What passes for translation commentary today can be pretty dismal—i.e., of questionable value as well as depressing, particularly if translation is your métier. Reviews, perhaps the most egregious example, make minimal acknowledgment of a translator’s intervention, even of their existence. Worse, when reviewers do comment on a translator’s work, their notion of translation is so simplistic as to be demoralizing.

Take the New Yorker’s 2020 review of Guido Morselli’s novel Dissipatio H.G., translated from Italian by Frederika Randall. The reviewer understands translation as seamless reproduction: “Randall … manages to get across, in English, the bleakness of Morselli’s restraint.” The translator is praised, clearly, but “the bleakness of Morselli’s restraint” isn’t exactly the Italian novel. The phrase evidently describes a meaning contained in the Italian, possibly linked to the author’s style, yet it is so abstract, not grounded on any linguistic or textual features, that it is obviously the reviewer’s interpretation. Hence the praise is self-congratulatory: the translator’s work is esteemed, but only insofar as it agrees with the reviewer’s reading (whether of the Italian text or the English version isn’t indicated). Translation is imagined as mechanical transfer, so transparent as to be invisible, not particularly resourceful or creative, certainly not an interpretive act in its own right.

Would we get a different view of translation, one that is both more illuminating and more appreciative, if we turned to translators themselves? Since the start of the new millennium, we’ve been given plenty of opportunities for an inside look, a veritable spate of books about translation written by professional translators, where “professional” means “with substantial lists of translated books to their credit,” some working at it full-time, others working in various other capacities as well—academics, poets, fiction writers, editors.

These translators are also pros, lest we forget, because they get paid for their translations. Some list translation among various sources of income; others translate as their livelihood, a situation that coincides with a certain precarity because English remains a language that translates relatively little, especially in the United States. Here translations make up a tiny fraction—far less than one percent—of total annual book output, which currently tops three million titles with self-published books far exceeding those published by trade and university presses.

Nevertheless, to be considered a pro, you must translate large quantities. Anthea Bell (1936–2018), the British translator of Asterix and W. G. Sebald, published approximately 250 book-length translations; the American poet Richard Howard (1929–2022), who translated Baudelaire and Proust, Camus and Sartre, Abdelkebir Khatibi and Gilles Deleuze, published over 200. Bell and Howard left little commentary about their voluminous work as translators, the odd essay or preface, some interviews, nothing compared to the sustained attention given to translation in Lydia Davis’s Essays Two: On Proust, Translation, Foreign Languages, and the City of Arles (2021) or Daniel Hahn’s Catching Fire: A Translation Diary (2022) or Damion Searls’s The Philosophy of Translation (2024).

All three translators qualify as professionals just in terms of productivity. Davis has written an acclaimed body of experimental short fiction over several decades, but she has also translated some 20 literary works from French and Dutch. Hahn and Searls have each Englished over 60 books from multiple languages—Spanish, Portuguese, and French in Hahn’s case, German, Norwegian, Dutch, and French in Searls’s. Despite how busy translating they are, they still managed to squeeze out an entire book-length account of it: they must feel driven to tell us about what they do. Can their ruminations have any impact on, say, the low level to which reviewing translations has sunk? Can knowing how they translate enhance our appreciation of their translation projects, maybe in an upbeat way that boosts the status of their profession (and their own)? Or will their revelations make us suspicious, if not paranoid, readers of their work, raising doubts about the interpretations they might be inscribing in their source texts sub rosa?

Lydia Davis’s Essays Two collects fascinating discussions of her translations of French fiction writers like Flaubert and Proust. She attends to specific verbal choices, although she is likely to pull up short when the analysis really starts to get interesting. In the first volume of Proust’s magnum opus, À la recherche du temps perdu, as the narrator elaborately describes a stained-glass window depicting a mountain of pink snow, he uses the phrase “des flocons éclairés par quelque aurore,” which Davis translates: “snowflakes illuminated by some aurora.” She pronounces the English word “aurora” to be “the perfect equivalent” of the French word “aurore,” which had been translated as either “sunrise” or “dawn” by Proust’s previous English translators: C. K. Scott Moncrieff (1924–30) and his revisers, Terence Kilmartin (1981) and D. J. Enright (1992). Davis points out that their choices, along with “daybreak,” usually translate a different French word, “aube.” She decided to use “aurora” because of definitions she found in a particular dictionary, Le Petit Robert:

the aube is the first light that begins to whiten the horizon; the aurore is the brilliant pink, rosy, or yellow-gold gleam that appears in the sky following the aube; then the sun itself appears.

The English word, Davis writes, “means the same as the French: the redness of the sky just before the sun rises.” I was disappointed that she didn’t cite an English dictionary to show the words are really “the same.” But this omission doesn’t stop her from concluding, confidently, that aurora “does add something else of its own to a text—its surprise, its novelty, and of course its perfect match to the French original.”

