The art worlds of gender performance: cosplay, embodiment, and the collective accomplishment of gender

In recent years, cosplay has gained global visibility as a performing art in which fans dress up as fictional characters from anime, comics/manga, or games. Although scholars contend that cosplay exemplifies gender performativity and may even offer a new heuristic for understanding social interactions in general, they rarely examine how gender is performed in cosplay. Taking together the production of culture perspective and interactionist theories of gender, I detail how cosplay participants project gendered sensibilities through conventionalized body movements and modifications. Contrary to the prevailing focus on individual cosplayers, I demonstrate how makeup artists, photographers, and photo editors all contribute to the success of gender play in cosplay. Contrary to simplistic accounts of donning hyper-masculinity/hyper-femininity, I argue that cosplay participants’ pursuit of authenticity makes singular orientation to sex category insufficient and demands a version of masculinity/femininity that also attends to the character’s personality. Situated in the art worlds of cosplay’s production and evaluation, my findings invite scholars to consider the collaboration between cosplayers and their supporting crew as a conceptual heuristic that attunes our attention to the collective accomplishment of gender embodiment, whose multi-authorship is often obscured.

Yang The Journal of Chinese Sociology (2022) 9:9 their "authenticity" according to commitment and adherence to their source characters (Matsuura and Okabe 2015;Rahman et al. 2012).
Notably, the "play" in cosplay involves not only role play but also gender play: the character that cosplayers strive to impersonate can be of a different gender, and gender-crossing is common in the cosplay community (Norris and Bainbridge 2009). Unlike other gender-bending communities, gender-crossing cosplayers-sometimes called "crossplayers"-bend their gender for their love of characters, rather than for an explicit political agenda or for externalizing an internal gender identity that does not match their sex assigned at birth (Lamerichs 2011;Leng 2013;Tompkins 2019). While noting differences in motivation, some scholars argue that cosplay shares apparent similarities with drag and thus exemplifies gender performativity (Bainbridge and Norris 2013;Lamerichs 2011;Leng 2013;Norris and Bainbridge 2009). Others take a more radical stance. Seeing cosplay as an alternative to drag, they suggest cosplay may help us find new heuristic tools for thinking about social interactions: Just as "performance" served as "a structuring trope" for human action in the age of broadcast media and service industry, "animation" arises as "an alternative model of and for human action" in the age of digital media and content industry (Silvio 2010, p. 423), a model that allows us to "see what … [the trope of ] performance has obscured" (Manning and Gershon 2013, p. 107). 1 How can cosplay, as a subcultural practice and an alternative heuristic, redirect our sociological attention to the hitherto overlooked aspects behind the accomplishment of gender? Answering this abstract theoretical question requires an empirical exploration of the production of gender embodiment in cosplay-only then can we take our case-specific findings to more familiar arenas of everyday life and examine whether similar processes are at work without being captured by our existing theoretical lenses. However, to date, the scholarly writing on cosplay provides few examples of how people cosplay, not to mention how cosplay participants stage gender performance despite constant authenticity assessment (for critiques, see Kirkpatrick 2015). Given the theoretical insights such studies could offer, this gap should not only concern cultural scholars and area specialists interested in the globalization of Japanese ACG culture, but also feminist sociologists and social theorists searching for new ways to conceptualize gender.
In this article, I draw upon interviews with cosplay participants in China-not only cosplayers but also makeup artists, photographers, and photo editors-to situate cosplay into its "art worlds" (Becker 2008) of production and evaluation. I examine how gendered images of fictional characters are reproduced with human bodies, paying close attention to the impression-managing tactics participants use to embody characters of different genders, as well as the division of labor in the production of such gendered impression. In contrast to the prevailing focus on cosplayers, I demonstrate how the supporting crew is also crucial for the success of gender play in cosplay: Not only do cosplayers need to convey gendered sensibilities with poses and facial expressions, the supporting crew also needs to imbue these messages into cosplayers' bodily features. These two types of body work-body movement and body modification-can establish a gendered impression in the audience based on the feelings that cosplayers' bodily cues evoke. But singular orientation to sex category alone is insufficient for "authentically" recreating a character. Instead of donning hyper-masculinity/hyper-femininity, participants display an additional orientation to personality category, and strive to stage a version of masculinity/femininity consistent with the character's placement into these categories.
While my empirical findings center around the collective production of gender embodiment in the Chinese cosplay scene and may not be directly transposable to contexts outside the ACG fandom, they offer a starting point to think theoretically about how cosplay as a heuristic can enrich social theories. As I will elaborate at the end of this article, insofar as theory can be understood as hermeneutics and ways of looking at the world (Abend 2008), looking at gender as a cosplay (after decades of looking at gender as a drag) invites us to theorize gender as an artwork-collectively manufactured in its art worlds-whose multi-authorship is often obscured by neoliberal individualism. Using cosplay as a heuristic thus requires further integration of theoretical insights from sociology of gender (Goffman 1976;West and Zimmerman 1987) and sociology of culture (Becker 2008;Peterson and Anand 2004). Doing so will alert us to the theoretical possibility that seemingly solo-authored gender presentation of "authentic self " may have a hidden team of collaborators-even though they may not take the same roles or perform the same kinds of body work as their counterparts in cosplay teams do.

