Video game localization
Video game localization is the preparation of video game software and hardware for sale in a new region or country. Although translating the text is a large part of localization, the process includes any changes made to a game, including altering art assets, creating new packaging and manuals, recording new audio, transforming hardware, cutting out whole portions of the game due to differing cultural sensitivities and/or local legal requirements, and even adding sections to replace cut content.
The decision to localize a game relies heavily on economic factors, such as the potential profits that could be made in a new country. As such, the process is usually undertaken either by the game developers themselves or by a third-party translating company, though unauthorized fan localizations can occur if a translation is poor quality or if a game is not going to be released in a specific language. As an industrial field, localization is still in development and lacks consistency in terms of implementation and importance. Gathering information about industrial localization practices can often be difficult because of the lack of consistency between companies, as well as non-disclosure agreements many translators have to sign.
The goal of localization is to create an enjoyable, non-confusing play experience for the end user by paying heed to their specific cultural context while being faithful to the source material. The suspension of disbelief is of utmost importance to the process; if a player feels as though the product was not meant for them, or if the localization creates confusion or difficulty in comprehension, this may break immersion and disrupt the player's ability to continue the game.
The founding concepts of game localization can be seen early in videogame history, as in the case of the localization of Pac-Man. The original transliteration of the Japanese title would be “Puck-Man,” but the decision was made to change the name when the game was imported to the United States out of fear that the word ‘Puck’ would be vandalized into an obscenity. In addition, the names of the ghosts were originally based on colors - roughly translating to “Reddie,” “Pinky,” “Bluey,” and “Slowly.” Rather than translate these names exactly, they were renamed to Blinky, Pinky, Inky, and Clyde. This choice maintained the odd-man-out style of the original names without adhering to their exact meaning. The change in cultural context between the two countries provoked a change in the game text that was not a precise translation.
An important concern for early localization was the limited amount of processing space available to house text strings that were longer than the originals, as was often the case with the NES and SNES. Ted Woolsey, translator of Final Fantasy VI, recounts having to continually cut down the English text due to limited capacity.
Often the budgets and production times for localizations were short, resulting in translations that were either confusing or entirely re-written. Early translations were sometimes “literally done by a 'programmer with a phrase book'.” For instance, the original translation for the Sega Genesis game Beyond Oasis (original Japanese title, Story of Thor) was discarded by the English editor because it was nonsensical. Instead, it was completely re-written without any input from the translator. Sometimes the poor quality of the translation helped make the game more notable, as in the case of the notoriously poor translation of Zero Wing, whose text “All Your Base Are Belong to Us” became an early Internet meme.
Technology in the early 2000s expanded to allow text to be stored in ASCII strings instead of in picture format, allowing for more efficient processing and more storage space for housing text. Better audio capabilities and reliance on voice acting created new challenges and avenues for translation, allowing the use of dubbing while also adding the burden of translating and recording new audio. As graphics improved and games relied more on cinematics, more attention had to be paid to lip-syncing as well as visual gestures that might be culturally-specific.
In present times, there has been significant uptake in the amount of text and dialogue in a game, especially for triple-A RPGs. For instance, the team in charge of localizing Fable II into five languages consisted of 270 actors and 130 personnel- a far cry from the lone programmer with a phrase book. Likewise, the dialogue scripts for Star Wars: The Old Republic contained over 40 novels worth of text. Director of audio and localization Shauna Perry reports that the game had as much audio as ten Knights of the Old Republic recorded back-to-back. The length and intensity of these projects presents never-before-seen complexity in the localization process.
Levels of localization
Depending on the financial viability of importing a game to a new locale, a number of different levels of localization may be undertaken. The first level is no localization. A game may still be imported into a region in the original language if there is a potential market for it, but no efforts to localize the game will occur, to cut back on costs. The second level is box and documentation localization, in which the box and documentation or manuals for the game will be translated, but not the game itself. This tactic may be used if the game has little text or story, such as early arcade games, or if the target locale is expected to have a decent command of the original language, as in the importation of some English-language games such as Secret of Mana or Terranigma into Scandinavian countries. The third level of localization is partial localization, in which game text is translated but voiceover files aren’t. This helps cut down on the cost of hiring actors and re-recording all of the dialogue, while still making the game comprehensible in another language. Voiceover dialogue that doesn’t appear on screen may be subtitled. The final level of localization is full localization, where in all of a game's assets will be translated, including voiceover, game text, graphics, and manuals. This is the most expensive option, and is usually only undertaken by AAA game companies.
