I started to translate Mallarme's poems because I wasn't content with the English versions I could find. In translation, poets who preceded and followed Mallarme are accessible to us. Baudelaire is accessible to us, Apollinaire is accessible to us. I couldn't find a way into Mallarme in the English versions I found, and I suspected it wasn't his fault. Mallarme was an English teacher, and he did translations of the poems of his beloved Edgar Allan Poe, in prose, of course. But his poems in French were metrical and rhymed. Sometimes the meaning of his poems is guided, let us say, by the rhyme and meter he chose, "ceding the initiative to words" as he wrote in "Variations sur un sujet." There are alternate homophonic readings of some of his lines as well, that could never be rendered into another language as poetry, only as notes.

My inability to reproduce the multiple layers of Mallarme's poems in French into American verse is a disappointment, but an inevitable one. Those layers rely on sound similarities that aren't available in our language. Most translations of Mallarme into English rhyme and use traditional meters. This seemed to me to be the wrong approach. We have seen how Mallarme approached translating Poe, after all. Mallarme translated this way moves even further away from the meaning of the poem as a second rhyme scheme, this time in English, imposes an alien framework over the poem. These translations, no matter how carefully constructed, often sound academic to me. I come from a poetry tradition that learned a great deal from 20th century French poetry, Apollinaire, Reverdy, Desnos and all the rest of them.

Some of the great pleasures in music in 20th century American poetry come from responses to French poetry, for instance, John Ashbery's "The Tennis Court Oath," or Charles Olson's Rimbaud takeoff "Variations done for Gerald Van Der Wiele," or Ted Berrigan's translations of French poetry in "Bean Spasms." Hopefully the music I heard there echoes a little in these versions.".