Yet if the English translation “adds something else of its own,” can it really be called “the perfect equivalent”? Wouldn’t some addition mean that the words don’t match, there’s some overspill or remainder in the translating language, a ragged edge between the translation and its source text? If “something else” is added, it goes beyond any strict equivalence, and the English is doing something different from the French. Davis doesn’t seem aware that her choice has fixed the meaning of the French word by excluding other semantic possibilities, especially after consulting a dictionary. If she had consulted a variety of dictionaries, she might have interpreted the word differently. French-English dictionaries vary. So do French ones. The Trésor de la langue française informatisé (1971–94) gives virtually the same meaning as Davis for “aurore”: “Moment qui suit l’aube et précède immédiatement le lever du soleil, où l’horizon présente des lueurs brillantes et rosées” (the moment that follows dawn and immediately precedes sunrise, when the horizon shows brilliant pink glimmers). Cambridge’s Global French-English Learner’s Dictionary (2018), however, defines “aurore” simply as “moment òu le soleil se lève” (the moment when the sun rises). Then it translates the word as “dawn.”

What makes Davis’s choice interesting happens only in English. “Aurora” is a poetical archaism that dates back to the 15th century, whereas the French “aurore” is not archaic but current usage, whether now or in Proust’s period. The English word carries a range of resonances, mythological as well as astronomical, although they include the generic “dawn” as well as Davis’s meaning: “aurora” can signify “the colour of the sky at the point of sun-rise; a rich orange hue,” according to the OED, which cites the Elements of Dyeing (1791): “silks to be dyed of an aurora or orange colour.” This text is itself an English translation from French, William Hamilton’s version of the chemist Claude-Louis Berthollet’s dyeing manual, so that Davis has in effect adopted an 18th century translator’s solution for the French word.

The age of “aurora” must surely be part of the surprise and novelty that Davis mentions, a lexical obsolescence that suddenly turns into newness, starting with its deviation from the choices made by the previous Proust translators (“sunrise,” “dawn”). “Aurora,” Davis believes, “will not be very expressive” because “it has not accumulated the same emotional and metaphorical associations for us as dawn.” But “aurora” feels new precisely because it is old and because it circumvents those “associations,” which now seem so banal, drained of feeling or resonance by overuse, reduced to the romantic or the sensational. “Aurora” is just so much more evocative than “dawn” would be, especially in the context: a breathlessly intricate account of a stained-glass church window that depicts a mountain of pink snow, “snowflakes illuminated by some aurora.” The word not only names an atmospheric phenomenon, and a moment in the day, but it’s a color, a particular shade of glass, aestheticized to the point of preciosity in this passage, the church setting giving it an ethereal or spiritual quality. Davis’s poetical translation participates in the image of Proust as the gay aesthete, a fin de siècle sexual stereotype.

She doesn’t talk much about the effects of “aurora,” the way it nuances the narrator’s tone or voice, tracing a personality. She is distracted by the idea of establishing a perfect match. So she can’t explain what is so striking about her choice. A poetical archaism can only seem like a surprising novelty when the cultural norm imposed on translation consists of the current standard dialect of the translating language, the form of the language that is the most commonly used and therefore the most familiar. Davis’s “aurora” breaks that norm of everyday English to stage a reading of Proust that wasn’t new when she did it (2004), but that hadn’t yet been done in English translation.

Daniel Hahn’s “diary” covers roughly three months during which he translated a novel by the Chilean writer Diamela Eltit, Never Did the Fire (2022). He describes it as drawing back the curtain on his “process,” mostly his writing decisions and why he made them. He offers a glimpse of what it means to be a professional translator today, moving from one book to the next, revising one translation while proofreading the galleys for another, keeping an eye on the bottom line: cost efficiency. “In practical terms,” he makes clear, “my working process cannot possibly afford my having a diligent little conversation with myself about each individual word, weighing up every pro and con, etc.” His diary resembles what used to be called “think-aloud protocols,” empirical research that had translators verbalize their thoughts about their translation decisions as they were being made. Hahn’s exposition has a blow-by-blow quality, as if we’re getting the straight dope without embroidery, immediately, while it’s happening. The whole thing was posted online, apparently, and Hahn mentions other translators chiming in on the translation problems he was encountering. His book represents the world of professional literary translation, or a sizable segment of it.