Cosplay: entering the scene
Despite the increasing popularity of cosplay on a global scale, scholarly writing on cosplay is limited (Kirkpatrick 2015;Leng 2013;Rahman et al. 2012). Although some scholars wondered if animation and cosplay offer us new ways to conceptualize and theorize human activity (Silvio 2010;Manning and Gershon 2013), much of the existing work on cosplay, as Kirkpatrick (2015) acutely observes, is preoccupied with why people cosplay (Bainbridge and Norris 2013;Lamerichs 2011;Peirson-Smith 2013;Rahman et al. 2012;Tompkins 2019;Wang 2010;Winge 2006), rather than how people cosplay. This unbalanced attention bears a theoretical cost, because it is only by studying the latter can we start to explore the similarities and differences between the process of cosplaying a fictional character like Uchiha Sasuke from Naruto and the process of enacting an abstract quality like "hegemonic masculinity" in everyday encounters. As Becker (1953) famously shows, studying "why" risks pathologizing both the activity under study and its participants, whereas turning attention to "how" allows us to identify social processes that can be useful for understanding participation in other, seemingly unrelated, activities.
To be fair, the issue of gender performance is a notable theme in this small body of work. This is hardly surprising, as gender-crossing cosplay is not only common but wellrespected-if done successfully (Leng 2013). Yet much of the academic debate focuses on whether cosplay subverts gender norms (Bainbridge and Norris 2013;Gn 2011;Lamerichs 2011;Loke 2016;Norris and Bainbridge 2009;Tompkins 2019). Many of these studies do highlight cosplay as an example of Butler's (1990) theory of gender performativity and point to its embodied aspects (e.g., Bainbridge and Norris 2013;Gn 2011;Lamerichs 2014). Still, they tend to be reticent about how exactly gender is performed/embodied in cosplay. Sometimes, they simply characterize the performance as individual, hyperfeminine/masculine displays. For example, in her study of male-to-female cosplay, Leng (2013, p. 90) asserts, "When men crossplay as women, they are not merely donning femininity, but hyper-femininity, revealing the socially constructed nature of gender roles yet concomitantly reinforcing them. " My study takes up this missed opportunity for reconceptualizing gender accomplishment by delving into the processes through which cosplayers perform gender, and the labor division that enables their gender embodiment.
Motivations for cosplay are not irrelevant to understanding these processes. As Winge (2006) notes, most people participate in cosplay for their love of characters and for socializing with other ACG fans. These widely shared motivations shape cosplay in profound ways. First, as cosplay is supposed to express cosplayer's love of the source character, it requires "inhabiting the role of a character both physically and mentally" (Norris and Bainbridge 2009), not just dressing up in exotic costumes (Kirkpatrick 2015;Leng 2013). Second, because cosplay facilitates socialization among fans, participants constantly learn from and evaluate each other's cosplay (Bainbridge and Norris 2013;Lamerichs 2011;Matsuura and Okabe 2015;Rahman et al. 2012). Hence, cosplay is not only embedded in the dyad between a character and a cosplayer, but in a network of performers and spectators (Lamerichs 2014).
One central concern in this performer-spectator network is authenticity. As a proxy for cosplayers' love for their characters, authenticity is a key criterion for assessing cosplay performance. Cultural sociologists have long argued that authenticity is not inherent, but socially constructed, recognized, and achieved (Peterson 1999;Hughes 2000). As such, it must be "performed, staged, fabricated, crafted, or otherwise imagined" (Grazian 2010, p. 192). In cosplay, because authenticity is evaluated by "commitment and adherence to the original character, " participants need to attend to both the character's physical attributes (e.g., costumes, makeup) and mimetic attributes (e.g., facial/bodily expressions) in their performance (Rahman et al. 2012, 326;Norris and Bainbridge 2009). Otherwise, they may be seen as not making enough effort to reproduce the appearance and mannerism of the original character (Lamerichs 2014). When a cosplayer fails to adequately present themselves as the same gender as their chosen character, their performance may even be read as a parody that pokes fun at the character (Leng 2013). Given the importance of authenticity, studying gender play in cosplay requires an interactionist lens that captures not only how one performs gender, but how one does so collectively and reflexively with an anticipation of how their performance might be evaluated.

Gender performance: from individuals to art worlds
Interactionist theories of gender are hugely indebted to Goffman's dramaturgical theory. Whereas some scholars argue cosplay is akin to performance art Bainbridge 2009), Goffman (1959) argues all interactions are performances, involving a performer and an audience. Since people have an interest in controlling how others treat them, they project their desired impression to others via self-presentation. A classic example is "gender display": During interactions, people present "conventionalized portrayals" of "culturally established correlates of sex" (Goffman 1976, p. 69), so that others will read such displays as signs of masculinity or femininity and react in appropriate ways. The body is a crucial resource for such impression management, as people can imbue symbolic cues of masculinity/femininity into their bodily features via the "body work" of managing and modifying physical appearances (Gimlin 2007;Hoang 2015;Wacquant 1995). Thus, femininity and masculinity are not expressions of essential sex difference, but performances that honor and reinforce sex difference (Goffman 1976;1977).
