When it comes to media making the transition from country to country, games are without peer. More are localized into English and marketed as mainstream than either books or films. How many chart topping films or books originate in a non-English language? According to Metacritic, far fewer than those found amongst videogames. Without a doubt, localization plays an important role the modern game industry.
The localization industry started when Japanese companies began creating games with more complex and dialogue-heavy plots for systems like the Famicom, Mark III and MSX. But it was only with the 16-bit generation, on the SNES, TurboGrafx-CD and Sega CD, that localization really entered public awareness.
The genre at the frontline has always been the RPG. But Japanese companies were originally reluctant to release domestically-produced RPGs in the West. As Jeremy Blaustein recalls of Konami during the early 1990s, “Originally the perception was, ‘RPGs will never be popular in America’. Common wisdom was that RPGs were for the Japanese and action games were for the West. I thought all along that was bullshit. The problem was one of translation; the players just weren’t getting it, and if they got it they’d love it.”
Plugging this hole required small publishers willing to take risks. The first of many was Working Designs, who published games for NEC’s TurboGrafx systems and the Sega CD. Loved and hated for their methods, they brought over games that would have otherwise been ignored. With cartridge based games, the leader was developer and publisher Square, but they were more cautious about which RPGs saw release in the U.S.
When Final Fantasy IV eventually reached America, Square was disappointed with the sales. Ted Woolsey, formerly of Square, explains: “Back in the 1990s, Japanese developers were frustrated they weren’t seeing acceptable sales in the U.S. At the time anime wasn’t as popular, and it was thought that the look of Japanese games – combined with the fact that Americans didn’t have earlier games as examples, as Japan did – would mean they wouldn’t be familiar or satisfied with a game’s content. To fix this, it was suggested that something similar to U.S. comics, including scatological humor and other familiar references, would resonate better.”
Square’s desire to cater specifically to U.S. audiences led to the development of Mystic Quest. It wasn’t the success they’d wanted, and based on this disappointment, Square canceled the American release of Final Fantasy V, despite Woolsey nearly finishing the translation. In Square’s view, the U.S. market simply wasn’t ready. Later, Square again attempted to cater to the U.S. with Secret of Evermore.
When a foreign title does make the journey west, there is a huge amount of work to be done, as shown in Atlus’ fascinating production diaries. There are also plenty of potential snags. In the 16-bit era, localizers ran up against challenges like non-contiguous text and a lack of ROM space. Woolsey remembers that for Secret of Mana, “Probably 40 percent or more of the text was nuked – there just wasn’t space. Story elements, nuance and personality had to be stripped out. It was, in some ways, the hardest game I’d worked on. I loved that game, but am probably most dissatisfied with the result. Certainly tried my best, but that thing nearly killed me.”
There are other challenges as well: One person trying to make hundreds of characters sound unique (as with Suikoden); dealing with dialect differences, because an Osakan style of speech can’t necessarily be replaced with a Southern drawl; and doing a ton of background research. Blaustein recalls that “you could find the best translator and he will do a terrible job. Why? Because you have to do research, maybe in the field of dragons and knights, and if you care enough, you’ll look into European history and find out how things reference each other. There’s games where I had to go deep into the arcane issue of the five element theory, Chinese Yin Yang and Shintoism. One involved so much Chinese mythology, it was practically Chinese itself.”
Beyond the obvious requirement for natural-sounding dialogue, coherence and respecting the original creators’ vision, everyone has their own interpretation of what makes a good localization. Generally everyone agrees on the need to create a parallel experience for English-speaking players, but as Blaustein reiterates, “People mistakenly think if you translate the words or names directly, somehow you’re going to be conveying more of the original essence.” He explains how certain words have an exotic echo to the Japanese ear; to replicate this in English, the sound needs to be changed, such as when Woolsey renamed Tina in Final Fantasy VI to Terra. Blaustein agrees with the decision: “I’d say Ted was right on the money with Terra. I had this exact conversation with Kojima on Metal Gear Solid. We were talking about characters, and I said, for a Japanese person it has a foreign quality. ‘Decoy Octopus,’ however you slice it, sounds hokey in English. Looking back, it worked, but not for the same reason the Japanese worked – to them, it comes across as impenetrable.”
Translation also needs to be seamless. Leaving some things unexplained alienates those unfamiliar with the culture. Bill Swartz of Mastiff Games is critical of sometimes lazy localizations, using the kappa, a mythological Japanese creature, as an example: “I would argue that if you were to just say ‘kappa,’ then you haven’t localized it. You’ve done a mish-mashed job which is going to be dependent on pre-existing knowledge. If it’s there seamlessly for the Japanese because they have a knowledge of what ‘kappa’ is, then it’s incumbent upon you to either explain it or insert a word so that it’s seamlessly apparent to the target audience.”