Officially produced localizations generally fit into one of two shipping models: post-gold or sim-ship. The post-gold localization model is undertaken when the original game has already been completed, and usually has already been released. As a result, there is a lag between the release of the original and the release of the localized versions. The post-gold model allows localizers to access a completed game rather than working with incomplete bits and pieces, and generally allow for more time to complete translations, so fewer translation errors occur with this method. This model is used commonly by Japanese AAA producers, though these companies are starting to move towards a sim-ship method for marketing reasons.
The sim-ship model, short for simultaneous-shipment, allows a game to be released in multiple regions at the same time. Because games have a short shelf-life and are prone to be pirated, there is a profit incentive to release games simultaneously across the globe. However, with this method, a completed version of the game is unlikely to be made available for localization workers, resulting in a greater risk for translation errors, as crucial context and game information may be missing. Most Western games follow the sim-ship method of production.
In addition to these shipping models, different production methods may be used to create the localization, usually either outsourced or in-house production. Most game companies in North America and Europe rely on an outsourcing model of production, and this model is popular amongst emerging game development markets such as Chile, Russia, and China. In the outsourcing model, a company that specializes in game translation is hired to undertake the entire process. Oftentimes outsourced companies do not have the full game available to work with, and are dealing with only portions of the game's text or art, resulting in a “blind localization.”  In a blind localization, only a limited amount of information about the final game is available, resulting in a lack of context which can hinder productive localization. Sometimes, even if a game is incomplete, the developer may send a mostly-finished version of the game so that translators can play through the game and get a better sense of the text they’re working with.
In the outsourcing model, developers and publishers will usually provide the translation company with a localization kit. A localization kit may contain elements such as general information about the project (including deadlines, contact information, software details), resources about the game itself (a walkthrough, plot or character descriptions, cheat codes), reference materials (glossaries of terms used in the game world or used for the specific hardware), software (such as computer-aided translation tools), code, and the assets to be translated (text files, graphics, audio, and so forth). An insufficient localization kit can severely hamper translation efforts.
Alternate to the outsourcing model, translators may control the localization in-house. This model is more common for Japanese developers, most notably Square Enix. In the in-house model, the localization process is completely controlled by the developer, though it is common for freelance translators to be hired on a project-basis. Translators working in this model still usually receive a localization kit, but also have greater access to the original game and to the original artists and authors. Because Japanese developers rely on the post-gold model, the in-house translators favored by these companies usually have full access to the completed game. This allows the translation to have fewer context mistakes and results in an overall smoother localization. The downside is the long delay between the game's release in the original country and the subsequent release of the localized version, which is of concern in a global market. Companies like Square Enix are beginning to move towards a simultaneous shipping model, with a shorter release time between different versions.
Finally, a game can be localized through the unauthorized efforts of fans. Fans may be willing to put forth a huge amount of unpaid labor in order to localize a game if it would be otherwise unavailable. If a game is not going to be released in a specific territory, for instance due to doubts about making sufficient profits, fans may take up the slack and release a translation on their own. For instance, the Game Boy Advance game Mother 3 (2006) was not going to be released in North America, possible due to poor sales of the previous installment of the series, Mother 2 (renamed EarthBound in North America). Fans petitioned Nintendo to localize the game, and when that failed, they undertook the process themselves, resulting in a fan-helmed English-language release of the game in 2008.
Inferior localization may also prompt fan action, as in the case of the fan community Clan DLAN. The group has undertaken the work of localizing many games, mods, cheats, guides, and more into Castilian Spanish when the official versions were of poor quality, such as with The Elder Scrolls IV: Oblivion.
Tasks and challenges
The major types of localization are as follows.
- Linguistic and cultural: the translation of language and cultural references maintaining the feel of the game but making it more appealing for the receiving locale.
- Hardware and software: for example the change between PAL and NTSC, re-mapping of hotkeys, gameplay modifications.
