Article body

Background

Localization practices emerged in the 1980s in order to enable computer software, chiefly of US origin, to be functional in a given user environment with new language settings and within target culture conventions. The hallmark of localization is that it deals with the content in electronic form, requiring special engineering processes. Localization takes an adaptive approach to converting the source content to gain a look and feel that are indistinguishable from the equivalent local products (Fry, 2003). During the 1990s, the localization sector grew into an industry dealing with an increasing range of products, including video games. Most major publishers of video games draw their profits from international markets by localizing games into six to eight languages (Chandler, 2005, p. xxi), and more. Whereas software localization deals mainly with English as the source language, Japanese is one of the main source languages for games localization. Although the video game was an American invention, with Higinbotham’s first rudimentary but innovative tennis game created on an oscilloscope and Spacewar by MIT researchers in the 1960s (Poole, 2005, p. 15), Japanese video games have been distributed in the global market since the appearance of early games such as Space Invaders in the 1970s, PacMan in the early 1980s, followed by the best selling Super Mario Brothers in the mid‑1980s. Into the 1990s, global hit titles such as the Pokémon series sold 140 million units worldwide (Nakamura and Onouchi, 2006, p. 30). These global successes raise the question of the role played by translation and localization in games of Japanese origin.

Early Japanese‑developed games did not have many translatable assets beyond a few simple sentences, or any significant culture‑specific content needing to be modified. From a player’s perspective, games such as Space Invaders and PacMan did not come across as Japanese games as such, although behind-the-scenes descriptive Japanese names portraying the personality of each ghost character in PacMan did have to be reworked into English. The original Japanese names, “Oikake [Chase],” “Machibuse [Ambush],” “Kimagure [Capricious]” and “Otoboke [Playful]” were rendered as “Shadow,” “Speedy,” “Bashful” and “Pokey” respectively, in a localized version (Kohler, 2005, p. 212), somewhat changing the original meaning. In games, attributing names that are punchy and memorable is often more important than a faithful rendition of the original meaning; even proper names can be drastically changed for the sake of added entertainment value, contrary to norms in literary or screen translation. This arises from the skopos of games localization, which lies in maximizing the entertainment value of the product, thereby allowing players a smooth gameplay experience without the jarring effects that can result from foreignization strategies in translation (O’Hagan and Mangiron, 2004a).

Advances in game technology have led to recent game consoles such as Xbox360, PlayStation3 and Wii, which are, in effect, high performance computers that provide a compelling gameplay feel with high resolution graphics or an innovative game controller design. The domain of video games is driven by technology, which in turn affects localization. Role‑playing games (RPGs), such as Final Fantasy XII, released in 2006, with a massive volume of text,[1] including dialogue for a movie within the game, present a totally different localization environment from that of the early puzzle and platform games, which had little in‑game text, let alone cinematic elements, to be translated. Today’s sophisticated games are complex pieces of software with interactive multimedia components, each of which may be subject to localization. Unlike content consisting only of written text for print media, a video game typically has translatable assets such as in‑game text, art files (text in graphics and graphics per se), in‑game audio, cinematic assets in addition to printed materials (Chandler, 2005, p. 51). Games are designed to induce pleasure by a complex use of these different elements, which, in turn, need to be understood by the translator and the localizer.

This article examines video games from a Translation Studies perspective with a focus on the localization of Japanese games. Despite the fact that Japanese games have gained international popularity over the past three decades, little translation research has been done on video games in general and Japanese games in particular. Building on earlier case studies on the Japanese Final Fantasy (FF) games (Mangiron and O’Hagan, 2006; O’Hagan, 2005a; O’Hagan and Mangiron, 2004a), this paper provides further evidence of how the entertainment value of the end‑product dictates the skopos of games localization. This investigation into the unique localization model used in the FF series demonstrates a case where the process of localization itself is leveraged to increase entertainment value.

