Literal and Liberal, Conscious Translation Style Choice

Literal translations are bad. Liberal translations are bad. Fidelity is good. Fluidity is good. Anything is better than nothing. Nothing is better than crap. Just about everyone has their own conception of what’s “acceptable and proper” for a translated work. But let’s just step back for a moment, and figure out where the major positions are coming from, and then try to figure out where we want to stand, when.

My feeling is that the exact method and style of translation a translator adopts depends primarily something that people seem to make a blind assumption about, the intent of the translator. Different types of texts, in different settings, value different styles of translation and it’s the job of the translator to adjust themselves to that setting. When given a choice, the translator must then pick what sort of relationship they want to have with their audience. Do they intend to create something accessible or something useful for rigorous examination, and so on. It’s all about knowing you have a choice, and deciding what’s best at the moment.

But, for the sake of illustration, let me first sketch out in very broad strokes the two ends of the spectrum that most of us play in, with different shades of meaning.

“Literal is bad”

“Literal” translation has become almost an insult in some places. It accuses the translator of sticking too close to the source text (ST), creating something that is difficult (or impossible) to read in the target text (TT). Perhaps some of this comes from the fact that people who know nothing about translation often think this is how translation is done, “translate word for word, and you have a translation. Duh?”

But putting negative associations aside, the notion of “formal equivalence” can be roughly summed up by saying that the translator makes the two texts, ST and TT, equivalent, with respect to their formal features. Pushing the characterization to the point of absurdity, it means bringing the ST over word for word, and if the TT sounds utterly bizarre and like aliens speaking, that’s just how it is, learn to live with it.

In reality, people who closer to the literal side of the spectrum still value readability and all that, but their intention towards the work dictates that they sacrifice the “curl up in a chair by the fire on a lazy afternoon” potential of the work in favor of rendering the ST accurately.

You could say that a literal translation provides more insight to a work for someone who is already familiar with the language and culture of the original. By preserving as much of the original, someone who has some familiarity can get more out of it than from another work.

“Liberal is bad”

Taking the opposite approach is the functional/dynamic equivalence position. I’m simplifying the world to the point of caricature, but the if you were to imagine formal equivalence as taking a very hard look at the words in the ST and forcing the words into the TT whether the target language screams or not, then dynamic equivalence is reading the ST, figuring out what the meaning and intent is for that text, then rewriting it in the target language.

And so on this far end of the spectrum, we can get the accusations that the translator has injected too much of their own views and interpretations into the text that weren’t in the original. After all, is there only one meaning to every piece of text? Is it possible that in selecting which meanings to bring over, some important ones were left out, or new ones added in? Finally, in the case of us amateur translators, have they misread the original to begin with?

Closer to reality, the range of possibilities for dynamic equivalence tends to be really wide, so it’s hard to say anything about it in practice. Perhaps the only thing I can say is that it can be used as a way to cover lack of ability to understand the original, by rewriting what’s going on from imagination. However, whether you’d like to call that a translation or something else entirely is another matter.

Just as going over the top with literalism can seem bad to a reader, going too far with paraphrasing can also seem bad to a reader, and ultimately, it’s the reader who decides if something is “good” or “bad”

Floating in the spectra

But great, so we have the exaggerated endpoints of the spectrum painted, everyone has an idea where they are, but that doesn’t help us decide what to do for the text we have sitting in front of us right now that needs translating.

This is why I mentioned “intent” earlier. Whether you’re conscious of it or not, you execute your translation with some kind of intention. It might be something as simple as “make it sound as if this were an original work in this language” or something as complex as “show how the very structure of every line is deeply related to the others in each stanza”. In the end, there’s two things to be considered, how much of your audience shares your intent (and so are looking for the things you did), and how well you accomplished what you intended to do.

Bringing things into focus here on visual novels, here are some things to consider.

How much cultural knowledge expectation do you have for the audience?

Do you think some peculiar language is a detriment to the enjoyment/use of the work?

Do you think that the text has merits simply as text itself, such as rhyme, meter, length, or word order?

Would you rather bring the text to the reader, shielding them from cultural/linguistic differences, or would you rather bring the reader to the text, making them aware of the differences.

The list of course goes on and on, but as you start thinking about such things, you should be able to figure out what sort of intention you want to bring to your work.

As for myself, I’ve been characterized as being something of a literalist, but not quite. I believe that in the works that I’ve done, there is artistic value embedded in the order that phrases and ideas (not simple words) come in. If a term is mentioned, I’d like to keep it in, if something isn’t mentioned, like perhaps gender or number, or even tense, I’d like to avoid it if possible. All this even if English squeaks a little here and there from the strain. It’s difficult to do, and I get my enjoyment by trying to come up with something that fits as closely as possible. It also tends to take me longer since I’m busy considering alternatives where a simple and obvious rewrite would do.

More rambling in roughly this area of thought to come sometime soon…


  1. clarious
    12/02/2008 08:40 PM | #

    Interesting, I have thought about this for a while.

    So in your opinion, which style is better for translating a visual novel? I think the main purpose of translating a visual novel is to let the reader to understand and ‘feel’ it, that mean we must retain all the original intention of the author, and choosing translation style is like choosing a method to do it.

    Oh, and about something unrelated, I just read about insani al|together 2008 in which they said they reject the idea of grading translation quality. Any comment about that?

    As you may have noted already, I am not a native english speaker, but I hope you still can understand what I write :)

  2. 12/02/2008 10:53 PM | #

    Mmm. I’m still exploring the options and getting a feel for the landscape. However, one thing I do feel strongly is that there’s no “one” way. All the things you listed that a translation should ‘do’ are fine, but the question at hand is how you go about doing it.

    All translators need to make concrete choices, to make a sentence work here means breaking something there. Languages don’t map 1:1, this is an unavoidable fact. Literal or liberal, it all breaks things everywhere.

    So pick your priorities, be very clear about them to the reader, then go do the best you can with them.

    —As for ratings, I’ve got enough material for another article, so I’ll just say, ‘wait until next time’

Commenting is closed for this article.