The introductory chapter, written after the diary was completed, gathers Hahn’s ideas about translation and writing. It makes conceptual statements, but it isn’t really theoretical speculation and doesn’t cite any theorists, at least not explicitly. Readers familiar with the history of translation theory and commentary might notice covert references to commonplace notions. When Hahn opines, “You might think of translating as writing the book I believe the author would have written if they’d been writing a book in English,” he echoes early modern translators like John Dryden who in his 1697 version of the Aeneid “endeavor’d to make Virgil speak such English as he would himself have spoken, if he had been born in England, and in this present age.” Here the translator presides improbably over some sort of weird ventriloquism, or reincarnation, or transmigration of souls (while Hahn channels Dryden?), magically erasing differences of language, identity, culture, history. And when Hahn declares that “If fidelity is a useful notion at all, what I am seeking is fidelity to what I imagine to be the source text’s effect” (his italics), he insinuates Eugene Nida’s “principle of equivalent effect” (circa 1960s), where the goal is to ensure (somehow) that a reader’s response to a translation is the same as a reader’s response to the source text—regardless of when, where, how, and by whom the reading is done.

Hahn seems to know that these theoretical clichés are false: perfect equivalence between two languages doesn’t exist. When he asks, “Is the work’s re-expression really the same thing as its source?” he answers, “No, nor could it ever be,” treating what he does as “sleight of hand,” an illusion of linguistic transparency that enables a translation to pass for the original composition it translates. He insists that “readers should feel they’re getting unmediated access to a work of art, even if they know—once you’ve brought the house lights back up—that they aren’t.”

Would we get a different view of translation, one that is both more illuminating and more appreciative, if we turned to translators themselves?

Why, according to Hahn, doesn’t a translation give readers direct access to the source text? “Since every language works differently,” he points out, “every language encodes slightly different information into its words, beyond their simple meaning.” This makes translation “impossible,” in his view. But untranslatability doesn’t repress his blind ambition to translate or, as he puts it, to keep “aspiring to a pre-existing perfection” (pre-Babelian?), writing translations that “aspire to be impossibly the same as another text.” He expresses this frustrated idealism with affable, even teddy-bearish resignation. “I create a new thing,” he writes, “one that’s identical to the original book, except for all the words.”

Hahn considers translators to be “individual interpretative readers and individual creative writers” (his italics). His verbal choices, he asserts, “will be based mostly on the specific context (how the word fits into the narrator’s train of thought, the rhythm of her sentences and things like that).” Yet he describes Eltit’s narrative “voice” as “curiously hard-to-pin-down,” and he confesses that her novel is one “I’m not sure I do understand.” Not surprisingly, then, his diary presents only vague, elliptical interpretations of the Spanish text. Whether they result in a translation that is creatively written would be difficult to assess.

Hahn’s translation process, by his own admission, is not so much intended action as automatic writing, largely unconscious:

I don’t stop to think about it—I go with what feels right, and while I probably could explain the decision if I chose to turn back and look at it, most of the time I don’t allow it any such deliberate intention.

This frank account inevitably makes one wonder what Hahn’s translations have done to their source texts. Whatever transformations he might have wrought would, in any case, be concealed beneath the illusionistic transparency he prizes in a translation, giving the impression of unmediated access to the original. If he ever got called out, he could always say he didn’t “allow it any such deliberate intention.”

In Damion Searls’s The Philosophy of Translation, the professional translator is just the opposite, exceedingly deliberate in both reading and writing. Searls’s own deliberations, however, are filled with peculiar demurrals over what kind of book he is writing and how he defines and practices translation. He claims to be doing “philosophy,” speculation that is more personal than “theory,” which he sees as “academic” and “naturally” in “tension with practice.” He includes himself among the “practitioners” who “don’t want to be told they have an ‘implicit’ theory in their head that some theorist knows more about than they do.” Yet he weaves a discourse about translation that convincingly synthesizes phenomenology and Russian formalism (Maurice Merleau-Ponty, Viktor Shklovsky) with Mikhail Bakhtin’s concept of dialogism. In other words, Searls’s engagement with translation is couched in High Theory, and his exposition sometimes reads like an article in an academic journal, circa 1980s, when an influx of Continental theoretical discourses washed over humanities departments in US colleges and universities. This apparent contradiction makes me curious about why Searls is so prickly about theory. Is his distress over theoretical accounts of translation caused by some that were aimed at his own work? Or is he just in denial about the translator’s unconscious?

Searls locates his thinking about translation in a specific theoretical tradition, “the so-called Western tradition,” or “basically Greek-Latin-French-English-German-Russian.” He feels this entails an acceptance of the “German Romantic understanding” of translation, which he criticizes for viewing language as coterminous with ethnicity, so that it judges translations as either ethnocentric—assimilating the source text to the receiving culture (“domesticating”)—or ethnodeviant—registering linguistic and cultural differences in receptors (“foreignizing”). Deep down, however, Searls is really a foreignizer: he connects Friedrich Schleiermacher’s notion of the translator (“who is well acquainted with the foreign language, yet to whom it remains nonetheless foreign”—in Susan Bernofsky’s version) with Shklovsky’s defamiliarization in art. In the end, Searls confesses to “a wavering adherence to the German Romantic model.”

