Karly

I first read Don Quixote when I was thirteen, which, looking back, feels like a slightly deranged thing to ask of a child, but I did read it, I understood enough of it to pass whatever assessment followed, I knew about the windmills and Rocinante and Sancho Panza, and then I carried on with my life without giving much thought to the fact that I’d just read a novel published more than four hundred years ago in a version of my own language that was both recognisably Spanish and very obviously not quite the Spanish I was using to gossip with my friends at school.

Years later, I went on to get a degree in languages and linguistics, during which I took an entire module on the history of the Spanish language, so I’ve actually sat in a university classroom learning how Latin became Spanish, why certain sounds disappeared, why others changed, how vocabulary travelled and evolved, and why the language Cervantes was writing at the beginning of the seventeenth century looks the way it does. I’ve studied semantics and pragmatics, I’ve spent an embarrassing amount of my adult life thinking about why people choose one word instead of another, and Spanish is, rather importantly, my first language, which would suggest that I should be reasonably well equipped to translate Don Quixote.

I’m not, because if somebody put the original in front of me tomorrow and told me to produce an English translation that was accurate, readable, funny, stylistically faithful and capable of doing even a fraction of what Cervantes is doing in Spanish, I think I’d simply hand the book back and wish them well.

Knowing two languages is unfortunately not enough

There’s a wonderfully uncomplicated idea of translation that exists when you’ve never had to think very hard about translation, which is that the translator reads something in one language, understands what it means and then writes the same thing in another language, so Language A goes in, Language B comes out and everybody goes home, except languages, rather inconveniently, refuse to work like that.

The problem is that words don’t exist in neat bilingual pairs, and the moment you’re dealing with literature, particularly literature written four centuries ago by somebody who was deliberately playing with language, register and genre, knowing what the words mean is only the beginning of the problem. Cervantes published the first part of Don Quixote in 1605 and the second in 1615, at a point when Spanish was still undergoing changes that would eventually give us many of the forms we recognise today, and there are features of Cervantes’s Spanish that sound distinctly old to a contemporary speaker.

There’s another complication, though, because Don Quixote himself deliberately reaches backwards and adopts the language of the chivalric romances he’s spent far too much time reading, meaning that some of the language is old because the novel is old, while some of it is old because Cervantes is writing a man who is intentionally trying to sound even older. Don Quixote’s linguistic problem is, in some ways, spectacularly contemporary: the man has consumed so much of one particular kind of media that it’s altered his idiolect.

His chivalric romances haven’t merely given him ideas about how the world works, they’ve given him a register through which to perform those ideas, complete with deliberately antiquated vocabulary, constructions and forms that already sounded strange to Cervantes’s contemporaries. Today’s children have 6 7; Don Quixote had Amadís de Gaula. The mechanisms of brainrot are apparently eternal.

This is extremely funny until you’re the person who has to translate it and suddenly have to decide what, exactly, the English equivalent is of a seventeenth-century Spanish man attempting to imitate an already antiquated literary register. Do you make him Shakespearean? Do you start throwing thee and thou around? Do you make him sound vaguely medieval and hope nobody asks too many questions? More importantly, if you make the entire novel sound generically old because Don Quixote itself is old, what happens to the joke?

This is where literary translation stops looking like replacing words and starts looking suspiciously like being asked to rebuild an entire linguistic system in another language, and, incidentally, this isn’t an entirely hypothetical nightmare because students of translation have actually been asked to translate passages of Don Quixote as part of university coursework. My sincerest well wishes, therefore, to every student currently being asked to translate Don Quixote, every student who has translated Don Quixote in the past and every student who will one day open an assignment brief to discover that it is their turn. I have a language degree and enough knowledge of historical Spanish to understand the scale of the problem, and I’d like you to know that I believe in you, although I cannot help you.

En un lugar de la Mancha

We don’t even have to venture particularly far into the novel before everything goes wrong, because En un lugar de la Mancha, de cuyo nombre no quiero acordarme… is possibly the most famous opening in Spanish literature and, to a Spanish speaker, not a particularly mysterious sentence. I know every word in it, I know what the sentence means and, if you asked me to explain it to you, I could do so without hesitation, but ask me to decide what Cervantes should sound like in English and suddenly we have a problem.

