Sentimental Avatars: Gender Identification and Vehicles of Selfhood in Popular Media From Nineteenth-Century Novels to Modern Video Games

Theorizing Gender Differences in Media Experience

The gender differences apparent in digital games and sentimental literature help to highlight the way in which social systems of power and control become structured through media and interaction. As we detail below, sentimental novels were widely written by women for women, and yet they worked to reinforce patriarchal social models. Such gendered standards continue to play out in the context of video games given their masculine, competitive associations (Fox & Tang, 2014, 2017; Gabbiadini, Riva, Andrighetto, Volpato, & Bushman, 2016; N. Taylor, Jenson, & de Castell, 2009) and male-dominated creators (Johnson, 2014). Our choice to compare these two media forms stems from their reciprocal extremes that nonetheless bear an uncanny resemblance. Put differently, as these respective forms of embodied media are and were envisioned as created by and for a specifically gendered cohort, the discourse surrounding them exhibits the entrenched assumptions of sex and gender that persist despite significant technological and cultural changes. While these cultural norms can seem ubiquitous, scholars increasingly recognize the need to account for the diversity of individual experience and the possibility of alternative modes of engagement. As we will show, the ways women seem to experience these media forms share core similarities (between then and now) and yet are consistently more complex and nuanced than stereotypes project.

In the games context, observed gendered distinctions extend from game-genre selection and avatar construction to avatar roles and behaviors. One interpretation of these findings is that the ways we operate in digital spaces resonate with cultural stereotypes of who plays what and how we play it—that the forms and functions we adopt online reflect the forms and functions of gender identity in the nondigital world. However, emerging scholarship on games and sentimental literature complicates previously understood patterns of how men and women interact with these media in gendered ways; identification, we realize, is complicated and highly individual.

Video games in particular have recently been identified as a potential realization of divergent queer experiences related to gender and sexuality, due to how their narrative structures can challenge heteronormative ideals of progress (Chess, 2016). Chess contends that the medium of video games has the potential to queer narrative trajectories by encouraging free exploration and play rather than solely emphasizing defined in-game goals, linear advance, and narrow models of progressive development. An ethnography of queer lived engagements with World of Warcraft similarly documents how games can afford powerful opportunities for nonnormative experiences of gender and sexuality, even as these technological spaces embody traditional understandings of these elements of identity (Sundén, 2012). Recent studies have likewise reconsidered how we define and interpret the cultural context within which these diverse experiences of gender and avatar use exist. For instance, Shaw, Lauteria, Yang, Persaud, and Cole’s (2019) unpacking of the discourses of the so-called video game culture identifies how mainstream media’s “othering” of games culture reinforces existing societal power dynamics, including those related to gender. As such compelling scholarship contends, individual media experiences, “gaming culture,” and studies of these phenomena are embedded in a cultural landscape that largely conceptualizes gender identification in binary terms, even as we increasingly recognize that identification is multifaceted and reflects continuously shifting factors including cultural context and individual differences.

We argue that the existing methodological tools of current quantitative scholarship in gaming studies have limited their means of inquiry to a binary model of gender and a static interpretation of personal identification. Restricting theoretical and empirical work in gaming studies to gender, often simply classified as “man” or “woman,” overlooks crucial explanatory possibilities. Following recent qualitative work calling for more nuanced research on games and gender (de Castell & Jenson, 2003; Jenson & de Castell, 2014; Ruberg & Shaw, 2017; Shaw, 2012), we argue that scholars working in quantitative game studies would benefit from integrating more fluid and dialogic approaches to gender and identification into their methodologies. In this regard, recent scholarship on sentimental literature that revalues individual experience and questions inherited critical understandings—of reading and embodiment as well as of gender and genre—can help point the way. By more closely examining the diverse ways in which gamers interface with these evolving technologies, especially through incorporating situation-oriented, fluid concepts of gender from qualitative games work, we can increase the precision and validity of conclusions about gender and game experiences. Rather than reproducing a narrow, binary logic of nondigital gender identity in gamer culture, we propose that quantitative scholars of game culture might become better equipped to produce meaningful, generalizable findings by considering how gaming culture interfaces with the increasingly diverse and nonbinary constructions of gender and identity in the nondigital world.

