2. Bishōjo Games and Embedded Narratives
The development of the romance game genre is considered unique to Japan “so much so that 24 percent of [over 1500] respondents [in the US survey on romance games] automatically associate the term ‘romance game’ with Japanese products” (McDonald 2015, p. 39
). In addition, romance games or romantic components in games are widely considered to be “for women” (McDonald 2015, pp. 38–39
) in North America and Europe, whereas Japan’s romance games were born from the male-dominant culture surrounding early computer systems such as DOS and PCE in the 1990s and grew with fascinations with high technology. Starting with successful game releases such as Classmates
) (Elf 1992) and Tokimeki Memorial
(Konami 1994), the romance game genre quickly grew into a much larger market. The increased accessibility to computers thanks to the introduction of Microsoft Windows spawned long-run hits such as To Heart
(Leaf 1997) and Kanon
(Key 1999). The trend soon spread to family-oriented game platforms such as Sega Saturn and Sony PlayStation. Originating from text-based mystery and escape games in closest proximity to adventure games in the 1980s, visual novels were not initially associated with romance until these representative titles redefined the game genre as a mix of novels and dating simulations. As the market matured through the 2000s, the division between the Windows version and the family platform version came to serve as a standard marketing scheme to separate the original adult edition containing pornographic contents from the all-age edition. The screen composition consists of a visibly marked text box in front of stationary 2D characters and background panels, usually accompanied by character voices. As shown in Figure 1
, the screen layout of the visual novel games is highly characteristic to the degree that bishōjo games are immediately identifiable. Many of these titles recorded profitable sales in anime adaptations as well, as best exemplified by Fate/stay night
(Type-Moon 2004) and its long-run anime franchise (Studio Deen 2006–2010; ufotable 2014–2020) as well as Clannad
(Key 2004), whose anime series (Kyoto Animation 2007–2009) marked a breakthrough to a large television audience. Among these games, many can be identified as visual novels, although they include other gaming components such as strategies and RPGs.1
Their plot structure is best described by the narrative formula of a tree diagram or parallel plot with branching scenarios so that each scenario corresponds to each of the player character’s multiple love interests or taishō
(targets, objects) of gameplay. In this essay, the player character, or the protagonist/narrator character whose viewpoint is shared by the player, will be called PC, and the characters embedded in the game system as “target” characters corresponding to branching scenarios will be called love interests or LI. Following the existing discussions, I will first define the bishōjo games by several characteristics identified in and outside scholarship, which can be outlined in three terms—the decline of gaming elements, narrative strategies of immersion, and games’ proximity to literature.
The bishōjo game genre’s development is frequently equated with the decrease of gaming elements or the interactive and coincidental factors that change narratives according to each player’s gameplay. Visual novels’ narrative experiences may vary to some extent due to the “decision points,” which are occasional multiple-choice decisions to determine which scenario route the player will follow, but the player’s interaction with the game is limited to pushing a button to turn pages and read. The narrative intensity and complexity resembling novels made visual novels a rich source for “visually appealing and diegetically complex anime titles” (Cavallaro 2010, p. 8
), in contrast to the custom of game releases as promotional materials for anime and manga. The more the plot pursued depth and integrity, the less the text became interactive. As Marie-Laure Ryan argues, interactivity in gaming more likely conflicts with aesthetic design and the reader’s immersive experience; interactivity conflicts with the creation of sustained narrative development, and consequently, with the experience of temporal immersion (Ryan 2001, p. 258
However, early productions of romance games did not rely so much on narrative intensity and integrity. Gaming components were vital to dating simulation games in which the player controls the male avatar and uses numerical parameters to measure romantic relationships. Winning in the game was often judged by how the player balanced multiple abilities such as study, sports, and romance, resembling strategy games. Soon some game productions employed the technique of alternating between the “gaming” mode, such as RPG battles or parameter-driven dating sims, and the “romantic story” mode formulated as a novel-like reading experience, as represented by the Tokimeki Memorial
series (1994–2009) and the Sakura Wars
) series (Sega 1996–2005). Eventually, game elements turned into extras attached to the novel part, such as To Heart (Leaf 1997), which simply added “mini-games,” including shooting, puzzle, and action, to the romance plots. Azuma Hiroki points out that games in the 1990s, especially Shizuku
(Leaf 1996), changed the conceptual foundation of the bishōjo game, which “were once designed for [the user’s] communication [with a computer], into the content-oriented media” so that games can be consumed as narrative contents (Azuma 2007, p. 203
, written in proximity to horror and mystery novels, reduced the number of decision points and endings for the sake of plot coherency and emotional impact. These games retain the flowchart structure, but the multiple scenario/ending framework does not signify that the game is open to many endings and choices. Typically, there are only as many scenarios as LIs, and the multiple choices, while they may make minor differences here and there in character reactions, are intended to direct the PC–player to one of the branching prescripted scenarios. Playing bishōjo games eventually came to signify spending hours by passively reading the scenario that appears on the screen, while “an interaction between the player and the game system is limited to the act of selecting one of the two choices, which occurs only once every few hours” (Azuma 2007, p. 199
A lower degree of interaction in no way signifies that the player’s psychological engagement in reading is less active or emotional: rather, many bishōjo games are known as nakigē
(crying game), or games that make the player cry by provoking intensely emotional reactions to the story. The player’s self-referential projection on PC intensifies, although (or rather, “because”) interactive components are reduced. The interactions between the player and the game came to be internally embedded in the narrative itself, which is one of the most prominent features of bishōjo game scenarios. In other words, the player’s own experience of gameplay appears to be already embedded in the written script, which enables the PC to act as if he is simulating the player’s gameplay inside the game. The most prominent functional role shared by the player and PC is the repetition of the story. While reading a novel is modeled as a chronologically linear experience with a beginning and an end, nearly all romance games, equipped with multiple endings and parallel scenarios, are designed on the premise of repetitive play or repetitive reading. The player repeats gameplay to discover multiple endings and their corresponding LIs as if PC repeats the same timeline to meet his promised love. In fact, repeating the same scene, motion, and narration is an essential element in video games across genres as gaming often equally means for the player to learn and anticipate the pattern of the computer-programmed moves and behaviors. While other narrative media may expect the audience to read or view once, “video games are designed to be played multiple times by a player” and “[l]earning the patterns of behavior and working around them is usually itself part of the game” (Wolf 2002, p. 81
). With romance games, replayability holds special importance because gameplay equals the player’s repeated returns to saved points or decision points of the story to make different choices in an attempt to unlock all LI scenarios and endings.