John Ormsby’s nineteenth-century translation gives us “In a village of La Mancha, the name of which I have no desire to call to mind”, while John D. Rutherford opts for “In a village in La Mancha, whose name I have no desire to recall”, and Edith Grossman opens her 2003 translation with “Somewhere in La Mancha, in a place whose name I do not care to remember”. These translators are looking at the same Spanish words and producing English sentences that communicate essentially the same information while sounding noticeably different, which is precisely where the illusion that translation is merely a matter of knowing vocabulary begins to collapse.

I can look at lugar and tell you that it means place, but that doesn’t tell me whether place is the best word in English. I can tell you what querer means and I can tell you what acordarse means, but apparently that still leaves us with “have no desire to call to mind”, “have no desire to recall” and “do not care to remember”, all of which give us slightly different narrators before we’ve even finished the first sentence.

Grossman’s “Somewhere in La Mancha” is particularly interesting because somewhere isn’t sitting there waiting for us in the Spanish. Cervantes gives us en un lugar, and literally we could quite happily produce “in a place”, yet “somewhere” immediately sounds like somebody beginning a story, which means we’re no longer asking which English words correspond to the Spanish words but what Cervantes is doing when he begins his novel and how an English sentence might be able to do something similar. That is translation, apparently, and I already hate it here.

And then you have to make Cervantes funny

Semantic equivalence is already enough to give me a headache, but Don Quixote has the additional inconvenience of being funny, and humour is where any remaining illusion that translation might be a tidy operation completely collapses because Cervantes plays constantly with register, proverbs, social class, literary convention and the different ways his characters speak. Don Quixote doesn’t sound like Sancho Panza, and neither of them simply sounds like “a person from 1605”; the differences between them are part of their characterisation, and the ridiculousness of Don Quixote’s chivalric performance depends partly on everybody else existing around him as evidence that normal human beings don’t speak like this.

A translator therefore has to preserve not only what these people are saying but the relationships between the ways they say it, and this is the point at which I begin to feel profoundly affectionate towards literary translators because I genuinely don’t know what I’d do. You can’t simply translate a Spanish proverb word for word if the resulting English sounds like somebody experiencing a neurological event, but replacing it with an English proverb may introduce cultural associations that were never present in the original, while a joke built around a word may cease to be a joke when the English equivalent has none of the same secondary meanings.

A sentence whose humour depends on register may require an entirely different linguistic mechanism to make an English reader perceive the same absurdity, so you’re no longer asking what something means but what it does, why it does it and how on earth you can persuade English to do it too.

Unfortunately, we also need to talk about literacy

All of this came back into my head recently because a content creator took a passage from Don Quixote onto the street and asked ordinary people to read it, and some of them couldn’t, which would perhaps be less alarming if the task had been to translate it, analyse Cervantes’s manipulation of chivalric register or explain an obscure seventeenth-century lexical item, but the people being asked were native Spanish speakers and the task was simply to read the passage.

There are several enormous caveats that have to come before I become the world’s most irritating literacy campaigner, because being approached on the street with a camera is not an academic assessment, a selectively edited social media video is not a representative linguistic study, and Don Quixote contains unfamiliar vocabulary, constructions and references that a contemporary Spanish speaker can’t reasonably be expected to understand immediately. The expectation should never be that somebody can open a seventeenth-century novel and instantly understand every word on the page, because I certainly can’t despite the aforementioned degree, historical linguistics module and thirteen-year-old encounter with the poor man fighting windmills, but there is a considerable difference between not knowing every lexical item and being unable to navigate a text the moment it becomes unfamiliar.

There’s a skill involved in reading a difficult sentence and continuing anyway, in encountering an unfamiliar word and using morphology or context to make a reasonable guess, and in reaching the end of a paragraph, realising you didn’t quite follow it and committing the apparently revolutionary act of reading it again. You don’t need to know precisely what an adarga is the first time Cervantes gives you one in order to continue reading the sentence, because uncertainty isn’t a sign that reading has failed and, for most of human literary history, readers have managed to survive it rather successfully.