- Legal: age ratings may differ depending on the country of release. They are controlled by national or international bodies like PEGI (for Europe), ESRB (for US and Canada), ACB (for Australia), or CERO (for Japan).
- Graphics and music: Some games may exhibit different characters, or the same ones with a slightly different appearance in order to facilitate players identification with their avatar. Music may also vary according to national trends or the preferences of major fan communities.
When games are more story- than action-driven, culturalising them can be challenging because of all the premises the designers are taking for granted in the development of the plot. Asian gamers seem to prefer more childlike characters, while Western countries might emphasize adult features. An example of the changes that are likely to happen during localization is Fatal Frame (known in Japan as Zero and known in Europe as Project Zero) (Tecmo 2001). In the original Japanese version the female protagonist, Miku, was a frightened seventeen-year-old girl looking for her brother Mafuyu who disappeared after entering a haunted mansion. In the US and European versions Miku is nineteen, has Western features, and is not wearing the original Japanese school uniform. Unfortunately, developers did not think necessary to change her brother's appearance, so when players do find Mafuyu at the end of the game they do not seem to be blood-related.
A similar thing happens with the depiction of blood, and real historical events; many things have to be readjusted to fit the country's tolerance and taste in order not to hurt sensibilities. This is probably one of the reasons why so many games take place in imaginary worlds. This customisation effort draws on the knowledge of geopolitical strategists, like Kate Edwards from Englobe. During the 2006 Game Developers Conference in California she explained the importance of being culturally aware when internationalising games in a presentation called "Fun vs. Offensive: Balancing the 'Cultural Edge' of Content for Global Games" (Edwards 2006). Both developers and publishers want to please their clients. Gamers are not particularly interested in where the game comes from, or who created it any more than someone buying a new car or DVD player. A product for mass consumption only keeps the branding features of the trademark; all the other characteristics might be subject to customisation due to the need to appeal to the local market. Therefore, the translation will be in some cases an actual recreation, or, to put it in the words of Mangiron & O’Hagan (2006), a "transcreation", where translators will be expected to produce a text with the right "feel" for the target market. It is important for translators to be aware of the logic behind this. Video games are a software product, and as such, they will have manuals and instructions, as well as interactive menus and help files. This will call for technical translation. On the other hand, we will also find narration and dialogue closer to literary texts or film scripts where a more creative translation would be expected. However, unlike most forms of translation, video games can adapt or even change the original script, as long as it is in the search of enhanced fun and playability of the target culture. We can only find a parallel of this type of practice in the translation of children's literature where professionals often adapt or alter the original text to improve children's understanding and enjoyment of the book.
SCEE David Reeves, has stated that the main reason that Europe is often affected by significant content delays is because of language localization. He stated "the problem is that there isn't enough incentive for developers to work on multiple language translations during development. Hence, Europeans suffer delays and may never see a particular title". He also commented on why the UK and Ireland which are English speaking countries, also experience the same delays as those in continental Europe with many different languages despite little or no modification. He stated "With PlayStation Store we could probably go in the UK almost day and date. But then what are the Germans and the French going to say to me? That I'm Anglo-centric" indicating that the reason that these countries also must wait is to avoid criticism from other large European gaming countries such as Germany and France.
Oftentimes localization changes include adjusting a game to consider specific cultural sensitivities. These changes may be self-enforced by the developers themselves, or enacted by national or regional rating boards (Video game content rating system). However, games are still sometimes released with controversial or insensitive material, which can lead to controversy or recall of the product.
Games localized for import into Germany often have significant changes made due to the Unterhaltungssoftware Selbstkontrolle’s (USK) strict policies against blood and gore, profanity, and symbols associated with racial hatred, such as Nazi symbolism.
For instance, the German version of Team Fortress 2 (2007) has no blood or detached body parts as a result of this regulation, which can cause difficulty for players as it is hard to tell if an enemy has been hit or taken damage (218). As a result, mods known as “bloodpatches” have been created for this and many German games that allow the blood and gore of the original game to be unlocked. Despite a significant overhaul of the graphics, the German localization of the World War II game Wolfenstein (2009) contained a single visible swastika on an art asset. As a result, Raven Software recalled the game.