As Woolsey explained, the Japanese can have a different method of conveyance. “Some Japanese writers … link themes, ideas and images together in ways that are less story driven, and more suggestive or lyrical. There are times where a translator, without access to footnotes, needs to pad a bit to get a point across. Some central themes in RPGs, such as filial piety, duty and obedience to teachers, elders and authority don’t play quite as well outside of Japan, and need additional verbiage to explain.”
Sometimes it’s not just the script but a game’s overall structure which needs to be changed. “The Japanese version of Tenchu was a fantastic concept with some technical issues,” says Swartz. “Working with the original developer we added levels, fixed camera issues, frame issues, did a bunch of cleaning-up and created an amazing product.”
And there are times when change is crippling. Working Designs infamously rendered Exile 2 and other games impossible due to increased difficulty. As Working Designs explained at the time, this was due to the liberal return policies in the U.S. and the concern that leaving the game as-is would result in a large portion of “extended rentals” at their expense.
Fandom for Japanese games is often intense, and while some changes during localization are for the best, fan expectations can limit a company’s options. Swartz recounts an interesting story while working on La Pucelle: “There were a couple of scenes where there were these cross-like structures, and the whole thing was the most incredible temptress in a teapot, in that we were accused of censoring it. We even ran the screenshots side by side. But some random fan got hold of it and it turned into a firestorm. That to me served as a reminder of how sensitive the hardcore market is. I read a lot of the postings, and people really spent time, and there’d be one guy with some Japanese trying to rate our translation.” Despite the scrutiny, Swartz doesn’t regret the experience. “It was amusing and gratifying. Mostly gratifying, because we spent a ton of time on that. It wasn’t like you get the assets and pass it to the translator and then he comes back and you’re done. It was hundreds of hours of my time, of the producer’s time, sitting around the table, reading stuff out loud, asking ourselves ‘does this express motivation’ and so on.”
With negative fan reactions as intense as they are, what happens when a localization isn’t liked or, worse still, doesn’t happen at all? Oftentimes, the fans do it themselves. The first landmark translation was in 1993 with SD Snatcher on the MSX, by a Dutch group called Oasis who sold the work at MSX fairs. Years later, the rallying cry for internet-based groups became “FFV!” which led to a scene explosion and countless translation patches, even on systems like the Saturn and PlayStation 2. While these are scattered across the internet, many are collected on sites dedicated to ROM hacks.
Professional reactions to fan work are mixed, notes Woolsey: “I’ve been amazed. Some have been quite good, done by sharp people who know Japanese and how to write. Some have been almost comical. I recall one site, with some fairly vocal contributors, who were advocating translating using Japanese grammar (subject, object, verb). They claimed to have the most faithful translations, but when you read them it sounded like Yoda.”
Even so, Blaustein also says that professionals can benefit from the fan community’s work. “One of the most shocking things in my career has been to see the extent to which fans go. On any number of occasions I’ve been given the task to translate a game, I’ve gone on the internet to gather my research, and how many times do I find that a fan has already played the Japanese game, done a translation and put up JPEGs of everything? It’s absolutely frightening, because they will spend more time than I am allowed for the official translation. Sometimes I may even use their website, because if it’s a series, and I jump in at two or three, I need to go back and learn the series’ history. Phoenix Wright is a perfect example. I relied on this incredible fansite where he showed items – I needed to know, when translating, what does that thing look like? I’ll find out things[from sites] that the client should have told me.”
With the resurgence of classic games via download services like the Wii’s Virtual Console and the PlayStation Network, it raises the question of whether fan translations are harming the possible release of previous Japan-only titles. A few have come out, but why aren’t more available? The consensus among publishers was that it comes down to expense. “The juice has to be worth the squeeze,” admits Swartz. “It’s certainly possible to localize games for download. There’s no inherent reason why it couldn’t be done. But it goes without saying that a good localization is an expensive, time-consuming operation, and if you haven’t seen a lot of it yet, it’s because the business doesn’t make sense.”
Blaustein offers an idealistic solution; “It’s cost prohibitive … but nothing would lend itself better to input from the fan community. It’s perfect! Have contests; harness the power of the fans who would do it happily.”
Some might argue, however, that fan translations are no longer needed – that this is the golden age of localization. So what does the future hold, then, with growing budgets and an increasing divide as Japan becomes isolated through Western developers’ success? Swartz is optimistic: “There’s no doubt the Japanese industry is in decline – it’s both consolidated and shrunk, so there’s fewer games in total, and games are coming from fewer publishers. So the industry is definitely getting more American- and more European-centric. [But] there are teams and groups that put a lot of effort into it. The bigger companies routinely produce wonderful localizations. I think overall as games have moved more and more mainstream, quality has gone up, and really good localizations have gone from being pretty rare to being pretty common. These are interesting times.”
John Szczepaniak is a South African-born copy editor for a Time Warner subsidiary. Currently he’s disappointed no one is localizing Yuusha no Kuse ni Namaiki or Ryu ga Gotoku Kenzan.