The player’s repetitive play can be accordingly projected as simulations of PC traveling time to return to the past so as to make different choices and discover alternative courses of events or divergent worlds. It is no coincidence that the visual novel genre originally began as horror games such as Otogirisō
(Chunsoft 1992) and Kamaitachi no yoru
(Chunsoft 1994), in which the player makes multiple attempts to escape from, or prevent, the doomed death of PC or their partners. The branching plot structure is also an effective device as it allows the player multiple attempts to solve a mystery by accumulating information from repeatedly reaching “bad end” scenarios. Bishōjo games refined this model by weaving together parallel plots and romantic narratives so that the PC–player’s romantic (and sexual) interests in LIs effectively lead to multiple endings. The popularity of the replay plot in bishōjo games allegedly peaked with video/computer games such as Gun Parade March
(Alfa System 2000), Air
(Key 2000), Higurashi: When They Cry
(Higurashi no naku koro ni
) (07th Expansion 2006), Ever 17 (KID 2002), and Clannad
. Many of these narrative-heavy games were soon adapted into television animation series, and the popularity of the looped plot became more globally visible and accessible through anime such as Steins;Gate
(2011 anime adaptation from Nitroplus 2009) and Puella Magi Madoka Magica
(directed by Shinbō Akiyuki in 2011). Azuma locates these cyclic narrative traits employed in novels themselves, especially Sakurazaka Hiroshi’s novel All You Need Is Kill
which was published in 2004 and adapted into the Warner film Edge of Tomorrow
in 2014. In All You Need Is Kill
, the plot is entirely conceived based on the framework of gameplay, in which “play” must be repeated until the protagonist, who is virtually a game player trapped in the repeated timeline, wins the deadly battle with invasive aliens and survives to live the next day. In this narrative model, the plot typically centers on the protagonist who repeats the same timeline until he finds the right ending, e.g., one of the parallel worlds where he survives the “fate,” or the tragic end leading to PC’s or LI’s death. Another representative example is Higurashi
, a novel in the game format that is no longer equipped with occasional choices for branching paths and instead realigns supposedly synchronic parallel plots into a linear diachronic narrative structure. Once the “gameplay” reaches a tragic end in which mass murders happen, and the character who “plays the game” in the scenario is killed, she resets and restarts the same timeline from day one. The gameplay accordingly simulates the player’s own gameplay, including accumulating knowledge and skills, making different choices, and ultimately moves the clock forward to find a breakthrough to a happy conclusion. Steins;Gate
also prioritizes plot coherency and emotional drama over contingency by the player’s own choice, which successfully builds “the equation between the protagonist’s and the player’s viewpoints by means of synchronizing the protagonist’s and the player’s memories” (Ishii and Shimokura 2013
The cyclic narrative experience in and outside of the game generates the illusion of the player–PC identification and thereby immerses the player in the diegetic world of the story. This immersive effect of bishōjo games contradicts many of the premises proposed in game studies. In the narrative game experience, these pre-scripted narratives are called embedded narratives, which hold “pre-generated narrative content that exists prior to a player’s interaction” (Salen and Zimmerman 2003, p. 383
). In contrast, emergent narrative “arises from the set of rules governing interaction with the game system” (Salen and Zimmerman 2003 p. 383
) and is thus open to the player’s choice of action based on the way games function as systems. Game scholars have argued that the game’s openness to one’s choices and the freedom of control over actions have provided a sense of freedom and thereby reinvented the concept of reality in games. For instance, “reality consists of a mapping not of appearance but of control” (Rehak 2003, p. 107
), namely, the reality of interfaciality between the player’s manipulation and the game avatar or game world’s obedience to it. In terms of narrative structure, the simulation of narrative, especially how much the player can control and influence the course of events and narratives, determines the degree of realism. Conversely, in old media narratives such as literature and cinema, realism relies on traditional values, including mimesis, chronological order, and cause–effect relationship. Video games often seek “social realism, which is achieved by designing the world to match the real one, with streets and stores and parks, as well as organizing rituals and ceremonies that enable players to identify their social place in the world” (McMahan 2003, p. 75
). The sand box-type games that allow “mods” (modifications) to intervene in the original programming and alter appearances or rules of the game world provide a further advanced sense of freedom and reality thanks to the higher level of control over how one can play the game. In contemporary gaming experiences, this heightened sense of reality configured by the game’s openness to modifications promotes the player’s immersion in the game world.