China also has strict censorship rules, and forbids content that endangers the “unity, sovereignty and territorial integrity of the state” or the “social moralities or fine national cultural traditions,” amongst other qualifications. As a result, the Swedish PC game Hearts of Iron (2002), set during World War II, was banned because the historically accurate maps depicted Manchuria, West Xinjiang, and Tibet as independent states. Additionally, Taiwan was shown to be a territory of Japan, as was accurate for the time period. However, these inclusions were considered harmful to China's territorial integrity, so the game was forbidden from being legally imported. The localization of Football Manager (2005) was similarly banned because Tibet, Taiwan, Hong Kong, and China were all treated as separate teams, putting them on equal footing.
Video games come accompanied by a variety of texts, for example manuals, dubbing scripts, and subtitles that need translating, but they also have other type of texts in a format only common to utility software, like a word processor, or an internet browser. All these programs have one thing in common: information and commands are readily available at the click of a button. It is what we call ‘interactivity’. The interactive element of computer programs has serious consequences for translators because it means that access to texts and information is random, i.e., each user will activate a particular message or command at a different point, or not at all. An arbitrary sequence of events does not allow for linear texts and contextual information, therefore, translators lose two of the most important sources needed in the decision making process: co-text and context. When the program is still unfinished or no localization kit has been prepared, some information is still available, although difficult to obtain, from similar manuals, the localization manager, or the actual technical team responsible for the software. Esselink (2000) is probably one of the best references for the localization of utility software and web pages.
Unfortunately, the localization software industry has not been able to create a GUI (graphical user interface) localization tool for translators to use with video games, like the ones used in the translation of utility software and web pages. These programs (like Catalyst and Passolo) allow users to work directly but safely with the game code, generating a visual representation of the final product, which means that translators can see exactly what the end result will look like and adjust the text or the interface to suit the space available as well as the general look. The LRC (Localization Research Center) and LISA (The Localization Industry Standards Association) has ample information on these programs.
Linguistic assets will be utilised in a variety of ways at different times throughout the creation, development, and launch of the game, mainly: the game itself, which has a variety of texts in multiple formats, the official web site of the game, promotional articles, game patches and updates.
Textual types and file formats
Within these products there are different textual types, each of which has its own characteristics and purpose. Because we are dealing with a multimedia product, the challenges translators have to face are also multimedia. Within the same project they have to deal with a wide variety of issues like reproducing the oral quality of dialogue in writing, lip-synching for dubbing, space and time constraints for subtitling, number of characters for subtitle, UI, etc. The following paragraphs are an attempt to classify the several textual types that accompany the standard PC video game:
(Written form. May be a Pagemaker or a Word format) Although it always has some attractive and engaging creative writing, partly promotional partly literary, most of the manual would normally be filled with didactic texts when telling players the instructions to be followed to fully enjoy the game. Manuals would also include technical texts with the appropriate hardware and software specifications to be able to run the game application. In addition players will always find corporative and legal texts, informing users of their rights and responsibilities attached to the acquisition of an entertainment software product.
(Written form. Pagemaker or Word format) Like manuals, game boxes and packaging present a mixture of textual types, the difference being the space provided, limited not only by the size of it but also by images of the game, logos of the companies involved and legal labelling requirements. It mixes an alluring promotional text, together with concise technical information and legal notices.
(Written form. WordPad/Rich text format) This small .txt file is probably the last thing in the development process. It is used to inform users of all the last-minute adjustments and how to make sure that the product runs smoothly, as well as to correct mistakes and typos in the printed material, such as manual and packaging. It is mainly a technical text.
(Written form. HTML or Java format) It mixes a promotional text with a journalistic one, but it will also have technical details like minimum requirements, etc. A lot of the information offered through the official web will be similar to the one that was shipped with the game. But websites tend to include previews and reviews of the product, notice boards, customer support and downloadable files to fix specific problems, or patches with new language versions, as well as screenshots, concept art, thematic screen savers, merchandising, and fan blogs.
(Spoken form. There will be a separate sound file per utterance. Written scripts will normally be in spreadsheets or Word tables.) Speech delivered by game characters where registers, accents, and idiosyncrasies have to be conveyed into other languages. Sometimes an extra column is included to add inflection comments for the dubbing director. A part of the dubbing script may include atmospheric utterances also in a spoken form. Many games might feature characters talking or reacting to players’ actions. These characters may have little or no relevance to the plot, but their inclusion and to the immersion of the player in the virtual world. No synchronisation is normally required, but orality has to be maintained.