I do wonder whether our increasingly fragmented relationship with written language is making that kind of uncertainty feel less natural, because so much of what we read every day is designed to surrender its meaning instantly, whether it’s a caption, a headline, a notification or a sentence written specifically because somebody knows you might scroll away after seven words. Difficult prose asks something completely different of us because it requires us to stay with a sentence that hasn’t immediately given us everything we wanted from it, and that ability, like every other aspect of literacy, has to be practised.

So yes, hooray for good translators, genuinely, because I’d quite like somebody else to make the impossible decisions about lugar and acordarme on my behalf, but also, and I can’t believe this is apparently where my life has taken me, please bring back literacy.

Every translation is also an interpretation

What fascinates me most when I put different translations of Don Quixote next to each other is not deciding which translator is correct, because increasingly I think that’s the least interesting question we could ask when they all understood the Spanish and the differences exist precisely because understanding the Spanish doesn’t automatically produce an English sentence.

A translator has to decide what matters most in every particular moment, whether that’s rhythm, humour, the literal image, register or making sure the reader understands a cultural reference without turning the novel into an annotated hostage situation, and sometimes preserving one thing inevitably means loosening your grip on another. The final English text becomes the accumulation of thousands upon thousands of those decisions, which is why I’ve become increasingly suspicious of the phrase “faithful translation”, because we still have to ask what exactly we’re being faithful to.

Are we being faithful to the vocabulary, the syntax, the rhythm, the historical distance, the joke or the reading experience? John Ormsby himself addressed this problem in the nineteenth century, arguing in his introduction that Cervantes shouldn’t simply be rendered into artificially antiquated English because Cervantes wasn’t, for the most part, writing artificially antiquated Spanish, and that decision alone tells us something important about translation because refusing to make the English sound old can actually be an attempt to remain more faithful to an old text.

My linguist brain loves this argument considerably more than my hypothetical translator brain does, because every answer seems to create three additional questions and at some point somebody still has to commit to an English sentence and send the manuscript to an editor.

Perhaps the degree worked after all

The strange conclusion to all of this is that I was probably more confident in my ability to translate Don Quixote when I was thirteen than I am now, because at thirteen I knew Spanish and I knew English and therefore probably would’ve assumed that translating between the two was simply a matter of knowing enough words. After a language degree, a linguistics education and an entire module devoted to understanding how Spanish became the language I speak today, I now know enough to understand how catastrophically insufficient that would be.

I know that register carries meaning, that pragmatics carries meaning and that the history of a word carries meaning, but I also know that two expressions can have almost identical semantic content and still do completely different things in conversation, and that sometimes the most literal possible translation is precisely the one that fails to communicate what the original was doing. This is deeply annoying from the perspective of anybody hoping for a tidy answer, but it’s also what makes translation such an extraordinary form of reading.

The great trick performed by a good literary translator is that we’re not supposed to spend every sentence thinking about any of this. Somebody else has already sat there agonising over whether lugar should become place, village or somewhere, and we simply turn the page; somebody has decided how ridiculous Don Quixote should sound, how Sancho should answer him, whether the joke can survive intact and how much of the seventeenth century an English-speaking reader needs to feel around them, and when it works, all of that labour disappears into the experience of reading.

So I have a language degree, Spanish is my first language, I read Don Quixote at thirteen and I have quite literally studied the history of the language Cervantes used to write it, yet I still don’t know how the fuck I would translate Don Quixote, and perhaps knowing enough about language to understand exactly why I have no idea where the fuck I’d begin was the qualification after all.

If, after all of this, you’re inexplicably interested in reading Don Quixote rather than running away from it, Yale has made Roberto González Echevarría’s entire course on Cervantes available through Open Yale Courses. It consists of twenty-four lectures devoted to a close reading of the novel in the context of Renaissance and Baroque Spain, and, perhaps appropriately after everything we’ve just discussed, the course itself uses John Rutherford’s English translation while recommending Francisco Rico’s edition for readers tackling Cervantes in Spanish.

As for me, I’ve apparently decided that reading Don Quixote at thirteen wasn’t enough punishment and I’m taking on the adult version now, this time aloud and alongside my mother and grandfather. Between three generations of native Spanish speakers, a language degree and the accumulated wisdom of several centuries of Cervantes scholarship, I’m cautiously optimistic that we can defeat the cursed book eventually. If nothing else, at least this time nobody is grading me.