Ron Silliman has a post about the odd phenomenon of “homophonic translation”—rendering a poem into English not (primarily) by dictionary meaning but by phonetic similarity. As far as I know (and Silliman agrees), the first practitioner of the art was Louis Zukofsky in his 1969 book Catullus (here are a couple of samples, paired with more normal translations; you’ll get the idea from the rendering of Catullus’s “Minister uetuli puer Falerni/ inger mi calices amariores” into “Minister wet to lee, pour the Falernian/ and gear me chalices, ah by bitterest”). Silliman brings to my attention a more recent example, David Melnick’s 1983 version of the Iliad, Men in Aïda, which begins as follows:
Men in Aïda, they appeal, eh? A day, 0 Achilles!
Allow men in, emery Achaians. All gay ethic, eh?
Paul asked if tea mousse suck, as Aïda, pro, yaps in.
Here on a Tuesday. ‘Hello,’ Rhea to cake Eunice in.
‘Hojo’ noisy tap as hideous debt to lay at a bully.
Ex you, day. Tap wrote a ‘D,’ a stay. Tenor is Sunday.
Atreides stain axe and Ron and ideas ‘ll kill you.
(You can see more of Melnick’s book here, and a transliteration of the Greek here.) I don’t like it as well as the Zukofsky, but that’s a matter of taste. I agree with Silliman that both are more successful than Chris Tysh’s version of a passage of Lautréamont’s Les Chants de Maldoror—not for its own sake (it’s quite nice: “All on, Sultan, evoke two languages, debar us, my dizzy song key, sail it o’er the parquet. The bandage is fine…”), but as an equivalent of the French (“Allons,Sultan, avec ta langue, débarrasse-moi de ce sang qui salit le parquet. Le bandage est fini…”): it simply doesn’t sound like it, and Silliman’s conclusion is that “the mapping of sound patterns from French to English” doesn’t work. I suspect he’s right, and some languages are better suited to the technique than others. (You can also read Silliman’s own homophonic version of Rilke and Charles Bernstein’s versions from Basque and Portuguese.)