for avoiding rhyme has been the results achieved by all those who have used rhyme in translating The Divine Comedy: they have shown that the price paid was disastrously high. I believe that all those who have offered rhymed translations of Dante could have produced far better poems if they had not used rhyme.’

I AM THE WAY INTO THE DOLEFUL CITY,
I AM THE WAY INTO ETERNAL GRIEF,
I AM THE WAY TO A FORSAKEN RACE.

JUSTICE IT WAS THAT MOVED MY GREAT CREATOR;
DIVINE OMNIPOTENCE CREATED ME,
AND HIGHEST WISDOM JOINED WITH PRIMAL LOVE.

BEFORE ME NOTHING BUT ETERNAL THINGS
WERE MADE, AND I SHALL LAST ETERNALLY.
ABANDON EVERY HOPE, ALL YOU WHO ENTER.

I saw these words spelled out in somber colors
inscribed along the ledge above a gate;
‘Master,’ I said, ‘these words I see are cruel.’

Hofstadter observes that ‘although it does not rhyme at all, I find it of far greater appeal than the preceding two, firstly because it does not show any evidence of having been a struggle (though it no doubt was), and secondly because it comes far closer to mirroring the precision of Dante’s meter’ (pp. 536–7).
It is curious that Hofstadter does not consider the translation of Dorothy L. Sayers11 (1949) that, in my opinion, does the best in at least partly preserving the hendecasyllables and the rhyme:

THROUGH ME THE ROAD TO THE CITY OF DESOLATION,
THROUGH ME THE ROAD TO SORROWS DIUTURNAL,
THROUGH ME THE ROAD AMONG THE LOST CREATION.

JUSTICE MOVED MY GREAT MAKER; GOD ETERNAL
WROUGHT ME: THE POWER, AND THE UNSEARCHABLY
HIGH WISDOM, AND THE PRIMAL LOVE SUPERNAL.

NOTHING ERE I WAS MADE TO BE
SAVE THINGS ETERNE, AND I ETERNE ABIDE;
LAY DOWN ALL HOPE, YOU THAT GO IN BY ME.

These words, of sombre colour, I descried
Writ on the lintel of a gateway; ‘Sir,
This sentence is right hard for me,’ I cried.

However, sometimes to respect the rhyme is not enough to produce an equivalent effect.

In Eliot’s ‘The Love Song of J. Alfred Prufrock’ there are the famous two lines:

In the room the women come and go
Talking of Michelangelo.

There is, if not a rhyme, an assonance, and Italian translations (both by Luigi Berti and Roberto Sanesi)12 did not respect it (by the way, they translate the whole poem without rhyme):

Nella stanza le donne vanno e vengono
Parlando di Michelangelo.

In the French translation,13 Pierre Leyris attempts to maintain an effect of rhyme by changing the meaning of the source expression:
Dans la pièce les femmes vont et viennent
En parlant des maîtres de Sienne.

A recent Italian translation,14 rather clumsily, has a false rhyme and a bad assonance:

Le donne vanno e vengono nei salotti
Parlando di Michelangelo Buonarroti.

This is one of those cases in which the translator feels entitled to change the reference of the source text in order to preserve its effect. But is there a real preservation of the effect? Even though the rhyme has been saved, we have lost the wit of the original assonance, which also presupposes an English accent in the pronunciation of the Italian name (not Michelan’gelo but Michelangelo’), and thus acquires a debasing effect. Moreover, speaking of the Masters of Siena (let alone knowing the family name of Michelangelo) presupposes some competence in the history of Italian painting, while ‘Michelangelo’ is popular and available enough for a kitsch usage. I suspect that these ladies speaking of Michelangelo appear more bas bleu or blue-stocking than if they were speaking of Duccio di Buoninsegna or Simone Martini. At this point I decide that a rhymeless solution is better than a rhymed one. However, neither Berti nor Sanesi were insensitive to other problems of substance: the nine syllables of Eliot’s first line and the eight of the second became in Italian, respectively, a dodecasyllable and a nine-syllable line. The distich thus keeps a sort of lapidary gnomic dignity.

As a matter of fact the original ‘Prufrock’ is entirely rhymed (Let us go then, you and I, / When the evening is spread out against the sky / Like a patient etherised upon a table; / Let us go, through certain half-deserted streets, / The muttering retreats . . . and so on). The Italian translations are not. Was it a case of mere laziness? Once I tried to retranslate part of the poem finding acceptable rhymes, and I quote only a few lines for reasons of humility, to show to what an extent I consider my attempt unsuccessful: Tu ed io, è già l’ora, andiamo nella sera / che nel cielo si spande in ombra nera / come un malato già in anestesia. / Andiam per certe strade desolate / nel brusio polveroso / di certi alberghi ad ore, in cui folate / senti di notti insonni, e l’acre odore / di ristoranti pregni di sudore . . .

