As a professional translator – and as a fan translator – there are two questions I regularly hear:
- Why does anything need to be “localized”? Can’t you just translate all the words into other words with a dictionary?
- Why does that thing need to be re-translated? There’s already a decent translation of it!
I hope to write a super-detailed article about these questions someday, but I thought I’d share what other translators have said about these topics over the years and centuries.
Basically, the term “localization” is relatively new in the translation industry – I’m not even sure people started using it until the early 2000s. Even so, the idea behind localization (that focusing on the original intent is just as important as the original words, if not more important) has probably been around for as long as people have been translating.
Similarly, the idea of re-translating works that have already been translated is nothing new. It’s probably been around as long as people have been translating too.
So, with that explanation out of the way, let’s see what acclaimed literary translators have to say.
Incidentally, this 1954 quote from John Ciardi is probably my favorite explanation of localization that I’ve seen so far!
St. Jerome (Letter to Pammachius, 395 AD)
In this letter to the Roman senator Pammachius, St. Jerome explains his translation philosophy of “dynamic equivalence”, which is what we’d probably call “localization” today:
I not only admit, but freely proclaim that in translation from the Greek – except in the case of Sacred Scripture, where the very order of the words is a mystery – I render not word for word, but sense for sense.
Samuel Butler (The Odyssey, 1900 Translation)
In the preface to my translation of the Iliad I have given my views as to the main principles by which a translator should be guided, and need not repeat them here, beyond pointing out that the initial liberty of translating poetry into prose involves the continual taking of more or less liberty throughout the translation; for much that is right in poetry is wrong in prose, and the exigencies of readable prose are the first things to be considered in a prose translation.
[…]
The Odyssey (as everyone knows) abounds in passages borrowed from the Iliad; I had wished to print these in a slightly different type, with marginal references to the Iliad, and had marked them to this end in my MS. I found, however, that the translation would be hopelessly scholasticized, and abandoned my intention. I would nevertheless again urge on those who have the management of University presses, that they would render a great service to students if they would publish a Greek text of the Odyssey with Iliadic passages printed in a different type, and with marginal references.
John Ciardi (The Inferno, 1954 Translation)
When the violin repeats what the piano has just played, it cannot make the same sounds and it can only approximate the same chords. It can, however, make recognizably the same “music,” the same air. But it can do so only when it is as faithful to the self-logic of the violin as it is to the self-logic of the piano.
Language too is an instrument, and each language has its own logic. I believe that the process of rendering from language to language is better conceived as a “transposition” than as a “translation,” for “translation” implies a series of word-for-word equivalents that do not exist across language boundaries any more than piano sounds exist in the violin.
The notion of word-for-word equivalents also strikes me as false to the nature of poetry. Poetry is not made of words but of word-complexes, elaborate structures involving, among other things, denotations, connotations, rhythms, puns, juxtapositions, and echoes of the tradition in which the poet is writing. It is difficult in prose and impossible in poetry to juggle such a complex intact across the barrier of language. What must be saved, even at the expense of making four strings do for eighty-eight keys, is the total feeling of the complex, its gestalt.
The only way I could see of trying to preserve that gestalt was to try for a language as close as possible to Dante’s, which is in essence a sparse, direct, and idiomatic language, distinguishable from prose only in that it transcends every known notion of prose. I do not imply that Dante’s is the language of common speech. It is a much better thing than that: it is what common speech would be if it were made perfect.
Max Hayward & Manya Harari (Doctor Zhivago, 1958 Translation)
Pasternak’s prose has astonishing power, subtlety and range. While always remaining simple and colloquial, it is exceptionally rich and poetic. Indeed, he makes use of sound and word association in the manner of a poet of genius. His language has a vitality which must be rare in the literature of any country and is perhaps unique in that of Russia.
Needless to say, these very qualities face the translators with difficulties which are almost insurmountable, and we have no illusions that we have done justice, even remotely, to the original.
