When I met Robert Pinsky, at the first meeting of the Association of Literary Scholars and Critics, I congratulated him warmly on his splendid translation of the Inferno, which I had reviewed in these pages. I told him I looked forward with eager anticipation to his versions of the Purgatorio and Paradiso. He shook his head. He had spent too much time, he said, on the Inferno; he had neglected his own poetry and his responsibilities as Poet Laureate. He was leaving the rest of the poem to others.
In an interview published in The New York Times (which I read later) he gave a graphic account of his days and nights with Dante: “I called the translation a feat of metrical engineering and I worked obsessively. It’s the only writing I’ve ever done where it’s like reading yourself to sleep each night. We have pillowcases stained with ink where my wife took the pen out of my hand at night—you know, one more tercet, one more tercet.”
The “feat of metrical engineering” which distinguishes his translation from all its predecessors and sent him to bed pen in hand was a novel solution of the problem facing all translators who tackle Dante’s Divina Commedia: to find an English equivalent of the terza rima, Dante’s innovative verse form: tercets in which the two outer lines of the first rhyme and the ending of the inner line rhymes with the two outer lines of the second, and so on until a single line, rhyming with the inner line of the preceding tercet, closes the canto. It is a system which imposes a forceful forward movement on the verse. But it is a difficult meter to handle in English. As one modern translator of the Inferno, John Ciardi, has pointed out, “It requires 1500 triple rhymes to render the Inferno, and even granted that many of these combinations can be used and re-used, English has no such resources of rhyme.” It has been used occasionally in English—Shelley’s “Ode to the West Wind,” MacNeice’s “Autumn Sequel”—but it is one thing to compose your own poem in this demanding pattern and quite another to use it to translate with any degree of fidelity a long poem by somebody else. Most modern translators of Dante have, understandably, renounced any attempt to reproduce Dante’s rhyme pattern, though Ciardi offered a compromise: tercets in which the outer lines rhymed but the end of the inner line had no effect on the following tercet.
Pinsky’s solution was to make “a more flexible definition of rhyme…a fairly systematic norm that defines rhyme as the same consonant sounds—however much vowels may differ—at the ends of words,” a system that he says “happens to correspond to some preference of my own ear, a personal taste.” He also points out that the examples he gives as illustrations—swans/stones, gibe/club, faces/houses, flavor/quiver—come from poems …