to miss Dante's point altogether, by ignoring the past tense and the simplicity of the verb s'unto. Here again Mandelbaum gives us Dante's meaning with appropriate diction.
Further on in his review Carne-Ross, taking Mandelbaum to task for what indeed is an unfortunate epithet in "Pay honor to the estimable poet" (Inferno, IV, 80) wisely refrains from referring us to the same passage in Ciardi, where we would have read

Honor the Prince of Poets, the soul and glory
that went from us returns. He is here! He is here!

where both the glory and the two exclamations are the translator's inventions. Indeed, Ciardi never hesitates to revise and embellish Dante to fill out a pentameter. The reader may want to see what he has done to three famous passages: Inferno V, 127ff (Francesca's narrative), Purgatorio XXIV, 52–54 (Dante's self-definition), and Purgatorio XXX, 55–57 (Beatrice's first words).
Mr. Carne-Ross derides Mandelbaum's metrics, but seems to swallow Ciardi's. It is true, of course, that pentameter lords it in both translations (not difficult to see why), and that both struggle to escape, each in its own way, the terrible temptation of iambics. Mr. Carne-Ross knows as well as anyone the problems a translator must face in trying to render many-syllabled Italian in clipped English, yet he treats every slightest padding or repetition in Mandelbaum as though it were a lapse of taste, while on the other hand he tolerates Ciardi's inexcusable and often ludicrous interpolations. He acknowledges no dignity or elegance or any felicity at all in Mandelbaum's work. Worst of all, in the dogged consistency of his fault-finding he leaves the impression that he sets little value on those underlying, indirectly presented meanings and pressures which determine the character of Dante's poem and the bent of Mandelbaum's translation.
Irma Brandeis
Bard College
Annandale-on-Hudson, New York


D.S Carne-Ross replies:

Irma Brandeis's defense of Mandelbaum, and indirectly perhaps of herself since she contributed to the chorus of unseemly praise from which I quoted ("marvelously good, and a joy to read aloud"), takes the odd form of an attack on Ciardi—as though I had lavished on his version of Dante's poem the sort of commendation which she and others saw fit to lavish on Mandelbaum's. I did nothing of the sort. If I had to read Dante in English I would always go to Binyon, but his translation is so far superior to Mandelbaum's that any comparison would have been absurd, so I went instead to Ciardi's more modest, decent performance. Not a masterpiece, but well enough written to allow one to read on without being continually distracted by the translator's infelicities.

To complain, as Brandeis does, that Ciardi fails to render the full sense of the original here and there misses the thrust of my review, which is no great matter. What does matter is the misunderstanding about the nature of translation which her letter reveals. She is evidently one of those, all too common in the academic community, who believe that bad or at best mediocre verse is an acceptable stand-in for great poetry so long as it provides the point-for-point correspondence to the original which allows the instructor to hold forth. For that, you must go to a straight prose version, which in the case of Dante means Singleton or the Temple Classics edition. A poetic translation is always going to take its liberties, of omission, addition, substitution. It must do so to live. However fine Mandelbaum's understanding of Dante may be (Freccero seemed rather less convinced of this), he simply does not write well enough, as I tried to show, to carry a major text. One may recall Ernest Newman's Wagnerian tenor who understood the role of Tristan perfectly but unfortunately had very little voice.

Brandeis's position is too far from my own to make argument fruitful. She has in mind the classroom situation where monoglot students must be conned into believing that the tawdry paperback they are studying is "really" by Dante or Sophocles or whoever, a practice that plants a lie in their souls and is calculated to deaden whatever native sense of poetry, of language, they may possess. I was writing with the general reader in mind (who may sometimes be a student), someone who cares for poetry and lacking languages must rely on translation. If he wants to read Dante, he had best go to Binyon, but since Binyon's English is sometimes as difficult and contorted as Dante's Italian, he may prefer Ciardi. If he is wise enough to check the translation against a good prose crib, something essential will come across.

Shall We Dante? - The New York Review of Books

[Sybarite]

Translating prose must be fraught with enough difficulties, but goodness knows how anyone even starts translating poetry.

I have the utmost respect for those that do so.