In bishōjo games, by contrast, the illusion of identification between the player and PC is created through the frustrated experience of the narrator/protagonist trapped in the literary text. The simulation of repetitive gameplay, often embedded in the written scenario itself, illustrates a suffocating closed system of the narrative world where characters are controlled by the pre-determined plot or “fate” from PC’s viewpoint. In literary texts, characters under the author’s control are often ignorant of the meta-world outside the text where they live, but PC in bishōjo games may theorize and dream about the outside of his diegetic world. PC and the player, who share the same functions across the boundary of the game world, builds sympathy through the shared perception of parallel plots and thereby come to equate the game system with the force of fate. The parallel plot structure simultaneously provokes a threat to the player because the narrative structure itself reminds him that there is a “real” world outside the game that transcends all parallel plots. In theory, any game with a player character by default implants the double consciousness of play; the player–protagonist relationship can be intense and emotional, but the player is simultaneously fully aware of the character as an artificial construct (Salen and Zimmerman 2003, p. 453
). To counteract this internal narrative defect, some games tactically embed the PC–player gap into the pre-generated scenario itself. For instance, One: kagayaku kisetsu e
(Tactics 1998) incorporates the plot device that forcefully takes the PC into “the world of eternity,” a simulated reality where the PC becomes forgotten by game characters and restart his life. This simulative “outside” inside the game allows the player to project his identity on PC in the form of immersion in the game (Honda 2005, p. 97
) and helps the player misperceive his gameplay to be “gameplay as a metaphor” within the game (Azuma 2007, p. 212
). The scripts by the bishōjo game scenario writer guru Maeda Jun, such as Air
, and One
, equally utilize this internalized meta-world to amplify the player’s emotional immersion in the scenario. In Clannad
, for instance, multiple scenes of a transcendental dream-like world are inserted to imply and repeatedly remind the player that there already exists an external sphere where the PC is equipped with an ability to restart his life. PC must repeat his life to save each LI, which will eventually lead to the main LI’s recovery from illness. More recently, the isekai
(otherworld) plot formula more efficiently fulfills the same function by omitting the process to invite the player to the game world immersion. In the isekai
genre novels, the protagonist immediately dies in a car accident (or by a heart failure) at the start and reincarnates into a game world. The immersive narrative effects of bishōjo games have developed by internalizing game functions into literary texts in the form of plot and metaphor. Once the narrative formula of gaming fully transforms into novels, the bishōjo game genre may simply lose its raison d’être
as a game and yield itself to novels that employ the rules of the game.
The cyclic narratives further extend the identification process to the player/reader by immersing them in the illusion of gaining the power to overwrite the text. This mode of repetitive reading explicitly stems from new practices of computer-generated gameplay but simultaneously revives and reaffirms a highly masculine model of the empowered self by means of romance narratives. This reading practice establishes the illusion of the player–PC overriding the power of the author and thereby becoming the author himself, or more precisely, the programmer or administrator of the game world. Simultaneously, the player as a transcendental being observing the parallel scenarios becomes inevitably aware that these scenarios cannot coexist—they are often parallel worlds that are mutually exclusive and contingent on PC’s love choice. If the PC saves one of the heroines, the others, although quickly removed from the PC’s sight once they are deemed unselected, may quietly undergo their tragic ends because of illness, accident, battle, or the like. This is also how bishōjo games coincided with the so-called “world-type” or sekai-kei
stories of the 2000s. The plot commonly centers on the male protagonist, an average middle or high school boy, who meets a girl. She is, in fact, doomed to face a global or galactic crisis and fight in order to save the world. The personal, romantic relationship is thus projected onto the world crisis without contextualized settings such as government, army, or society. Although it does not overlap in entirety with the romance game genre, this setting characteristic of the 2000s grew in close parallel with romance game narratives (Azuma 2004, pp. 15–16
). Against the disappearance of social contexts, the story exclusively zooms in on the romantic relationship of “I” (boku
) and “you” (kimi
). As Kasai Kiyoshi and Genkai Shosetsu Kenkyukai attest, the trend is considered to have emerged with TV anime Neon Genesis Evangelion
(Anno Hideaki 1995–1996), peaked in the 2000s with such works as Shinkai Makoto’s 2002 anime film Hoshi no koe
[Voices of a Distant Star]
and Akiyama Mizuhito’s novel Iriya no Sora UFO no Natsu
(2001–2003) (Kasai 2009, p. 21
). The world-type narrative ideology of choosing one girl he truly loves while sacrificing others, sometimes even abandoning the entire world, although the girl he “truly loves” may change in each gameplay. This choice can be ethically questionable and radical because the decision would not only confirm a world (and a love interest) but also results in the destruction of other worlds and other potential loves (Howard 2014, p. 384
). The PC enjoys the privilege of ignorance about tragedies in parallel worlds while acquiring the sense of achievement with overcoming the fated scenario.
As PC replays and chooses the path to save the girl he loves, he increasingly gains the sense of recovering his agency through decisions he makes on the future of the world, however passive and automated the decisions in games may be. As Murray states, the computer user finds pleasure in obtaining agency as “agency is the satisfying power to take meaningful action and see the results of our decisions and choices (Murray 1998, p. 126
). The PC–player seems to gain the sense of being an autonomous and unique individual by sharing the game form and the thematic narrative of the game, as “meaning emerges from a game’s mechanics—the set of decision and consequences unique to each one” (Johnson 2012, p. 33
). Ötsuka Eiji also argues that in the narrative consumption, especially in the post-1970s loss of the political grand narrative in Japan, video games have played a significant role as they provide game players with the illusion of becoming an active agency of creating a new and original story while passively playing and consuming the prepared plot (Ōtsuka 2012, pp. 13–14
). In Ōsawa Masachi’s argument, the gameplay, conceived as a struggle for the end of the story, further reflects today’s social climate where the loss of the ending in public discourses corresponds to fandom where fan-based productions repeat settings and characters to defer ending (Ōsawa 2008, pp. 200–1
). Azuma compares bishōjo games to a simulation of growth by means of conquering girls, emotionally and often sexually, so as to recover paternal authority (Azuma 2007, pp. 310–11
). Against postmodern challenges, from the disappearance of narrative to the decline of belief in individual freedom, the PC–player rediscovers and reaffirms the autonomous self as a unique individual who is capable of choosing a love by means of choosing the right world among all alternatives. The process of repetition produces a sense of control over history and narrative: by sharing the viewpoint with the player, PC’s transcendental ability to interrupt the authorship of the story provides the reader with the illusion that he is actually creating the story by consuming stories. This postmodern illusion of the modern agency further contributes to the sense of conquering obstacles and attaining his object of love.
Bishōjo games reveal the way computer-generated game narratives have both revolutionized the literary reading practice and simultaneously reaffirmed the metaphorical power of the modern self through the rhetoric of romantic love. The mode of story consumption reveals postmodern characters framed by repetitive gameplay, whereas the ideological messages deliver by the scenarios largely return to the idea of romantic love as an ideological tool to reaffirm personal freedom and patriarchal control. The intensely literary experience of gameplay generates the illusion of modern realism as it resorts to the recovery of male subjectivity through narrative technology of romantic love.