The poems present an even greater problem than the prose. Rather than delay the publication of the book until a version has been made by an accomplished English poet who knows Russian—a work which would inevitably take a very long time—we have adopted the expedient of merely giving a literal translation of the verse without making any attempt to convey its form (other than the length of the stanzas). But clearly, in the case, for instance, of ‘The Wedding Party’, where the metre is that of the Russian peasant dance tune, the chatushki, and this conveys the mood, a literal translation has only the most limited use.
For the convenience of the reader, a cast of the principal characters is given at the beginning of the book. In addition to surnames, Russians address one another by their name and patronymic as well as by diminutives. The cast, which lists the various forms of each name, may help to prevent confusion.
A word should perhaps be added about place names. Many of those occurring in the novel have meanings which are relevant to their context. Usually such names are translated in the footnotes, but occasionally, where the style of the passage seems to require it, we have used the reverse method, putting the Russian form of the name in the footnote and its English equivalent in the text.
Edward G. Seidensticker (The Tale of Genji, 1976 Translation)
New translations of great classics need not seek to justify themselves. There have been translations of very great writers by very great writers, and they have been superseded. Since there is probably no such thing as a perfect translation of a complex literary work, the more translations, one would think, the better. Arthur Waley’s translation of The Tale of Genji has been so important to me over the years, however, that I feel impelled to remark briefly on my reasons for undertaking a new translation. It was my introduction to Japanese literature, and its power upon repeated readings—I could not give their total number—has continued to be so great that the process of preparing a new translation has felt like sacrilege.
Yet the fact remains that the Waley translation is very free. He cuts and expurgates very boldly. He omits one whole chapter, the thirty-eighth, and close scrutiny reveals that the titles of at least two chapters, the thirtieth and forty-first, are meaningless in his translation because he has omitted the passages from which they derive. It may be argued that he tidies things up by cutting, and therefore “improves.” In some cases he probably does. One shares his impatience with clothes and meals and ceremonies, and may sometimes wish that Murasaki Shikibu had shared it too. On the whole, however, his excisions seem merely arbitrary. One is at a loss to know why he can cut Chapter 38, which is quietly, meditatively beautiful, and then went on to translation Chapter 44, in every respect its inferior.
More complex, and perhaps more interesting, is the matter of amplification. Waley embroiders marvelously, sometimes changing the tone of an episode or the psychological attributes of a character. Perhaps here too he sometimes “improves,” but the process of amplifying and embroidering is continuous, and one is very reluctant indeed to conclude that Murasaki Shikibu has the worst of it all the way. The whole of the new translation is implicit comment upon the process. For explicit comment, it will suffice to note that the new translation, which may be called complete, contains fewer words than the boldly abridged Waley translation. What this fact says, essentially, is that however wonderful may be the effects which Waley achieves, and I for one have always found them to be wonderful, his rhythms are rather different from those of the original, which is brisker and more laconic, more economical of words and less given to elaboration. If it should be the aim of a translation to imitate the original in all important matters, including the matter of rhythm, then it may be said that the translation offered here has set itself a fuller set of aims than did that of Waley. How full have been its successes is for others to say.
Robin Buss (The Count of Monte Cristo, 1996 Translation)
On the broader question of translation, I have tried above all to produce a version that is accurate and readable. A great deal of nonsense is written about translation, particularly by academics who approach it either as terrain for theoretical debate or, worse still, as a moral issue: ‘the translator must always be faithful to his original,’ Leonard Tancock wrote, oddly assuming that translation is a masculine activity, even though on this occasion he was prefacing Nancy Mitford’s translation of La Princesse de Clèves (Penguin, 1978). ‘…he has no right whatever to take liberties with it…Nor has he any right to try to smooth the reader’s path by the omission of “dull” bits, short-circuitings, explanatory editions, radical transferences or changes of order.’ Why? And who says? Is it the reader who is demanding this perfection, this absence of explanatory additions, and so on?