Dialogue for subtitling
(Written form. Spreadsheets and tables are preferred for this although subtitles might be hard-coded in order to synchronize them with video and animations). Oral text in written form. The dubbing script may be applied directly in the subtitling of the game, which results in cluttered and fast subtitles with no character limit per line, nor lines per subtitle. In addition, translators may be faced with the fact that not all languages allow for the same freedom when writing subtitles. Often translators will have to apply techniques used in the translation of children's literature and comic-books to convey certain characteristics that would otherwise be lost. Time and space constraints are very relevant here.
(Written form. Table format, sometimes hard-coded text file due to the interactivity of each item). Space in menus, pop-up windows and hint captions is at a premium and redesigning is rarely an option, so translators will have to maintain a similar number of characters to that of the original label. Similarly to what happens in software localization, video games may have very detailed and crowded menu options to control different features of the game such as difficulty level, as well as graphic display selection, mouse sensitivity, or feedback preferences.
Graphic art with words
(Written form. A multi-layered graphic format will be needed). Players will normally find this type of graphic-text in game names but they can often be seen throughout the game as part of the branding of the product, as well as in advertisements.
In the 2010s there has been debate about how Japanese games are localized, particularly for Nintendo platforms. Some fans consider resulting changes to plot and characterization as marring the original artistic vision, and some object to sexual content being removed or bowlderized. Localization of Nintendo games is commonly handled by a Nintendo division called the Treehouse. In the face of Nintendo's unwillingness to communicate about localization, speculation and conspiracy theories circulated among enthusiasts, and several employees of the Treehouse were alleged to be responsible for unpopular changes.
Allison Rapp, a Treehouse employee not directly involved in localization, garnered controversy due to her comments on Twitter. Attention on Rapp was heightened as part of the Gamergate controversy by the circulation of an undergraduate essay by Rapp which favored cultural relativism regarding sexualization of minors in Japanese media. The essay argued against the sort of censorship that the Treehouse's critics decried. Some however interpreted the essay as defending the exploitation of children, and readers of The Daily Stormer organized a letter-writing campaign to have her fired. That initiative was controversial within the Gamergate movement, with some supporters considering it justifiable treatment of an ideological opponent, while others considered the campaign against Rapp to be unethical or not aligned with the movement's goals. Rapp was subsequently fired, though Nintendo issued a statement that the reason was that Rapp had held a second job against company policy. She maintains that her controversial online presence was the true cause.
- O’Hagan and Mangiron, p. 111
- Bernal-Merino 2008
- O’Hagan and Mangiron, p. 211
- O’Hagan and Mangiron, p. 49
- Kohler 2005, p. 226
- Corliss 2007
- O’Hagan and Mangiron, p. 327
- O’Hagan and Mangiron, p. 58
- Chandler and Deming 2012, p. 317
- Fahey 2008
- Chandler 2005, p. 12-14
- O’Hagan and Mangiron, p. 117
- O’Hagan and Mangiron, p. 234
- Chandler 2005, p. 46-47
- O’Hagan and Mangiron, p. 118
- Dietz 2006
- Dietz 2007
- O’Hagan and Mangiron, p. 121
- O’Hagan and Mangiron, p. 119-21
- O’Hagan and Mangiron, p. 56
- O’Hagan and Mangiron, p. 116
- Diaz Montón 2007
- PS3 News: Reeves: Euro Gamers Wait Because
- Mangiron, Carme; Orero, Pilar; O'Hagan, Minako (2014-01-01). Fun for all: translation and accessibility practices in video games. ISBN 9783034314503.
- O’Hagan and Mangiron, p. 220
- Dietz 2006, p. 131
- Schnittberichte - Wolfenstein
- Good 2009
- Zhang 2012
- Klepek, Patrick (December 14, 2015). "From Japan, With Changes: The Endless Debate Over Video Game 'Censorship'". Kotaku.
- Singal, Jesse (March 31, 2016). "Gamergate May Have Just Won Its Most Disgraceful Scalp Yet". New York Magazine.