3. Otome Games and Emergent Narratives
Since existing academic discussions of visual novel narratives have exclusively focused on male-targeted games, those game narratives have been theorized as if female players of romance games do not exist. Accordingly, the discussions presented above precluded many of the characteristics relevant to gameplay mechanisms in women’s romance games. Indeed, the female-oriented game market was a rather marginal category until the game industry rapidly expanded to portables and smartphones in the 2000s. The game genre with reverse gender codes to bishōjo games, namely those with a female PC and multiple male LIs who have their respective scenarios, came to be commonly called otome games. The symbiosis of computer culture and romance, as seen in bishōjo games, limited the genre’s marketable demographic to men with their own room and computer for the luxury of enjoying sexual content and did not transfer well to women’s side. Early otome games such as Angelique
(Koei 1994) and Haruka: Beyond the Stream of Time
(Harukanaru toki no naka de
) (Koei 2000) were released for video game consoles, including Super Nintendo and PlayStation, without any pornographic “original” editions, which were released preceding family versions in male romance games. Even without sexual content, these consoles, often placed in a room open to attention from family, restricted romance gaming to limited types of occasions and consumers. The growth of otome games thus depended on portable games with higher resolution images that emerged around 2005, namely Play Station Portable and Nintendo DSLite. Later, Play Station Vita became the main medium of otome games, leading smaller productions to major cross-media marketing successes such as Hakuōki
(Idea Factory 2008) and Uta no ☆ purinsusama ♪
(Broccoli 2010). The miniaturization of the game platform has further stimulated the market to expand to different age groups of women. Recent surveys on the gaming population in Japan report the fast increase of middle-aged female gamers, evidencing the shift of the gaming population from teenage male players in the 1990s to female players consisting 40% of the gaming population in the 2000s.2
In 2014, Sega Networks reported that the smartphone game market had possibly reached its saturation point, now starting to decrease usage. The only exception is the emergent group of female gamers in the age group of 35–49, known as “around 40,” who are casual players of romance simulation games (Sega Networks Game Style Research Lab 2014
). According to otaku market research by Yano Research Institute, the men’s romance game market has shrunk by half since 2007 and is currently a quarter or less of its peak, whereas the women’s romance game market has steadily grown with an annual 10% increase.3
The otome games, being a latecomer to the market, emerged as an experimental amalgam of heterogeneous game genres, some new and some borrowed from established recipes in men’s genres. On the one hand, they closely followed the designs of male-centered dating games by means of imitating both the game format and narrative structure while converting the PC’s gender to female and LIs’ to male. Tokimeki Memorial Girl’s Side
(Konami, PS2, 2002–) and its serialization continuing to present is a typical example of the simple reverse gender approach. Oddly enough, the “girl’s side” has sustained popularity while the original “male side” had ended with the last title in 2009. Some visual novel games in the otome genre seem more closely modeled after the narrative structures of bishōjo games’ cyclic time, such as Amnesia
(Idea Factory 2011) and Clock Zero
(Idea Factory 2010). On the other hand, otome games employ gaming and narrative schemes, which had to be created from a blank slate. According to Erikawa Keiko, one of the founders of Koei as well as of Koei’s otome game label known as Neo Romance, the making of women’s romance games began by first hiring female workers and training them as programmers and game makers (Famitsū 2015
). Since the 1980s, Koei is a pioneer of the strategy games and samurai war tactical simulation games set in historical Japan, such as Nobunaga’s Ambition
series (1983–) and Kessen
series (2000–2004). The borrowing of military war simulation into women’s romance resulted in a unique reinvention of dating games in the conceptual framework of battle-driven historical dramas, simulating military strategies and political conflicts. Although it is difficult to judge whether Koei’s approach triggered the trend, romance in otome games tends to take place in historical settings; or more precisely, the diegetic game world of otome games tends to be filled with war, ranging from military combats to political terrorisms, into which the female PC is thrown unintendedly. This tendency reveals a clear contrast to bishōjo games usually staged in Japanese school students’ contemporary everyday lives.
The formal features of otome games can be best described by the persistence of the emergent elements. Antithetic to embedded narratives, emergent narratives are more ruled by the mechanic functions of the game system and contingent on each player’s choice of action. For instance, the player may choose to train PC for more skills, buy an item at a shop, move the PC to the next destination, or simply speak to one of the LIs. The feature was standardized in the earliest dating simulation games for men but quickly became obsolete as the visual novel style with embedded narratives soon dominated the genre. Sustaining the emergent gaming elements in otome games resulted in partitioning the gameplay into multiple modes between the story mode and the play mode in mixtures of visual novels, RPG battles, and strategies. The disconnection between different segments of gameplay disturbs the interrelationship between the embedded and the emergent narratives, as the alternations of modes regularly remind the player of the simple fact that this is a game and the player exists outside the game. For example, in La Corda d’Oro
(Koei 2003), PC’s major task is to earn points from practicing the violin and impressing the audience, which can be used to purchase items such as presents to LIs and dresses for concerts. The PC must use her time efficiently to discover music scores, develop artistic interpretations, find a partner to play them in ensembles, and finally, compete in contests. In this model, numerical romance parameters are scaled to the PC’s musical efforts, which directly mediate the player’s labor to gameplay. The second title in the Neo Romance series, Haruka: Beyond the Stream of Time
, and its long-lasting series to date, commonly employ the style of alternating scenario-based drama parts in the visual novel format with turn-based role-play battles. A map screen mediates them to allow the player to move the troop of PC and Lis, similar to strategy games. The intermittent battle sequences force PC to be visually and vocally represented on the screen, which reminds the player of their un-immersive presence or the player’s lonely existence outside the game. Some game titles employ narrative techniques shunned as taboos in men’s romance games. For instance, Norn9
(Idea Factory 2013) provides three different PCs and their respective sets of LIs, which enables the player to experience multi-subjective narrations and parallel plots. While innovative in its approach, the narrative structure simultaneously prevents the player from projecting oneself on a single PC. Adding voice to PC’s own lines is another approach that accentuates PC’s fundamental difference from the player. As seen in Figure 2
, it is also common in otome games to show PC’s own character face in the text box to accompany PC’s narration and dialogues, which is hardly acceptable in bishōjo games, especially in those where PC is designed as a faceless character to invite the player’s identification. The gaps of sexual or romantic knowledge between PC and the player further contribute to the dissociation as PC is overly innocent and ignorant of romance matters, let alone sexuality, in contrast to the player whose gameplay is primarily driven by a desire for romance. These formal characteristics to retain the PC–player divide are not simply sustaining distance between the two but are also likely to re-present PC as an object of affection and visual pleasure for the player rather than PC as a locus for the subject position. They indicate crucial differences from the male-centered model of romance gaming, which affirms that narrative features converge on the player–PC identification. Otome games seem to take advantage of the player’s detachment from PC and her game world inhabited by LIs to decenter the player’s identity and desire, thereby resisting the narrative effects of immersion.