Such academic theorists insist that translation must read like a translation – it is somehow immoral to conceal the process that has gone into making it. ‘Ordinary’ readers usually demand the opposite, and reviewers in quite respectable papers sometimes show little appreciation of what the process means and involves: ‘Not all of this material works in translation,’ said one serious review of a book by Umberto Eco; and another: ‘…the stories [of Viktoria Tokareva] are well served by their translator, who hardly ever gets in the way’.
In philosophical terms I am quite willing to admit the impossibility of translation, while still having in practical terms to engage in It and to believe that everything must, to some extent, be translatable. I feel no obligation to avoid smoothing the reader’s path and none, on the other hand, to ‘getting in the way’ from time to time. Above all, I want to convey some of the pleasure of reading Dumas to those who cannot do so in the original language and, through my one, particular version (since no translation can ever be definitive), to reveal aspects of his work that are not to be found in any of the other existing versions. This is a new translation and consequently a new interpretation of a great – and great popular – novel. If nothing else, most people would surely agree that it is long overdue.
Breon Mitchell (Tin Drum, 2009 Translation)
The most common question I was faced while working on this new Tin Drum was, “What was wrong with the old one?” This question reveals a fundamental misunderstanding about the nature of literary translation. It is precisely the mark of a great work of art that it demands to be retranslated. What impels us toward new versions is not the weakness of existing translations, but the strength and richness of certain works of literature. The works that are never retranslated are those we only care to read once.
We translated great works because they deserve it—because the power and depth of the text can never be fully revealed by a single translation, however inspired. A translation is a reading, and every reading is necessarily personal, perhaps even idiosyncratic. Each new version offers, not a better reading, but a different one, one that foregrounds new aspects of the text, that sees it through new eyes, that makes it new.
Emily Wilson (The Odyssey, 2017 Translation)
There is often a notion, especially in the Anglo-American world, that a translation is good insofar as it disguises its own existence as a translation; translations are praised for being “natural.” I hope that my translation is readable and fluent, but that its literary artifice is clearly apparent.
The translator, in an interview with the New York Times about translating The Odyssey, also offered this personal experience:
If you have any other good examples to add to this list, let me know! And if you'd like to read more about translation, localization, and such, see my other articles here!
If you read a lot of classic foreign literature, you end up seeing a lot of forwards that comment on the experience. Sometimes the works can’t be translated literally because they’d be meaningless gibberish if you did, which you inevitably have the translators explaining every time.
Beyond the difficulties of translating rhyming poetry The Divine Comedy is so rooted in specific Italian politic figures of the time and religious beliefs that you need to have an explanation of who half of the people are. On top of that, you have things like how he portrayed a lot of popes in hell because of a legal document that he was upset by – which we’ve since found out is was a forgery to begin with.
Tale of Genji is another fun one, since it became unreadable in Japanese so quickly. Part of that was that it was written without kanji, and part of that is that Japanese aristocrats of the time spoke with poetic references that no one would recognize now.
“it became unreadable in Japanese so quickly. Part of that was that it was written without kanji”
Japanese is so full of homophones that trying to translate text that is pure kana is indeed a very frustrating experience.
And yet, titles like Earthbound proves that translating kana.only text is possible. All that’s needed is the proper context for the words.
The big problem is that Japanese people don’t believe in putting spaces between words.
True that, you gotta imagine the spaces yourself most of the time. At least Mother 2’s one of the few titles to actually have spaces in text, likely as part of Itoi’s mission to have the text resemble spoken speech. That aside, the most common occurrence of Kana-only speech is to convey the idea that speaker’s dumb as rocks basically. Or a robot.
“Itoi’s mission to have the text resemble spoken speech”
That can’t be the case. Spoken speach doesn’t have ‘space’ between words. You. Don’t. Read. Words. With. Separation/Pauses. Like. This. YouSpeakWordsInASentenceAsIfItIsOneBigWord. In that sense the japanese writing system is actually more coherent with how we actually speak.
Itoi probably just did it so for it to be readable and not a big mess.
