Skip to content
Menu
Life and Legends
  • About Us
  • Masthead
  • Submissions
  • Our Authors
    • 2014 Authors – Inaugural Edition
    • 2015 Authors
    • 2016 Authors
    • 2017 Authors
    • 2018 Authors
    • 2019 Authors
    • 2020 Authors
    • 2021 Authors
    • 2022 Authors
  • Read
    • Seventeenth Edition
    • Sixteenth Edition
    • Fifteenth Edition
    • Fourteenth Edition
    • Thirteenth Edition
  • Archives
    • Twelfth Edition
    • Eleventh Edition
    • Tenth Edition
    • Ninth Edition
    • Eighth Edition
    • Seventh Edition
    • Sixth Edition
    • Fifth Edition
    • Fourth Edition
    • Third Edition
    • Second Edition
    • First Edition
  • Be Part of the Legend
  • Contact
Life and Legends
November 15, 2026November 15, 2026

The Untranslatable, Translated: An Essay by Elton Uliana


Translators are forever being told that certain things cannot be translated but the curious thing is that we usually translate them anyway. The difficulty may lie in a word for which there is no equivalent, in a joke that refuses to travel, or in a sentence so dependent on the peculiar resources of its language that moving it elsewhere may be impossible at first glance; yet it is in the very act of confronting that obstacle that the translator finds a way of moving it, re-creating it, transformed by the passage but no less translated for having being changed. Translation never quite promises safe passage between languages without transformation, and the so-called ‘untranslatable’ has an oddly unsuccessful history of preventing translation.

Finnegans Wake ought to be able to put up a better fight than most. Terry Eagleton has called it “the most hermetic work in the English language”, which is perhaps another way of saying that it approaches the untranslatable. Except that even “the English language” becomes a complicated designation here. The book seems at times to be written in several languages at once, inventing words, corrupting others, making puns across linguistic borders and turning the primordial confusion of tongues into a literary method; it demands a kind of translation within the language before we even reach the territory of translation between languages. Yet this most notoriously hermetic of novels has been translated and retranslated, several times into Brazilian Portuguese, but also into languages as distant from Joyce’s linguistic apparatus as Japanese, Korean and Chinese, while parts of the Wake have travelled still further, into Esperanto, Georgian and even Ancient Egyptian. The supposedly untranslatable text, in other words, has generated an unusually vigorous history of translation. Translators have worked independently and collectively on the apparently impossible task of making Joyce happen elsewhere, only for other translators to come along and do it differently. The point is not that one of them must finally have cracked the code. Quite the opposite: every new version makes the idea of a single code harder to sustain. What looks like evidence of untranslatability becomes, in the very act of translation, evidence of how many translations are possible. Perhaps the ‘untranslatable’ does not describe something that cannot cross from one language into another so much as something whose crossing reveals just how much research, experimentation, trying and trying again happens along the way.

This is partly because the problem may lie less with translation than with what we expect translation to do. To believe that words are essentially the names of things is, amongst other things, what makes the translator’s task seem impossible. Translation does not simply replace one thing with another; it mediates between them, and mediation necessarily alters the terms of the encounter. I take the view that it is translation that poses a greater obstacle to the ‘untranslatable’, and not the other way round.

There is, of course, a serious argument behind the idea of the untranslatable. Edward Sapir, and later Benjamin Lee Whorf, proposed in different ways that language does more than provide names for a world already neatly divided into things and concepts. The language we speak, they tell us, participates in how we organise and perceive that world. Push that proposition far enough and the very possibility of translation looks precarious. If two languages divide experience differently, how could what is thought in one ever be reproduced exactly in the other? But ‘exactly’ is doing suspiciously heavy work here since difference need not make understanding impossible any more than understanding requires two people to inhabit identical worlds. And if the shape of the world is determined entirely by language, it is not clear whether or not without language it would not have any shape at all.  

In any case, whatever languages may do to our perception of the world, translators encounter their differences rather more concretely. Sometimes there simply is no word in English for the word sitting in front of me in Portuguese, while at other times there are several and none quite does the job, in the same way that a joke may disappear with the sound on which it depends, or a sentence may be doing something that English grammar politely refuses to reproduce. Even the most ordinary word can arrive carrying a social or emotional history for which no dictionary could reasonably be held responsible. These are not imaginary difficulties, nor do I think translators gain much by pretending that languages are interchangeable. But absence of equivalence is not absence of possibility, and the interesting question is what the translator does when the equivalent cannot be found. Sometimes we explain or compensate, sometimes we invent, leave something behind or make something strange newly familiar. Stubbornly, translation proceeds because difference gives it something to work on, even when that difference cannot be overcome.

What is curious, though, is that all this difficulty can produce a flourishing of semantic material rather than its extinction. Something may be lost, certainly, but loss has enjoyed a rather too privileged place in the language we use about translation, to my mind, and we sometimes forget to notice what turns up in its place. Once a word enters another language, associations begin to gather around it that were not there before, and sometimes an entire sentence discovers a rhythm its first language could not give it. Even a joke that refuses to make the journey intact can reappear somewhere else, by entirely different means. Translation is, in Roland Barthes’ useful formulation, a matter of ‘decomposing in order to recompose’. What crosses the linguistic border has already been changed by the crossing, acquiring another life through the very changes supposedly threatening its survival..

