Found in Translation: Evolving Approaches for the Localization of Japanese Video Games

Abstract

Japanese video games have entertained players around the world and played an important role in the video game industry since its origins. In order to export Japanese games overseas, they need to be localized, i.e., they need to be technically, linguistically, and culturally adapted for the territories where they will be sold. This article hopes to shed light onto the current localization practices for Japanese games, their reception in North America, and how users’ feedback can contribute to fine-tuning localization strategies. After briefly defining what game localization entails, an overview of the localization practices followed by Japanese developers and publishers is provided. Next, the paper presents three brief case studies of the strategies applied to the localization into English of three renowned Japanese video game sagas set in Japan: Persona (1996–present), Phoenix Wright: Ace Attorney (2005–present), and Yakuza (2005–present). The objective of the paper is to analyze how localization practices for these series have evolved over time by looking at industry perspectives on localization, as well as the target market expectations, in order to examine how the dialogue between industry and consumers occurs. Special attention is given to how players’ feedback impacted on localization practices. A descriptive, participant-oriented, and documentary approach was used to collect information from specialized websites, blogs, and forums regarding localization strategies and the reception of the localized English versions. The analysis indicates that localization strategies for Japanese games have evolved over time from a higher to a lower degree of cultural adaptation in order to meet target markets’ expectations. However, it was also noted that despite the increasing tendency to preserve the sociocultural content of the original, the language used in the translations needs to be vivid and idiomatic in order to reach a wider audience and provide an enjoyable gameplay experience.

1. Introduction

Over the last six decades video games have consolidated as one of the most popular entertainment options worldwide (Bernal-Merino 2014) and gaming has become the most lucrative entertainment industry, generating $145.7 billion in 2019, in comparison with $42.5 billion generated by the cinema box office and $20.2 billion from the music industry (Richter 2020). Since the origins of the game industry in the 1960s, Japan has been one of the key players in this sector (Picard 2013; Consalvo 2016; Hutchinson 2019), with companies such as Nintendo, Sony Computer Entertainment, and SEGA playing a significant role in the industry’s development, both as a hardware manufacturer and as a software developer and publisher (O’Hagan 2012). Japan is the third largest market for video games, after China and the United States (Newzoo 2020), and the second country in the world, after China, in terms of the dollar value worth of exports of video games-related equipment in 2019, with $2.12 billion worth of exports (Workman 2020). However, in order to move across different boundaries and borders, video games need to undergo a series of transformations (Newman 2019). Adapting a Japanese game to be sold overseas is complex and costly, as they have to be localized, i.e., adapted technically, linguistically, and culturally to meet the target territories expectations (Chandler 2005; O’Hagan and Mangiron 2013; Bernal-Merino 2014).

This paper examines how strategies for the localization of Japanese games for the North American market have evolved over the years through the analysis of industry perspectives on localization and target market expectations, as well as how the dialogue between industry and consumers takes place. Special attention is paid to the way in which players’ feedback can impact on localization practices. After briefly describing what game localization entails, an overview of the localization practices followed by Japanese developers and publishers is provided. Next, the paper presents three brief case studies of the strategies applied to the localization into English for three renowned Japanese video game sagas set in Japan: Persona (1996–present), Phoenix Wright: Ace Attorney (2005–present), and Yakuza (2005–present). This paper maps out how localization strategies for these series have evolved over time and how sociocultural, audience, and industrial factors may have impacted on the development of localization practices and, therefore, on the transnational circulation of Japanese video games.

2. Methodology

This paper was written with a descriptive and qualitative Translation Studies perspective, although it also drew on existing literature from other disciplines, such as Japanese Studies, Game Studies, and Media Studies. First, I provide a brief review of scholarly research on game localization and, more specifically, the localization of Japanese video games. Then I present three case studies of Japanese video game sagas. The criteria for selecting the sagas were the following:

• In order to be able to track any possible changes in localization strategies, it had to be a saga that spanned over at least 10 years, as from the author’s professional experience as a game translator, there have been changes in the approach to the localization of Japanese games over this period of time.
• It had to be a saga of video games that is set in Japan or includes a significant amount of references to Japanese culture, in order to examine how such cultural markers had been dealt with for the release of the North American version.
• The video games had to contain a high amount of text, in order to be able to explore the translation and adaptation processes they underwent in detail, so games containing features of role playing games (RPGs) or visual novels were deemed most appropriate.
• The localization of at least some of the games in the saga had been discussed widely on the Internet by translators, editors, or producers of the localization team, as well as reviewed or discussed in specialized sites, blogs or players’ forums.

With all these criteria and due to the scope of the paper, the selection was narrowed down to the following three sagas: Atlus’s Persona1 (1996–present), Capcom’s Phoenix Wright: Ace Attorney (2005–present), and Yakuza (2005–present) by SEGA and Ryu Ga Gotoku Studio.

The paper also adopts what Saldanha and O’Brien (2013) describe as a participant-oriented approach, focusing on the agents involved in the localization process (e.g., translators, editors, producers) and the end users (mainly players and specialized journalists). A documentary approach (Newman 2019) was also applied, whereby the opinions of localization practitioners, game critics, and players were collected from secondary resources publicly available on the Internet, such as interviews, blogs, specialized websites, and discussion forums.2 With this approach I hope to expand the scope of studies about localization by including agents who are involved in the production and the reception of the localized versions and by widening the lens to examine the communicative exchanges that take place over the Internet, which can provide a wealth of resources for documentary qualitative analysis of online communication (Bryman 2012).

