MCT: Michael Riffaterre

An entirely new name to me, Riffaterre seems to have remained fairly loyal to the semiotic approach. This essay, from 1985, after a heavy-going start turns out to be a clever and detailed examination of how translation practice highlights the "presuppositions" which are present in any literary text. This leads to the essay's title "Transposing presuppositions on the semiotics of literary translation", which really needs more work.

Let's start with the final sentence:

Perhaps the simplest way to state the difference between literary and non-literary translation is to say that the latter translates what is in the text, whereas the former must translate what the text only implies. 

The things the literary text implies are, broadly speaking, what Riffaterre means when he talks of presuppositions. For example there is an "intertext", which is the background of knowledge that a native reader may have.

The first example he looks at comes from Milton's translation of an ode of Horace. I'll skip over that to the second, which is a passage from Catullus. Depicting the moment when the Nereid Thetis sees and falls in love with Peleus, he argues that the word used to describe her partial nakedness (nutricum) has in Latin unavoidable connotations of breastfeeding, so puts motherhood forward as a concern of the poem. Nutricum literally means nurse; but in Latin it can (just about) be used metonynically to mean breast. In English, you can't do that; there just isn't a word that can hold both meanings. Riffaterre doesn't offer his own solution to this, but in the second example, he does

and this is where the essay really started to grab me. He looks at a tiny extract from a poem by Maurice Fombeure (again, never heard of him. Apparently a mid-20th century lyric poet.) One word, septembrales, is considered. Every educated Frenchman would know that this is a coinage by Rabelais, is the only adjective derived from a month's name, and connotes with wine (Rabelais's usage is in the phrase purée septembrale, a euphemism for wine.)

There's no equivalent word in English that contains both a month and the connotation of wine, so he translates "brumes septembrales" into "mists of the vintage-season".

Both examples, then, are evidence of the importance of the reader's context to understand the full implications of a literary work. I think that's the most useful thing to be taken from the essay, and you could make the point without using the term presuppositions, which seems to me too vague while appearing to be explanatory. The piece is also useful for the view on translation it gives. As always, close examination of difficult passages is enlightening.

Note
I've now looked at how Peter Green translates the Catullus passage. In a quite apposite demonstration of Riffaterre's view, he has the Nereids "mother-naked to breasts and below", expressing the maternal theme in a slightly transposed way, and also, I suppose, keeping some of the awkwardness that "nutricum" has.