The decentering approach further extends from the game form to the narrative mechanism. The key plot factor among commercially successful otome games that set the standard for the genre is the historical setting simulating battles and political dramas of the premodern or early modern eras. Conversely, some titles do not fit this description, I will focus on the two game titles that widely defined the otome genre, which are the Haruka series and Hakuōki. These will show how romantic love as a tactical measurement is aligned, or more precisely not aligned, with romance in the narrative form. Throughout the Haruka series, PC is a contemporary schoolgirl summoned to another world that resembles historical Japan. The periods she lands on range from the Heian era (794–1185) to the Taisho era (1912–1926). She learns that her identity is the priestess of the Dragon God, whose mission is to help her Eight Guardians to battle the evil force. The Eight Guardians, who are the game’s LIs, represent different elements of the constellation commensurate with the five elements in the Taoist cosmology of Ying/Yang. For example, the third game in the series draws the main plot and setting from The Tale of the Heike (Heike monogatari), a Japanese literary classic compiled around the 14th century in the form of a war epic narrating the historical battles between Taira and Minamoto clans in 1180–1185. The LIs are major legendary characters from the tale as well as imaginary ones across the Taira–Minamoto divide, such as Yoshitsune and Benkei, two popular figures on the Minamoto side in the epic, and Atsumori (1169–1184), a young samurai of Taira whose death was repeatedly adapted to Noh and Kabuki plays for tragic beauty. Hakuōki is another otome game example formulated by the narrative structure of history, but unlike the otome game examples I have presented above, the game strictly employs the visual novel format, which limits emergent narrative factors to occasional decision points. Set in the Bakumatsu era (the 1860s), the turbulent time of Japan’s modernization at the end of the country’s two-century-long isolation policy, military and political conflicts occurred between the pro-shogunate Tokugawa forces and the imperial nationalist forces that advocate Japan’s Westernization. PC travels to Kyoto in male disguise looking for her father, where she encounters zombie-like samurai called rasetsu and soon finds herself becoming a member of Shinsengumi. Shinsengumi is known as a military police force representing the pro-shogunate old-school samurai doctrine against the modernizing imperial government. LI characters are adapted from historical figures of Shinsengumi widely popularized through historical novels and films, such as Hijikata Toshizō (1835–1869) and Okita Sōji (1842–1868).
Due to the commonly known historical facts, the romance scenarios struggle between history and romance. On the one hand, the known course of history forces LIs to accept their doom to lose and die at a point in the war. The Guardians in Haruka
3, if siding with the losing Taira clan, are destined to perish, but even if siding with the winning Minamoto clan, such as Yoshitsune and Benkei, they are eventually hunted down and killed by forces sent by Yoshitsune’s half-brother Shogun Yoritomo. Similarly, Hakuōki
’s LIs, being Shinsengumi commanders and captains, must yield to the historical necessity that all but a few die in battles before modern Japan rises. On the other hand, similar to other otome games, these are equipped with alternative endings that lead to happy conclusions in which LIs survive to live a modest happy life with PC. PC in Haruka
3 is capable of using a dragon scale to leap time and redo the past to evade the target LI’s death. Similarly, Hakuōki bears multiple endings to ensure that the player finds the couple’s happy incognito life. The multiple endings raise issues on how endings should be valorized: the happy ending is the game’s right ending but cannot be fully regarded as the “correct” ending due to the dilemmatic narrative tensions between historical correctness and romance. LI’s death leading to “bad ends” must be overridden by the player–PC’s choice but does not nullify the historical value of the heroic samurai’s loss. In fact, Hakuōki adds a fictional setting to resolve this dilemma in a curious manner. Shinsengumi is not only a militia with old samurai doctrines but also a façade to conduct experiments with an elixir in order to artificially produce empowered samurai under the shogunate’s order. This elixir is made from the blood of the superhuman race known as oni
in Japan or as vampires in Europe. Once the game advances to a particular LI’s branching scenario, the LI may face an unbeatable enemy and choose to take the elixir to win the battle or protect PC, which enables him to gain power temporarily but consumes his life energy profusely. The addition of this setting ensures that the ending with a happy life with LI hints at LI’s imminent death due to his deteriorating health. It is important to note that these narratives take advantage of the common cultural essentialist view of samurai aesthetics that transience of life, or rather, death in a word, is more beautiful than continuing to live. The common taste for the loser’s side has long been identified in Japan’s traditional tales and films, as suggested by the hangan biiki
culture (Satō 1958, pp. 60–62
) and the mono no aware
aesthetics of the nostalgic samurai drama (Desser 1992, p. 148
These characters continue to be heroes for the Japanese because they are losers of history, embodying the lost tradition and the melancholy over the samurai spirit that disappeared from Japan.