Gender performance is not only embodied but reflexive, with interactional consequences that performers must consider. Building on the ethnomethodological literature on membership categorization (Sacks 1972), West and Zimmerman (1987, p. 133) argue that in a society organized around the gender binary, people are predisposed to sex-categorize themselves and each other into "indigenous categories such as 'girl' or 'boy, ' or 'woman' or 'man, '" as long as the category seems "relevant" and "appropriate. " 2 Although a person can be correctly described in infinite ways, and "[a] single category from any membership categorization device can be referentially adequate" for framing an interaction (Sacks 1972, p. 333), 3 the omnirelevance of sex category makes it a readily available resource for designing and interpreting actions vis-à-vis normative assumptions about gender (West and Zimmerman 1987). As such, gender is done when we manage and interpret conducts with reference to one's placement in a relevant sex category.
But in some contexts, multiple categories are potentially relevant to doing gender. In the US, West and Fenstermaker (1995) observe that the trilogy of gender, race, and class is omnirelevant to many interactions. In the Chinese ACG fandom, I noticed that personality category is also particularly salient. In such contexts, doing gender can be intertwined with doing other differences (West and Fenstermaker 1995). Importantly, because a single category could have been "referentially adequate" (Sacks 1972, p. 333), there is likely a motivation behind invoking additional categories-it accomplishes something extra in the interaction that singular categorical orientation cannot accomplish (West and Fenstermaker 2002). In the case of cosplay, as I will argue, orientation to characters' simultaneous placement in sex category and personality category helps participants stage what the ACG community would see as a more authentic recreation of fictional characters in the real world: Whereas singular orientation to a sex category flattens the character's unique persona, double orientation to sex category and personality category can both introduce additional resources for designing the cosplay performance and signal participants' deep understanding and sincere appreciation of the character.
While the "doing gender" framework is useful for analyzing the tension between gender play and authenticity assessment, it has its own limitations-not only for the case of cosplay but as a theory of gender in general. Xiao (2011, p. 610), for example, argues that the theory remains individual-centric in overlooking "the active contribution, beyond sanctioning, that others may make to the production of a particular gendered self. " Indeed, while West and Zimmerman (1987) note that gender performance is staged for, and thus evaluated by, an audience, in empirical analysis each gender performance is still understood as the product of a single actor. But as Becker (2008, p. 7) shows, performing arts, as with arts in general, are never products of individual artists: "all the arts we know, like all the human activities we know, involve the cooperation of others. " An implicit lesson in Becker's likening of artistic activities and other human activities is that, to fully understand the performance of gender, we must take Goffman's dramaturgical metaphor literally and seriously, subjecting it to the "production of culture perspective" (Peterson and Anand 2004) that cultural sociologists have been using for studying artworks of all kinds, including performing arts. This perspective "focuses on how the symbolic elements of culture are shaped by the systems within which they are created, distributed, evaluated, taught, and preserved" (Peterson and Anand 2004, p. 311). Such focus requires a shift of attention from individual isolated artists to what Becker (2008, p. 34) calls "art worlds"-"all the people whose activities are necessary to the production of the characteristic works which that world, and perhaps others as well, define as art. " In short, to the extent that gender is a performance, it is a joint activity involving "an extensive division of labor" (Becker 2008, 13) conditioned by such factors as technology and regulation (Peterson and Anand 2004).

Cosplay in China: an overview
The Chinese cosplay scene is a particularly fruitful case for integrating insights from sociology of gender and sociology of culture. While research on cosplay tends to focus on cosplayers alone (e.g., Leng 2013; Loke 2016), cosplay photos that circulate in China's online ACG community are often produced through explicit collaboration between cosplayers and a supporting crew of makeup artists, photographers, and photo editors. This does not mean cosplay in other countries is devoid of teamwork or digital mediation. Nor does it mean cosplay communities are clearly divided by national boundaries. After all, the internet played a major role in involving fans into transnational recreations of anime/manga characters, and the ACG fandom often foregrounds such multiplicity of authorships (Manning and Gershon 2013;Wood 2006). Insofar as digital mediation and explicit collaboration are two of "[the] few characteristics of animation which make it work particularly well as a trope for thinking about culture in a world of globalized computer technology" (Silvio 2010, p. 428), their overtness in China makes the Chinese cosplay scene a strategic case for exploring how cosplay as a new heuristic can benefit social theory. A cursory review of the history of cosplay in China allows us to speculate on a few factors that may account for such overtness.
Cosplay appeared in mainland China only after 1998 (Liu 2006). Most participants were born after the 1980s and grew up immersed in Japanese ACG products made available by China's market reform (Wang 2010). Despite its late start, cosplay in China developed rapidly in the 2000s (Liu 2006). State and markets played a critical role in this rapid growth. Seeing the huge profitability of the ACG subculture, the Chinese state decided to invest in its domestic ACG industry, while limiting the broadcast of foreign anime on television (Wang 2010). Coincidently, the emergence of online video-streaming around that time enabled fans to continue accessing Japanese anime despite local protectionism. Unsurprisingly, while university-based groups offer a niche for college-age fans (Ruan 2018), the Chinese ACG community has become largely internet-based, with photography being the predominant medium of cosplay (Wang 2010). Virtual spaces like online forums and Weibo serve as highly interactive channels for building friendships and reviewing cosplay photos (Hua 2019). Even when cosplay participants and fans socialize or perform at ACG conventions, they often refer to each other by their online avatar names and take cosplay photos for online circulation (Hua 2019;Jacobs 2013).