3. What Is Game Localization?

Localization consists of a series of complex technical, linguistic, cultural, legal, and marketing processes in order to sell a game in a different territory while also maintaining the “look and feel” of the original game and providing target players with a similar gameplay experience to that of the original players (Mangiron and O’Hagan 2006). It is a type of functional and user-centered translation, where priority is given to the players and their gameplay experience. The skopos or main brief of game localization is to provide a version that allows target players to experience the game as if it had been originally developed for them (Mangiron and O’Hagan 2006; Fernández-Costales 2012; O’Hagan and Mangiron 2013; Bernal-Merino 2014). Therefore, the process of localizing a video game does not only involve translating the textual assets; it also implies modifying the visuals, the storyline, the gameplay mechanics (rules), or even the music of the game, so that the localized version meets target players’ expectations (Chandler 2005; O’Hagan and Mangiron 2013; Bernal-Merino 2014). For the localization industry, translation is one of the final steps in the production of the target versions, although a number of translation scholars have been critical of such a narrow view of translation (see Cronin 2003; Pym 2004; O’Hagan and Mangiron 2013). In addition, the terms “translation” and “localization” are often used interchangeably in the industry and academia, and their conceptualization is still blurred and requires further epistemological debate (O’Hagan and Mangiron 2013, p. 104). In this paper, following authors such as O’Hagan (2012); O’Hagan and Mangiron (2013) and Bernal-Merino (2020), the most common term “game localization” is used to refer to the overall process of adapting a video game to sell in target territories, while “translation” is also used to refer to the textual operations that occur during the localization process.

In order to guarantee a smooth localization process, it is important that games are designed with localization in mind from the early stages of development, so that the game can be easily adapted to other territories without having to make substantial changes to its design (Chandler 2005). This is known as “internationalization”, which consists of developing assets that are flexible and can be easily replaced in each target locale (country/region and language) (Chandler 2005, p. 11). For example, the game code should be compatible with different formats for date, time, numbers, and special characters. In the case of Asian languages, such as Japanese, it is necessary to plan for the conversion from double-byte to single-byte character sets. It is also important to consider space limitations. In order to overcome them, it is recommended to allow for at least 30% extra space in the localized versions in order to avoid truncations and overlaps (Chandler 2005, p. 9). Other solutions to overcome space limitations are the use of scrollable menus and expandable text boxes.

In addition to technical issues, during the internationalization phase, developers are encouraged to design the game with the widest possible appeal and avoid culture-specific references (Chandler 2005, p. 11). In particular, any cultural issues that could be obscure or offensive for players or that could affect the age rating of the game should ideally be taken into account (O’Hagan and Mangiron 2013). For game industry expert Kate Edwards (2011, pp. 20–21), the process of “culturalization” is essential in game localization:

Culturalization is going a step further beyond localization as it takes a deeper look into a game’s fundamental assumptions and content choices, and then gauges their viability in both the broad, multicultural marketplace as well as in specific geographic locales. Localization helps gamers simply comprehend the game’s content (primarily through translation), but culturalization helps gamers to potentially engage with the game’s content at a much deeper, more meaningful level. Or conversely, culturalization ensures that gamers will not be disengaged from the game by a piece of content that is considered incongruent or even offensive.

While it is a moot point whether culturalization is going a step beyond localization or if it should be embedded in the localization process, it is important to present a coherent gameplay experience and not to break players’ engagement with the game with “culture bumps” (Leppihalme 1997) that may hinder their comprehension, touch on sensitive issues, or be considered offensive. As an example, in the Japanese version of the game Persona 5 (2016), one of the characters wears white shoes with the Japanese imperial flag. In the Korean version of the game, the flag was removed because many Koreans still associate the flag with World War II and may have found it offensive (Schleijpen 2019, p. 22). Therefore, as Consalvo (2016, p. 123) states, localizers act as “culture brokers”, adapting any required cultural elements “into something not just comprehensible but also enjoyable for players in other countries who speak other languages”. Cultural adaptation is also essential to guarantee the playability of the localized versions, as players need to interact with the game in order to advance and enjoy the gameplay experience. In Bernal-Merino’s words (2020, p. 301), “[p]ertinent game pragmatics can only occur when players understand the signs of each particular game and choose the right combination of responses to converse with each game-machine”.

However, the industrial workflows of game development and publishing tend to favor streamlined localization solutions that focus on written and recorded text, sometimes leaving unaltered other semiotic signs that also carry meaning and contribute to the cohesiveness of the game as a whole (Bernal-Merino 2020, pp. 299–300). Therefore, often “residual traces of a video game’s national and cultural origins” can be found in video games due to an oversight or limited funding (Mandiberg 2015, p. 112), although as Mandiberg argues, such moments of pragmatic incongruity can be positive, as they facilitate cultural circulation and “help players learn how to become hybrid citizens” (2015, p. 112).

The key issue seems to be how to find the delicate balance between foreignization and domestication, in Venuti’s (1995) terms. Foreignization implies preserving most of the original sociocultural content of the game in order to help target players become acquainted with the source culture (Fernández-Costales 2012; O’Hagan and Mangiron 2013). In other words, it provides a “fragrant” cultural product (Iwabuchi 2002) that can be identified as foreign, and its foreignness may be precisely one of its main appeals (Consalvo 2016; Hutchinson 2019). On the other hand, a domesticating or, in Iwabuchi’s (2002) terms, mukokuseki (literally “nationless”) or “odorless” approach is used to adapt the source content as required, making it more local in order to bring it closer to the target audience. It should be highlighted that domestication and foreignization are not totally opposing poles. During the course of any project, the localization team has to negotiate with both domesticating and foreignizing strategies (Carlson and Corliss 2011) to identify those cultural markers that may hinder comprehension or be offensive and those that can be preserved to help target players learn about the original culture. This can be done, for example, by means of adding additional information in the localized version, as in the case of the Phoenix Wright series (2001–present), where background narrative elements were included in the story to help players of the North American version understand the abundant Japanese visual and narrative references present in the game (Tsu 2014), as is described in more detail in Section 6.