Localization aside, literal translations, even when they are perfectly grammatically correct, can often be misleading. For example, generic plural nouns take the definite article in Romance languages, but the zero article in Germanic languages. So you say, “I like cats,” in English, but, “Me gustan *los* gatos,” in Spanish and, “Mi piacciono *i* gatti,” in Italian. If you translated this literally as, “I like *the* cats,” it would create the false impression that the speaker is referring to a specific group of cats rather than to cats in general.
Another example: There is an Italian film called “AlbaKiara.” It’s a terrible movie, but one of its first lines illustrates quite nicely why you shouldn’t translate literally. A girl called Chiara says, “Ero con Micky e Zippo, completamente nudi.” Word-for-word, that translates to, “I was with Micky and Zippo, completely naked.” At first sight, there seems to be nothing wrong with that. But if I asked you who was naked, you would probably answer, “Chiara.” And that is incorrect — they all were. How do we know? Because the adjective “nudi” is plural.
People usually use Japanese as an example of a language that loses a lot of nuance when translated into English, but it can happen with English’s closest relatives, too. For example, German and French both have two words for “you” – a formal one and an informal one – which English hasn’t had for centuries. To give another example, some languages’ words for grandparents always specify which side of the family they’re on, so translators into those languages might have to guess if a grandparent’s side of the family isn’t specified in the original text.
What nuances are there in English that few other languages have?
I wouldn’t say “few other languages”, but a fair number of nearby European languages don’t distinguish between the simple present (I go) and the present progressive (I am going).
In any case, one of the weirdest things about English is that it has gendered pronouns, but not gendered nouns. Most languages have both or neither.
Which languages do you mean? Romance languages do distinguish between the simple present and the present progressive. They just don’t make it obligatory to mark the progressive aspect.
French is unusual among Romance languages, in that it has no present progressive (“je vais” means both “I go” and “I am going”).
I’ve studied a bit of Dutch and Danish, and both also lack a present progressive.
Of course, we could start a whole Language Lawyer debate here and argue whether “être en train de” or “zijn aan het” could be described as equivalent to the present progressive. IMO they’re roughly similar but not perfect (hah!) matches: they’re idiosyncratic grammatical constructions that have no direct equivalent in English, much like “I am going” has no direct equivalent in French or Dutch or Danish.
Yes, I’m aware of that. I asked Kakahan44 which languages they were referring to because I was afraid they might believe languages like Spanish don’t distinguish between the simple present and the present progressive.
You’re right, that was my mistake.
It’s true that French doesn’t have a progressive tense per se, but it does often distinguish an action in the process of happening through other formulations, especially if not doing so could cause confusion.
I can’t speak with much authority about the French spoken in France, but in Québec French the formulation “je suis en train de [verb]” for “I am [doing verb]” is extremely common. “Je vais” is one case where you would pretty much never use the simple present if you mean you’re doing it right now (unless the context is very clear, e.g. someone just asked you what you’re doing right now), because it’d be likely to create confusion.
Obviously this is the case to different degrees in many, many languages that lack some common tenses, but French is one case where I think clarifications that basically amount to the present progressive tense are very common.
In English, there is an implicit tonal distinction between Germanic words and Latin words. To use a classic example, consider the difference between:
1) a cordial reception
2) a hearty welcome
They’re synonimous, but #1 uses Latin words to transmit a formal tone while #2 uses Germanic words to denote intimacy and friendliness. This tonal distinction arises from the unique blend of language and culture in England’s history (where French was considered classier than English) and is not common elsewhere AFAIK.
Now you’ve got me wondering if this is also true of Japanese, except with native Japanese words vs Sino-Japanese words.
I’ve noticed that the more formal you want to sound in japanese, the more “chinese” you use, and other japanese learners far more advanced than me claim to have noticed the same thing. Off the top of my head, I think about the difference between the phrases 見ての通り and 御覧の通り, which both mean “as you can see”, but 御覧 itself is a word I rarely see being used by people other than hotel receptionists, shop clerks, etc. I wish I could think of a better example, though.
Also, “military language” so to speak seems to use nothing but sino-japanese words with grammatical agents like particles being very scarce. Then when something unexpected happends the soldiers get so startled they forget to “speak like professionals” and switch to “normal” japanese to react to the situation, where the difference is even more apparent. At least that’s the impression I’m given.