Moreover, some businesses capitalized on the visual impact of cosplay as a promotional tool (Hua 2019;Liu 2006). With state investment and business appropriation, cosplay became "vulnerable to commercialization" (Fung and Pun 2016, p. 92). News media often focus on how eye-dazzling cosplay events attract consumers, overlooking what cosplay means for youth and adolescents (Hu 2012). As cultural sociologists have shown, commercialization can trigger the search for authenticity (Duffy 2013;Grazian 2005), and the open border of online community can make authenticity work more urgent (Peterson 2005;Williams and Copes 2005). In a sense, the quest for authentic recreation of characters may help Chinese cosplay participants distinguish themselves as someone who truly loves the character, rather than using cosplay for other ends.
If the pursuit for authenticity raises the quality standards of cosplay, the collaborative labor division allows participants to make the best use of their expertise and deliver a high-quality performance (Ruan 2018). Close cooperation between cosplayers and their supporting crew is common across cosplay groups in China (Fung and Pun 2016;Ruan 2018). It is also common to credit the supporting crew when posting cosplay photos online. Indeed, explicit acknowledgment of multi-authorship is a notable feature of the global ACG subculture that fascinates cultural theorists: Whereas conventional cinema viewers (and scholars) usually see the auteur as the sole creator of a film and assume a one-to-one correspondence between characters and their actors, ACG fans explicitly acknowledge that characters are created by collectives, not just individual auteurs or actors (Silvio 2010;Manning and Gershon 2013). Mirroring such many-to-one creator/ character ratio (Silvio 2010), the emphatic valorization of teamwork in the Chinese cosplay scene makes it an ideal case to theorize from-not just explaining peculiarities, but exploring new ways to conceptualize gender accomplishment.

Methods
Because photography serves both as a venue of cosplay in itself and a means of linking local, offline ACG events with the online ACG fandom, 4 this article focuses primarily on the production of cosplay photos. While literary scholars might prefer interpretive analysis of such photos and communication scholars often conduct quantitative analysis of gender stereotypes in images (e.g., Bell and Milic 2002;Kohrs and Gill 2021;Kuipers et al. 2017), both methods risk imposing the analyst's view as the ultimate meaning or effects of photographic portrayal. Yet for sociologists, it is the meaning that participants seek to convey with, and take up from, the bodily cues and gestures portrayed in cosplay photos that matters. Thus, instead of doing a visual content analysis, this study draws upon interviews with 21 cosplay participants.
Since no population parameter has been established for the Chinese cosplay community, I used internet-facilitated "purposive social network sampling" (Pfeffer 2018). I started by posting my call for participants in online cosplay forums in Baidu Tieba and on my social media. 5 I also asked respondents to introduce me to other cosplay participants they knew, and recruited cosplayers at an ACG convention. In this process, I followed the production of culture perspective (Becker 2008;Peterson and Anand 2004) to purposively recruit participants with different expertise and continued the recruitment till theoretical saturation. The final sample consists of 13 women and 8 men. 12 respondents indicated experiences with makeup, 6 with photography, and 9 with photo editing. 19 were cosplayers themselves, and the other two were photographers who work with cosplayers. The majority of my respondents had cosplayed characters of a different gender than their own, or helped others do so.
Because the Chinese cosplay community is heavily internet-based, I conducted the interviews online via Tencent QQ, an instant message software commonly used by ACG fans for socializing. Doing so allowed my respondents to feel at home during our semi-structured interview and send me photos occasionally for illustration. With a nonprobability sample, my goal is not to statistically generalize my findings, but dive into information-rich cases that "offer useful manifestations of the phenomenon of interest" (Patton 2001, p. 40)-the gender play in cosplay. During analysis, I paid special attention to the tactics that facilitate cosplayers' gender performance and participants' rationales for using them, while also attending to different insights from various roles in the division of labor. It should be noted that this labor division is not as rigid as that of an assembly line. Just as some music composers also act as performers (Becker 2008), some cosplayers do makeup for themselves and serve as photographers and/or makeup artists for others. The empirical goal of this article is not to create a typology of participants but explore the different kinds of tasks required for the production of gender embodiment in cosplay. In what follows, I first focus on the general templates for cosplaying female and male characters and the division of body work involved. I then complicate this male/ female binary by showing how participants' quest for authenticity requires them to produce different masculinities for different male characters. After concluding my empirical findings, I will flip my "case" and "theory" to sketch out the theoretical insights that can be gained by using cosplay as heuristic to look at gender as a "cosplay. "

Cosplayers and body movements: poses and facial expressions
Half a century ago, Birdwhistell (1970) pointed out that the critical human "gender marker" is neither genitals, which are mostly hidden, nor secondary sexual characteristics, according to which men and women are more unimorphic than dimorphic. Rather, it is the "tertiary sexual characteristics"-postures, movements, and facial expressions which are "patterned social-behavioral in form" and "learned and … situationally produced"-that distinguish men from women (Birdwhistell 1970, p. 42). During cosplay, cosplayers often design their poses carefully to embody their source characters' masculinities or femininities, sometimes in consultation with photographers and other members of their supporting crew.