As others have argued, quantitative methods typically offer deductive and generalizable contributions, while qualitative methods offer inductive and contextual understandings (Lingard, Albert, & Levinson, 2008). Digital games researchers have employed ethnographic (Boellstorff, Nardi, Pearce, & Taylor, 2012) and dramaturgical (Milik, 2015) perspectives to assess quantitative data, offering valuable insights that complicate the nature of gender-based decisions in games. Our work with literary scholarship participates in these ongoing efforts to develop new strategies for quantitative scholars to undertake games work. Thinking across disciplines, we argue, can afford a more robust understanding of gender in games and popular media writ large. We undertake this cross-disciplinary approach by examining how gendered experience has been articulated in sentimental literature and game avatars, beginning with an examination of how sentimentalism has been understood to reproduce female gender roles.

Feeling Gendered: Characters and Corporeality

Initially, a gender-neutral strain of enlightenment philosophy and aesthetic practice, sentimentalism can be defined as a mode that emphasizes feeling—both visceral sensation and emotional intuition—as a primary source of knowledge, meaning, and interpersonal connection. In its popularized 19th-century American form, sentimentalism describes a highly conventionalized, even clichéd, aesthetic practice associated with women writers and an intensely emotional reading modality associated with women readers (Berlant, 2008; Douglas, 1977; Samuels, 1992; J. P. Tompkins, 1986). Sentiment’s affiliated categories include a cultural ideology, a rhetorical style, and an individual sensibility. While these categories inevitably shade into one another, we focus on the literary tradition that fans and critics, both then and now, widely agree hinges on the reader’s identification with its characters and especially with a given novel’s protagonist.

Nineteenth-century sentimental characterization largely comports with contemporaneous standards of sex difference. (Gender, as Schuller, 2018 and others point out, does not exist as a concept until after World War II.) Female protagonists learn to be chaste, pious, submissive, domestically directed, and otherwise virtuous; their male counterparts learn to be brave, active, honest, loyal, and God-fearing (e.g., Douglas, 1977; J. P. Tompkins, 1986). In runaway best-sellers like Warner’s The Wide, Wide World (1987) and Cummins’s The Lamplighter (2017), an impoverished, orphaned young woman strikes out in a world that tests her moral compass and rewards her resilient virtue with the conclusive safe harbor of marriage to a man whose piety rivals his masculine capacity. While much of the recent scholarship complicates these historical narratives of clear-cut opposition and highlights ample exceptions to these patterns, the understanding of the general cultural imperatives remains the standard by which we measure such individual differences at the level of the character and of the novel.

Sentimental reading has likewise been understood in gendered corporeal terms. The feminized sentimental novel consumer was thought to project herself sympathetically into characters’ minds and bodies and so mimetically reproduce their emotions and emulate their physiology—not just to cry when the characters cried or faint when they fainted but also to experience minor sensation (e.g., a quickening heart, goose bumps) and to undergo more extreme psychophysical transitions (e.g., increasing psychic and moral discipline, adopting piety and passivity, rejecting self-indulgence; Noble, 2000; Sanchez-Eppler, 1988; Schuller, 2012). This phenomenology of reading was imagined to enact a set of desired social aims—to cultivate the reader’s gender-normative self-formation in tandem with the protagonist she imagined herself to be. At the same time, however, this psychological and sensorial stimulation was thought to potentially engender excessive pleasure and a dangerous preference for or addiction to vicarious experience. Sentimental novelists navigated this conflict in part by doubling down on the disciplinary possibilities of readerly engagement (in perhaps the most famous example of this logic; Stowe’s [2009] Uncle Tom’s Cabin makes a pitch for the reader to “feel right” as a political act). Popular sentimental novels thus typically socialized women into proper codes of behavior, including the understanding that consuming books for pleasure was not an appropriate behavior.