These historiographical facts or samurai philosophy are not thematically important in themselves, but more significantly, they serve to persistently marginalize the PC, and the player by extension, from the game world. Historical accuracy in the scenario disapproves the player’s possible illusion of owning control over the scenario of the game world, nor over the historical events of premodern Japan that had ended before modern Japan. In other words, this narrative structure leaves very little room for the player’s illusion of taking over the scenario’s authorial voice, thereby making the location of authorial control obscure. This dispersed authorial voice of the scenario invites the player to perceive oneself as a separate entity from history, from romances between the PC and LIs, and even from the game world itself. The otome game model makes use of “history” as a powerful counter-discourse to personal romance scenarios and thereby resists the integration of narrations among multiple subject positions and stories. In addition, the world of samurai LIs constitutes a male-centric world of camaraderie in which PC is a marginal participant, guest, and observer. In this respect, the game can be effectively set in another world than samurai Japan where male dominance is structurally ingrained as aesthetics such as gangs and aristocracies. The historical setting, however, more profoundly plays into the gender division, which endorses the male dominance of the world filled with battles, while the PC remains marginal to the men’s world. The PC’s appearance faithfully embodies her role, namely, a cute girl in male attire: the PC acts as a member of the male society and plays an active role (Kim 2009, pp. 179–80
), but men move history forward, and the PC’s agency in the scenario is often limited to be an object of love for game characters, and for the player as well. Regarding bishōjo games, I argued that LIs (and their worlds) are often mutually exclusive, which may raise an ethical question of how the unchosen LIs may die or disappear in parallel worlds. By contrast, LIs in otome games seem to coexist as a team whether PC chooses to love one of the LIs or the other, as they form a systemic constellation comprised of different elements. LIs as a team also leaves room for the player’s freedom to imagine homosocial (and homoerotic) relationships among them while the player simultaneously enjoys potentially lesbian affection to the PC.
PC’s marginality does not only bind her to the passivity of the feminine gender but also bestows power on her, often by birth. For example, PC in Hakuōki is later revealed to be an oni herself, whose blood contains the same mystical ingredients as the elixir that transforms humans into superhuman fighters. Her feeding of blood to some LIs serves as symbolically erotic scenes highlighting romantic relationships. PC in the Haruka series is a shaman figure who mediates power from the Dragon God to her circle of guardians. This approach is not limited to historical otome games but can be found in those with contemporary settings, among which Hiiro no kakera (Idea Factory 2012) deserves mention as the game probably refined this archetype for the otome genre. In this visual novel game, LIs embody the guardian roles representing different subhuman species and obey PC because the ancient contract binds them to PC, the hereditary shaman who owns mystical power. As one character says, the “blood,” namely a system of contract they simply obey, defines their roles: it is not a question of accepting or comprehending their duties. This powerless yet shamanic center figure is rare in bishōjo games, but one character particularly comparable is Rika in Higurashi, the shaman of the village who bears the ability to repeat time yet passively executes it by acting cute and depending on her team of friends to escape the repetition. The PC’s role as a mediator of power and contractor of the relationship suggests a significant premise about otome game, namely, that PC is functionally comparable to the game player. The PC simulates the role of a game player—a casual and unskilled gamer—as she feeds special skills and spiritual energy to characters who fight. This is how the historical setting itself may not be as important as the mechanical similarities between the player and the PC. For instance, a popular title with a contemporary setting such as Uta no ☆ purinsusama ♪ efficiently builds the operational PC–player similarities, although LIs are not a squad of samurai but a team of male idol singers to whom PC feeds songs for competitions. The PC is a seemingly powerless center of the network of fighters but executes the power of mediation that permeates through LIs who continue to fight on the frontline. The PC and the player of otome games may play on their own gender inferiority in the male-dominant world so as to be protected and loved by LIs but functionally feeds psychological and spiritual energy to LIs.
These otome game characteristics highlight their tendencies to decenter locations of desire, decompose identifications, and reject the player’s immersion. In this paradigm realigned by a new mode of gameplay, romantic love opposes the modern model of the autonomous self as generated by bishōjo games. If bishōjo games pursued strong narrativity and immersion, otome games show lenient inclinations toward what Ryan calls interactivity of digital texts. The otome games’ adherence to emergent narratives attempts to dismiss the authorial control from the scenario and seek narrative meanings in the complexity of interaction with the game system itself. The PC’s agency in romance is provided as a simulation of the game player who may lack control, power, skills, or experiences but owns the ability to distribute skill points and magical energy to the fighting heroes. In this respect, romance serves as an interface with the game system itself, as this model assumes the PC simulates the position of the game player and reminds the player of the untransparency of the medium. It is likely that the operational interaction with the system is more complex and rewarding and sustains the misperception of the game system as romantic relationships.
4. Voice and the Visual Form
The otome game model I discussed above explicates mechanisms behind the formal and narrative character of the genre but further mystifies the location of pleasure in women’s gameplay. Understanding otome games is particularly challenging since contemporary theorizations of games and fan culture are largely framed by the male-oriented desire for retrieving (the illusion of) power, as well as for obtaining ownership of characters as sexual objects. Uno Tsunehiro points out, for instance, that the bishōjo games and surrounding sekai-kei culture protect the consumers’ fantasy of sexually conquering female characters who unconditionally accept the male protagonist (Uno 2008, p. 205
). Saito Tamaki argues that the fundamental difference of otaku, or fans of anime and games, from other fans, is sexual desire directed toward fictional characters (Saitō 2000, pp. 50–53
). Without the illusion of the PC–player identification and of the subsequent sexual rewards, how does the player find pleasure in playing romance? If the player’s labor, or interfacial dedications to the game system and LIs, generates the illusion of romance, how does the player receive a reward when the player remains foreclosed from the game? Based on the hypothesis that otome games shift weight from the consumption of the story about the self to that which remains after the disintegration of the components that constituted the story, I will shed light on the audiovisual aspect of the games. Despite the different narrative schemes discussed above, otome games and bishōjo games share the visual design, which makes the genre immediately identifiable due to their unique screen composition. They also both draw upon the fetishism of characters and their voices as the primary parameter to drive game sales and motivate gameplay. In this section, I focus on the audiovisual aspect of the game in order to examine how it constitutes the pleasure of gameplay and thereby investigate the function of romance in otome games.