With regard to the different levels of localization, they are usually determined by marketing strategies and market size (O’Hagan and Mangiron 2013, p. 141). Chandler (2005, pp. 12–14) identified three levels of localization:

• Box and docs: this refers to translating the packaging of the game and the accompanying documentation, such as the manual, health and safety warning, etc. It is done for games with little text, games which are not expected to sell many copies, or for territories with a good knowledge of the original language.
• Partial localization: this means translating the box and docs texts, as well as all the in-game text, but maintaining audio and cinematic assets in the language of the original game, not revoicing them, and only providing subtitles.
• Full localization: this implies translating all the game assets, including the revoicing into the target language. This is usually done for AAA titles with high budgets that are expected to do well in the territories they are localized for.

After having briefly presented the main features and levels of game localization, the next section focuses specifically on the strategies that Japanese developers and publishers have adopted over the years to make their games available for international markets.

4. An Overview of Japanese Video Game Localization Practices

The decision to localize a Japanese game for international markets is complex, as it not only involves a global business strategy, but also local economic, geopolitical, and cultural factors (Picard 2013). In the early days of the game industry, during the 1980s and the early 1990s, Japanese games that contained a large amount of text were not localized, as this involved a considerable amount of work, and genres with little text, such as platform and fighting games, were prioritized (Baird 2019). Localization was only done into English, and it was usually an afterthought once the game was successful in Japan and had aroused a certain level of interest overseas (O’Hagan and Mangiron 2013). The translation was often made by the developers and some of the games contained errors that have since become memes, such as the notorious “All your base are belong to us” from the localized English version of the game Zero Wing (1989) (Kohler 2005). The games were then shipped in English to other countries, without localizing them or just localizing the box and the manual (O’Hagan and Mangiron 2013). In addition, during this period Japanese games were subject to extensive adaptation, because there were still no age rating boards and most of the games that came from Japan were addressed to a young audience (O’Hagan 2012). Interestingly, however, as O’Hagan (2012, p. 187) points out, such changes were made by Nintendo of America, a company that was setup to oversee Nintendo’s business in North America. This highlights the hybrid and transnational nature of the gaming industry since its origins, as pointed out by scholars such as Consalvo (2006) and Picard (2013), and illustrates the complex nature of game localization. Nintendo of America would systematically remove references to religion, tobacco, alcohol, violence, and vulgar language (McCullough n.d.). For example, in the game Super Castlevania 4 (1991) a naked statue was clothed and in Earthbound (1994) all red crosses that appeared in hospitals were removed. In Punch Out! (Nintendo 1987), the name of the character Vodka Drunkensky became Soda Popinsky, in order to avoid the reference to alcohol (McCullough n.d.).

The release of the PlayStation console in Japan in 1994 represents a landmark both in the game industry and in game localization, as it used the CD-ROM as a storage medium, which allowed for the inclusion of more text, music, and higher quality graphics (O’Hagan and Mangiron 2013, pp. 55–56). Some iconic games were produced and localized in the second half of the mid-1990s, such as Final Fantasy VII (1997) and Metal Gear Solid (1998), although translation quality was in many cases poor, as they contained numerous translation errors (O’Hagan and Mangiron 2013, p. 56).

As game technology evolved, games became more sophisticated with cutscenes and voiced dialogues included with the arrival of the PlayStation 2 in 1999 (Kohler 2005). Producing games became more costly and in order to maximize their return-on-investment Japanese developers and publishers started to localize their games into French, Italian, German, and Spanish (FIGS) (Bernal-Merino 2014). As the English-speaking market, and particularly the North American market, has traditionally been the main target for Japanese game companies (Bernal-Merino 2011; O’Hagan 2012), full localization was only done into English (O’Hagan and Mangiron 2013). For FIGS, partial localization was often adopted, using English as a pivot language: all the text was translated into the target language, but the game audio was kept in English and subtitles in FIGS were provided (O’Hagan and Mangiron 2013). Nowadays, the tendency is to localize Japanese games into more languages, such as Eastern European Languages and other Asian languages, such as Chinese and Korean (O’Hagan and Mangiron 2013). Together with Japan, China and South Korea lead the Asia Pacific region gaming market (Moore 2020) and there is a “yearning” for Japan in East Asian countries since the 1990s, so Japanese producers of popular culture are turning their eyes to that region (Iwabuchi 2006, p. 20).

One of the most common internationalization strategies adopted by Japanese companies through the years has been to develop what Iwabuchi (2002) calls mukokuseki video games, “odorless” products without cultural elements that identify them as distinctly Japanese in order to ensure a better reception in other countries (Picard 2013; Consalvo 2016; Hutchinson 2019). This is the case of games set in imaginary worlds, such as Animal Crossing (2001—to present), The Legend of Zelda (1986—present), or the Pokémon (1996—present) series, the localization of which has tended towards the adaptation and domestication of cultural references and humor.3 Another internationalization strategy used by Japanese developers consists of designing culturally hybrid games that mix elements from both Japanese and United States cultures, such as Final Fantasy X, where the two main characters are Tidus, who looks like a young Western surfer-type, and Yuna, who dresses with a kimono and behaves in a very Japanese way (Consalvo 2006).4 Finally, there are also video games that exploit their Japaneseness as one of their appeals for international audiences (Navarro Remesal and Loriguillo López 2015; Hutchinson 2019). For such games, the trend has been to preserve a strong Japanese sociocultural component and address the localization to a traditionally niche audience interested in Japanese culture (Mangiron 2012; Hutchinson 2019). This is the case, for example, of Okami (2006), which is deeply imbued with Japanese mythology (Hutchinson 2019).

