Translation & Language

Eco is both a semiotician and a novelist translated into dozens of languages, and the two roles make his conclusion unusually cool-headed: there is no perfect equivalence between two languages, and literal word-for-word translation only produces a monster nobody can read. The real shape of translation is negotiation — like diplomacy, both sides must concede, and the translator keeps calculating between "faithful to the source" and "comprehensible to the reader," deciding where to give and where to compensate.

The title comes from a concrete problem. In Hamlet, just before the prince runs his sword through the arras, he cries "How now! a rat?" — he thinks a lurking rat (a sneaking spy) is hiding behind it. But the Italian word topo covers both mouse and rat, so the translator is forced to decide on Shakespeare's behalf: does Hamlet see a contemptible little mouse, or a dangerous rat? Pick the wrong word and the menace leaks out. This kind of fork — where the source draws no distinction but the target language forces one — is exactly where negotiation happens.

From this Eco replaces "literal equivalence" with "equivalence of effect." Fidelity is not fidelity to the word but to the effect the original aims to produce in the reader. Sometimes the translator must change the surface — swap an English proverb for a functionally equivalent local one, replace a local joke with another joke that actually lands — seemingly "unfaithful," yet faithful at the deeper level. His own translations are full of this "change in order to stay the same."

But negotiation is not a free-for-all. Eco draws a line with "reversibility": a good translation could in principle be back-translated by another translator to something close to the original. Changes exist to compensate for loss, not as an excuse to improvise. This line separates "translation" from "rewriting" — negotiation has a boundary, and cross it and you no longer have the same text.

The book grew out of academic lectures, and its examples cluster heavily among European languages (English, French, Italian, German), with almost nothing on languages as distant from Indo-European as Chinese or Japanese. Eco mostly discusses literary translation; the negotiation logic doesn't transfer cleanly to technical, legal, or interpreting contexts. Heavy on speculation, light on "how to do it."

Bellos is a first-rate translator (of Perec and Kadare into English), and this book is a demolition job on "translation common sense." The first myth he topples is "word-for-word literalism." No strict literal translation exists, because words carve up the world differently in two languages; forcing one-to-one matches only yields nonsense. What we call "literal" is already the product of countless invisible choices.

The second and biggest myth is "untranslatability." People love to say some word "can't be translated" (German Schadenfreude, Portuguese saudade, Chinese yuánfèn), as if it locked away a secret outsiders can't enter. Bellos answers: anything that can be said can be translated — maybe with more words, or a clause of explanation, but the meaning never truly jams. The romance of "untranslatable" usually mistakes "no ready-made equivalent word" for "cannot be conveyed."

His deepest move is to relocate where meaning lives. Meaning isn't a specimen sealed in a dictionary; it is generated in use, in context, in the back-and-forth between people. The dictionary is itself a product of translation, not its precondition — humans had the need and practice of cross-language communication first, and only then compiled dictionaries. Translation is possible precisely because the essence of language is communication, and communication never required identical word-meanings to work.

The book is full of counterintuitive facts: the Nuremberg trials gave birth to modern simultaneous interpreting; global translation flows are wildly asymmetric (huge numbers of books are translated out of English, pitifully few into it); the EU runs on more than twenty languages translating into each other and hasn't collapsed. These aren't trivia — they're Bellos's evidence that translation isn't a scholar's parlor pastime but the load-bearing infrastructure of civilization, everywhere, just usually invisible.

The book runs on myth-busting polemic: strong at overturning received ideas, weak at offering a system — you finish disenchanted about translation but without a workable "how to translate" procedure. Some chapters simplify thorny linguistic problems for readability, leaving specialists wanting more; the examples still lean European.

The whole book circles one minuscule poem: Ma mignonne ("My Sweet"), written in 1537 by the French poet Clément Marot to a sick young woman — twenty-eight lines, just three syllables each, rhyming in couplets. Hofstadter invited dozens of friends and readers to translate it, translated it himself over and over, and gathered hundreds of versions. How can a three-syllable poem hold up 600 pages? Because each translation is a stance on "what translation should actually keep."