[Adrian]

I'm currently dipping into Douglas Hofstadter's Le ton beau de Marot, which addresses the subject of poetry in translation with respect to this poem, A une Damoyselle malade:

Ma mignonne,
Je vous donne
Le bon jour;
Le séjour
C'est prison.
Guérison
Recouvrez,
Puis ouvrez
Votre porte
Et qu'on sorte
Vitement,
Car Clément
Le vous mande.
Va, friande
De ta bouche,
Qui se couche
En danger
Pour manger
Confitures;
Si tu dures
Trop malade,
Couleur fade
tu prendras,
Et perdras
L'embonpoint.
Dieu te doint
Santé bonne,
Ma mignonne.

by Clemént Marot.

It's written with some quite specific rules that must be kept in any translation:

1. The poem is 28 lines long.
2. Each line consists of three syllables.
3. Each line's main stress falls on its final syllable.
4. The poem is a string of rhymed couplets: AA, BB, CC,...
5. Midway, the tone changes from formal ("vous") to informal ("tu").
6. The poem's opening line is echoed precisely at the very bottom.
7. The poet puts his own name directly into his poem.


Hofstadter tries his hand at "replicating" the poem in English and then moving on to other discussions of translation in general and especially the problems (and joy) of translating poetry. Well recommended.

[nnyhav]

I looked for verities in the last 2 grafs of Carne-Ross' response but found only transparent rhetorical imputations that don't rise to the level of argument:

"She is evidently one of those, all too common in the academic community, who believe that bad or at best mediocre verse is an acceptable stand-in for great poetry so long as it provides the point-for-point correspondence to the original which allows the instructor to hold forth."

"She has in mind the classroom situation where monoglot students must be conned into believing that the tawdry paperback they are studying is "really" by Dante or Sophocles or whoever, a practice that plants a lie in their souls and is calculated to deaden whatever native sense of poetry, of language, they may possess."

Utter folderol, all this. He is evidently one of those, all too common in the critical community, who believe that bad or at best mediocre commentary is an acceptable stand-in for strong argument so long as it provides a tit-for-tat correspondance with the interlocutor which allows the self-promoter to hold forth.

No Dante translation captures, or can capture, its greatness. The best that can be achieved is a distant appreciation of that greatness.

[Mirabell]

No, I wasn't referring to sound arguments when I mentioned truths. Yr right that he doesn't appear to like to use reason.

However, you saying

No Dante translation captures, or can capture, its greatness. The best that can be achieved is a distant appreciation of that greatness.

is not particularly far from Carne-Ross' words. I, personally, emotionally, like his refutation of dry poetry translations which emphasize the closeness to the original over poetical power. I am partial, for instance to Rudolf Borchardt's translation of Dante, which is totally wacked-out, artsy, unredable in places and gorgeous everywhere. Haven't finished it yet.

However, I tend to think that the better the poetry of a poetry translation is, the further away it is from the original. Borchardt's isn't a good translation. It is an utterly astonishing work of poetry, but as a translation...?

I vehemently dislike HOfstadter, and Le ton beau de Marot is no exception. He is, if I remember correctly, part of that Benjamin school of translation which values the sounds/metre/poeticality above the content.

[nnyhav]

Each poet makes his/her own language. And making poems does not a poet make; knowing something of the craft, but little of the art, I'm still trying to work out what comprises successful translation, and why certain poets are amenable to it and others immune. Dante seems the most difficult to English, but it certainly hasn't been for lack of trying (I only know the Ciardi, and Pinsky's Inferno) ...

A bit over a year ago TheValve tried a collective enterprise of translating poems; the results weren't too awful considering that they emerged from committee ... and seem on-topic, particularly the finalé:

Translating Mallarmé

Toute l’âme résumée
Quand lente nous l’expirons
Dans plusieurs ronds de fumée
Abolis en autres ronds

Atteste quelque cigare
Brûlant savamment pour peu
Que la cendre se sépare
De son clair baiser de feu

Ainsi le chœur des romances
À la lèvre vole-t-il
Exclus-en si tu commences
Le réel parce que vil

Le sens trop précis rature
Ta vague littérature.
___________________________

The entire soul invoked
Through our slow exhalings
In several rings of smoke
Dissolved in other rings

Attest to some cigar
Whose clever burn inspires
As long as ash stays far
From the clear kiss of fire

In the romantic song
That flies to lips beguiled
The real does not belong
Exclude it, it is vile

Exactness is erasure
Of cloudy literature.