Due to the “Cool Japan” phenomenon (McGray 2002), global audiences’ interest in Japanese culture has increased considerably (Kimura and Harris 2020), so references to Japanese media and popular food no longer need to be adapted for localized versions (Consalvo 2016). However, as stated by Hutchinson (2019, p. 35), the majority of Japanese gaming companies continue to develop games with “mukokuseki characters and visual designs for universal relatability”. The mukokuseki design approach shares the same principles as the internationalization phase of game localization: “designing flexible or ambiguous assets (whether they are characters, stories, images, or even gameplay mechanics) to appeal to the broadest possible audience” (Carlson and Corliss 2011, p. 75). Such an approach has been adopted by other Japanese media industries, such as anime and manga (Iwabuchi 2002), and it has also been present in the form of “pre-translation” in Japanese literature (Hijiya-Kirschnereit 2012). Some modern Japanese authors, such as Haruki Murakami, make their texts more easily accessible to an international audience by avoiding cultural specifics and/or by elaborating on cultural specifics that are self-evident to readers of Japanese (Hijiya-Kirschnereit 2012). However, some traces of Japanese culture are usually retained in cultural and media products, as they cannot be completely culturally neutral (Iwabuchi 2002) and an exotic “fragrance” of Japaneseness may appeal to international audiences, who have an idealized vision of Japan (Hutchinson 2019, p. 35).

Nowadays, the big question Japanese developers have to assess carefully when they decide to export their products is whether their game is going to be addressed to the mainstream audience, who may not be very familiar with Japanese culture, or to a niche market of Japanese culture fans, who are knowledgeable about Japan and want to experience Japanese culture as closely as possible (Ranford 2017). They also need to decide on the most appropriate localization strategy: domestication or foreignization. As already mentioned, domestication and foreignization are not mutually exclusive, although there is usually one predominant strategy. For example, in a localization project where a foreignizing strategy is applied to the visuals, the storyline and the sociocultural content of the game, domestication may be applied in certain cases when dealing with humor or cultural references that could be obscure for target players, be it visual or textual. Factors such as theme, storyline, genre, how similar games have been received in the target territory, and the intended target audience should be taken into account when defining the localization approach, as well as the desired age rating in the target territories, as this may differ in Japan and other countries (O’Hagan and Mangiron 2013). Many Japanese games with a low age rating contain jokes with sexual innuendo, while in Europe and North America such content is not deemed appropriate for a younger audience (Di Marco 2007). As a result, either the age rating needs to be changed, which results in a reduced market for the localized version, or any potentially inappropriate references needs to be removed or adapted (Di Marco 2007). According to Sam Mullen, localization director at SEGA of America, adapting Japanese content that is problematic or may feel out of place in a different region is always a difficult decision (Mullen cited in Khan 2019). If it is not adapted, target players will not understand it and may feel uncomfortable, but if it is heavily adapted, it may be viewed as censorship (Mullen cited in Khan 2019). The solution seems to be finding a middle ground where some elements are lightly adapted as required, so that they will not “pop out quite as hard” (Mullen, cited in Khan 2019). However, fans may not be happy with this option, and they may still view it as censorship (Mandiberg 2017). This was the case of the controversial localization of Fire Emblem: Fates (2015), which was harshly criticized by fans because some of the characters’ names and personalities were changed in the English language version (Mandiberg 2017). One of the mini games, where characters could caress other character’s faces in order to develop a stronger bond, was also adapted and characters simply visited other characters in the North American and European versions, something that angered fans because they perceived it as censorship (Mandiberg 2017; O’Hagan 2017).5

With regard to the preferred localization model, many Japanese developers localize most of their games in-house, and translators can access the original game or at least have access to contextual information, such as screenshots, walkthroughs, etc., if the game is still under development (O’Hagan and Mangiron 2013). While in the early days Japanese companies used to localize their games after they had been released in Japan, today the general trend in the industry, even for Japanese companies, is to opt for the sim-ship (simultaneous shipment) model, releasing the game at the same time all over the world in order to capitalize on the hype created by a global marketing campaign and to prevent piracy and grey import copies (Bernal-Merino 2020). However, localizing a game while it is still under development poses many challenges, as the text is constantly subject to changes and updates, and access to the visual context of the game is limited. This workflow can lead to errors, as translators only work with text and have no access to the visuals of the game, which can easily lead to mistranslations, especially when translating from Japanese language, which does not have grammatical gender nor number. For this reason, if there is not much contextual information available and translators have no access to the visuals of the game, it can be difficult to know if a character is speaking to one person or more, and also to know whether the addressee(s) is (are) male or female. For these reasons, a thorough linguistic quality assessment is required to ensure the cohesiveness between all the semiotic elements that make up a game, such as text, visuals, audio, and music (Bernal-Merino 2020).