Lvhui 6 highlighted the importance of poses by inviting me to imagine a male and a female character making the same pose: "Don't you think it's weird for a man to hold his cheeks and pretend to be cute? But if a cute girl does this, isn't it nothing strange at all?" While acknowledging that the specific poses and gestures depend on the character's personality and the scenario reproduced in the photo, she identified a few general templates for poses. Poses for men, including the way they sit and stand, tend to be hearty, whereas poses for girls tend to be shyer. "Of course, that's just cute girls, " she added. She used drinking as an example, suggesting that holding a cup with one hand looks heartier and is usually used for male characters, while holding a cup with both hands can be used for girls. Yinwu agreed with the idea that the general goal for posing as male characters is to make the audience feel the character/cosplayer is unrestrained, free to take up more space (and thereby dominate others), while for female characters it is the opposite. "For example, when girls sit on a chair, they usually keep their legs together and angle them; [if you are cosplaying] men, just keep your legs apart. Doing so makes you look more domineering. " These general templates reveal a set of "conventions" in the making of cosplay. According to Becker (2008, p. 29), people do not produce artworks totally from scratch, but rather rely on conventions-"earlier agreements now become customary"-to collaborate with each other and invoke emotional and aesthetic responses in the audience. These conventions are not only technical, but also gendered, both reflecting and constituting our understandings of gender (Mears 2010;Newton-Francis and Young 2015). Indeed, the general templates my respondents identified resonate with both sociological studies of gender double standards in interaction, where women and girls are expected to be more modest than men and boys (K. A. Martin 1998;Weinberg and Williams 2005), and with communication studies of gender stereotypes in advertisements, where women are often portrayed as passive, making gentle or deferential gestures (Goffman 1979;Kohrs and Gill 2021;Kuipers et al. 2017). Such modesty, reflected in women's selfconstraint, makes them an object to whom men can extend their helping hand (Goffman 1977). On the other hand, men's freedom in their body movement makes them an active subject, able to act upon others. In other words, through domineering body movements, female cosplayers can distance themselves from submissive feminine attributes and enact a "dominating masculinity" that involves "commanding and controlling specific interactions and exercising power and control over people and events" (Messerschmidt 2015, p. 33). By posing their bodies in ways that make the viewer feel dominated, female cosplayers bolster their membership claims to the male sex category (West and Zimmerman 1987).
In addition to poses, facial expressions are also used to convey different feelings. Several respondents noted general gender differences in facial expressions that they paid attention to in their performance. For instance, both Ningxun and Luoyu commented that keeping one's eyes "wide open" and puckering one's lips would make one "look like a girl. " Such facial expressions are rarely used for male characters but are often used for female characters. Although these facial expressions could contradict societal ideas about female withdraw and passivity with their active demand for a relationship with the viewer (Bell and Milic 2002;Kuipers et al. 2017), their seductive overtones may attract the audience's erotic attention, posit the character as an object of desire in the cosplayerviewer relationship so established and thus can still be disruptive for cosplaying domineering male characters.
Instead of keeping her eyes wide open, Luoyu would try to squint her eyes when cosplaying male characters "because men's eyes aren't very big. " She added that because many of the male characters she cosplayed were bossy or angry-looking, doing so helped her express "the toughness unique to men, " present her as "more aggressive and manly, " and make people feel she is "cold and unapproachable. " Ningxun, on the other hand, said, "If I'm cosplaying a cute girl, there's no way that I will frown my eyebrows as if I'm going to kill somebody, right?" This rhetorical question makes clear two things. First, he saw such facial expression as destroying the performance of a "cute girl. " As the "killing" metaphor suggests, frowning eyebrows convey a threat, making the expression more associated with men-the violence associated with dominating masculinity can be sufficiently conveyed by facial expressions. Second, he assumed his audience also knows the symbolic meaning of frowning eyebrows and its incompatibility with the image of a "cute girl. " To borrow Becker's words (2008, p. 30), "Only because artists and audience share knowledge of and experience with the conventions invoked does the art work produce an emotional effect. " It is on this common ground that cosplayers like Ningxun and Luoyu can utilize facial expressions to facilitate their gender play in cosplay. In short, like the movement of legs and arms, the movement of facial muscles is also coded with symbolic meanings that can evoke certain feelings in the audience, which then inform a distinct gendered impression.

