The issue of translating (and peripherally, transliterating) has weighed on my mind ever since I undertook a certain school project, translating the first chapter of a Classical Chinese novel into contemporary English. This led to two main problems: the issue of transliterating character names; and how faithfully to preserve the rhetorical structure, since it differs so wildly from English literature.

But before we go further, we should address the (annoying) lie of “we don’t have a word for X in our language, and therefore those people are cognitively impoverished”. From my Linguistics 101 lecture notes:

Everything in any language can always be translated into another language. There is no lexeme that is beyond the cognitive capacity of any normal adult. Translations can only differ in their brevity(length), and elegance(structure).

So sure, maybe some languages have no word for snow, especially if those languages are spoken by equatorial tribesmen. But that doesn’t mean that you can’t explain it to them. Even without videos or photographs, it is possible to describe snowflakes and snowploughs, and snowmen and snow angels. You could say “it’s like white sand, but very light, and very cold”.

Since Chinese employs tones, and an incredibly complex logographic writing system, projecting all that complexity into an intonational alphabetic writing system fails to preserve and reflect the range of surnames. “Chou”, “Chu” and “Chao” may as well be the same person; no English reader is going to easily remember the distinction between these three hypothetical characters. In order to better disambiguate them, I was tempted to actually translate them into English. After all, surnames like “Green” and “Forest” already exist in English. However, if I did that too well, I ran the risk of making the work sound like it was ‘set in merry ol’ England,’ in the words of my supervisor. So, begrudgingly, I was forced to stick with the dull and phonetic transliteration of the people’s names.

The second issue is a little more complicated, because although there are established conventions within both Chinese and English literature, each author is also unique and brings something of their own style to the mix. The question, then, was how (and whether) to bring this uniqueness to conventional English prose, or whether to simply preserve the rhetorical conventions of the work. In the interest of time, I ended up basically treating it like a translation rather than a localization project. So my work ended up reading a lot like a translation. You could just hear the Chinese accent behind the [grammatically correct] English prose.

That’s not to say that someday I might not undertake the task of doing a reverse-Kurosawa, and genuinely translate the entire work from Dynastic China to Imperial England. But for now, such a project is extremely low on the list of grand dreams.

More recently, however, with my continued studies in Japanese (and my persistent guilty pleasure of scanlations), the aforementioned issues still persist. In the case of certain stories (e.g. historical dramas) there’s really no way of getting around the issue of presenting the fact that you’re set in a time and a culture that’s far removed from most of the English-speaking audience. And yet, the task of the translator is to gloss everything, without relying heavily on footnotes in order to explain the culturally appropriate exchanges that differ from our own. After all, even on a basic level, the “thank you – you’re welcome” pair is extremely opaque, and such translations into other languages (“arigatou – douitasimasite”) don’t really match, on a semantic level. However, it’s the functionality of these phrases that really makes the difference. Sure, “arigatou” has a more literal sense of “I’m having difficulty”, but that doesn’t stop any native (or indeed non-native) Japanese speaker from understanding that utterance as a conveyance of gratitude.

In a way that relates to my last two posts, the phrase “ore no kachi da” is literally translated as “[it’s] my victory/win”. However, by an overwhelming margin (~70%) “I win” is the preferred English utterance to express one’s victory immediately after a tournament/trial/fight/etc. Incidentally, from what my Japanese native speakers tell me, 「俺の勝ちだ」 is more natural than 「俺が勝った」. Since translation goes both ways, adopting a literalist policy would result in reinterpreting “I win” as “it’s my win” — the exact same kind of disparity that the Japanese would find. (Too bad you don’t see scanlation comics of like, DC/Marvel on Japanese sites. That would have been some nice linguistic fodder to pad my data set.)

Of course, there are times when such glib glosses are impossible, such as when two characters are explicitly deconstructing the disparity between the functional and literal meanings of their own culturally embedded phrases. Or more commonly, when they’re punning, which is inherently predicated on the (phonetic, syntactic, morphological, etc) structure of their language. In such cases, it’s really hard to find comparable puns that also suit the meaning and function in the target language. And yet, what else can the diligent translator do? The quick answer is, the translator underlines the problematic text, and moves on.

This just barely scratches the surface, but I hope it at least presents some of the issues involved in translation, and why it’s not as easy as just “going word for word”. (ie, for those Google Translate enthusiasts, machine translation still isn’t there yet.)

An excellent and relevant quote I encountered while reading a certain article online gets to the heart of this matter rather succinctly, written/said by David Hoon Kim, a bilingual and tricultural author:

The translation must read not like a translation, but as though the writer had written the same work in another language. I think the best translators of literature always rewrite rather than translate, all the while remaining faithful to the original text.

New Yorker Online, Q&A: Living Language
(emphasis added)

Therefore, with specific reference to the “I win” vs “my win”, the clearly better choice for English would be “I win”, since that’s what a native speaker would employ, or indeed what the original author would have used, if he were writing in English. Even if the characters are ostensibly ‘foreign’ (e.g. an English novel, set in France, with french characters), the work is never written like an ESL essay. The idea that one should ‘preserve’ the exotic nature of the source language into the target language for translation is not only infeasible, it also subverts the legitimacy of both languages.

However, as I hinted to earlier, periphrastic constructions aren’t always the most elegant answer to translating a word that has little experiential relevance to the target language. In cases like those, I don’t see why such words shouldn’t enter the language as a transliteration, with perhaps a footnote. That’s essentially how languages acquire vocabulary anyways.

But wait. Did I just contradict myself by suggesting we should introduce “snow” to Kenya, but ban “my win” from English? Not exactly. The difference is that English has a succinct and dominant expression for the same sentiment already, so “I win” is probably going to outlast “my win”. On the other hand, if you’re trying to translate something, your task is not to simultaneously introduce language reform. Stick to the conventions of your language as much as possible, and only go beyond to introducing new constructions/vocab when necessary. To me, these are the principles that make the most effective translator (modulo common sense). If you’re handling proper nouns, it’s probably best to leave those untranslated, and transliterated instead. As much as I’d like to read a translated Chinese novel with Mr Plum and Mr Forest (Li and Lin, respectively) — both which, by the way, are legitimate English surnames as well — such zealous localization can only lead to a translation that fails to retain the setting, and the inherent quality of the original writing.