Digital avatars are likewise understood to be individual users’ vehicles of self within a given medium, usually virtual worlds and video games. In the 1980s, game developers borrowed the term “avatar” from Hinduism—meaning an earthly manifestation of a deity—to instill a greater sense of personal and moral connection to game characters (Bailenson & Blascovich, 2004). Such a decision itself resonates with nineteenth-century sentimental logic, given the latter’s commitment to the reader’s moral education via characterization. Further, research suggests that the psychological relationship between users and avatars involves traditional aspects of identification, such as perceived and hoped-for similarity (J. Cohen, 2001; Van Looy, Courtois, De Vocht, & De Marez, 2012), ideas about identification as a merger of character-self (Klimmt, Hefner, & Vorderer, 2009), and other approaches that emphasize virtual embodiment, such as self-presence, or the sense of the self as present in the avatar’s body, emotion, or identity (Ratan, 2012). Indeed, relative presence with other avatars has been shown to indicate a level of attraction or romantic connection between players (Bergstrom, Jenson, de Castell, & Taylor, 2017). As we elaborate below, in much the same way that sentimental literature has been understood to be produced by and for women, video games are still heavily marketed and culturally oriented toward male consumers (Ivory, 2006), with women often serving in secondary roles (T. Lynch, Tompkins, van Driel, & Fritz, 2016; Ratan, Taylor, Hogan, Kennedy, & Williams, 2015) or as little more than salacious ornamentation (H. D. Fisher, 2015). Female-gendered games, meanwhile, are widely marketed and oriented toward women by emphasizing stereotypically feminine pursuits and interests, such as building social relationships (Cassell & Jenkins, 2000).

The different avatar profiles that men and women have been seen to create and the different functions these avatars fulfill echo the cultural work performed by sentimentalism’s gendered characterization. Avatar functions tend to be gendered even though in many popular contexts of use (e.g., massively multiplayer online role-playing games), sex does not influence the avatar’s functional attributes. For example, fighting-function characters (e.g., warriors) tend to be customized as male, while supportive-function characters (e.g., healers) tend to be customized as female—tendencies that occur despite avatar sex choice having no functional effect on the avatar’s performance in these roles (Bergstrom, Jenson, & de Castell, 2012; Yee et al., 2011). The perceived gender specificity of avatars parallels the previously perceived sex specificity of sentimental characters and reproduces, at the level of individual election, the same notions of biological essentialism that naturalize such distinctions within the research in these fields.

The physiology of digital gaming likewise echoes narratives of sentimental reading. Much like the young Edith Wharton was forbidden from reading novels owing to their sensational workings (Hoeller, 2000), digital games have been argued to cause problems ranging from moral decay and anger to depression and suicide (Agosto & Abbas, 2017; Garretts, 2006). This cultural perception is reinforced by scholarship that demonstrates the diverse ways in which games stimulate a range of physiological responses in players including skin conductance, heart rate, systolic blood pressure, and diastolic blood pressure (Anderson & Bushman, 2001; Lim & Lee, 2009). Research further suggests that people who experience greater avatar identification also exhibit stronger physiological responses during avatar use including measures of arousal, attention, and positive affect (Lim & Reeves, 2010; Ratan & Dawson, 2015). Similarly, games have been found to impact players’ emotional states (Ravaja, Saari, Salminen, Laarni, & Kallinen, 2006) and to alter the neurological systems in their brains, leading to particular concerns about rewards structures and addiction in children (Cole, Yoo, & Knutson, 2012). As with sentimental literature, these physiological dynamics occasion an enjoyment and immersion that provokes cultural anxieties about the dangers of overuse, addiction, and aggression associated with digital gaming (Agger, 2012).

The gendered understanding of the phenomenology of sentimental reading echoes certain findings about the corporeality of digital media use. As Schuller (2018) has recently demonstrated, sentimentalism was bound up in the logic of impressibility—meaning “the capacity of a substance (e.g., a reader) to receive impressions from external objects that thereby change its characteristics” (p. 7). Women were understood to be more impressionable than men, a quality that made them more emotional and sensorially responsive than their more rational, largely disembodied male counterparts. Women readers were imagined to be more susceptible to the vicarious experience offered by sentimental literature, to prefer subjective collapse with characters, and to read for and with feeling (as opposed to pragmatic rationality) regardless of text. Schuller (2018) has recently unsettled such biological essentialism from within, by detailing how period science naturalized such gendered (or, in that time, sexed) logic. Research on digital media likewise draws on contemporary scientific models to complicate the gendering of media consumption’s visceral workings. While studies have found that individuals who identify more with their avatars also exhibit greater physiological arousal during use, in at least one study, men’s physiological arousal was found to differ during gameplay according to avatar characteristics, whereas women’s was not (Lim & Reeves, 2010). However, this study was conducted using World of Warcraft, a game marketed to men and whose content—involving fighting and online interactions with potentially hostile strangers—aligns with gender-normative masculine behavior. In the game Swordplay for the Nintendo Wii, which was marketed to women and families and whose content involves cartoon-style sword-fighting matches, women’s physiological arousal was found to increase with avatar identification (Ratan & Dawson, 2015). In other words, the measured differences in physical arousal in each study’s cohorts of men and women have been identified as influenced by factors other than gender.