His core claim is sharp: form is content. The clipped three-syllable lines, the paired rhymes, the intimate tone — these aren't packaging you can strip off; they are the poem's "music," the whole of what moves you. The so-called "fidelity" that renders only the "literal meaning" while dropping the meter is, in his eyes, the greatest betrayal: you've kept the poem's corpse and lost its life.

From here he opens fire on Nabokov. In translating Eugene Onegin, Nabokov, chasing literal precision, deliberately gave up meter and rhyme and compensated with mountains of footnotes. Hofstadter strongly objects: a poem rendered as a poem is a translation; rendered as annotated prose, it is surrender. He prefers Falen's kind of translation, which rebuilds the meter and rhyme even at a greater cost in literal wording. The core of this fight is a trade-off that runs through the whole issue: hug the literal, or hug the form and effect.

Deeper still, Hofstadter (author of Gödel, Escher, Bach) wires translation into his lifelong thesis: analogy is the core of cognition. Searching for "correspondences" between two languages is, at bottom, the brain making analogies and mappings — so translation isn't a marginal language skill but a miniature of thought itself. And constraint breeds creativity: precisely because they're boxed in by three syllables and rhyme, translators are forced into unexpected ingenuity; total freedom produces nothing good.

It is also a private book. His wife Carol died of illness while he was writing it, and grief seeps into the pages — so that little poem to a sick young woman gains a double echo. This gives the book a warmth no other treatise on translation has: translation isn't only technique, it's how one person takes another's language, and life, seriously.

Among the 600 pages sit long stretches of personal musing and delightful digression; the structure is loose, and impatient readers will find it "much ado about one little poem." Its stance is extreme — almost no quarter for Nabokov-style literalism — and his "form above all" doctrine doesn't obviously apply to long narrative texts. It is almost entirely about poetry.

Si Guo was an essayist and a man who honed translation as a craft for a lifetime. This slim 1972 book has been hugely influential in the Chinese-speaking world; the poet Yu Guangzhong wrote a long preface, "The Art of Adaptation," for it. Its concern is single but lethal: rescuing Chinese. In his eyes, bad translation was mass-producing a kind of "translationese / Europeanized Chinese" — sentences made of Chinese words but built on an English skeleton, awkward, convoluted, not how people actually speak.

His criterion is plain: the ideal translation should read as if the original author had written it directly in Chinese, not like a translation. To pull this off, the translator needs the courage to leave the source's syntax and reorganize in idiomatic Chinese, rather than dragging words across in English word order. Translation isn't hauling words; it is rewriting in another language.

The book's most practical part is a list of "translationese" symptoms that still cure ailments today. The flood of "de" ("a having-rich-experience engineer" — clause stacked on clause); the overuse of the plural marker "men" (Chinese plurals don't need a suffix — "audiences," "students" usually add nothing); the overuse of the passive "bei" ("he is considered to be," "the problem was solved," where natural Chinese is active); cramming in "a / a kind of" like an article ("this is a very good example" → "this is a good example"); and empty scaffolding like "as," "carry out," "for ... speaking." Each one is a fingerprint left by English grammar inside Chinese.

Remarkably, Si Guo is no fossilized purist. He grants that languages change and absorb foreign elements; what he opposes isn't "change" but "incoherence" — the lazy translation that uses "fidelity to the source" as an excuse for not bothering to make the Chinese flow. He wants not classical elegance but the idiom and clarity of contemporary Chinese. That yardstick still measures any stretch of Chinese accurately today — especially machine-translated Chinese.

Its examples come from English-to-Chinese practice of decades ago, and some judgments of usage and feel (such as certain blanket bans on "men" and "de") run strict by today's standards and shouldn't all be applied verbatim. It treats specifically the Europeanization of English-into-Chinese, with little help for translating out of Chinese or for non-English languages. But as training in the attitude that "a translation should read like Chinese," it has hardly dated.