Supporting crew and body modification: from poses to faces
While different body movements can help cosplayers establish different gendered impressions in their audience, many of my respondents considered bodily features a major constraining factor for cosplay. Liumei commented that she personally thinks people born with a round face are better suited for female characters. "Have you ever seen a cool guy in manga with a round face?" she asked. The same, Liumei continued, could be said about people born with "very large eyes, " which make them look cute. "Have you ever seen a cool guy in manga with cutesy large round eyes?" In asking these questions, Liumei implied there are established conventions about what "a cool guy" looks like in manga, and participants need to take these conventions into consideration during cosplay. Notably, the bodily constraints she discussed center on facial features rather than the "socially agreed upon biological criteria for classifying persons as females or males" (West and Zimmerman 1987, p. 127). For some female cosplayers, their bodies made it easier for them to cosplay certain male characters than certain female characters. Lvhui, for example, told me that because her nasal bridge looks "kind of high" and her chin looks "kind of flat, " when she showed someone a photo of her cosplaying a female virtual singer, she was told that "this is a boy. " But just as there are conventions around characters' looks, there are also conventions, or general templates, for dealing with cosplayer's bodily constraints, even though their application is not rigidly formulaic. While cosplayers could adjust some of their bodily features through dieting or other means, the supporting crew plays a critical role in helping cosplayers overcome these constraints through body work. Indeed, one of the most immediate responses to my question about how to appear as a character of a different gender was "makeup. " During the interview, many of my respondents spent a lot of time talking about the symbolic meanings of modified bodily features achieved by using eyeliners, false eyelashes, highlighting powders, and contour shadows. Eyes and eyebrows are particularly important objects of body work. While cosplayers sometimes squint their eyes when cosplaying male characters, makeup artists manipulate the size of cosplayers' eyes and the shape of their eyebrows to help them convey feelings of aggressiveness and toughness via eye contact. Using her appearance as an example, Lingdian commented, "My eyes are round, so I need to make them look longer, because men with round eyes are sissies. " To do so, she "opens" the inner corner of her eyes and extends her eyeliners. Similarly, Ange told me that she usually draws straight eyebrows going upward from brow ridges for male characters and softly curved eyebrows for female characters. When asked why she does so, she said it was for the visual effect-"To describe it in our term … [straight eyebrows have] an 'aggressive aura. '" Many respondents agreed that slender eyes are more "penetrating" and convey aggressiveness, whereas large round eyes indicate softness. Similarly, curved eyebrows look "soft, " while a prominent brow ridge looks "sturdy. " Suggesting a penetrating power to subordinate others, slender eyes enable cosplayers to embody dominating masculinity even at the absence of physical aggression. In one interview, Ruoqian asked me, "If someone with very large [round] eyes stares at you, do you think he's actually going to be fierce?" In other words, the slender eyes produced by the body work of makeup artists, often in spite of physiological constraints of the cosplayer, are designed to trigger feelings in the audience and impress them with the character/cosplayer's aggressiveness-under the backdrop of symbolic conventions shared by cosplay participants and their audience. These bodily features influence the way we interpret actions. While staring on its own may be a dominating act, having large round eyes can make the cosplayer appear too soft to claim the power expected to execute the implied physical aggression afterward.
Other facial features are also crucial for gender performance. One of the most conventional techniques the supporting crew use to adjust facial features is balancing light and shadow. Qianmo, who studies makeup professionally, tried to explain this technique to me with as little jargon as possible: "For male characters, nose contouring needs heavier use of color. In addition, cheekbones need to be contoured as well. Those facial parts that need highlighting also need to be made whiter than usual … Dark colors look concave, and highlight looks convex, so [together] they make facial features look more prominent. " Ange echoed with Qianmo's makeup tactics. Indicating her personal gender identity as a girl, she suggested that her lack of sturdy facial feature evinced femininity already: "Girls' facial features tend to be soft. Since I'm a girl, I don't need to emphasize those. Doing some light makeup will be enough. " In contrast, when cosplaying male characters, she has to do makeup "heavily, " applying shadow to her cheeks, eye sockets, forehead, nose, etc., to make these facial features "more prominent and angular. " Photographers can also engender cosplayers' facial feature with the help of lighting techniques. Qingkong uses even and "soft" lighting for female characters to "eliminate shadow without making the photo lose its stereoscopic feeling. " For male characters, he uses a combination of light and shadow to make the cosplayer's face appear "sturdy and angular": the contrast of light and shadow gives a visual impact that will make cosplayers look more powerful, he explained. By replacing the soft impression of cosplayer's face with an angular and sturdy illusion, such lighting technique helps cosplayers manifest the power essential to dominating masculinity. Again, this is important because softness in facial features is understood as an iconic index (Gal and Irvine 2019) of softness in interaction. An artistic convention that guides the supporting crew's body work, as these comments suggest, is the symbolic association of prominence and sturdiness with sense of power on the one hand, and softness and curviness with sense of weakness on the other. By infusing these symbolic meanings into cosplayers' bodies, photographers and makeup artists help cosplayers bolster their membership in the sex category of their chosen character.
Photographers' body work does not focus solely on cosplayers' face, as their choice of camera angle shapes how cosplayers' body appears in the photo. When shooting photographs for female cosplayers that are cosplaying male characters, Yeguang would position his camera in angles that would not "feminize" cosplayers. "If you shoot from the side, " he said, "her breast would look very obvious, wouldn't it?" Luoyu, on the other hand, pointed out that even body size is malleable with careful choice of camera angle-"Shooting from top down works better for female characters because [the cosplayer] will look petite. For male characters, [shooting from] bottom up will make the cosplayer look taller and bigger. " She added that when a female cosplayer is cosplaying a male character, "photographers should make sure the cosplayer is not starkly contrasted with the surrounding scene, such as desks and chairs, [because] this [contrast] will make her look petite. " Photo-shooting can be even more complicated if multiple cosplayers are involved. Luoyu recalled a time when she and another girl were cosplaying a heterosexual couple. Although Luoyu was cosplaying the male character, the other girl was four to five centimeters taller than her. "It's such a pain when we shoot the photos, " Luoyu said, "[I wore] elevator shoes with both height-increasing insole and outsole, and [stood on] a small stool. " Still, the stool only worked for half body shot. For full body shot, they had to "take advantage of where they stand" while the photographer tried to "find the right angle" to "make their height difference look normal. " Such pursuit for "normal" height difference between male and female characters resonates with Goffman's (1979, p. 28) analysis of gender in advertisements, where men's power and authority over women are "expressible in his greater girth and height. " Whereas Goffman elaborates such dimorphism with selective mating and other social selections, Luoyu's experience illustrates what can happen after selections. With careful choice of camera angle, photographers contribute to the production of such bodily dimorphism, helping female cosplayers to signal "natural" physical dominance as male characters.