These findings suggest that other data sets may likewise be the result of factors that are simply assumed to be gendered. To state what may now seem all too obvious, and as the evolving scholarship on sentimentalism also argues, men and women are equally prone to the workings of identification and equally psychologically, emotionally, and viscerally responsive to the objects with which they identify (e.g., Cassuto, 2009; Chapman & Hendler, 1999). The question, then, is what factors produce these gendered patterns of physical sensation. Some of the most significant contributors, we propose, are the gendered cultural expectations of how individuals consume media.

The Social Lives of Gendered Avatars: Or, Constructing Sex Difference, Then and Now

In a now-famous 1855 letter to his publisher, Nathaniel Hawthorne complained,

America is now wholly given over to a damned mob of scribbling women, and I should have no chance of success while the public taste is occupied with their trash—and should be ashamed of myself if I did succeed. What is the mystery of those innumerable editions of The Lamplighter, and other books neither better nor worse?—worse they could not be, and better they need not be, when they sell by the 100,000. (as cited in Baym, 1985, p. 138)

Gendered notions of popular gaming have been likewise well-established within both the gaming culture and the industry itself. The popular and commercial profile of a gamer has long reflected a stereotype of geek culture as comprised of young, White, socially awkward males (Kowert, Festl, & Quandt, 2014; T. L. Taylor, 2012). While this stereotype has faded somewhat in recent years, its underlying assumptions of race and gender persist in many ways today in both the gaming industry and the popular culture (Kerr, 2003; Nakamura, 2009). This stereotype also persists despite the fact that more women are playing games than ever before, with women over 18 representing 46% of all gamers in the United States (Entertainment Software Association, 2019). It is thus perhaps unsurprising that the most prominent figures in popular gaming culture are almost exclusively male (Paaßen, Morgenroth, & Stratemeyer, 2017). According to Shaw (2015), this biased view of gaming culture stems from the gaming industry’s targeting of a “hardcore” demographic of teen to young adult males as its primary market. As a result, individuals who choose not to play certain games or who do not dedicate a sufficient amount of time to games do not earn “true” gamer status in both the eyes of the gaming industry and in popular gaming culture (Paul, 2010; Shaw, 2012; D. Williams, Martins, Consalvo, & Ivory, 2009).

Much like female sentimental readers were cast as feelingful consumers of vapid novels rather than sophisticated aesthetes who sought intellectually challenging literary works, women are typecast as “casual” gamers who are less likely to enjoy more action-oriented and competitive styles of games (Hartmann & Klimmt, 2006). The industry’s gendered narratives are most evident in the creation of “pink” and “purple” games which attempt to target the “girl gamer” with gender-stereotypical activities such as fashion and building relationships (Cote & Raz, 2015; Kafai, Heeter, Denner, & Sun, 2008). Although “pink” games have been demonstrated to appeal to some women due to the presence of these stereotypically feminine activities (van Reijmersdal, Jansz, Peters, & van Noort, 2013), the explicit designation of games that are “for women” reinforces existing cultural divides about what genders are legitimate in mainstream gamer spaces. Additionally, designing games with certain features “for” certain genders taxonomizes games based on their level of suitability for (assumed) male or female experiences and reduces consideration of other important engagement factors, such as game genre. This “hegemony of play” reinforces a stereotypical gendering of the medium which conceptualizes women as inherently different in their gaming style and as thereby not belonging in mainstream gamer culture (S. Fisher & Jenson, 2017).

When a woman engages with nonapproved digital games, she can encounter great resistance in the form of verbal abuse in and out of the game (Consalvo, 2012). Hawthorne’s less-examined follow-up letter suggests the historical consistency of this dynamic. In it, he acknowledges his recent vilification of his female contemporaries and “admires” a novel he has since “enjoy[ed],” Fanny Fern’s roman à clef Ruth Hall (1854)—an unconventional sentimental novel about a widowed mother who becomes a successful newspaper writer (as cited in Baym, 1985, p. 139). Hawthorne, however, offers the backhanded compliment,

The woman [Fern] writes as if the devil was in her; and that is the only condition under which a woman ever writes anything worth reading. Generally women write like emasculated men, and are only distinguished from male authors by greater feebleness and folly. (as cited in Baym, 1985, p. 139)

Quantitative academic and industry-sponsored research on virtual worlds and identity have often historically, albeit unknowingly, reinforced these issues of reductive gender categorization. Women are typically understood to be a distinct subset of gamers who exist in contrast to the “standard” (male) player. This logic leads to questions of how women play differently (Huh & Williams, 2010; T. L. Taylor, 2003), why women seem less interested in games (Lucas & Sherry, 2004; Ogletree & Drake, 2007), and how to design games for women (Ochsner, 2015). The resulting research understands women as a unified group, with coherent interests and distinct preferences, and neglects more nuanced understandings of how and why different individuals interact with games (S. Fisher & Jenson, 2017).

