Kurt Andersen recently interviewed Natasha Wimmer, the translator of Chilean writer Roberto Bolaño, on Studio 360. Her eighth Bolaño novel in translation, The Spirit of Science Fiction (which is not a work of science fiction!), is out on February 5th. It was a great conversation, and I thought I’d share a few things from it that I found interesting.

Run-On Sentences

Wimmer mentioned something about Spanish that I have also noticed when translating from French: there is a much higher tolerance for run-on sentences. So, when translating, the translator will usually choose not to carry over all of these run-on sentences into English. As Wimmer puts it, the run-on is not necessarily a stylistic choice by the author, but something that is part of the texture of Spanish.

Does the Literary Translator Improve the Author’s Work?

It seemed odd to me that Andersen kept on asking if Wimmer feels she makes the writer’s work better. On the one hand, as a literary translator, I found this question incredibly flattering, since people usually assume translation is a rote task and that words can be brought over, one by one, into the target language, like moving a stack of crates from one side of the room to the other. And certainly, when translating a human resources announcement or a report full of jargon, I have felt that I’ve improved the style of the original in English. But I’m not sure any literary translator would have so much hubris as to claim to have “improved” a writer of talent. Now perhaps Andersen is just tossing out questions that ask for a negative answer, as interviewers sometimes do. And of course Wimmer explained that it’s not a question of “improving,” but of using the resources of English, and compensating for what can’t be done the same way as it is in Spanish. Translation is not improving, though it is a creative act; it’s writing, or rewriting, the same text in a different language.

Google Translate as Literary Translator

At the end of the segment, Andersen compared Wimmer’s translations of two sentences from the novel to Google Translate versions. As would be expected, the Google Translate versions were lacking in linguistic nuance (using the adjective “proletarian” instead of “working-class,” for instance) and, in the second, lengthier sentence, Google was unable to capture the cadence of the writing. When asked about the differences, Wimmer pointed out that the literary translator deals with writing that occurs in a “stylistically unexpected way.” Google Translate bases its translations on a corpus of already existing words and phrases. And in many contexts, that is sufficient and very useful. But it cannot “anticipate the new.” Surprisingly, despite his admiration for Wimmer’s skill, Andersen thinks Google Translate will learn to do this. It is hard to imagine machine translation understanding the range of meanings of each word, how that meaning may be changed by the words around it, the cultural or literary resonance a word may evoke, the history of how the word has been used, and how it contributes to the musical qualities of the prose. If machines one day translate novels, then I guess they will be writing them, too.

Photo of the Palacio de Bellas Artes in Mexico City by Ricardo Esquivel from Pexels