← Back to Arc Publications

That stupid expression "translation loss"


Personally, I think the idea of “translation loss” is a bit of a red herring, since every text will gain something from being translated. The idea of translation loss is bandied about by people who haven’t a clue what literary translation is; it’s akin to talking about “creative loss” if I write a poem about a tree and it doesn’t look like a tree.

Translation creates changes, obviously, and there might be cases where we feel on balance that that change has resulted in some sort of a loss. But surely, in a good translation, this is not the point: the point is what the text has gained.


It really depresses me when I go to events or lectures and somebody mentions, usually with a smile, that poetry is what is lost in translation or that to translate is to betray (following that Italian proverb). It shows that translators and translation scholars still have a lot of work ahead in showing people what it is that they do.

The notion of poetry being lost in translation is based on seeing meaning as something that is hidden behind a text, something that the text decodes, whereas, as Wittgenstein said, nothing is hidden. The meaning is in the form, not in some Platonic Form. Therefore if we pay attention to form, we can write a target text that will not clone the source text, which is an impossible expectation, but may well be a work of relevant similarity, to quote Andrew Chesterman. And as the above comment states, there may well be gains.


Hm, but maybe it’s just more that people think the original is by, say, Victor Hugo, so must be brilliant, and so the poor old translator (say Tim Ades) can’t be as good and is bound to lose something. That’s not what I think, since I think Tim Ades allows more people to read Victor Hugo, so that’s the point really.


What slightly troubles me is how do you know whether you are reading a ‘good’ translation or not?. If you don’t understand the language of the original (and probably don’t have the source text to hand anyway) and if the translation sounds perfectly acceptable and doesn’t jar in any way, how do you know that it could sound better in a better translation? And how can a translator convincingly convey the idiosyncrasies strangeness – even weirdness – of an original text without being thought of as a bad translator?


I would say if it sounds perfectly acceptable it must be a bad translation. If it grabs you, makes you think, if it works as poetry, if it does everything that poetry is supposed to do, then it is likely to be a good translation. So you trust that sense of its poetry first of all. If it is good as poetry it’s only then that you know it’s worth going through the comparison with the original, asking the opinion of native speakers and poets of the original language if necessary.


I can’t help thinking that if a poem in translation “sounds perfectly acceptable” it need not necessarily be a “bad translation”, but maybe a translation of a work that in itself was only acceptable or even mediocre. Surely though one could read what one thought was a wonderful translation only to find the translator was a very fine poet and brought to the final text rather more than the source had intended; a good translation? How is the non linguist to know or even be reassured?


I note that in a review of a new translation of Homer in the TLS, there was yet another reference to poetry being lost in translation. It is clearly an important issue to be debated, and it seems to me that the only way that non-linguists can be reassured is to be brought into the debate e.g. by reviewers drawing attention to the fact that the words we read in [English] are not those of the source text author but of the target text author. Bilingual editions are really important here. Michael Hoffman rightly criticised a new edition of Brecht’s love poems for being monolingual.


Hello. Partly I have to wonder whether the objection to this idea is the feeling that it is a bad thing - but must it be? I could perhaps write a dissertation on this but I will keep it very clear, in a world overrun by globalism the idea that everything should be able to pass across borders is perhaps the underlying mistake behind the objection to translation loss. Really for me, Spanish occupies an entirely different world to English, for example, and it can do many many things that are impossible in English, just as English can do things that are impossible in Spanish, and not just because of their cultures. Italian words have a completely different feel to them than Spanish words do (Italian words are knives, Spanish words are shadows), so even though the words and sentences may be rather easily translated from Spanish into Italian due to the linguistic proximity of the two, the result is something not even approximately the same, at least in many cases, the more nuanced the poetry is, the better the poet understands her language.


Hi @amy , welcome to the forum! This is a perspective I hadn’t considered – that we expect things to cross borders and when they don’t, we describe it as a loss. Especially with the development of AI translation processes such as Google Translate, the general expectation seems to be that at some point in the future, all translation will be instant and automatic, and it is frustrating to encounter a text that contradicts that in any way. Perhaps poetry will gain a new status at that future moment of technological achievement, as the one thing that can’t cross the border automatically.

Or perhaps artificial-intelligence-based translation of poetry is not impossible, only particularly difficult, and will follow soon after.

Incidentally I love your metaphors for Spanish and Italian. I wonder if you could elaborate on them…?


Further to what Amy said, I agree, and I (and others) have said many times that translation is not loss, but gain. Poems in particular gain from a new perspective, a new culture, a new interpretation. It’s what poems do. But language also gains from the fact that it crosses a border into a language where the perspective is different. Because that’s what languages do. Change and adaptation, and being seen in a different light - that’s how texts and words, even, survive. So I would say that not only is the inability to cross easily into another language and culture not bad, but it is crucially important.


I too love the idea that Italian words are knives and Spanish words are shadows. Are French words fireworks, then? And German words bits of Lego? But what about Russian ones? Or Chinese ones? I’d love some more metaphors of this kind. (Even if it is just a bit of a poetic game.)


I agree the metaphors about languages are interesting, but are they helpful? Are they just based on rather stereotypical ideas we have about those cultures? For example, lots of people who don’t speak German have rather reductive ideas about the ‘personality’ of the language, which differ wildly depending on whether their encounter with it has been via Schumann or via Anglo-American films about Nazis.
Can a language really have a ‘personality’, when we have to use it to express our whole spectrum of ideas and emotions? And if so, can we judge the personality of our native language, or only other languages? Presumably genuinely bilingual people would be best placed to comment on this, but even then, could we be sure the different ‘personalities’ belonged to the languages and not, say, to the parent who spoke each one? I think a lot of people do feel they themselves are different in different languages, but I’m not sure this phenomenon has much to do with the languages themselves.


Well, they’re maybe helpful if we know they are just a game. If we take them too seriously then, I agree, they can be very limiting. And also annoying to people to whose languages we apply them, just as any stereotypes are. But I am often told by members of my bilingual family that I have a different personality in German and English: I am bossier in German. A phonologist friend says that I just have slightly lower pitch in German. So this suggests there maybe is something about the languages that makes us speak differently. I, too, would be wary of calling it the personality of the language, though.