Starting with this post, as part of our ongoing interview with Carpe Fulgur, the localization team behind doujin title Recettear, we’re going to be dealing with questions regarding the game’s localization, and putting the process under a microscope to see just how the English version was approached.
Concerning the localization process, what do you see as the focus of your efforts, your rationale or motto when it comes to changing the raw text? What lines do you draw when making the decisions on providing focuses like accents, catch phrases and the like?
Carpe Fulgur co-founder, Andrew Dice: Interesting you should ask this. The primary mantra when editing the text is: “Keep it true to the work itself”. That is, whatever the text ends up reading like, make sure it makes sense in the context of the world provided in the story and that characters and plot elements are consistent throughout. A good example is Tear saying “Merde” at times; we learned from EGS that the setting of Recettear is essentially Fantasy Not-France, so several characters break the “translation barrier” as it were and drop a few recognizable idioms in French (Tear’s in particular to replace a similar “good grief” catchphrase she had in Japanese).
It added a little extra flavor to the text, and it’s true to the world Recettear provides and makes it seem a little more “real” to the player. I also work to make sure that all of the characters maintain a roughly consistent voice throughout – Tear’s very formal language which practically lacks any contractions at all, Recette’s oddball phraseology, Caillou’s occasional linguistic slip-up wherein he speaks more like someone his age “should” speak, you have to make sure all of this is consistent or the characters come across as flat.
Another example of “being true to the work” might lie in how I’d approach something like Muramasa. A lot of people (rightly) harped on the cuts made to the game’s dialogue (cuts mostly brought about by hideously inept translation), but I know some people complained that Japanese was the only vocal option in that game. To be honest, had I been in charge of the project, I’d have also left Japanese as the only vocal option even if the budget for an English cast was available.
Muramasa is already so incredibly drenched in classic Japanese culture, that it would actually feel right, I think, for the voicework to remain Japanese, almost like watching a subtitled Kurosawa film or somesuch. Many people interested in Muramasa would already be looking for that very Japanese cultural experience, so it’d be worth it to remain as true to the spirit of the work as possible.
Occasionally this requires a bit of creative editing as well — a lot of Recette’s sayings like “Capitalism, ho” or “Yayifications” are, technically, inventions of the localization, but they were done in order to get Recette’s boundless enthusiasm for her job across the language barrier. She absolutely loves what she does, and she drops sayings like “Yatta!” constantly, but just translating it “straight” doesn’t really get her voice across properly. She pretty much needs the “quirkiness” to get her character across in English; without it, she doesn’t come across correctly and reads far too flatly (and yes, we tried it).
Now, obviously you need to apply some logic to all this; we don’t want everyone to have OUTRRRRRAGEOUS French accents, for example, as the text actually needs to be legible. And naturally you have to be real careful to not change any plot elements — you’re bringing the game over in part because you like the plot in the first place (or we hope, at least), so as editor you need to strike a delicate balance of making sure the script is engaging in English without actually mangling anything.
Above all, though, it needs to be true to the spirit of what you are working on — it needs to make sense internally to the person experiencing it, it needs to sound consistent, and it needs to sound authentic.
How much research did you have to put in to try and cast the characters correctly for the localization process? Can you touch upon what the localization process is like for a text-heavy game like Recettear?
There was quite a bit of back-and-forth with EGS about various characters. Recettear’s setting (which is shared with Chantelise and the upcoming Territoire) has quite a bit of backstory to it, and we wanted to make sure we understood the character histories and the like as we went ahead and crafted what each character sounded like in English. There wasn’t a lot of external research — EGS ultimately defined the characters, after all — but we did talk over a lot of things with EGS to ensure we understood the characters and their world as best we could.
The way we approached Recettear was to break the game into chunks for raw: do one block of related text, then the next, then perhaps the item list, then the next event block, etc., then do a rough draft of the localized script for each, and once the whole was complete we’d go back and look over the script again, and see if any changes needed to be made to help the various parts flow together or connect better. It helped that the Recettear text is organized into major events and arcs internally, so we were able to organize things internally fairly quickly.
Beyond that, you simply take it one step at a time — work on the script, play it through, read it, keep the rest of the script in mind and what this scene or that item needs to do in the wider picture. EGS was hugely helpful in that they provided us with several of their dev tools that allowed us to edit individual scenes and then immediately reload that scene to view the changes (and do similarly for Recettear’s massive item list) — I can’t imagine trying to do this project without it.
Stay tuned for more talk of Recettear’s localization over the next few days. Next time, we’ll discuss translating humour and Carpe Fulgur’s future plans.