Followup exercise was translating Verlaine.

[nnyhav]

Blogspotting: Today's NYer Bookbench:

This issue of the magazine features the poem “Here the Birds’ Journey Ends,” by Mahmoud Darwish, who passed away August 9th. The poem was translated, from the Arabic, by Fady Joudah, Darwish’s friend and frequent translator (a collection, “The Butterfly’s Burden,” was published in 2006). Joudah is also a doctor who worked for a time in Africa with Doctors Without Borders and currently practices in one of the hospitals in the Houston, Texas, medical center where Darwish died following heart surgery. Joudah kindly agreed to talk with me about Darwish, his work, and his legacy.

more ...

[Eric]

A few comments:

1)
The Brandeis versus Carne-Ross discussion is valuable in that it brings up bones of contention, but often these discussions are held by people who have something personal against one another. So, the whole dicussion becomes a kind of duel, which we outsiders are shut out of. Because it all becomes point-scoring, rather than elegant appraisal.

The argument tends to be won by people who have themselves actually attempted translation and can explain, pedantically if needs be, all the things they took into consideration when wrestling with the text. Critics love to cavil, without actually taking the risk of being criticised themselves. So a non-translating critic must be especially even-handed.

2)
I can't stand this clever-dickery that means that people write poems especially to make them hard to translate. This reduces poetry and prose to crossword puzzle level, where the main thing is to use 28 lines of 3 syllables, or leave out the the letter "e", so that people will pat the translator on the back and tell them how clever they are (as opposed to skilful, erudite, knowledgeable, cultured, etc.).

3)
I agree with Mirabell about content of style when you first translate a text. I always maintain that the translator should start with a dry, wooden, pedantic, etc., translation of the poem. This means that the translator has understood the basic words, meaning, puns, ambiguities, etc. Only then should s/he move away from the original, and try to make an elegant version in the new language. There are really atrocious translations of poems, where the translator makes all manner of excuses about their "interpretation", "version" or "rendering" of a poem. This usually means they neither understood the original properly, nor had the inventiveness to try to actually translate the poem.

4)
If you know no Italian, but can read, say, German and French, you can get nearer the original by reading translations of Dante in various languages, as opposed to only comparing two English ones.

5)
With the Mallarmé, this thing about the pedantic, wooden start to grasp the meaning is important. Some translators plunge in there, without first getting a clear picture of the images the poet is using.

For instance:

i) the ash-glow-cigar image is very clear in the original. You've got y'r long brown bit, y'r burning glowy bit, and y'r ash, which has a tendency to fall off. The key image here is the way that the ash separates from the clear kiss of the fire behind it. So "se sépare" is crucial. "Stays far" means something different to "separates". Staying is motionless, separating involves movement.

ii) Also, "dissolved" is a gradual, slow movement; "abolis" gives a more abrupt image.

iii) Maybe, "attest" does not have the same nuance as "atteste". But you would need a native-speaker of French to tell you.

iv) One tricky verb is "voler", which can mean "to fly" and "to steal". "Steal away" may be a solution.

v) And "vague" can suggest or hint at "wave" or "vague".

vi) Another subtle difference is that "rature" means "crossing out", i.e. a line on top of something still visible; "erasure" means you can't see it any more.

Most of the translation is good, accurate. But the rhyme has been forced in places ("beguiled" with "vile"; "song" with "belong"). And, with ambiguous words, or ideas, translators are sometimes forced to come down on one side or the other. Translation is compromise. But the translator should be aware what he is discarding.

As I always claim, translating rhymed poetry is trickier than prose, because everything is so concentrated.

[Adrian]

Originally Posted by Eric

I can't stand this clever-dickery that means that people write poems especially to make them hard to translate. This reduces poetry and prose to crossword puzzle level, where the main thing is to use 28 lines of 3 syllables, or leave out the the letter "e", so that people will pat the translator on the back and tell them how clever they are (as opposed to skilful, erudite, knowledgeable, cultured, etc.).