Prior Work on Japanese Video Games and Localization

The field of game studies is rapidly growing internationally as a legitimate area of academic research (Egenfeldt‑Nielsen, et al., 2008, p. 2) with research bodies such as the Digital Games Research Association (DIGRA)[2] being established in 2002. After a long period of stigmatization and prejudice (Newman, 2004, pp. 5‑7) the field is attracting interdisciplinary research interest from a variety of perspectives, ranging from more traditional narrative‑focused studies to approaches based on ludic considerations to embrace aspects of gameplay, as well as game design and programming and many more (O’Hagan, 2005b). Among an increasing number of scholarly works and popular writing on video games available in English, however, only a fairly limited number focus on Japanese games and localization issues. Kohler (2005) provides a detailed analysis of Japanese video games from a historical perspective of games development, focusing on Japanese game design characteristics, with some commentary on translation and localization issues. He argues that the insertion of cinematic elements, which distinguishes today’s video games from early ones, began with Japanese video games (Kohler, 2005, p. 5). Sophisticated cinematic sequences, made possible by the enhanced capacity of game machines, form what Kohler calls “Japanese cinematic video games.” Many of today’s games contain these cinematic audiovisual sequences, called “cut‑scenes,” to enhance the game with compelling graphics. Discussion of cut‑scenes, as in Newman (2004),[3] is relevant to the consideration of dubbing and subtitling strategies for such audiovisual elements.

The focus of this paper on Japanese games also prompted a search for literature from Japanese sources. Despite the high profile of Japan both as a producer and as the world’s second largest consumer of video games (Newman, 2004, p. 3),[4] games studies as an academic discipline has only just emerged there with the Japanese chapter of DIGRA, which was set up in 2006. Most scholarly works published on video games in Japan are not directly relevant to the present study where the focus is limited to localization and translation issues.[5] The scarcity of relevant work can also be attributed to the emergent state of Translation Studies in Japan (Mizuno, 2007). Of the Japanese sources, Yahiro’s comparative study (2005) of 22 Japanese RPG titles and their American localized versions published between 1989 and 1996 is relevant to the localization concerns of the present paper. On the basis that games localization reflects socio‑cultural considerations (Yahiro, 2005, p. 10), Yahiro’s study sought to highlight the socio-cultural specificity of Japanese games by contrasting them with the localized versions in terms of graphics, in‑game text, as well as text used as part of the game’s setting. This study brought home the fact that graphics in games reflect certain socio-cultural views and therefore are frequently subject to modification during the localization process. Yahiro’s study not only confirms the significance of visual representation in games, but also illustrates the scope of localization as encompassing nonverbal elements beyond the written words. The latter element is in contrast with screen translation for cinema, with the exception of some animation films, where the original image is sacred and cannot be touched for the purpose of language transfer.

In the area of games localization, Chandler (2005) gives a comprehensive description of the tasks involved in localizing modern video games, highlighting linguistic and cultural issues as well as problems of a technical nature typically arising during the process of localization. It provides a practical guide to the requirements for the successful globalization of a video game, including issues such as the file types to be localized and country-specific ratings (Chandler, 2005, pp. 27‑35). The latter may impose removal of or alterations to graphics and symbols. Many similarities to software localization are illustrated, such as the need for internationalization, so that the game is designed to make it amenable to subsequent localization in terms of (a) game code which should support different character sets, (b) a user interface allowing for the expansion of translated strings to avoid truncation, and (c) neutralizing hard‑to‑translate culture‑specific references. Many of the issues raised by Chandler are echoed in Dietz (2006), who highlights game specific aspects in comparison to software localization mainly from the point of view of localizing US games. Industry sources (Darolle, 2004) also stress the unique characteristics of games localization, and the way they differ from the characteristics of software localization, highlighting the former’s creative dimensions, whereby each game product delivers a unique experience. This seems to have led games localization to develop largely independently of software localization, which focused on more utilitarian objectives.