News headlines also seems to follow this style.
Wow, that’s super interesting!
Gosh I love the instruments simile to explain translation and localization.
I know right? I never really thought of different languages as being like different instruments and how you can’t really make them sound the same even if they play the same notes.
It’s brilliant.
It also captures how masters of a language are masters precisely because they know the specific attributes of a language. Shakepeare is so loaded with double meanings and nuances that you lose them just by pronouncing them with a modern English accent.
https://www.youtube.com/watch?v=YiblRSqhL04
Or look at Franz Kafka, who starts a book with the line that would most literally translate to “As Gregor Samsa one morning from uneasy dreams awoke, he found himself in his bed into a gigantic insect-like creature transformed.” Sounds a little strange in English, but it’s built on features of German, one of which is the ability to create gigantic sentences which don’t really mean anything until you hit the end.
“features of German, one of which is the ability to create gigantic sentences which don’t really mean anything until you hit the end.”
So… like every language in existence?
No.
The entire point of that sentence is that it’s used a form where the verb is pushed to the end. The entire line is adding context to the transformation, but we don’t know the transformation has happened until the very last word.
Not all languages tend to be equally concerned about run-on sentences, for example.
The verb is not “transformed” — that’s a participle — but “found,” which is in the second position of the main clause, where it belongs.
Also, the term “run-on sentence” refers to several independent clauses separated by commas rather than periods, exclamation marks or semicolons. It has nothing to do with how long a sentence is.
Actually, the line literally reads, “found he himself,” not, “he found himself,” so the verb is actually in the FIRST position of the main clause.
To me, it’s like saying that English sentences don’t make sense until the end, where the object is placed. Note that Subject-Object-Verb order is the most common on Earth.
This is why people (especially the usual) who insist on “super-literal” translations (exaggerating a bit for effect, but the general point’s the same) irritate me.
I think it’s why I find the translations of Pokémon Black/White and Black 2/White 2 to be so off. It seems like the new translators (the old one was kicked out after Platinum) favoured this more literal approach, and as such, various texts (dialogue especially, not good for a supposedly story-focused game) sound really stilted.
The older one was kicked out after Pokemon Platinum after a particularly disastrous blunder. He was a member of Something Awful, a website that’s anything but politically correct, taking the suggestions of his forum peers to insert lines and forum incident references in a children’s game that has nothing to do with it.
That speaks to a complete lack of professionalism, and his antics were a liability to the game. (The recent Pokemon leaks provided an interesting insight into the localization process, such as when the game grew in popularity it became an attractive litigation target, so they had to change even mentions of Boys Scouts in Gen 2… And now defamation suits?)
Moreover, after Nintendo Treehouse’s work on Fire Emblem Fates, a notoriously awful localization, where under the guise of “cleaning it up from fanservice” they respected neither tone nor much of the source material, and were clearly overselling their skills as scriptwriters who can “improve the shitty FE:if story”… and in a stunning display of history repeating itself, one of the localization editors, an ex-Atlus guy at that, bragged that he was consulting with a tumblr blogger who disseminated fake news about the original Japanese version containing RAPE scenes, and inserting her lines to change various scenes other than the ones he suggested.
What happened after that, that The Pokemon Company went ahead and put out a PR announcement that “Treehouse isn’t involved in the localization of Pokemon Sun & Moon.” No, not “TPC is handling localization”, but outright distancing themselves from Treehouse. Then NCL announces they’re becoming more involved in the localization process, and suddenly Nintendo games no longer get visual “Japanese cultural references” removed from localized versions anymore.
Translation implies some minimum degree of trust between the audience, the original creator, and the translator.
One could say “what’s so bad about a little bit of fun here and there with the script?”