The visual novel’s game format is characterized by the absence of visual sensation because animation factors, whether in character motion or background surrounding, tend to be limited to the minimum. These formal characteristics modeled after classic adventure games of the 1980s to early 90s have not changed their foundations, although illustrations and character designs continuously become updated and more detailed. The characters in the frontal medium shot stand against a painted landscape, leaving visual disparity from the foreground characters. The character figure standing behind the text box is hardly animated, usually consisting of a few stills of illustrations. The character’s motion is signaled by the change of the posture, the movement of hands or eyes that occasionally blink or shift the direction of gaze. While these animations may somewhat convey the character’s mood, visual information that contributes to the description of the story and the character’s psyches is minimal. These visual traits common across visual novels and romance games pose critical challenges to some theoretical frameworks of technological evolution and game. They function in different modes of pleasure production than games with movie-like spectacles and three-dimensional open stages. The romance games resisted the technological advancement of game systems, animation of movements, and image rendering. Lev Manovich, for example, envisions continuity and modularity as new visual media production in games. He argues that “where old media relied on montage, new media substitutes the aesthetics of continuity” (Manovich 2002, p. 135
). In his prospect, visual media will gradually shift from cinematic deception known as continuity editing toward spatial presentation as a coherent whole in the VR-style representation of the world. While in traditional cinematic production, the object, i.e., the end product, is a single-stream output in which constituent elements are no longer accessible, digital compositing means to select and assemble components that can remain separate without losing individual elements’ independence. This model presents the visual integrity and controllability of module components as important factors for the perception of reality. The visual narrative method of the romance games was initially a result of the technological and financial constraints of video game production but thereafter maintained the same style of layering illustrated stills and text boxes regardless of technological advancement. The characters also closely mirror the unmimetic style with flat compositions and unrealistic facial designs common in manga and anime. Cavallaro points out that the self-conscious artificiality of anime-like character design in visual novels promotes immersion (Cavallaro 2010, p. 3
). Azuma argues that the romance games’ resistance to technological innovations occurred due to the game’s need to simulate the act of reading novels (Azuma 2007, pp. 200–4
It is worth noting that these arguments on visual novels are solely based on bishōjo games or romance games with a male PC and female LIs, excluding otome games in their entirety. Although they share similar visual forms, I have argued that the stories and game systems of otome games show rather different narrative mechanisms from bishōjo games. If bishōjo games’ visual form and scenarios converge on the effects of immersion in and identification with the game world, the reward of the game already resides in the literary mode, i.e., simulating PC’s cathartic act of saving and conquering the girl he loves and reclaiming power as the author and as the father. If otome games resist this form of male pleasure and instead pursue the fluid maneuverability of mediation and disintegration among multiple layers of gameplay, a term that may best express this elusive pleasure is juiciness. For example, unlike hard-core players, those more attracted to fiction and interaction are keen on receiving positive feedback from the game as a reward for the player. The reward does not contribute to character growth, such as bonus skills or high scores in the game, nor to benefits in the player’s reality. According to Jesper Juul, juiciness is an immediate pleasurable experience that gives excessive positive feedback, which makes the player feel powerful and in control (Juul 2009, p. 45
). What distinguishes this pleasure from established ideas of reward is the fact that this pleasure is non-diegetic or pleasure felt by the player in the form of an interface with the game. Therefore, the praise given to the player does not affect growth or winning inside the gameplay and assumes the player’s externality. Due to these features, juiciness may be an “enigma” (Juul 2009, p. 49
) in comparison to existing concepts of pleasure, but this notion helps lead our discussion to the mysterious component of romance games that has vigorously driven the market, i.e., the voice.
What constitutes the narrative foundation of the game is not only the visual language and the verbal text but also the voice, which has carried a vital role in the game narratives. As part of the overall media-crossing industry encompassing anime and its related franchises, romance games have primarily relied on the commercial power of voice due to the intermedial nature of the game products. This voice-centered culture of Japanese anime and video games may have originated from the low-budget and cross-media condition in which voice is often the sole factor that remains unchanged across different media formats, visual styles, and story settings. Even when no budget for animation is available, voices and scripts may suffice the need for the production of the visual novel. In other words, voice undertakes the function of the imagined character continuity across the visual and narrative discontinuities, thereby compensating for the lack of cross-text integrity. The voice industry has dramatically grown in Japan in close correlations with virtual idols, vocaloids, and voice actresses/actors who perform in concerts and on radio, resulting in the branding of voices in a similar manner to the idol industry. Aside from these industrial factors of characters’ voices, however, the functions of voice in romance games pose significant questions in narrative media.
The problem with the role of the voice in romance games is the fact that voice demonstrates little significance in narrative functions themselves—in other words, the voice carries surplus values that exceed semantic necessity. Indeed, many bishōjo games, especially those in the 1990s and early 2000s, which were first released for computers, came out without vocal audio. Unlike moving picture media, including cinema, anime, and even video games with cinematic visual composition, romance games narrate the story in the written text, often in a combination of the PC’s first-person narration and dialogues with/among other characters. Customarily voices are generously assigned to LI characters, while the PC’s own utterance or narration is rarely voiced, which generates the condition that PC’s uttered words, let alone narrations and monologues, are silent written texts, while an LI’s words are doubly represented in both voice and text. As far as the narrative function is concerned, LIs’ voices should be regarded as redundancy in the narrative plane of the story or expressing character personalities, although they may be useful in augmenting emotional expressions. This surplus value attached to LI’s voice can be best explained by the notion of fetishism. The power of voice has been most passionately theorized in Western philosophies and psychoanalysis, among which Derrida’s theory of logocentrism in Of Grammatology stands out due to its scale of impacts. In this phenomenal work, Derrida criticizes the Western belief in the truth value of logos or the phoneticization of words (Derrida 1998, pp. 3–4
). In the Western metaphysical tradition of speech as a divine presence, voice has enjoyed a special privileged position assigned for its role as a vehicle for a purer form of meaning, thereby locating the voice as an integral center of the Word and humanity. According to Derrida, the logos represents the immediacy of mental experience in speech, as it conveys the life and presence of the speaker. Theories surrounding the tradition of vocal speech in Western philosophies have unequivocally joined this poststructuralist direction.