I stopped because I had the impression of having written an Italian poem of the late nineteenth century. My Eliot sounds too similar to certain verses published by the Italian poet Lorenzo Stecchetti in 1877 (Sbadigliando languir solo e soletto – Lunghi e tedosi giorni, – Dormire e ricader disteso in letto – Finchè il lsonno ritorni, – Sentir la mente e il core in etisia, – Ecco la vita mia).

Perhaps my translation would have been acceptable to Italian readers if published in 1911 (when ‘Prufrock’ was written), but Berti’s translation is from the 1940s and Sanesi’s from the early 1960s. At that time Italian culture was receiving Eliot as a very contemporary poet and was reading him after having read the Italian experimental poets like the Hermetics. In this sense Eliot had a great influence on the further Italian neo-avant-garde. Italian readers enjoyed a certain dryness in Eliot, a quasi-prosaic style, as a reaction against a lot of traditional poetry. We are here facing the phenomenon of the so-called horizon of the translator.15 Each translation is received within the framework (or ‘the horizon’) of literary conventions that inevitably influence the choices of the translator. Berti and Sanesi were moving in the Italian literary horizon of the mid-twentieth century and did not avoid rhyme out of laziness, but because they negotiated their version by taking into account the expectations of the Italian readers of their times. They decided that the rhyme, in Eliot, was less important than the representation of a ‘waste’ land and that it was of primary importance to save the reference to sawdust restaurants full of oystershells (that, by the way, recalled Montale’s Ossi di seppia, bones of cuttlefish for the Italian reader). To render rhymes at any cost risked making Eliot too ‘gentle’ and operatic, thus losing the feeling of that ‘handful of dust’ that pervades so many of his poems.

Italian translations of ‘Prufrock’ were thus determined by the historical moment at which they were made and by the ‘translator’s horizon’. This is the reason why translations, in general, age.

I do not want to spend any more words on the difficulty of translating poetry, nor to consider the many theories of how poetry is untranslatable and can only inspire imitation, rewriting, and so on. I want only to stress that poetic texts are a sort of touchstone for translation, because they make clear that a translation can be considered absolutely perfect only when it is able in some way to provide an equivalent of the physical substance of expression.
I shall provide more examples to illustrate this in the next chapter, before approaching the problem of intersemiotic translation.

NOTES

1. Roman Jakobson, ‘Linguistic aspects of translation’, in Reuben A. Brower, ed., On Translation (Cambridge: Harvard U.P., 1959), pp. 232–9. See also Roman Jakobson, ‘A few remarks on Peirce’, Modern Language Notes xciii (1977), pp. 1026–36.
2. Heraklit, in Gesamtausgabe (Frankfurt: Klostermann, 1987).
3. Georg Gadamer, Warheit und Methode (Tübingen: Mohr, III, 1960).
4. George Steiner, After Babel (London: Oxford U.P., 1975), iv, 3.
5. Quasi Come (Milano: Bompiani, 1976).
6. Jacques Derrida wrote in L’écriture et la différence (Paris: Seuil, 1967, p. 312): ‘A verbal body cannot be translated or transferred into another language. That’s exactly what a translation drops. Let the body drop, this is the essential energy of translation.’ But in the opening lecture of a symposium on translation held in Paris in December 1998 he said that a translation must be ‘quantitatively equivalent to the original [. . .] It is not the problem of counting the number of signs, of signifiers and signified, but rather of counting the number of words [. . .] There is an ideal law, even though inaccessible: not to translate word per word or word by word, but to keep close as far as it is possible to the equivalence of a word through another word.’
7. New York: Basic Books, 1997, chapter 17.
8. The Inferno of Dante, tr. Robert Pinsky (New York: Farrar, Straus & Giroux, 1994).
9. Dante’s Inferno. Translations from Twenty Contemporary Poets (New Jersey: Hopewell: Ecco Press, 1993).
10. The Portable Dante (New York: Penguin, 1995).
11. The Divine Comedy – 1. Hell, tr. Dorothy Sayers (Harmondsworth: Penguin, 1949–62).
12. Poesie di T. S. Eliot, tr. Luigi Berti (Parma: Guanda, 1949). T. S. Eliot, Poesi, tr. Roberto Sanesi (Milano: Bompiani, 1966).
13. T. S. Eliot, Poèmes, 1910–1930, Texte anglais présenté et traduit par Pierre Leyris (Paris: Seuil, 1947).
14. T. S. Eliot, Poesie 1905/1920, Italian tr. by Massimo Bagicalupo (Roma: Newton, 1996).
15. See for instance Itamar Even-Zohar, ed., ‘Polysystems Studies’, Poetics Today 11.1. (1990). See also Antoine Berman, Pour une critique des traductions: John Donne (Paris: Gallimard, 1995) and La traduction et la lettre ou l’auberge lointain, second edn (Paris: Seuil, 1999).

VIII. From substance to matter

Let me return to my personal experiences as a translator of Nerval’s Sylvie, to further our discussion of the role of substance in texts aiming at an aesthetic effect.

Hidden verses

Before translating it I had read and reread Sylvie over a period of at least forty years. In spite of this