But it’s getting so bad these days that the practice of localization as a whole is negatively perceived. Where localizers approach scripts in a game jam from the viewpoint “What should I remove in here? I already have a joke/reference, how do I stuff it in here? I don’t like this plot event, in the trash it goes. I hate this character, his lines are scrapped and he’s now my self insert.” It’s so consistent and frequent, and normalized, that localizers EXPECT changes and random references in their output or other localizer’s outputs (about that, so much for respecting translators… Some have this weird herd mentality where any translator not following their practices exactly is a hack, or a “socially irresponsible translator”, even with a client and audience who are satisfied with the work.) Game localizers openly cite Working Designs as their inspiration.
And with the sheer absurdity of games like Yakuza 4 Remastered (American version specifically) inconsistently removing rape references from an M-rated game about organized crime and murder, references that were okay in 2010 when “Sega’s more censorship-happy localization team” (in the words of the new localization team, which uses a hostess mode removal in Yakuza 3 to prop up their own work as the superior “uncut” version!) you can see the worst of the fan’s fears that “localization = 4Kids” is actually becoming true.
Video games were an immature medium, and sadly the unprofessional conduct of many North American localizers, either their work or their post-mortem comments about how awful the stories, the culture, the developers, even the audience are, and how good it is that they rewrote it on their own behind the backs of everyone; is perpetrating the idea that video games have no worthwhile stories to provide, and are at their best when used as an educational children PSA or a vehicle for the translator’s dubious jokes. This attitude is holding back the medium, no exaggeration.
Yeah, I agree. I’m not some literalist purist who wants everything to be a 1:1 translation with no creative license and no attempt to make it sound natural in English, but I’m just so much more suspicious nowadays when I know localizers approach it with attitudes like “this is boring, I’m going to make things up to spice it up” or “I disapprove of the content so I’m going to ‘fix’ it.”
I also can’t stand every JRPG translation trying to be a pale imitation of FFXII. The original intent of FFXII’s localization was to use accents and different speaking styles to preserve the different speaking styles in Japanese. But the imitations have the opposite effect where everyone speaks in the same overly formal pseudo-antiquated British accents that just makes everyone sound exactly the same and, to a modern American, adds a such a general aura of stuffy formality that even characters who are meant to be casual and unpretentious in the original come across as dignified.
I get that preserving the tone and meaning of the original can involve massively rephrasing things, but there’s a difference between saying essentially the same thing a different way and just making crap up because you want to “fix” the original. It’s such an insult to the original writer. If someone decided to “improve” my work by throwing in a bunch of random crap I never intended and taking out things they didn’t like, I’d be pretty upset.
That said, as with all writing, you can get away with almost anything if you do it well enough, but it’s so rarely well done.
What inappropriate content was put in? I just remember stuff like “the karate I learned on the internet”
It is somewhat tiring to read the n-th iteration of Intro to Loc injected into a debate that is not actually about the existence of localization, but its quality and execution. Localization is a matter of necessity, but there are levels to this game, and no one is helped by the compulsive shuffling of people participating in the debate into one of two binary camps.
On that note, I miss translators providing constructive criticism of other translators in the (online) public sphere. I refuse to believe they are all part of one happy confirmation bias bubble and assume they have other channels to voice their criticism – but having informed public voices dissecting controversial localization efforts would immensely help the discourse. Where are the Seidenstickers to the Waleys in Japanese pop media?
“The older one was kicked out after Pokemon Platinum after a particularly disastrous blunder. He was a member of Something Awful, a website that’s anything but politically correct, taking the suggestions of his forum peers to insert lines and forum incident references in a children’s game that has nothing to do with it.”
From what I gather all he did was insert in-jokes from the forum and name a couple characters after some guys LP, not that bad. That said, the reason he was let go had nothing to do with SA but because TPC switched to localizing games internally and Nob was a freelancer. What it is that made TPC decide that is unknown so I wouldn’t put conjecture as fact. https://www.pocketmonsters.net/articles/315
“The Pokemon Company went ahead and put out a PR announcement that “Treehouse isn’t involved in the localization of Pokemon Sun & Moon.” No, not “TPC is handling localization”, but outright distancing themselves from Treehouse.”