Another characteristic feature of the localization approach typically taken by Japanese companies is the dynamic and close relationship between the development and the localization team (O’Hagan and Mangiron 2013), which leads to what Bernal-Merino (2016) calls “co-creation” in game localization. An example of this can be found in the localization of the game Bayonetta (2009), where the translator, J. P. Kellams, collaborated closely with Hideki Kamiya, the director of the Japanese version (Kellams 2009). They reviewed together the translation work, discussed the lines, did rewrites, and the director approved or rejected additions or rearrangements in the localized version (Kellams 2009). In addition, one of the translation decisions for the localized English version led to a modification of the original Japanese game. Kellams was asked to come up with sounds for when Bayonetta summons demons for her attacks, but after thorough research he decided to use Enochian—an invented language of the 16th century that is said to be the language of angels—for summoning demons and to represent angels’ voices (Kellams 2009). As there are not many resources for Enochian, Kellams ended up creating a syntax and new words for the language, which he used in the localized version (Kellams 2009). This worked so well that Kamiya decided to rewrite the original Japanese lines to match the nuance of the lines in Enochian in the English version (Kellams 2009).

Another example of co-creation and culturalization can be found in the localization of Shadowverse (2016) by Cygames. The development team shared information with the localization team in advance in order to detect political, historical, religious, race, or gender expressions that could inadvertently be offensive for players of the target version (Dino 2020). In addition, the localization team proposed a change for one of the characters, which was going to be named Phantom in the original Japanese version, but translators thought that Specter was more fitting, so writers changed it in the source script as well (Dino 2020).

The above-mentioned examples highlight the collaborative nature of the localization of Japanese video games, as well as the authorial role translators may sometimes adopt. Such issues are explored in more detail in the next sections, where three brief case studies of renowned Japanese video game sagas—Persona, Phoenix Wright: Ace Attorney, and Yakuza—are presented, analyzing how their localization approach has evolved over time and how the localized versions have been received in English-speaking territories.

5. The Localization of Persona: Rethinking Localization Strategies to Meet Fans’ Expectations

The localization of the Persona series (1996–present), by Japanese developer Atlus, is a good example of how Japanese companies listen carefully to the feedback of players of the English version and rethink their localization strategy accordingly. The first game of the series, Revelations: Persona (1996), was adapted extensively for the North American release. The name of the title was changed from 女神異聞録ペルソナ(Megami Ibunroku: Persona, literally “Strange Tales of the Goddess: Persona”) to Revelations: Persona, and Japanese place and characters’ names were changed into English names. In addition, many of the characters were redesigned: their skin was made lighter, their hair style or color was changed, and the protagonist’s piercing was removed (Girard-Meli 2015; RPG Site 2019). A particularly controversial change was that Mark, one of the main characters, became African-American for the North American version for no obvious reason (Censored Gaming n.d.). The gameplay mechanics were also changed: the enemy encounter rate was lower and the amount of experience that could be gained was increased, possibly to make the game easier and more attractive to new players (Censored Gaming n.d.).

The localized English version received so much criticism because of all these changes that in subsequent releases the localization team decided to change the localization strategy in favor of foreignization. Starting with Persona 2: Eternal Punishment (2000), references to Japanese culture were preserved in the localized version in order to share Japanese culture with Western players (Girard-Meli 2015). It should be mentioned that this was in fact the third Persona game released in Japan, but the second one, Persona 2: Innocent Sin (1999), was not released internationally until 2011 for the PlayStation Portable (PSP). Although Atlus explained that the reason for not releasing the game was the workload of the localization team at the time, it is believed that it was also partially due to the abundant references to Japanese culture and the inclusion of Hitler and Nazi symbolism (Hilliard 2015; Did You Know Gaming? 2015), as well as the possibility of having a gay male relationship (Aether 2015). For the PSP release in Japan and overseas, Hitler’s name was changed to Führer, his design was modified to include sunglasses and a cape to hide his uniform, and swastikas were replaced by iron crosses (Did You Know Gaming? 2015). The lack of an official localized version led fans to take localization into their own hands6 and a translation patch was released in 2008. Although Atlus was aware of its existence, they took no action (TV Tropes n.d.).

Subsequent releases of the Persona saga, particularly from Persona 3 onwards, have also kept as much visual and textual Japanese references as possible (Namba cited in Ward 2015). In Persona 4 (2008), for example, the story takes place in a small rural town in Japan and Japanese cultural elements such as kotatsu (a low table that has an electric footwarmer) were kept in the English version (Girard-Meli 2015). In Persona Q: Shadows of the Labyrinth (2014), one of the dungeons is designed to resemble a Japanese festival and there is a group of men carrying a shrine, which was kept in the localized English version. In order to help English-speaking players understand the reference to Japanese culture, an important amount of dialogue was included in the localized game (Madnani 2015). This is another example of co-creation and emphasizes the authorial role game translators can have. In another scene, a character eats takoyaki (little balls made of flour and octopus) and instead of changing the graphics or adapting the reference, the original Japanese word was kept in English (Madnani 2015).

In Persona 3 and 4, much work was devoted to making the dialogues sound natural for the English-speaking audience (Girard-Meli 2015). Different translators were in charge of localizing the dialogue for specific characters in order to be able to provide them with the right voice in English (Girard-Meli 2015). There was also an editor, who worked on the first translation draft and made changes to the text so that it sounded more natural in English (Madnani 2015). The editor did not even play the game and just worked with the translated text (Madnani 2015). The role of an editor, who does not necessarily speak the Japanese language, is considered important when localizing Japanese games in order to make the game sound as natural as possible to the target audience (O’Hagan and Mangiron 2013).