Prior studies on Japanese games localization (Mangiron and O’Hagan, 2006; O’Hagan, 2005a; O’Hagan and Mangiron, 2004a, 2004b) have found that entertainment value is given top priority, and it is this factor that most influences the overall translation and localization strategies: the localized game is designed to deliver a gameplay experience similar to that enjoyed by the players of the original game. Furthermore, the focus on entertainment value affords the translator much more freedom although within a limited space compared to other types of translation, leading in some cases to the need to recreate the target text from scratch, rather than translating the given text string. This generally goes beyond the scope of translation and is thus described as transcreation to stress “the freedom granted to the translator, albeit within severe space limitations” (Mangiron and O’Hagan, 2006, p. 20).[6] These studies also suggest that while games are translated within the broad framework of a domestication approach, they also require a complex mixture of foreignization strategies to retain some of the unique Japanese flavour of the original game. The balance between domestication and foreignization in translation strategies seems to differ from game to game and among different genres, and also according to the different approaches taken by individual game publishers. The presence of this variation is one clear difference from the localization of business software (O’Hagan and Mangiron, 2004a, 2004b), and makes games localization less amenable to standardization.

The literature search suggests that there is a limited amount of relevant prior work, including Japanese publications, which presents a Translation studies perspective to localization of Japanese games. As an initial attempt to fill this gap, the present paper takes a descriptive approach by bringing relevant facts into light, in order to lay the foundation for further research on localization of Japanese games. Given the sheer variety of games available today in terms of genres and platforms, however, one must admit that any generalization would be difficult if conclusions are drawn from a limited number of game titles, from one genre and one platform. Keeping in mind the limited applicability of the findings, this study focuses on a specific localization model used by one particular Japanese publisher on certain games in their RPG series. First, the background to this particular RPG series is discussed in order to contextualize the research.

The History of the Final Fantasy Series and Its Localization

The Final Fantasy (FF) series is among the best selling Japanese RPGs on the international market,[7] renowned for its innovative game design and the use of cinematic elements, pioneered by Japanese game developer/publisher now consolidated as Square Enix (SE).[8] In contrast with its early beginnings (one translator, hired in 1991), SE now has an in‑house localization department in Tokyo with over 40 staff members, and it localizes its products into six languages (English, French, German, Italian, Spanish and Korean). SE is recognized as consistently producing some of the best Japanese‑to‑English localizations in the industry.[9] Some details on the localization approach taken by SE are reported in Mangiron (2004) and O’Hagan and Mangiron (2004a). Games localization encompasses a wide scope of adjustments beyond linguistic and cultural factors and includes such considerations as the appropriateness of game difficulty in light of the target audience. For example, the US version of FFIV (released as FFII) was not a direct localization from the original FFIV but of its easier version FFIV Easytype (see Table 1), as the original game was considered too hard for American players at the time, which was also the reason for not localizing FFV. The game was eventually released in 1999 on a different platform when the North American market became more familiar with Japanese RPGs.

The history of the FF series itself is a remarkable example of the ways in which games evolve, and of how this evolution can be directly relevant to issues of translation and localization. Following the first FF game on a cartridge, played on an 8‑bit Nintendo Entertainment System (NES or Famicom) in 1987, later releases of the game developed into the 16‑bit environment of Super NES (Super Famicom), still with a limited capacity. The limited storage of earlier games was a serious problem for accommodating translated text fully, as illustrated in the account by the translator of FFVI (released as FFIII in North America), who noted that the text suffered a reduction of 50% or more in volume when translated into English (Kohler, 2005, p. 226). While restrictions on the length of translations is still an issue today in games localization, this is due to screen space and game logistics rather than hardware capacity. Japanese games pose a particular translation challenge because of the need to express—in a target language that does not use ideographs—the meaning packed into a short string of kanji characters without increasing the length of the string (Mangiron and O’Hagan, 2006).

