Bits on Translation Projects

This afternoon, while on lunch break, a thread popped up on the Gemot about someone starting up a translation project for a game. It happens every now and again there, and on some level I’m happy that the visual novel community is generating what appears to be more and more interest in the English speaking world. I’m sure on this or that forum out there in the world, more such posts happen, but I more or less use the gemot as my home forum.

Anyways, as is often the case, the person who wants to do a translation is still learning (hey, I am too, I’ve just started walking down this road a few years earlier). As projects go, this one is a large one, a full-length modern game. That’s also somewhat expected from a new translator, since they don’t have a sense of scale and time yet. But that’s fine, we all make these mistakes, and if he buckles down and works at it, he may finish.

I hopped into the thread just to give some advice about good things to do before embarking on a huge translation project. Are these the best things? Who knows, I’ve only done this in my spare time for about 2-3 years now, hardly experience to stand on a pulpit and shout ‘best practice’ but even in that little time, I’ve made plenty of mistakes to learn from…

The first thing I suggested was to have a complete script in front of you. I mean, literally, the data file, that has all the script text, all the pointers that link them together, the whole shebang. I know it’s often the most convenient way to just “play through the game and translate what comes up!”, and you know, to start off, while you’re waiting for someone to build you the script extraction tool (or you’re building the tools yourself) then there’s not really a big problem. As a start.

But it can’t last, not if you intend on being thorough and translating the whole script. If you were making a synopsis of the game, a “walk-through guide” or something, something where you don’t have to be complete down to the last detail, then you might be able to get away with using a giant text file. Without looking at the script, you’ll never know what secret bits you might be missing.

The bigger problem is this, visual novels tend to branch, make a choice, and suddenly, “on day 3” you have 2 parallel lines of scripts, “day 4” and you might have 4, and so on. You’d have to manage your scripts somehow. “day 4 event 4 from route X”, which becomes pretty confusing if a game… doesn’t have days or other clear demarcations.

Worse, if you finish say, 1000 lines (5 words a line is already 5000 words, quite a number of pages if you printed it out by the way, and a ‘short’ 4-6-hour game like Narcissu Side 2nd is 6400 lines, while full games can easily take 10-20+ hours to play). At some point or other, you’re going to have to insert these 1000 lines back into the actual script files anyways. Depending on the engine’s script format, that can be as simple as a script replace, or as complicated as a horrific task to do (all together now) by hand.

And often it’s closer the latter.

My suggestion to work from the script was responded to with essentially “but if I translate with the voice acting, CG, and music, it helps me.” Well, there’s nothing wrong with that, in fact, I heartily agree that if voice acting, bgm and cgs are available, they offer extremely useful cues for framing your translation

For example, you see a line in the script “それ、しないわよ”, and it translates into: “I won’t do that.” But, you go back and listen to the voice clip, and the actress sounds pretty angry at the time, so, maybe you might want to consider “I’m not doing that!” or maybe she’s being aloof, and might lend itself to emphasis on ‘that’ so: “That, I won’t do.” The actress acted out the line under the director’s supervision, so it’s a better window into what the line means than just the text alone.

Depending on your ability with both languages, you can do some pretty interesting things. For example, if a key word that’s emphasized is in the front (or middle, or back) of a line, with some thinking, you might be able to get the emphasized word to come out highlighted in the sentence, like previously with the word “That”.

Is all this necessary? Not really. It’s not even feasible all the time, and sometimes shouldn’t be used. But, it adds more flavor to your writing, and I at least hope that people who understand the language just a little, can appreciate how things line up.

However, the most surprising thing that occurred on the thread, and what prompted this little editorial piece, was that he hadn’t played the game through yet and he’s translating as he goes.

Somehow, my message of “play through the game first before you translate seriously” didn’t go through. I have rather practical reasons for making that suggestion.

The first, is obvious. Why are you translating something you’re not familiar with. It’s almost like selecting a mate, while you can just go by first impressions and commit, most people would think you’re stark raving mad. Taking up a game with a few tens of thousands of lines of text, is committing to months upon months of continuous work. If you’re doing it right, you’d probably see the script in your dreams. It’s silly to say you’re committing to something and have ‘every intention of finishing’ when you don’t know. This is what drives vaporware people…

Also, that comparing line against the voice acting bit I mentioned earlier? Well, at least for me anyways, I have a good enough auditory memory to have a fair sense of the tone that a given line is spoken in. It’s not perfect, but it’s often enough that I don’t have to literally have the game running next to me as I work on a script.

And finally, if you don’t know how the story pans out, how can you include all the little details that foreshadow the upcoming plot development, and so on. Even something as simple as knowing that some lines come up over and over again and should be consistent.

This isn’t even a question of good writing, bad writing on the part of the original author. If you’re telling a story, more often than not, if you’re consciously telling it, you’ll include details that will be important later. You’ll mention a car parked by the road, if the car explodes when you walk outside again in a few hours, it’s only natural.

Well, can’t you go back and fix it all when you’re done? Well sure, but that’s probably alot to fix, considering plot changes tend to carry everything with it. But sure you can.

From past experience, you’d be crawling through at a pace that’s only a hair faster than if did it again from scratch. Why would that be? Because your English version wouldn’t be reliable, right? So, you’ll have to compare it against the original text, and then, translate it in your head at least to compare against what you have to check. Essentially, you’re redoing everything except for relatively simple things.

I suspect the source of this notion comes from a somewhat misguided notion of how translation works. Context, it’s all about context. Any language is very ambiguous, because of the least effort principle, so if you don’t steep yourself in the context, then you can’t do a good translation. Even professional translators have specializations, someone who handles electrical patents is steeped in the terminology and references for that field, sure, they can manage somewhat in a totally different field, but they’ve lost alot of that specialized edge that makes them good in an area.

Same goes for an interpreter, it might look like they’re “translating new material on the spot” but they have large amounts of experience in the settings that they work in, and use that to help make good interpretations. Dropping a casual business interpreter into a diplomatic negotiation would be a Bad Idea™.

People who first come into visual novel translation often seem to have a similar notion of how language works “The words are the same, so if I just get those right, then how can I be wrong even i I’m walking in cold!” If only it worked that way.

We translators are writers after all, we have to retell the story, not repeat the words because the words themselves don’t match our language exactly. We have to pick, and choose, and decide how to present things. That means we have to join in on the storytelling process itself. So read up on your context.


Commenting is closed for this article.