Eric, you have spectacularly misunderstood my post. The poem was written by Marot in 1537 and I doubt he wrote it with present-day translations in mind.

Secondly, Hofstadter chose this poem in part because of its formal structure. A prose-poem or a less structured poem would have given any translator too much freedom, one thing he wanted to avoid.

And finally, the whole gist of the book is how to translate the poem sensitively. He opens the book with a very literal translation (even the chosen title, "To a Sick Damsel" is awful) specifically to show how not to do it.

[Eric]

You are perfectly right, Adrian. A poem written so long ago has nothing to do with clever-dickery for translation. Just out of interest, the spelling is remarkably modern. If you look at a Rabelais text, you really need a glossary. This is very straightforward and simple. Are you 100% sure that this is not in itself a translation from the older French?

I personally shy away from translating most rhymed poetry because of the fact that if you're not careful you end up with rhyming doggerel.

Secondly, another crucial feature of poetry is the picture that the poet puts into your mind. If you forgive my mistake about Hofstadter and the rhymed Clément Marot poem, and look at my detailed comment about the Mallarmé one, my point there is that the translator(s) change too many small things, small pictures in your mind, so that the sum total is no longer quite the same poem.

Personally, I think the way to translate a poem sensitively is to actually have a shot yourself. Do you happen to know how much actual translation Douglas Hofstadter has done himself, and from which languages?

[Mirabell]

Originally Posted by Adrian

And finally, the whole gist of the book is how to translate the poem sensitively.

No, it's not. You said it yourself when you, true to the book (*barf*) claimed

It's written with some quite specific rules that must be kept in any translation:

(my emphasis)

[Eric]

Mirabell, you reminded me - the rules:

It's written with some quite specific rules that must be kept in any translation:

1. The poem is 28 lines long.
2. Each line consists of three syllables.
3. Each line's main stress falls on its final syllable.
4. The poem is a string of rhymed couplets: AA, BB, CC,...
5. Midway, the tone changes from formal ("vous") to informal ("tu").
6. The poem's opening line is echoed precisely at the very bottom.
7. The poet puts his own name directly into his poem.

What he says is perceptive enough about this specific poem. I agree that all these factors need to be taken into account by the translator. The rules are a starting point. Then the right vocabulary and nuance are needed.

But I hope you see the point I make, i.e. that translating rhymed poetry is tricky. Even if you follow all of Hofstadter's rules, you're not there yet.

And I'm still wondering about the modern spelling of a poem written centuries ago (now that I know that it was!).

[Mirabell]

they need to be taken into account, yes. they "must be kept in any translation"? hell no.

[Stewart]

Originally Posted by nnyhav

Followup exercise was translating Verlaine.

When I went to the Translation Workshop at the Edinburgh Book Festival, one of the things covered was poem by Verlaine. Not the one above, but Chanson d'Automne:

Les sanglots longs
Des violons
De l'automne
Blessent mon coeur
D'une langueur
Monotone.

Tout suffocant
Et blême, quand
Sonne l'heure,
Je me souviens
Des jours anciens
Et je pleure

Et je m'en vais
Au vent mauvais
Qui m'emporte
Deçà, delà,
Pareil à la
Feuille morte.

The workshop should have been called Translating French Workshop - that way the many of us who knew not a word of French wouldn't have bothered and they might actually have got some translation done. (Especially since the other examples were Queneau, Rimbaud, and Moliere.)

It's a really nice little poem, especially when spoken aloud, and from what the guy was saying, is, for how miniscule it is, a bitch to translate.

[Eric]

When I did my workshop at the Society of Authors in June, I assumed that not everybody would know all the languages. So I constructed activities which resembled detective work, and discussions of "what would you do, if faced with this dilemma?" type of things.

I wanted the participants to think and do things, not just sit and admire my knowledge and pronunciation of whatever language. Although you can't help showing off a bit. Reading a poem in Finland-Swedish dialect risked Pseuds Corner, perhaps, but doing the Boris Johnson excerpt [no name mentioned initially], and asking whether it was a translation or not, was fun.