He divides his book into two halves that turn out to be contradictory in their approach to translation. In the first half he develops a concept of literary language as an innovative deviation or “arc” away from the “baseline” of everyday language use. “Reading like a translator,” Searls’s mantra, is reading for this arc in the source text—and then reproducing it in the translation. Searls illustrates the idea with his own translation of Uwe Johnson’s 1,700-page novel Anniversaries (2021): “the translator has to tease out what’s an aspect of Johnson’s particular writing from what’s merely the default German baseline and then capture what Johnson is doing to and with the German language.” That word “capture” waves a red flag: if the translator must reproduce a specific stylistic “aspect” of the source text, then the translation process can’t avoid the appearance of mechanical substitution, a matter of engineering a perfect fit, a strict equivalence.

The second half of Searls’s book abandons this instrumentalism by opening up everything to variation. “Each translator translates a different thing,” he writes, “in precisely the same way that each reader of a given book reads a different book.” There can be as many translations of a source text as there can be interpretations of it. “No one translates a text,” Searls observes, “they translate their reading of the text, and everyone has different reading experiences.” This view assigns the translator an omnipotence that can seem brutal in its treatment of the source material: “What’s important to preserve,” Searls says flat-out, “depends on what the translator finds in the original.” The translator’s power takes the form of remaining “attentive to how the translating language works, overriding any demand for ‘equivalence’ with how the original language works.” In the long run, the play’s the thing, the impact of the translation, not its relation to the source text.

In emphasizing the translator as reader, Searls seeks to distinguish translation from “’analyzing’ or ‘understanding’ the original,” arguing that “reading” carries less “authority” than “interpretation.” But this view seems like an unwillingness to take responsibility for reading, for its relations to other, competing readings, its place in the world. Searls believes that “a translation is not so much an interpretation of the original text as its own special kind of strangeness-reinforcing writing.” But isn’t it rather the strength of the translator’s interpretation—gauged against other, more familiar readings—that creates a sense of foreignness in a translation, that arc of literary innovation jumping away from the normative baseline, as in Davis’s use of “aurora” instead of “dawn”?

Searls seems to assume a hermeneutic model of translation. He characterizes M. D. Herter Norton’s 1930s versions of Rilke’s poetry as laying down “her vision of who he is. For her, he is the canonical Great Man, one who speaks to us all.” What we want to know, however, is: By what verbal means did she inscribe her interpretation of the “monumental” Rilke in her translation? And why did it get marginalized in the flood of other versions that make him the most frequently (re)translated modern poet in English? Searls mentions different Rilke translations by the likes of the Sackville-Wests, J. B. Leishman, and Stephen Spender, but we never hear about their “vision” of Rilke, only the verse forms they chose. Translation, it would seem, is nothing if not a conflict of interpretations, some stranger than others. But even to make that point, a translator needs not just to enumerate verbal choices but to articulate precisely how a source text is being interpreted.

In a conversation about her work posted last year on the Granta website, Julia Sanches, a young translator from Portuguese, Spanish, and Catalan (with nearly 30 books so far), made a telling remark: “sometimes,” she said, “it’s hard to think about the thing you’re doing when you’re busy doing it.” Don’t be misled: Sanches is not suggesting that her translation process doesn’t involve thinking. She had just said, in fact, that “it depends less on the language than the book, the author and how they use the language in question.” For her, translation is more than a matter of matching words and phrases: the translator constructs contexts in which to interpret the source text, framing verbal choices not with the word, phrase, or sentence but with the entire “book,” placing it in the oeuvre or career of an “author,” and analyzing its “use” of language. What seems to trouble her is not the effort to think while doing, but the difficulty of moving back and forth between different kinds of thinking: practical intuition versus hermeneutic reasoning.

This movement also seems to trouble the professional translators who authored the books under review. They share an emphasis on translation practice that threatens to push issues of interpretation into the background, or merely suppress them. Even translators willing to comment at length, to draw on research and develop theoretical arguments, are reluctant to articulate the interpretive angles they take in their work. This elevates intuition over argument, discussion, debate. Readers, including reviewers, would appreciate translations more deeply if translators talked about what makes their translations different from—not similar to—the source texts they translate. That difference results from the translator’s interpretation, the act that enables a text to be rewritten in a different language for a different culture. This can’t be understood with an instrumental discourse, where the ideal translation is imagined to be the seamless reproduction of a source-text invariant, “preserving the essence of the original,” as it is commonly put, even when the essence is held to be untranslatable. No, to understand translation in a way that matters, we need to think of it as endless interpretation, endlessly variable, endlessly innovative.