There is another difficulty with the idea of untranslatability, and a rather important one, for history has been remarkably uncooperative with the proposition, people having, after all, spent several thousand years translating things they might reasonably have been told could not be translated. Nor is the evidence exactly scarce, since the Epic of Gilgamesh travelled restlessly between ancient languages, while the books of the Hebrew Bible crossed into Greek in the Septuagint and later into Latin and countless vernaculars, accumulating interpretations, disputes and retranslations along the way. Babel may have given Western culture its great myth of linguistic confusion, but confusion has never proved much of a deterrent. Throughout history, faced with the inconvenient fact that other people speak differently, human beings have translated them. They have also mistranslated them, adapted them, appropriated them and translated them again. None of this demonstrates that languages are transparent to one another; it suggests something more modest, and perhaps more interesting: incomprehension has repeatedly produced translation rather than prevented it.

Our present moment gives this old argument a peculiar new twist. For decades, translation theory has been moving away from the reassuring idea that translation consists of finding equivalents, yet contemporary translation technologies depend, in different and increasingly sophisticated ways, on discovering patterns of correspondence across enormous quantities of language. Alignment, similarity and algorithmic probability have made machines remarkably good at producing sentences that resemble sentences humans might produce, but the irony is difficult to miss because just as we have become increasingly suspicious of simple equivalence as an account of translation, we have built machines capable of translating on an unprecedented scale by calculating likely correspondences across languages. Anthony Pym calls equivalence a “presumption of interpretative resemblance”, and the word ‘presumption’ is key. Equivalence is something a translation proposes, rather than something already built into the structure of languages and waiting there to be discovered. AI has not made that problem disappear and, if anything, its fluency makes the old question harder to see, since when a translation sounds entirely at home, what kinds of difference have been negotiated, displaced or simply made invisible in the process of arriving in its new language?

There are good reasons, then, not to dispose of the idea of the untranslatable altogether. To declare everything translatable can produce its own kind of complacency, particularly when the traffic between languages is anything but smooth. Emily Apter has argued for the untranslatable not simply as a failure of communication but as a productive resistance to the easy circulation of meanings across cultures. The distinction is important enough since languages do not meet one another on neutral ground, and translation has always been entangled with movements of ideology, power, conquest, migration, commerce and cultural prestige. Today, more often than not, the lingua franca waiting at the other end of that journey is English, though increasingly an English spoken between people for whom it is not a first language at all. To insist that everything can be translated must therefore not become an insistence that everything can be made comfortably intelligible to anglophone readers, stripped of whatever resists assimilation. Some words should perhaps arrive carrying evidence of the distance they have travelled; some sentences are entitled to, and perhaps should, remain a little foreign. Untranslatability, understood this way, does not tell the translator to give up; it asks us to pay attention to what the act of translation might otherwise conceal. And perhaps this is also a good moment to remember what equivalence actually means. Equal value sounds reassuring enough, but value between languages is notoriously difficult to measure, and there is no linguistic scale on which we can place two words and watch the needle settle.

Among all the reasons I remain rather fond of the idea of untranslatability, despite having spent these pages arguing against its practical existence, is that it gives translators permission to admit that some things are extraordinarily difficult to move. We do not have to pretend that every word has its counterpart waiting patiently in another language, or that nothing gets bruised on the journey. Perhaps this is also why translators make such convenient fall guys amongst the leading actors. We become conspicuous when something appears not to have survived, although rather less attention is paid to everything that may have happened to it, or come into being, along the way. For me, then, the untranslatable names a difficulty worth lingering over, not a border at which translation is obliged to stop.

Which brings us back to Finnegans Wake, still obstinately untranslatable and, rather inconveniently, translated. Its several Portuguese incarnations, alongside versions in Japanese, Korean, Chinese and elsewhere, do not prove that nothing was lost, or that Joyce can somehow be made equivalent to himself in another language. What they demonstrate is stranger and more useful, since translation has never needed to abolish difference in order to cross it. Perhaps the mistake was to imagine the untranslatable as a locked door when it is closer to the resistance we feel as we push against one. We know it is there because something pushes back; the trouble is that, after several thousand years of translation, people keep getting through. 


Elton Uliana is a literary critic and translator. He is co-editor of the Brazilian Translation Club at University College London (UCL), where he contributes to the teaching of translation theory. His work has appeared in World Literature Today, Reading in Translation, The Massachusetts Review, Routledge’s Art in Translation, Asymptote, West Branch and the José Saramago Foundation’s Blimunda, as well as in the anthologiesDaughters of Latin America (HarperCollins, 2023) and Afro-Brazilian Short Fiction (UCL Press, 2024). In 2026, he received The Massachusetts Review’s Jules Chametzky Translation Prize for his co-translation, with Amanda Sarasien, of José Luís Peixoto’s “The Woman Who Dreamed”.


Loading

Share the Legend
©2026 Life and Legends | Powered by WordPress and Superb Themes!