The foreignizing approach adopted by Atlus led them to relocalize the first game of the series, Revelations: Persona, for its remastering for the PlayStation Portable in 2009 (Awkerman 2009). The localization team decided to improve the original localization, keeping the new translation closer to the Japanese original version by retaining the original Japanese names for characters and locations, looking for less Americanized names for personas7 and demons, and keeping skills names closer to the Japanese ones (Awkerman 2009). Interestingly, some of the translation errors of the first localized version were left unchanged as a wink to the players, because they could be easily recognized by players of the first version, who had discussed them in forums, and they could also be considered a homage to how far localization practices had improved at Atlus over time (Jeriaska 2009).

Atlus’s change of localization strategy towards a more foreignizing one that preserved the sociocultural content of the original was praised by specialized blogs, as illustrated in the reviews below:

By incorporating cultural differences rather than erasing them, imbuing the dialogue with enormous personality, and hiring actors that care about the characters they’re voicing, Atlus U.S.A. has helped bring the series to the top of the role-playing heap in the West

(Girard-Meli 2015).

…the lengths the localization team goes through to make gripping, enjoyable dialog, translated from the original Japanese into English, while still maintaining the feel of the game and the thoughts the original developers wanted to convey

(Awkerman 2009).

The continuous development of the Internet and other technologies has led to user empowerment and the visibilization of game fan communities (O’Hagan 2017). Players can be very vocal in game forums about what they like and what they do not like, but game companies do not always listen to them. Japanese game developers and publishers often base their business decisions on “how imagined players and player communities think about games and what ‘a game’ or even ‘a Japanese game’ should look and play like” (Consalvo 2016, p. 123). However, Atlus’s change in localization strategies shows how they did take into account players’ feedback and expectations, for which they have received praise in the specialized media:

Learning to market a new franchise, especially a dark one like Persona, is not easy. Atlus’ missteps are not surprising, but the company took public feedback to heart and adjusted its localization approach. Instead of running from its niche appeal, Atlus embraced it

(Wallace 2013).

However, it should be mentioned that the localization of Persona 5 (2016) received some criticism because of the quality of the translation (Lee 2017). Fans complained because it did not sound natural in English and contained several errors, such as mistranslations, typographical errors, and inconsistencies in the translation of Japanese honorifics (Krammer n.d.). It is possible that one of the main reasons for the decrease in quality is the fact that the time frame for localizing the project was short, so a big team of six translators and eight editors worked on it (Lee 2017). As a consequence, the translation work was fragmented, which resulted in inconsistency issues. It would also seem that there was not enough time allocated to perform a thorough quality assurance process, which resulted in an English version with several mistakes and led to a negative reception of the localized version.

6. The Localization of the Phoenix Wright Series: Creating a Hybrid Game World

The localization of the Phoenix Wright: Ace Attorney series (2001–present) offers another interesting case study.8 The first instalment of the series was released by Capcom in 2001 in Japan and in 2005 overseas. For its localization the game underwent intensive adaptation, as it contained many references to Japanese culture, as well as many puns and play on words, which had to be rendered naturally into English because the “player must engage with the language of the game as a primary means of gameplay” (Consalvo 2016, p. 126). The localized version was set in Los Angeles, the characters’ names were adapted to English, and many cultural references were domesticated. For example, one character’s love for ramen soup became a love for hamburgers, because the translator, Alexander O. Smith, thought at the time that his target audience might not know what ramen was (Mandiberg 2015).

The localization was done following a co-creation approach. Smith would propose names and accents for characters, and discuss them with the international localization team, the producer, the writer, and members of the development team in order to “satisfy good English reception as well as fidelity to the original” (Mandiberg 2015, p. 119). When the dialogue would not make sense to a Western audience, Smith would freely rewrite it, so that it would work well in that context (Mandiberg 2015). The localization team also requested that a number of graphics were adapted, but as the funding for localization was limited, only a small number of graphics were modified (Mandiberg 2015). Thus, many visual references to Japanese culture were kept in the localized versions, such as an image of a Japanese samurai, a temple, images of Japanese food, clothing, furniture, and Japanese characters, thereby resulting in a hybrid game world set in the United States but with several traces of Japanese culture (Mandiberg 2015).

The next installments of the Phoenix Wright series were translated by Janet Tsu, who often writes about localization in Capcom’s blog. For Tsu (2014), localization is about striking a balance between textual accuracy and forging an emotional link with target players, which allows them to enjoy the localized game:

[…] as a piece of entertainment, the stories in games are primarily concerned with the feelings and reactions, or the “emotional experience”, of the player in its original language, and therefore, any localization must strike a balance between what is “textually accurate” and what is what I call “emotionally accurate”

(Tsu 2014).

For this reason, and due to the change in setting to the United States despite the constant presence of Japanese culture references in the games, Tsu decided to introduce new narrative elements in the localized versions to explain elements of the backstory, such as the founding of the Nine-Tails Village near Los Angeles, which is steeped in Japanese mythology (Tsu 2014, online). This creative addition to the narrative was deemed necessary to help players understand Japanese cultural references and to prevent having to remove or modify larger parts of subsequent games (Baird 2019). As the game is set in an alternate-universe version of Los Angeles, where Japanese culture blends with North American culture, fans of the series have come to call the hybrid setting of the game “Japanifornia” (Tsu 2014).

With regard to the game’s reception, overall, the localization of the Phoenix Wright series has been acclaimed by players and critics because of its hybrid nature; it is rooted in Japanese culture, but also includes popular cultural references that Western audiences understand (Consalvo 2016). The use of idiomatic language and the successful rewriting of numerous jokes and puns into English has also been well received (Consalvo 2016). In game critic Scott Baird’s words (2019):

The English localization of the Ace Attorney series managed to capture the spirit of the original, while also replacing the Japanese humor and references with content that would be understandable to an international audience. A change like this is necessary to the success of a text-heavy game and it’s why the Ace Attorney series has gained such an international following.