After makeup artists groomed the cosplayer and photographers took the shots, production enters the post-processing stage, during which photo editors would adjust the photo's color tone and add special effects as necessary (e.g., if the character is using magical power). At this stage, cosplayer's gender presentation can be further polished. "Those who did an excellent job in cosplaying male characters online may not look like that offline, " Luoyu revealed to me, "Of course, some girls do have face that looks like boys' , but most girls don't have man's skeleton. So in addition to makeup, we also put effort into post-processing. " Tanwei's experience tellingly illustrates Luoyu's observation. Although she was rarely recognized as a man when she cosplayed male characters at ACG conventions, she recalled, "After my photos went through post-processing, people asked me if I'm a man or a girl. " Hence, even though she knew little about the technical details of post-processing, she was well aware of its importance for the success of cosplay.
Shenwu explicated such technical detail, telling me that photo editors need to adjust the size and shape of cosplayer's face, canthi, and body with filters in Photoshop. For example, whereas for female characters, "[cosplayer's] eyes need to be made bigger within a reasonable degree, " for male characters, cosplayer's canthi need to be made "slightly longer and sharper. " As in makeup, such bodily modification makes cosplayer's eyes more penetrating for male characters. Similarly, Ruoqian told me that, to make female cosplayers look like male characters, she uses the liquify filter to "make eyebrows lower, face sharper, eyes longer, and nose smaller. " Photo editors and makeup artists, therefore, share a similar goal and follow similar conventions in their body work-to imbue gendered sensibilities into bodily cues-albeit with different tools.
Although Photoshop offers an easy way to modify cosplayer's bodily features, some respondents saw it as supplementary to makeup. After Ange elaborated on her makeup tactics, I invited her to talk about her experience in post-processing by asking how she made female cosplayers look like male characters and vice versa during that stage. She gave me a rather brief answer. "There's not so much to talk about post-processing, " she said, "If the eyebrows you drew doesn't look aggressive enough, liquify it to [make it] look more aggressive … Basically it's a remedy [for makeup] to adjust facial features. " In other words, Photoshop is not a substitute for the body work done by makeup artists and photographers, but used to pursue perfection, and such coordination is made possible by their shared conventions about body modification. As Ruoqian commented, "If you rely on Photoshop for everything, why don't you go and draw the character instead?" Rahman et al. (2012) identified commitment as an important dimension of authenticity in cosplay. Albeit not referencing authenticity directly, Ruoqian's comment implies that over-reliance on Photoshop indicates a lack of commitment, which in turn renders the cosplay pointless. This is not to say that post-processing is unimportant. Rather, photo editors can help cosplayers resolve their bodily constraints that makeup artists cannot fully resolve. To borrow Yinwu's words, "After all, it's people's collective effort that makes a good cosplay. " It is interesting to note that most respondents rarely elaborated on changing visual markers like facial hair, breasts, or Adam's apple. Mostly, they briefly mentioned using shoulder pads, breast-binders, and Photoshop to change secondary sex markers, and quickly moved on, as if making these modifications is a matter of course. In contrast, my respondents discussed subtle features and their symbolic connotations in much greater detail. Albeit subject to body work, subtle bodily features like eye shape appear natural and immutable, while a secondary sex marker like facial hair can easily be shaved. In dramaturgical language, because we often expect "a confirming consistency between appearance and manner" (Goffman 1959, p. 15), the seemingly fixed nature of bodily features may serve as a naturalizing justification for cosplayer's masculinity/femininity. The body is thus actively involved in doing gender not only through movement but also modification, as modified bodies are also used to produce non-biological differences that are rendered as natural (West and Zimmerman 1987). In cosplay, such modification often relies on a supporting crew of makeup artists, photographers, and photo editors. Gender embodiment in cosplay, therefore, is a collective process.

Double categorization: recreating a soulful character
So far, my findings seem consistent with Leng's (2013) assertion that gender-crossing cosplayers are donning hyper-femininity/hyper-masculinity. The general templates presented above portray male and female characters as dichotomous, mutually exclusive stereotypes. But general templates are, after all, general. Simply following these templates is insufficient for cosplayers to "authentically" recreate their source characters. Just as Becker (2008, p. 29) notes that "a particular convention may be revised for a given work, " many respondents were quick to qualify their general observations as not applying to certain types of characters. Others insisted that every character is unique and there is no such a thing as a general template for cosplay.
When I asked Liumei if she discovered any commonalities while cosplaying certain types of characters, she asked me instead, somewhat impatiently, "Is there any commonality between you and me? If there is, wouldn't we be twins?" Liumei's impatience and sarcasm reflect a tension in our interview, between a sociologist trained to search for patterns and a cosplayer socialized to pursue authenticity. Obsession with patterns across characters risks obscuring the individuality of each character, which can make the cosplay inauthentic. Even Luoyu, who discussed these patterns in great detail, stressed that "cosplay without a deep understanding of the character is soulless. " "I can see a cute girl or a handsome guy, " she said, "but I can't see the character. " In other words, even though cosplay participants may develop a "folk sociology" of gender, they have more to consider than just the character's gender. Albeit fictional, each character has their own personality and may thus deviate from gender stereotypes. Successful cosplay, in which a portrayal is seen as authentic to the character, displays an orientation not only to the source character's membership in a sex category, but also to the character's simultaneous placement in a personality category, which further requires cosplay participants to modify, rather than just exaggerate, gendered attributes.