This is just blatantly false. https://nichegamer.com/2016/02/26/rest-easy-the-pokemon-company-is-localizing-pokemon-sun-and-moon/
I didn’t follow any of the Fire Emblem stuff when it was happening so I can’t comment on that. Sources would be very much appreciated.
I’ve been playing White 2 recently and I noticed how stilted and full of translation-isms the dialogue is.
Actually, when I’m playing, I keep commenting on how “genki” seems to be translated literally as “energy” every time it appears, and while it’s not wrong per se in English, it’s just bizarre how frequently people say stuff like “your Pokemon are looking energetic!” or something like that because “genki” is used in so many more contexts in Japanese than “energy” is in English.
It’s never a good idea to ignore how words can have many definitions. I like how the word “get” has more than a dozen distinct meanings.
I think Ken Liu’s comments on his translation of The Three-Body Problem by Liu Cixin would be a worthy addition to the page — though I don’t remember the exact wording, the comments were definitely of a similar sentiment to those added so far and I know I quite liked them (I remember one specific comment about how the best translations into English do not read as if the text was originally written in English, but instead convey a sense of the original cultural context).
I just remembered about my French version of “Through the Looking Glass”, one of my very first look at what translation is all about.
There’s not only have a translator’s note justifying why it’s a localization, but also an annex with a direct translation of every single pun in the book.
Just too prove why a litteral translation would not work.
Every single one? No sir, only the ones on the surface.
But that reminds me about one good practice: annotated editions. For in the end, aren’t they but a translation from a language to the same language?
It’s one thing to disapprove of using the generic ‘his’ instead of ‘his or her’, but Buss errs in claiming that Tancock *assumes* anything. Pretty worrying error for someone whose role involves a close knowledge of English usage.
She was mocking him by taking his words as literally as the translation style he prescribed.
Coming in about a year late, but I found the following interesting:
“If my count is correct, at least seventeen translators have preceded me in translating one or more of the fictions of Jorge Luis Borges. In most translator’s notes, the translator would feel obliged to justify his or her new translation of a classic, to tell the potential reader of this new version that the shortcomings and errors of those seventeen or so prior translations have been met and conquered, as though they were enemies. Borges has tried in his essays to teach us, however, that we should not translate “against” our predecessors; a new translation is always justified by the new voice given the old work, by the new life in a new land that the translation confers on it, by the “shock of the new” that both old and new readers will experience from this inevitably new (or renewed) work. What Borges teaches is that we should simply commend the translation to the reader, with the hope that the reader will find in it a literary experience that is rich and moving. I have listened to Borges’ advice as I have listened to Borges’ fictions, and I—like the translators who have preceded me—have rendered Borges in the style that I hear when I listen to him.”
Andrew Hurley, from Collection Fictions of Jorge Luis Borges
I can see at least two reasons to retranslate even when the first attempt was fine:
– Different goals for different audiences and circumstances. To go back to the first example cited in the comments, “The Divine Comedy”, the version I read had so many notes explaining who the characters were, the puns in the original Italian etc. that they had to put them as footnotes rather than at the end. Despite using a smaller font, they took about 40% of the pages’ surface… That was fine by me, but someone else could prioritize a smooth reading experience with fewer interruptions, or the book’s portability (the version I read was massive).
– Translations age, too. A few months ago I read “One Day in the Life of Ivan Denisovich” by Aleksandr Solzhenitsyn, in an edition published in 1973. Since the point of view character comes from a humble background, he thinks with a lot of slang (at least in that French translation). Some of it is still in use, some I knew from old movies, some had become so obscure I had to look it up…
The Divine Comedy is probably the worst example to use. Dante is writing for a specific audience from a specific time, and no translation can work around the fact that the reader is supposed to have a lot of specific pre-knowledge about subjects Dante isn’t going to explain in any kind of detail. You’re either going to get page up and page down of footnotes explaining what Dante is talking about, or you’re going to get an incomprehensible mess of references to people and events you’ve never heard of. It just isn’t a book meant for some random guy in 2023 to sit back and relax with.