However, while reviewers and players who are not familiar with Japanese culture believe the localization of this saga is exceptional, some of the more hardcore fans of Japanese culture consider it to be one of the worst translations ever made due to the extreme level of adaptation, as shown in some debates in specialized players forums.9 This proves how difficult it can be to strike the right balance when localizing Japanese games and highlights the need to monitor the reception of localized games to obtain information about how target players’ preferences and tastes evolve over time in order to assess the best localization approach to be applied.

7. The Localization of the Yakuza Series: Introducing Dual Localization

The localization of SEGA’s Yakuza series has also evolved over time, from adaptation and domestication to what I call “dual localization”, which aims at catering for the needs of both fans of Japanese culture and mainstream players who are not necessarily familiar with it. The first game of the Yakuza series was released in 2005 in Japan and in 2006 in North America and Europe for the PlayStation 2. One of the first things that was adapted for overseas markets was the name of the series, 龍が如く (ryū ga gotoku, “like a dragon”), into Yakuza, probably to help identify the game as Japanese and use its Otherness to appeal to Western players (Navarro Remesal and Loriguillo López 2015, p. 9). The game was fully localized and dubbed into English. A cast of famous actors, such as Mark Hamill and Michael Madsen, was used, but the reception by players was not positive (Cullen 2019). Players did not like the voice acting, thought that the characters were mischaracterized, and felt that there was too much casual swearing that did not fit in well given the context.10 Due to this bad reception, for the next releases of the series, SEGA opted for partial localization; they maintained the original Japanese voice cast and provided subtitles in English and FIGS (Siliconera 2008).

However, in the North American version of Yakuza 3 (2010), some Japanese content was cutout, such as mini games involving hostess clubs, as well as mahjong and dating sims, probably in an attempt to increase the game’s transnational appeal (Navarro Remesal and Loriguillo López 2015, p. 9). Gamers and critics objected to this and, as a result, the localized version of Yakuza 4 had no content cutout (Navarro Remesal and Loriguillo López 2015, p. 9). In addition, the localization adopted a more foreignizing approach in order to stay closer to the original cultural content (Cullen 2019). In an interview with specialized technology website Engadget, Yasuhiro Noguchi, producer of Yakuza 4, stated that they had changed their localization strategy after carefully taking into account the feedback of Western fans:

When I took on the project, I reviewed the Yakuza franchise history in the West as well as the valuable feedback we received from our fans on Yakuza 3. Based on consultations with the Yakuza team in Japan, we decided to bring a more complete localization that was more faithful to the source material

(Noguchi, cited in Fletcher 2011).

This illustrates the influence that fan communities can exert over gaming companies through their interaction via forums, websites, and social media. However, despite the change of strategy in the localization of Yakuza 4, the game did poorly in the United States and Atlus did not decide to localize Yakuza 5 (2012) until PlayStation announced they would fund the localization after receiving numerous fan requests (Khan 2020). In this case, the power of the fans had an impact on a business decision regarding the localization and transnational circulation of a game.

In 2013, SEGA acquired Atlus and the Atlus localization team, with the addition of Scott Stritchart, have been in charge of the localization of the Yakuza series since, applying a similar foreignizing strategy for cultural content (Khan 2020). When a number of Yakuza games were remastered, some of the elements that had been adapted in the previous editions were retranslated in order to keep them closer to the Japanese original. For example, in Yakuza 3 (2010) fish names, which were in Okinawan dialect in the original game and had been given English names in the first localized version, were reverted to the original names (Cullen 2019). Moreover, some substories and minigames that had been omitted were included in the remastered and relocalized version (Botts 2020).

A special mention should be made to two spin-off games—Judgment (2018) and Fist of the North Star: Lost Paradise (2018)—and the latest game of the saga at the time of writing, Yakuza: Like a Dragon (2020), which features English and Japanese language audio versions and English subtitles for both audio tracks (Hashimoto 2020). This dual approach to localization started when the localization team at SEGA decided to translate the script of Judgement (2108) twice and provide an English dubbed version after many years with just subtitles (Scarpinito 2019). The new characters and storyline in the game were considered a good opportunity to introduce the fully localized version with English voices in order to appeal to a wider audience (Scarpinito 2019). Therefore, the localized game can be played both with Japanese and English soundtracks. There are two sets of English subtitles: one presents an English translation of the Japanese audio, as in previous games of the series, and the other one is a verbatim reproduction of the English voiceover (Glagowski 2020). Such a dual localization approach is designed to attract as many target players as possible, providing them with the gameplay experience they are seeking, be it a more domesticating one, with English voices and heavily adapted, or a more foreignizing one, with Japanese voices and a translation that stays closer to Japanese language (Strichart 2019). Such customization caters for the two traditional target audiences of Japanese games: mainstream consumers with little or no knowledge of Japanese culture and Japanese culture fans who expect the games to be as close as possible to the original ones.

Scott Strichart (2019), localization producer of the Yakuza series at SEGA, describes the dual localization method they used as follows:

Essentially, we took a base translation and then pushed it out into two different directions for Japanese audio and English audio. The Japanese audio got our traditional “Yakuza” pass, listening intently to each line and crafting the dialogue to suit it. The English script was written for actors to perform it, with more of a focus on making sure it sounded like things people would actually say in English. Sometimes, the two versions are totally the same! Others, it’s totally different [...] the English is still a faithful localization of the story, and the Japanese subtitles were still crafted with all the considerations for a good read that always goes into a Yakuza title localization.

