A Comparison of Three Translations of Dante's Inferno

A point was raised recently in one of my classes concerning the validity of studying translated works. On the one hand, if it were not for translations, many of the world's most significant works would be inaccessible to anyone who could not read the language in which they were written. But on the other hand, a translation brings into the original work the translator's subjectivity. It is important to remember that a work in translation is not the original work, but the translator's rendering of it. Differences in ideas and opinions about the text may manifest themselves and influence the translation. Needless to say, then, translations will inevitably differ not only from the original, but also from other translations. Because of the possible consequences such disparities may have on the work and the reader's understanding of it, it would seem relevant to compare and contrast translations.

Juxtaposing three modern translations of Dante Alighieri's The Inferno yields three rather different readings of a single text. Each of the translators -- Allen Mandelbaum, Robert Hass, and John Ciardi -- approaches Dante in a different way. Mandelbaum tries to remain true to the original Italian lines, while Hass does to the text what could best be described as anglicizing, and Ciardi offers what he calls a "verse rendering" where he adapts Dante's terza rima scheme to a rhyming scheme of his own. These are the basic differences between the translations, but what difference do the differences make?

Mandelbaum's approach to translation is basically to adhere as closely as possible to the original text. His translation is published alongside the original Italian lines, and Mandelbaum's version reads more faithfully to the original Italian in his choice of syntax and diction than either of the others who take liberties with the text and thus deviate. Consider the Italian version of Canto XXXIII alongside the English:

Poi cominciò: "Tu vuo' chi'io rinovelli disperato dolor che 'l cor mi preme già pur pensando, pria ch'io ne favelli. (XXXIII.4-6)

Then he began: "You want me to renew despairing pain that presses at my heart even as I think back, before I speak. (XXXIII.4-6)

Word for word, he translates this passage with English words that correlate almost exactly with the medieval Italian ones. Mandelbaum is the only translator to include mention of the Hunger Tower also being known as the Eagles' Tower, and whereas both of the other translators complete Anselm's remark with the word "strange" which Dante did not use, Mandelbaum keeps his anxious thought unspoken when he says, "'Father, you look so ... What is wrong with you?'" (XXXIII.51). Such close attention to detail is important in translating works, since the translator's function should simply be to transcribe a work from one language to another as accurately and exactly as possible for the purpose of making a version of the work as similar as possible to the original accessible to readers of different languages. Even when Dante's phrasing becomes tortuous and complex, Mandelbaum follows Dante's lines and meanings almost zealously, which, while almost making his translation somewhat abstruse at times, also lends it scholarly merit.

Now consider Hass's version of the same stanza:

Then began: "You want me to call up
sorrow so desperate that the thought of it
wrings my heart before my tongue can speak.
(XXXIII.4-6)

Rather than trying to translate verbatim, Hass paraphrases the Italian words with English figures of speech. From the Italian "che 'l cor mi preme" which literally means what Mandelbaum translates as "that presses my heart," Hass translates "wrings my heart," using the English idiom in place of a literal translation. The syntax differs, too, from the original Italian. The approach he seems to have taken in translating a non-English work is to translate the idioms and metaphors as well as the words. The dictionary defines the verb anglicize as "to make English or similar to English in form, idiom, style, or character." This is precisely what Hass tries to do, though he does not conform completely to English standards; he uses sentence fragments -- "then began" (XXXIII.4), "and died" (XXXIII.70), "and he to me" (XXXIII.106) -- and ends sentences in prepositions -- "...the head he'd been gnawing at the back of" (XXXIII.3) and "...what my heart forewarned me of" (XXXIII.41). Originally, I had regarded this as some sort of scholarly statement that because this was a work in translation, English grammar need not apply to English words, but by the time I had completed his version of Canto XXXIII, it seemed more to be inconsistency. Anglicization, examples of which abound in translated versions of works, changes the tones of the original works. Granted, some allowances must be made for such factors as grammar and sentence structure, since, for example, English does not use the subject-object-verb structure common to many other languages. But translating non-English expressions into English ones changes what might be termed the cultural essence of the work as it was originally written, as well as the text itself.

Then there is Ciardi's rendering, in which he not only anglicizes Dante, but also modifies the terza rima rhyming pattern of ABA BCB CDC to an ABA CDC EFE scheme of his own. In order to accommodate such a scheme, he adapts the diction and syntax in his translation more than either of the other scholars. Any one whose only encounter with The Inferno happens to be through Ciardi's translation will be seriously misled about the original structure and organization of the poem. The second stanza of Canto XXXIII here reads:

he began to speak: "You ask me to renew
a grief so desperate that the very thought
of speaking of it tears my heart in two.
(XXXIII.4-6)

Here, the translation is taken further. "That presses my heart" now becomes "it tears my heart in two," another metaphor that deviates even farther from the original, in order to have the line rhyme with the first. The overall success of this translation is at best questionable, since it evokes a most likely unintentional comic element at times. Consider this absurd rendering:

I was Count Ugolino, I must explain;
this reverend grace is the Archbishop Ruggieri:
now I will tell you why I gnaw his brain.
(XXXIII.13-5)

This translation is, to its credit, more comprehensible than the others, but comprehensibility is not the only mark of a successful translation. What is sacrificed in its stead are reliability, accuracy, and faithfulness in the translation from one language to another by revising the original to such an extent in terms of rhyme and diction. This poses problems in judging the translation's validity as a translation. In this particular case, Ciardi's version can be considered more of a poetic adaptation than a translation.

The best recourse in approaching a work of literature would be to read and study it in its original form and language. But this is not always possible, and this is where translations come in. Because of the importance of the task, translators have certain responsibilities. In my opinion, the job of the translator should be to transcribe the work as closely as possible from the original text into another language and not to take liberties with it. The translator should present the text with its words translated but not its expressions, idioms, style, or character. Anglicizing not only detracts from a translation's scholarly merit, but it also constitutes a sort of ethnocentric conceit. Editing and modifying detract from the original work, which, if it has achieved success enough to warrant translating, has obviously done so for a reason. Therefore, I find Mandelbaum's version to possess the most scholarly merit. The one fault that could be accredited to him is his not being able to translate with a terza rima rhyme scheme. I have read translations of Molière, Baudelaire, and Racine that were both true to the text and incorporated suitable measures and rhyme schemes, but Mandelbaum manages to compensate for this shortcoming by juxtaposing his translation with the original text. His decision to dispense with the rhyme scheme is, in any case, preferable to Ciardi's attempt to impose an inappropriate one. Mandelbaum's "verse translation" far surpasses both Ciardi's "verse rendering for the modern reader" or Hass's anglicized version.