Contemporary critiques of voice, especially in postmodernist theater performance, fully absorbed this Derridian stance of speech as a failure of signification and have stated that “speech has no function except to show its failure as a medium of communication” (Martin 1991, p. 31
). For our discussion of romance games here, at least two concepts resist the Derridean theory of phonocentrism, which are fetishism and the Japanese language. Romance games display a form of intense voice fetishism, but it cannot be hastily equated with the vocal supremacy in the Derridean scheme, mainly because of script-centrism in other languages. As Derrida disclaims, his theory is meant to apply to phonocentric languages, a few exceptions from which are non-phonetic scripts outside logocentric civilizations such as Chinese ideographs (Derrida 1998, pp. 98–99
). As a language that combines non-phonetic Chinese characters and phonetic Japanese in the written form, the written Japanese text does not fit these theories very well. In fact, in order to reduce the physical textual length and save the limited text box space, Chinese scripts are highly efficient and welcomed in the text space on the game screen text. As discussed earlier, the voice’s narrative function in the romance game format resides in its redundancy as the exact textual script of voices; or worse, the voice may prove to be inferior to the written text because the visual reading pace in gaming may easily surpass the pace of the voice. Furthermore, many games presume repetitive gameplay and, therefore, repetitive reading of the same scenario, which commonly makes the player scan and skip scenes or lines by purely relying on information from the written text. In its subordination to, and deferral from, the written text, the voice becomes a residue of signification, a prolonged reverberation of the text.
In addition to this semantic “uselessness” in the narrative function, voice further generates discontinuity among elements that provide the illusion of life, i.e., of the integrated human presence, which Derrida closely associates with the pleasure of logocentrism. In his critique of the Western ethnocentric occult, for instance, he lists hieroglyphist prejudice as an example of the romantic faith in the sublime meaning coming from mystical language (Derrida 1998, p. 86
). In other words, fetishism may intensify regardless of, or rather because of, the crypticness of language. Mladen Dolar points out that classical opera best represents voice fetishism because the language is often incomprehensible, and therefore, the bearer of the voice carries a deeper meaning transcendental of cultural barriers (Dolar 2006, pp. 30–31
). The logocentric meaning intensifies and rises to the sublime scale with the imaginary presence of the utterer, the privileged position as a conduit of the transcendental source of meaning.
This opera model further complicates the voice fetishism in romance games. If opera is a performance enabled by the integration of multiple constituents of a character, such as physical appearance, acting, and voice, into a human presence on stage, the performance that helps us see the characteristics of the romance game, although I take the risk of cultural essentialism, is bunraku. A character representation in bunraku is functionally divided into four visibly separate groups of performers, who are the puppet body itself, its puppeteers visible on stage, the pair of a chanter who voices narration and characters (tayū
) and a shamisen (string instrument) player, and those who make sound effects off stage. Similarly, the audiovisual composition of romance games seems to accept the condition that the signifiers are disseminated on multiple planes of image and sound, which nurtures a love for the 2D characters as a patchwork of audiovisual segments. This game aesthetic simultaneously reveals significant differences between bishōjo games and otome games. Often originally designed as pornographic games, bishōjo games explicitly present LI characters as the imaginary body of object for enticing and fulfilling PC’s (and the player’s) sexual desire, however illusory it may be. In its representational capacity of embodying the gazer’s desire, the female body integrates the narrative components and thereby represents itself as the flesh of the voice. Pornographic games about male homosexual couples (so-called boys’ love) for female players, although a rather small market, similarly provides the illusion of “physical contact with the flesh that emits the voice” (Ishida 2019, p. 294
) through voice acting. Otome games’ decentered narrative form makes a problematic case in contrast to these games with a clear indicator of sexual desire due to the decentered nature of the game form, which variegates the player’s desires and goals. Otome games display a general absence of sexual acts, which is instead substituted by the dissemination of subtle romantic moments without a clear center or climax. If speech’s superiority over the written text derives from the presence of a human being as an integrated whole, then our hypothesis with otome games is that the voice provides fetishism and comfort to consumers all the more because they lack the material presence of body and life.
To further unravel this notion of voice as pure auditory pleasure, it will be useful to delve into what Dolar calls the object voice in his adaptation of the so-called object a
in Lacanian psychoanalysis. Theoretically, the small object a
signifies the breach in causality or the unattainable object of desire that drives us to imagine and seek in a box or behind the screen. This conceptualization of voice as a philosophical tool for narrative strategies helps us examine the structure of desire in otome games. Voice in these games is the sole live-action element directly drawn from the imaginary outside beyond the game and lures viewers to believe life behind the collage of disintegrated visual and textual representations. As Slavoj Žižek also defines the pleasure of the voice using the Lacanian concept of the object, voice is situated between lack and excess, as the lack returns in the form of the excess of pleasure (Žižek 1996, pp. 92–93, 105
). On the one hand, voice represents the lack of body and existence, a signifier of the disintegration from the source of meaning and life. On the other hand, it embodies an excessive compensation for the lack because it fails to signify meaning and thus imposes pleasure, thereby tracing our physical reality and a signifier for the presence of the lost object of desire. In Žižek’s words, the voice can serve as “an organ without a body” (Žižek 1996, p. 110
), a piece of the lack, or a signifier for the absence of (a piece of) the body. To use more realistic wording, the disintegration of visual planes, as well as the voice’s role as excess to generate the meaning or illusion of body, actively contributes to voice fetishism and production of desire. The pleasure, however, is detached from sexual signification as well as from the body, which enables otome games to desexualize voice and generate the pleasure of gameplay. Voice in romance games holds a unique status as a fetish object of pleasure because it eludes various semantic properties of storytelling, which are textual narrative, visual representations, and the imagined body of the voice beyond the game screen. The voice from LIs, now unleashed from the semantic role of signifying, emerges as a fetish voice per se or a piece of the real sent from the story.