Such studies frequently reproduce sentimental logics of desire. Nineteenth-century sentimental literature casts feminine aims as essentially generous and relational, in contrast to masculine ambitions that are highly individual, self-interested, and outwardly oriented (toward financial and professional success; Mendelman, 2019a). Games studies have found that women and girls tend to play fewer competitive, fast-paced, and action-oriented games than men and boys, opting for more puzzle and other “casual” genre games (Sherry, Lucas, Greenberg, & Holmstrom, 2013). Women who depart from such generic expectations are often assumed to do so for reasons that likewise reflect gender norms. One study, for example, found that of the relatively small portion of women who played the highly popular online game League of Legends, most tended to play support roles (e.g., healing), presumably to assist their male gaming partners (Ratan et al., 2019). “Girl gamers” are thus still understood to prefer cooperative play built on relationships and emotionally driven narratives (Cote & Raz, 2015; Ray, 2004). The cultural scripts surrounding avatars, like those of sentimental literature, thus essentialize female experience into one of subservience and prosocial behavior.

Sentimental novels famously promulgate such gendered roles in the context of gendered settings, distinguishing the public sphere as masculine and the domestic sphere as feminine. Whereas the male Bildungsroman, or coming of age novel, typically features a young man going out into the world and learning to make his way through such external adversity (à la Goethe’s Young Werther), the female equivalent tends to lead its young woman through a series of internal (psychic and emotional) conflicts and then, in reward for her successful self-discipline and pious virtue, back into the home via marriage. Classic American examples of this gendered plotline include the trajectories of Ellen Montgomery in The Wide, Wide World (1854/1987) and Gertrude Flint in The Lamplighter (1854/2017). Cohen (2002) and Fraiman (1993), for example, detail these generic expectations and gendered milieus in their respective studies of contemporaneous French and slightly later British sentimentalism.

Virtual worlds and avatar roles likewise reproduce such cultural scripts about gendered environments and activity. For example, studies of two of the most played avatar-based online games at the time, World of Warcraft and EverQuest II, found that avatar gender influences the ways that people behave in these spaces. Specifically, players were more likely to engage in stereotypically feminine activities (e.g., healing) when using a female avatar (Yee et al., 2011), women who used male avatars (less than one tenth of all women players) were found to prefer more player versus player combat and less chatting than the average male player (Huh & Williams, 2010), contrary to the more social and less competitive motivations found for the majority of women players (Yee, 2006). The nineteenth-century parallels to such role-play are evident: Healing’s subservient role and chatting’s interpersonal goal echo the position that sentimental novels encourage women to adopt, while competition and aggression appear as masculine/male traits from which women are expected to refrain. Lest there be any doubt about the endurance of such cultural logic: The above paper argues that the male-avatar-using, combat-oriented, nonchatting “women were ‘out-manning’ the male players” (Huh & Williams, 2010, p. 171).

In much the same way that sentimental literature continues to be associated with female readers and writers (Mendelman, 2019a), competitive and “hard-core” games continue to be associated with men (Chess & Paul, 2019). These assumptions lend themselves to straightforward analysis in quantitative methods. After all, historically, most digital games’ character creation screens contain a binary male/female toggle, if they have one at all. And even in games without avatars, players will often construct social systems and networks to allow for easier interaction, often categorizing characters with a gender (Woodford, 2012). This reciprocity between online and off-line life, between player and culture, has long shaped game studies. While efficient, these classifications come with additional meanings that place expectations of visual representation and behavior on the individual (Hester & Eglin, 1997). Our research methods often replicate the very social problems we seek to explore.

Beyond Binaries of Sex and Gender: Evolving Work in Literary and Games Studies

Scholarship in both literary and games studies has begun to rethink our methodologies in efforts to better account for the diversity of individual experience in media engagement. Much of this work interrogates its own critical assumptions and methodological limitations as well as unsettles earlier scholarly understandings of its objects of study. Here, we briefly trace several recent trends within studies of sentimental literature and then consider how these intellectual currents dovetail with recent work in game studies.