FFVII was released in 1997 to be played on PlayStation using CD-ROM technology, which provided increased capacity for text and allowed audiovisual content to be streamed directly from the disc. The use of CD‑ROM as a storage and retrieval medium in the console was considered to be the game medium’s most important aesthetic transformation (Newman, 2004, p. 93) with dramatic cut‑scenes in three‑dimensional graphics and far better symphonic music quality. By pushing RPG into the mainstream game genre in the US, FFVII sold more copies in the US market than in Japan, despite poor translation quality in the localized version (O’Hagan and Mangiron, 2004b). FFVII was also the first FF title to be made available on the European market, with translations into German, French, Italian and Spanish. These translations were made from the English version for the North American release rather than from the original Japanese version. After FFVII, the then FF publisher, Square, moved the localization operation in‑house in order to improve quality. To enhance translation quality, the subsequent FFVIII and FFIX were translated directly from Japanese instead of using the English version as an intermediary (O’Hagan and Mangiron, 2004a, p. 58).

Table 1 shows the release date of each FF game.[10] There have been other versions subsequently released, including releases for different game platforms, but the table presents the initial releases only for the original platform. The first FF game took nearly three years to become available in a North American version in 1990. Since FFVII, localized versions have been released more regularly, with a delay of approximately six months for the North American version, followed by European versions.

Table 1

FF Initial Release Timeline for Different Versions

FF Initial Release Timeline for Different Versions

-> See the list of tables

FFX was released in 2001 on PlayStation2, in DVD format, with 90% of the dialogue performed by a team of 100 voice actors (Kohler, 2005, p. 122), taking advantage of the increased storage capacity for recording human voices. This was the first time actual human voices were used in the series, prompting in turn the use of the dubbing mode for cut‑scenes for the North American version, which became the basis for subtitled European versions (Mangiron, 2004). To date, dubbing is only done in English for the largest North American market, which probably reflects the cost and time it takes to include a dubbed mode. FFXI was an online game, followed by FFXII in 2006 and FFXIII released in 2008.

Modern video games are interactive multimedia entertainment products that can be played on different platforms such as consoles, portable devices and on PCs. Enabling a game to be playable on different platforms, in different regions, languages and socio-cultural contexts entails multiple layers of adjustments in the localization process. By the same token, the technical platform on which games are built opens various avenues for players to seek gameplaying pleasure. Just as a computer software’s capability is enhanced with each updated version, the game medium seems to allow enhancement through localization. The next section focuses on the unique release version called the “International” published by SE for certain FF titles.

Japanese-Specific Reverse Localization Model

Of the FF series referred to in Table 1, FFVII, FFX, FFX2 and FFXII also have so called “International” versions. These International versions are released based on the localized North American version of the original Japanese game. The International version is intended exclusively for release in the Japanese domestic market. The main rationale behind this version is to make available a previously published Japanese game as an enhanced game with a somewhat different look and feel, incorporating various improvements made in the American version in terms of gameplay. Drawing on an earlier study (O’Hagan, 2005a), this section discusses this particular release version, taking the case of Final Fantasy X2 International + Last Mission (hereafter referred to as FFX2 International) published in 2004 as the International version of FFX‑2. It has all English in‑game voices, exactly as in the North American version, with the addition of Japanese subtitles, and contains newly developed gameplay features in addition to the enhancements that had been made in the North American FFX2 version. Since the Last Mission component is an entirely new addition, the voices for the latter are only available in Japanese. For the purpose of this paper, the Last Mission component is not discussed as it has not gone through a localization process. Unlike in the North American version of FFX2, however, in FFX2 International, all the in‑game written text, including menus, items list, non‑voiced conversations, etc., remains in Japanese, to the extent that it is difficult for non‑Japanese speakers to follow the game, although familiarity with other localized versions may help to a certain extent. Figure 1 illustrates how the International version relates to the original game via the American localized version.