- Clepek, Patrick (March 4, 2016). "The Ugly New Front In The Neverending Video Game Culture War". Kotaku.
- Stuart, Keith (March 31, 2016). "Nintendo denies Alison Rapp firing is linked to harassment campaign". The Guardian.
References and bibliography
- Bernal-Merino, M. 2006. "On the Translation of Video Games". The Journal of Specialised Translation, Issue 6: 22-36
- Bernal-Merino, M. 2007. “Training translators for the video game industry”, in J. Diaz-Cintas (ed.), The Didactics of Audiovisual Translation. Amsterdam / Philadelphia: John Benjamins.
- Bernal-Merino, M. 2007. "Localization and the Cultural Concept of Play". Game Career Guide
- Bernal-Merino, M. 2007. "Challenges in the Translation of Video Games". Tradumática, No 5.
- Bernal-Merino, Miguel. (2008). “Inside the Game Localisation Round Table.” Develop. Retrieved December 2nd, 2014.
- Chandler, H. 2005. The Game Localization Handbook. Massachusetts: Charles River Media
- Chandler, Heather M and Stephanie O'Malley Deming. (2012). The Game Localization Handbook (2nd ed.). Sudbury, MA; Ontario and London: Jones & Bartlett Learning.
- “Clan DLAN: Traducción de videojuegos, traducción y creación de mods, modding, revisiones, guías, rol y más. Todo en español.” (2014). Retrieved December 2nd, 2014
- Corliss, Jon. (2007). "All Your Base are Belong to Us! Videogame Localization and Thing Theory." Accessed July 15, 2012. Retrieved December 2nd, 2014
- Dietz, F. 2006. Issues in localizing computer games. Perspectives on Localization edited by Keiran J. Dunne. Amsterdam and Philadelphia: John Benjamins, 121-134.
- Dietz, Frank. (2007). "How Difficult Can That Be? The Work of Computer and Video Game Localization." Revista Tradumatica 5: "La localitzacio de videojocs." Accessed July 12, 2011.
- Diaz Montón, Diana. (2007). "It's a Funny Game." The Linguist 46 (3). Accessed July 12, 2011. Retrieved December 2nd, 2014
- Edwards, Kate. GDC 2006 presentation "Fun Vs. Offensive"
- Esselink, B. 2000. A Practical Guide to Localization. Amsterdam and Philadelphia: John Benjamins.
- Fahey, Mike. (2009). “Star Wars: The Old Republic Script More Than 40 Novels Long.” Kotaku. Retrieved December 2nd, 2014
- Good, Owen. (2009). “Swastika Gets Wolfenstein Pulled from German Shelves.” Kotaku. Retrieved December 2nd, 2014
- Heimburg, E, 2006. Localizing MMORPGs. Perspectives on Localization edited by Keiran J. Dunne. Amsterdam and Philadelphia: John Benjamins,135-154.
- Kohler, Chris. (2005). Power-up: How Japanese Video Games Gave the World an Extra Life. Indianapolis: Brady Games.
- Mangiron, C. & O’Hagan, M. 2006. “Game localization: unleashing imagination with ‘restricted’ translation”. The Journal of Specialised Translation 6: 10-21
- O’Hagan, Minako and Mangiron, Carme. (2013). Game Localization: translating for the global digital entertainment industry. Amsterdam/Philadelphia: John Benjamins Publishing Company.
- Sutton-Smith, B. 1997. The Ambiguity of Play. Cambridge/London: Harvard University Press.
- “The Mother 3 Fan Translation.” Retrieved December 2nd, 2014
- Zhang, Xiaochun. (2012). “Censorship and Digital Games Localisation in China.” Meta: journal des traducteurs, 57(2), 338-350. Retrieved December 2nd, 2014
- Localization Production Pitfalls - excerpt from 'The Game Localization Handbook'
- Game Localization and the Cultural Concept of Play
- Best practices for game localization
- You Spoony Bard!: An Analysis of Video Game Localization Practices
- 22 Ideas for Better Game Localization - Tips and ideas on how to improve the localization process for video games
- Localisation for gaming industry: adapting culturally - Games – Taking your game worldwide
- Videogame Localization and Internationalization