(Strichart 2019)

The idea for such a dual approach came from the localization team and was accepted by the company, which illustrates how SEGA is willing to modify their localization strategies in order to attract a major fan base in the West. This new approach has been perceived as something positive in specialized blogs (Cullen 2019; Davidson 2019; Scarpinito 2019; Tailby 2019; Botts 2020) and by fans,11 and has helped the Yakuza franchise transform from a niche experience during the mid-2000s to a genuine hit with millions of fans with its later games (Davidson 2019).

As in the localization of the Phoenix Wright and the Persona series, the localization team for Yakuza believes that using natural, good sounding, and fluid English is the best way to reproduce the “intent” of the original game in the localized version (Riesenbach, cited in Cullen 2019). In order to achieve idiomatic language, the localization team is made of translators and editors, as in the Persona series. Such an approach seems to be working and has been viewed positively by critics. As game critic Joey Davidson states: “fans want localization efforts to pack a translation that matches the core meaning of the original product in working English” (2019).

8. Conclusions

Game localization is a functional and user-centered type of translation that aims at reproducing a similar gameplay experience to that of the original game for target players (Fernández-Costales 2012; O’Hagan and Mangiron 2013; Bernal-Merino 2014). Due to the linguistic and cultural distance between Japan and North America and Europe, Japanese developers and publishers have traditionally been willing to develop culturally “odorless” games (Consalvo 2016; Hutchinson 2019) or adapt their creations as necessary in order to guarantee a good reception (and strong sales) overseas (O’Hagan and Mangiron 2013). However, adaptation is a double-edged sword, and several factors need to be taken into account when deciding the global localization strategy, such as budget, schedule, the genre, and the theme of the game, as well as the age ratings in different countries, how previous games of the same series or the same genre have performed abroad, and the intended target audience (O’Hagan and Mangiron 2013), either a mainstream audience or a niche audience (Ranford 2017). At the end of the day, as Mandiberg states, “[the] key to most game localization is staying safely within the restrictive forces of taste and budget” (2015, p. 124). Thus, companies need to balance all these factors when deciding their localization strategy in order to maximize revenue.

After briefly outlining what game localization entails and describing the main localization approaches that have traditionally been applied to Japanese games, this paper has presented three brief case studies of the localization of three renowned and long-standing sagas set in Japan and culturally marked as Japanese: the Persona, the Phoenix Wright: Ace Attorney, and the Yakuza series. The descriptive analysis, which mainly involved the collection of information coming from Internet sources, provided by members of the localization teams and specialized blogs and websites, illustrates how localization and adaptation approaches for these three sagas have evolved through the years, from a more domesticating, “odorless” approach, to a more foreignizing, “fragrant” approach, that uses the traces of Japanese culture and its Otherness as a selling point, in order to attract a wider audience. Today, people around the world are more familiar with Japanese culture thanks to the “Cool Japan” phenomenon (McGray 2002), digital technology, and the global success of Japanese anime and manga (Consalvo 2016; Hutchinson 2019). Players expect Japanese products to preserve traces of the original culture and not to be subject to extensive adaptation, which sometimes is perceived negatively as censorship (Mandiberg 2017). Therefore, the tendency seems to be towards localizations that are closer to the original Japanese content but using natural sounding language that does not interfere with the gameplay and fosters an emotional connection of a player with the game. The emphasis Japanese developers place on using idiomatic language, adding new dialogues, and editing the target text as required to make it sound natural, highlights the importance translation has within the broader game localization process. Textual translation is key in order to facilitate target players’ access to the sociocultural content of the original game, which is essential to foster playability. Further studies with similar Japanese video games set in Japan would be necessary to confirm if the trend detected in this small case study also applies more commonly in the industry. Moreover, reception studies with players of the localized versions would allow for more information to be obtained about their preferences, which could help game developers and publishers fine-tune their localization strategies and keep their fans satisfied.

Gamers are very active and vocal on Internet forums and they discuss the localization of their favorite games in detail, clearly stating whether they like them or not and why (O’Hagan 2017); therefore, paying attention to them can help identify the most suitable localization approach, as in the cases of Atlus and SEGA. Similarly, the localization of Japanese games is often reviewed in specialized websites and blogs, and game translators are often interviewed and can also avail of company blogs to explain their translation decisions. This promotes communication with the target audience and lets the fans know how complex adapting a Japanese game can be, making them aware of all the different restrictions that may impact on localization strategies, such as time, budget, and ratings. Therefore, it could be argued that game localization is one of the most discussed types of translation. The agents involved in the production and reception processes of a localized game talk about it extensively in blogs, interviews, reviews, specialized websites, and community forums, establishing a dialogue that is impacting on localization strategies, and consequently, on the transnational circulation of Japanese games.

SEGA’s dual approach to the latest instalments of the Yakuza franchise deserves a special mention, as it brings localization to the next level by offering two types of localization: full localization with English dub and subtitles and with a higher level of adaptation, and partial localization with Japanese audio and English subtitles that are closer to the original Japanese. While dual localization seems to be the ideal, it may not be an option for many developers due to the additional time and budgetary cost it involves. However, this seems the best approach for Japanese developers and publishers who would like to reach a wider audience while keeping their niche audience also satisfied.

To conclude, it should be highlighted that localizing Japanese games is a collaborative, creative, and adaptive process that allows players from different linguistic and cultural backgrounds to enjoy the games in their own language while getting a glimpse of Japanese culture. Localization provides a hybrid space where the creators of a game, with their cultural background, and the target players, with their own, different cultural background, can meet, by virtue of the careful craft of the localization team, in a space where they can find each other in translation.