Every act of translation is, in some irreducible sense, an act of betrayal. The Italian aphorism traduttore, traditore — translator, traitor — captures with epigrammatic economy a truth that translators themselves have acknowledged, debated, and agonised over for centuries: that the passage from one language to another inevitably entails the sacrifice of something essential, something that inheres so intimately in the original that it cannot survive the journey intact. What that something is, and whether its loss constitutes a genuine impoverishment or merely a transformation, remains one of the most vexed questions in literary and philosophical thought.
The most venerable position in this debate is the one associated with the Romantic tradition and articulated most forcefully by Wilhelm von Humboldt, who maintained that each language embodies a unique worldview — a distinctive mode of apprehending and organising reality that is coextensive with the grammatical and lexical resources at the speaker's disposal. If this thesis is correct, then translation is not merely difficult but, in the strictest sense, impossible, for to translate a text is to uproot it from the cognitive soil in which it grew and replant it in terrain whose chemical composition is fundamentally different. The translated text may resemble the original, but it can never be the original, any more than a photograph of a landscape can be the landscape itself.
This position, however compelling in the abstract, has always sat uncomfortably with the empirical reality that translation manifestly works. Readers of Constance Garnett's renderings of Dostoevsky, whatever their limitations, have acquired something genuine — an encounter with a sensibility, a narrative intelligence, a moral seriousness that is recognisably Dostoevskyan even in English garb. The experience may be imperfect, but it is not fraudulent, and to insist otherwise is to privilege theoretical purity over lived literary experience in a manner that few practising readers or writers would find persuasive.
The philosopher Walter Benjamin proposed an alternative framework that sought to transcend the sterile opposition between fidelity and betrayal. In his celebrated essay on the task of the translator, Benjamin argued that the purpose of translation was not to reproduce the meaning of the original but to release what he called the 'pure language' latent within it — a universal linguistic substratum that no single language could express in isolation but that all languages, taken together, gestured toward. Translation, on this account, does not diminish the original but completes it, revealing dimensions of significance that remained dormant so long as the text existed in only one linguistic form. The metaphor is characteristically mystical, but it contains a practical insight of considerable value: that the encounter between languages in translation can generate meanings that neither language could have produced alone.
Contemporary translation theory has largely abandoned the quest for a single correct approach and has instead embraced what might be called a situational ethics of translation. The appropriate strategy depends upon the nature of the text, the purpose of the translation, and the relationship between the source and target cultures. A legal document demands a different kind of fidelity than a lyric poem; a text being translated for a scholarly audience requires a different approach than one destined for the mass market. This pragmatic turn has been salutary in many respects, liberating translators from the paralysing dichotomy of literal versus free and allowing them to make contextually sensitive judgements about where their primary obligations lie.
Yet something is lost in this pragmatism, and it is worth pausing to identify what it is. The older debate, for all its occasional sterility, kept alive a question that the situational approach tends to dissolve: the question of what we owe to the otherness of the original text. A translation that domesticates the foreign — that smooths away its strangeness and renders it comfortably familiar — may serve the reader's convenience, but it does so at the cost of suppressing precisely the quality that made the text worth translating in the first instance. The best translations, I would suggest, are those that manage to be simultaneously hospitable and unsettling — that welcome the reader into an unfamiliar world without pretending that the world is less unfamiliar than it actually is.