The first of these trends is rethinking the oft-binaristic modeling of sentimentalism as an aesthetic category itself. Sentimentalism has long been defined by what it is not (e.g., not complicated, not self-conscious, not aesthetically innovative). Developing a more nuanced sense of how people actually engaged with sentimental literature has involved rethinking many of its negative definitions and recognizing its complexity, self-consciousness, and potential for innovation—all of which involves analyzing how social expectations of sex and gender inflect these dynamics and earlier assessments thereof. Such studies reevaluate sentimental literature by contextualizing it within its immediate historical moment and thus within the gender politics of its milieu (Burstein, 2002; Dillon, 2004; Dobson, 1997; Mendelman, 2019a; Merish, 2000). Efforts to rethink sentimentalism in terms of its cultural context also include more detailed studies of how sentimentalism was read and processed by “amateur” readers, as well as by cultural critics, public intellectuals, and other figures endowed with professional authority (Brady, 2011; Halpern, 2013). This latter work with letters, diaries, and other textual and material artifacts—what is known as “reception studies”—aligns with wider disciplinary efforts to reconsider scholarly methods and their implicit or explicit exclusions. One apposite line of this work seeks to revalue emotional identification with literature, questioning longstanding watchwords of readerly dispassion, detachment, and critique and highlighting the gendered, racialized, and socioeconomic assumptions of these standards (Felski, 2011; D. S. Lynch, 2014). Ongoing work in literary studies thus seeks to develop more accurate accounts of sentimental literature’s consumption from two directions: on the one hand, thickening the descriptions of individual readers’ lived experience and, on the other, detailing the genre’s gender-neutral experiential complexity.

Recent literary scholarship also enriches models of sentimental literature, and of gender and reading more broadly, by participating in the growing body of work known as affect studies. This multidisciplinary line of inquiry explores dynamics of feeling, physicality, and other elements of embodiment and, within literary studies, contemplates the relationship between emotion and the aesthetic encounter. For example, prominent affect studies scholar Lauren Berlant has written a “national sentimentality” series that tracks the gendering of the genre and of affective models of being in the world from the 18th century into the present (Berlant, 1991, 1997, 2008). New work on affect in game studies similarly emphasizes the importance of identity and representation within casual games and gaming history (Anable, 2018). While affect studies is noteworthy for its efforts to avoid reductive gender politics (and the politics of other aspects of embodied subjectivity, such as distinctions of race, class, sexuality, and region), the field has been faulted for several of its own binaristic tendencies. One criticism is that affect’s cultural theorists often reinforce a Cartesian dualism they ostensibly reject, separating sensing from experience (defining affect as “unqualified intensity,” distinct from emotion’s “subjective content”; Massumi, 2002, p. 28), and rely on tenuous psychobiological claims to do so (Frank, 2015; Leys, 2011; Papoulias & Callard, 2010; Wetherell, 2012). A second criticism is that efforts to avoid this pitfall reverse these subtle priorities and advance a cognitivism that is no less invested in problematic science (Cromby & Willis, 2016). A third concern is that considering affect as a presociolinguistic form reduces its politics to a good–bad, liberating–coercive binary and precludes more nuanced interrogations of how it plays out in individual lives and social bodies (Cooter, 2014; Hemmings, 2005; Leys, 2011; Papoulias & Callard, 2010). New work on sentimental literature grapples with these criticisms by foregrounding feeling as a process that integrates affect, historicizing sentimental characterization in terms of period science, and drawing on other recent cross-disciplinary efforts to think more dynamically about the relays between thought/cognition/mentality and feeling/emotion/corporeality (Mendelman, 2019a, 2019b).

Game scholars have likewise recently begun to complicate binaristic discourse through mixed-method studies. As many of these newer studies note, the body of quantitative research on avatars often presents sex and gender interchangeably or with little detailed analysis of larger notions of gender. Women are consistently presented within larger gendered assumptions of how and why women play games (Kerr, 2003). While there is work examining how communities can function in online game spaces as a form of rebellion against hegemonic gender expectations (Sundén, 2009), there is a need for deeper investigations of gendered identities in relation to how game avatars are used and why. In understanding how women play solely in relation to female avatar use, researchers reinforce stereotypical notions of gendered play, lose sight of other influences such as hit-box size or availability of avatar role-play options, and overlook other motivations such as gameplay style or online community membership (Jenson & de Castell, 2018; Shaw, 2012).