Figure 1

FF X-2 International in Relation to Its Original FF X-2

FF X-2 International in Relation to Its Original FF X-2

-> See the list of figures

The International version primarily constitutes a reverse-localized version and presents some interesting features from a translation perspective. The most obvious difference from the original FFX2 is that all voices have been dubbed in English. The player can turn on Japanese subtitles for the English dialogue, which is the translation of the Japanese scripts of the original FFX2. The new Japanese subtitles were necessary because the original Japanese does not always match the English voice due to liberties taken in the English translation process for the North American version. For example, in one scene the American translator invented the name “Dullwings” as a play on words of the group called “Gullwings.” This invented name made its way into the Japanese subtitles of the International version, resulting in a similar but new play on words: Dullwings from カモメ団 (kamomedan) for Gullwings becomes バカメ団 (bakamedan) [Sillywings], creating the same intended effect of making fun of Gullwings (Square Enix, 2004, p. 40). At the same time, some changes made in the North American version, such as proper names, reverted back to the original names in the Japanese subtitles to avoid confusion for Japanese players who had played the original game. The International version has the look and feel of a foreign film, which appeals to certain Japanese players. The linguistically hybrid nature of the International version can also be seen in recurring phrases, known as “battle quotes,” which are uttered repeatedly by the main characters in the game just before, during or after a battle. Although most battle quotes are translated into English, the International version leaves some of them in Japanese such as “Yuna Mairimasu” [“Yuna at your service”], “Zenkaide Ikuyo” [“Let’s go full throttle”], “Simete Yaru” [“I’m gonna get you”], etc. (Square Enix, 2004, p. 349), albeit all spoken with an American accent, giving an impression of pseudo Japanese.

Other translation-related changes made in the International version are of a subtle nature, yet relevant to translation issues. In one scene, a nonverbal cue by the main character Yuna was changed from nodding her head to shaking her head to match the English dialogue; in this scene Yuna is giving the negative answer “No” to a negative question: “Aren’t you gonna return it?”[16] (Square Enix, 2004, p. 42). In Japanese, negative answers to negative questions are given by an affirmative response, accompanied by a nodding gesture, as shown by Yuna’s gesture in the original FFX2. The change made was an apparent attempt to match the English language convention, which the International version follows. There are several other examples of subtle changes made in the images in cut‑scenes for different camera angles, cuts, etc., conjuring up a somewhat different feel from the original FFX2.

Attention to these details demonstrates how the game is presented as a new version to deliver a new pleasure to the player. The International version is perceived as an enhanced product worthy of purchase, even if the player has already played the original game, not only because of the added new gameplay features, but also because of the different look and feel with a touch of foreignness conveyed through the spoken English dialogue. The publisher suggests that part of the enjoyment of the International release is to spot various changes in scripts via Japanese subtitles, which reflect how the American version had been rendered (Square Enix, 2004, p. 40). The International version illustrates a case where the entertainment value is drawn out of the localization process itself. This unusual take on a localization model in turn prompts the question of the significance of the original. The use of digital technology allows continuous content updates of computer software and web pages, and the International version concept here attempts such upgrades through a different language version. To this end, in this model of the International version, the original source text takes on a different meaning from that in other types of translation in that the function of the original here is merely to set off a chain of enhancements through the localization process. Ultimately the production of such a version is market-driven and, in this case, driven by the appeal to Japanese players who are looking for the “foreign feel” of playing a reverse localized game aside from new gameplay features. In the International version, the translation concept itself provides the very inspiration for new entertainment value.

Conclusion

The challenge posed by video game translation and localization is complex and multi‑faceted, reflecting in part the novel nature of video games as a new form of interactive digital entertainment. Games localization prioritizes entertainment value to convey a similar gameplay experience to the player of the original game. In fact, the above FF series case study suggests that “enhanced” game experience surpassing the original is sought in a localized version. In the case of the International version, the original merely provides a working base rather than work completed and set in stone. The digital medium of games seems to invite a continuous chain of improvements through localization rather than localization as a means of re‑production of the original. The question of what gameplay pleasure means to different players playing different localized versions of allegedly the same game is then crucial to the understanding of the skopos and the nature of game localization. Research questions such as this beg the involvement of players themselves. The study on the reception of localized games will contribute fresh insights into game studies, which is currently largely limited to a single language mode. The focus on the localization process of Japanese games is productive in bringing in fresh perspectives to broaden the horizons of games studies as well as Translation Studies.