Workshop leaders have a duty to advertise their wares clearly, so that you're not conned.

The basic problem with that poem that Stewart quotes here is that there are a lot of "o" sounds, and quite a few "eu"s, etc. If you reproduce the rhyme, internal rhyme, alliteration and so on, you might end up with doggerel. Yet this whole poem is built on rhyme. What to do?

[Stewart]

Originally Posted by Eric

The basic problem with that poem that Stewart quotes here is that there are a lot of "o" sounds, and quite a few "eu"s, etc. If you reproduce the rhyme, internal rhyme, alliteration and so on, you might end up with doggerel. Yet this whole poem is built on rhyme. What to do?

And also the fact that in French the word monotone is precisely that: monotone. Indeed the syllables are balanced in French poetry, unlike in English where you get stresses on certain syllables.

[Mirabell]

Indeed, stewart. This is a major point. German hacks translating English poetry often overlook the fact that english is an easier language to rhyme in, that pentameter is almost a natural rhythm in english (cf. Steele). rhymes in German invariably sound contrived.

and, to drag bejamin out again. it is more important to translate the form than to do justice to the words? Benjamin's translations have changed the very meaning of some Baudelaire poems.

I also saw a number of English Rilke translations and was dumbfounded that people would think that they were in any way equivalents...

[Adrian]

Originally Posted by Eric

You are perfectly right, Adrian. A poem written so long ago has nothing to do with clever-dickery for translation. Just out of interest, the spelling is remarkably modern. If you look at a Rabelais text, you really need a glossary. This is very straightforward and simple. Are you 100% sure that this is not in itself a translation from the older French?

Not 100% sure, but if you say that Rabelais wrote very differently, I imagine the text I posted isn't the actual original version.

Secondly, another crucial feature of poetry is the picture that the poet puts into your mind. If you forgive my mistake about Hofstadter and the rhymed Clément Marot poem, and look at my detailed comment about the Mallarmé one, my point there is that the translator(s) change too many small things, small pictures in your mind, so that the sum total is no longer quite the same poem.

I agree with you. The essence of the poem can easily be lost when you're familiar with the original poem. Any translation never seems to live up to the original.

Personally, I think the way to translate a poem sensitively is to actually have a shot yourself. Do you happen to know how much actual translation Douglas Hofstadter has done himself, and from which languages?

I know he's fluent in languages other than his native English, but I don't know his expertise in translation. Pretty experienced, I'm guessing.

[Adrian]

Originally Posted by Mirabell

No, it's not. You said it yourself when you, true to the book (*barf*) claimed

...

(my emphasis)

I think the crux of the problem is that if you don't have those rules, the poem would be changed so much it wouldn't be recognisable as a translation but would just appear to be an original poem.

[Eric]

Hofstadter must have a good grasp of at least a couple of foreign languages, otherwise he wouldn't stick his neck out and start listing rules. But as so few Brits and Americans have any insights at all into languages, he will soon get people saying "wow" at his erudition. The guru effect can be dangerous.

When I Googled for that Clément Marot poem, I couldn't unfortunately find one instance that claimed it was the original version. Yet I cannot believe that it was spelt like that in 14-hundred-and-something.

Many rhymes sound contrived in English too. My pet hate, whose efforts I will have mentioned many times, is Ezra Pound as translator. He uses the most ludicrously obscure and archaïc words, just to get the rhyme, while original is quite ordinary, even when written centuries ago. For instance:

Sentir non può di lui spirito vile,
Di cotanta vertù spirito appare.
Questo è lo spiritel, che fa tremare
Lo spiritel, che fa la donna umile.

Now, I'm no expert on Old Italian, but the words don't look that complex or sophisticated. But by the time Pound's finished with this verse (part of Sonetto XIII), it has been transmogrified into a ludicrous piece of nonsense:

No vile spirit to discern his vertu is able
So great is the might of it,
He is the spryte that putteth a trembling fyt
On spirit that maketh a woman mercyable.

Yea! Both syntax and vocabulary obscure rather than explain. He did get the rhyme, but Procrustes has been lurking again.