One subset of games research that affords an especially rich topic for more diverse gender interpretations is the so-called avatar sex swapping, that is, men choosing female characters or women choosing male characters. A number of studies refer to this practice as gender swapping, but this practice would be better understood as sex swapping because the vast majority of games force the player to choose between male or female (Barnett & Coulson, 2010). Avatar sex swapping in games is quite common, with one study reporting 20% of female and 32% of male World of Warcraft players using at least one sex-swapped avatar (Jenson et al., 2015). Qualitative studies that consider gendered avatar identification beyond the behavioral assumptions of sex selection found that even those players who seemingly fit into gendered roles held various reasonings which did not fit with typical assumptions (Hayes, 2007, 2008; T. L. Taylor, 2003). Avatar identification research similarly demonstrates that men and women report sex-swapping avatars for reasons ranging from identity exploration (Baldwin, 2019) to navigating specific social interactions, such as deflecting sexual advances or soliciting gifts (Hussain & Griffiths, 2008; Martey, Stromer-Galley, Banks, Wu, & Consalvo, 2014).

These findings demonstrate the need to consider the multitude of environmental and internal factors that contribute to player decision-making when it comes to avatar sex selection and gendered behavior rather than just assuming that avatar sex selection has an inherent and unchanging connection to a person’s gender identity. More diverse approaches to these decisions—along the lines of reception studies in literary criticism—would help to elucidate and more accurately frame the motivations behind these decisions and their contributions to gender experience. This style of player-centric research has been present in game studies (Behrenshausen, 2013), yet there is little work within this subfield that focuses on avatars and identity. At the same time, we would benefit from reconsidering the complexity of decisions that may not be conscious and the choices that people make without knowing the exact reasons why they do so. In that arena, more detailed studies of the games themselves and their broader cultural contexts would be illuminating. While much of the games work centered on expanding and diversifying these approaches to gender has been qualitative, some quantitative work has begun exploring these topics.

One important trend in this research involves complicating the way studies define avatar identification, enabling a more in-depth analysis of avatar choice motivations. For example, Banks and Bowman’s (2016) Player–Avatar Interaction Scale attempts to understand the player–avatar relationship using social dimensions such as “closeness,” a quality of experiential intimacy determined by (1) a player’s emotional investment in the avatar, (2) the imagined anthropomorphic autonomy of the avatar, (3) the player’s suspension of disbelief in engaging with the avatar and game world, and (4) a sense of control over the avatar. These four factors predicted a player’s “closeness” experiences with the avatar. In another example of work along these lines, Kafai, Fields, and Cook’s (2010) analysis of avatar creation resources and motivations within the virtual world of Whyville illustrated the complexity of many avatar creation tools and players’ common desire to learn these systems in order to create better avatars. Similarly, the avatar affordance framework advanced by McArthur, Teather, and Jenson (2015) tasks researchers with analyzing character creation interfaces (CCIs) as actors in their own right. These authors assert that “the problematic mediation of gender and ethnicity via CCIs act as ‘regulatory regimes,’ uncritically participating in the cycle of socially exclusive values” (p. 238). Many of the recent efforts to integrate diversity in games work focus on computational approaches to increasing diversity into avatar experiences (Harrell & Lim, 2017; Vermeulen, Van Looy, De Grove, & Courtois, 2011) rather than using quantitative approaches to examine how the diversity of those experiences play out in real gaming contexts. As we specify below, incorporating strategies utilized by qualitative scholarship into quantitative work will allow more diverse interpretations of gender to be measured and validated in generalizable terms, and will thus deepen our understanding of gendered experience across time and space, as well as in terms of individual variance.

Conclusion

The worlds of the imagination, woven through the words of nineteenth-century American literature or through the design of modern digital games, seek to inspire the individual and to allow for flights of fantasy. Nonetheless, our assumptions about the reality of gendered experience prescribe—and limit—the possibilities of individuals to embody these fantasy worlds before they even open the book or pick up a game controller. In using the systems that have been established in quantitative methods, many researchers further these cultural assumptions.

This article has provided examples both in literary analysis and in digital games research where more nuanced meanings of gender are applied to qualitative methodologies. In our discussion, we highlighted ways in which nineteenth-century understandings of the ways people embody media echoes understandings of sex and gender within gaming culture in a binary, prescriptive fashion. We highlighted a range of topics for future quantitative games work to consider when examining gender. Specifically, we discussed analyzing social expectations of characterization, scrutinizing physical sensation’s relationship to embodiment, complicating binary models and avoiding biological essentialism, and contemplating the gendered nature of space and participation. We also identified the through lines between several recent trends within literary criticism and games studies, particularly assessing how individuals understand their media experience rather than reifying popular narratives or prior scholarly understanding. If quantitative studies of digital games and gamers were to incorporate some of these suggestions, future digital games research would have a stronger theoretical foundation from which to challenge long-established gender-normative stereotypes in the gaming community.

We conclude with three strategies for quantitative games scholars to consider in their own work. First, rather than simply relying on a male/female binary in participant gender identification, researchers might utilize available scales of gender identification (Martey et al., 2014). Currently, available scales include the Bem Sex Role Inventory (Holt & Ellis, 1998) and the Traditional Masculinity–Femininity Scale (Kachel, Steffens, & Niedlich, 2016). Utilizing these scales would allow scholars to identify what subconstructs or other factors (potentially even those seemingly unrelated to gender) may be influencing identification. In a parallel vein, Spencer’s (2007) Comfort and Conformity of Gender Expression Scale works to investigate some of the complexities of gender identification, such as how comfortable an individual feels with certain types of gender expression. Emerging quantitative work on gender identity has recognized the importance of evaluating femininity and masculinity in degrees (Magliozzi, Saperstein, & Westbrook, 2016), but there remain limited inquiries into ways to approach this topic without continuing to use a binary model.

Our second and third strategies negotiate this current limitation. Our second recommendation is that scholars might attempt to more closely consider how experiences with gender change over time. Most existing quantitative work centers on surveys and experiments at a single point in time or during a single instance. Our analysis demonstrates that experiences related to gender and avatar identification are dynamic, context-dependent processes. A player’s experiences related to gender will vary based on the social pressures and expectations of that unique game world, the individual’s experience with the game, social ties within in the game, and potentially many other factors that will vary over time. We therefore encourage scholars to incorporate more longitudinal methodology into their studies. Such methods could include repeated measurements of subjects, including interviews centered on internal changes and physical studies centered on physiological shifts. These measures could be supplemented with repeated surveys centered on gender identification that involve self-report and self-assessment as well as physiological analyses. Such approaches to data gathering would help researchers to better understand whether certain features of a given individual’s gender identification appear more or less stable over time and how these features potentially intersect with other aspects of the gaming experience.

Finally, quantitative scholars could more critically examine the assumptions about gender that underpin research hypotheses and study design. Rather than treating gender as a static and often predictive factor, quantitative methodologies can better capture the diversity of gender experiences by considering their dynamic interplay with cultural context, interpersonal social dynamics, other social identities, and physical arousal, among other variables. Approaches along these lines would involve constructing studies with greater openness to explanations of behavior rather than simple gender-based correlations, and incorporating network-theory methods that consider players’ not-necessarily-gendered connections and context by mapping social ties and relationships. Friendship groups, school cliques, and categorical membership, for example, have been found to significantly influence individual behaviors (for more on membership categorization analysis; see Hester & Eglin, 1997). Players who join with friends are more likely to play roles that benefit cooperative play rather than pursue individual preferences (Milik, 2017). Similarly, players are far more likely to select certain types of game characters (such as rogues or spies) if they wish to “grief” or cause disruptions to others’ play (Lin & Sun, 2005). Attending to how these social and environmental factors intersect with, and perhaps unsettle assumptions about, apparently gendered phenomena would help quantitative researchers capture higher quality data and lead to more complex and meaningful results.

In an ideal world, the young woman joining an online game today would not be bound by the same gender norms as the young woman reading the sentimental novel in 1800s America. At the very least, scholars can work to develop our methodological systems and academic standards along with our evolving media. Gender influences how people engage with media, but in order to better understand how it matters, we need to examine it beyond the binary and interpret it more broadly than as a context independent predictor. When individuals open a book or pick up a game controller, their gender identity influences how they experience the imagined world they confront and help construct. Often, they imagine themselves in ways that reflect their gender identity. But that identity is malleable, context-specific, and shape-shifting. When people embody imaginary spaces, their experiences unsurprisingly reflect the multifaceted ways in which they view themselves. Why should they be anyone else?