As I mentioned earlier, from the ethnomethodological perspective that informed the doing gender framework, sex and personality are two of the many membership categorization devices that enable us to describe a character with a category contained in them (Sacks 1972;Schegloff 2007;Heritage 1984). Like "man" and "woman" (West and Zimmerman 1987), each category from such a device serves as a "store house" of conventional knowledge that helps us interpret and construct actions for its members (Schegloff 2007, p. 469). Yet because the permutation of "correct" categories is infinite (Heritage 1984), and one category from any device can be "referentially adequate" (Sacks 1972, p. 333), people must decide which category or combination of categories is the most situationally relevant one(s) for constructing their actions.
While West and Zimmerman (1987) argue sex category is omnirelevant in our society, personality category is no less salient than sex category in the ACG subculture. For instance, when consuming Boys' Love manga, in which male characters are often depicted as beautiful and in homoerotic relationships, fans regularly categorize characters as seme (top, literally to attack) or uke (bottom, literally to receive). 7 Despite internal variations and subdivisions, seme "designates a dominant personality" whereas uke "designates a submissive personality" (Galbraith 2011, p. 222). With the global circulation of Boys' Love manga and slash-like fan fictions, terms like seme and uke have "become part of the collective jargon of this particular counterpublic discourse, regardless of an individual's native language" (Wood 2006, p. 405). Many ACG fans even apply these categories to explicitly straight characters when consuming stories outside this genre, thus "making explicit homoerotic subtexts to originally 'straight' works" (F. Martin 2012, p. 368). Seme and uke are not the only personality categories. Others include tsundere (cold/hostile on the outside but warm/kind on the inside), mukuchi (reticent, literally mouthless), shōta (youthful boys), onē-sama (mature and domineering ladies), etc. In short, when seeing characters, fans not only place them into a sex category, but also into a personality category.
In the context of cosplay, this means that fans would become disconcerted when a submissive man (uke) is not cosplayed like a submissive man, or a domineering lady (onē-sama) is not cosplayed like a domineering lady (cf. West and Fenstermaker 1995). Youqing confirmed the importance of personality categories by pointing out that the gender boundary is not the only barrier for what kinds of characters one can successfully cosplay. Noting that many male uke characters in manga look very effeminate, she said, "Because I'm pretty big and strong, I'm not suitable for cosplaying these characters just like I can't cosplay many female characters. " Youqing suggested that even though in her personal life, she had short hair and a flat chest, was often called a "boy" by others, and behaved like a gentleman in front of girls (and even had girls fall in love with her as a result), these traits and experiences were not always advantageous for cosplaying male characters, because not all male characters are depicted as "big and strong. " Personality matters in cosplay because intra-gender differences in ACG subculture is not smaller than inter-gender differences.
To the extent that the success of a performance is evaluated by how much the cosplay team authentically recreated the source character's physical and mimetic attributes (Norris and Bainbridge 2009;Rahman et al. 2012), one cannot successfully perform gender without simultaneously performing personality. On the flip side, since attaining a deep understanding of a character's personality often requires fans to immerse themselves into the anime/manga/game where the source character comes from, displaying an additional orientation to the character's personality category can potentially signal one's sincere appreciation of the character and set up an implicit contrast with those who understand the character only superficially and stage a "soulless" cosplay. After all, while we presume one's gender should be evident at a glance (Kessler and McKenna 1978;West and Zimmerman 1987), categorizing one's personality demands a deeper level of engagement.
This in-depth comprehension of the character not only can serve as a resource for evaluating the authenticity of cosplay photos circulated online, but can also be a resource for designing one's own cosplay performance. As archetypes, personality categories introduce additional conventions that can assist the management of poses, facial expressions, and bodily features when ACG fans recreate a fictional character in the real world. Whereas singular orientation to a character's sex category risks flattening the character into what Luoyu called "a cute girl or a handsome guy, " attending to the relevance of both sex category and personality category can produce a diverse range of masculinities and femininities, from which participants may choose practices and attributes that best resonate with the source character's "soul. " As gender scholars have long argued, masculinities are plural, and there is no single and universal conception of manhood (Connell 2005). Although some practices and representations of men may overlap with the "quality contents" commonly associated with femininity (Schippers 2007), they can nevertheless present alternative ways of "being men" without being dominating and should thus be recognized as masculinity (Connell 2005;Messerschmidt 2015). Assessing whether the diverse range of gender representation in the ACG subculture can challenge the gender binary or continues to reproduce heteronormativity is far beyond the scope of this article. There is already a bourgeoning literature growing out of this concern (Bainbridge and Norris 2013;Gn 2011;Lamerichs 2011;Loke 2016;Norris and Bainbridge 2009;Tompkins 2019). However, the multiplicity of masculinities is key to cosplay because it calls for different body movements and modifications for different male characters. In what follows, I mainly focus on how cosplay participants stage distinct kinds of masculinities for uke and seme characters to make their performance authentic in the eyes of their audience.
Although the general templates for cosplaying male/female characters often center on embodying gendered sensibilities like aggressiveness/softness, when cosplaying uke characters, cosplayers also target a soft aura to present a non-dominating masculinity. After telling me that "seme has a more aggressive aura, whereas uke has a softer aura, " Qianmo suggested the facial expression for uke tends to be softer than seme. She elaborated this difference in facial expression with a metaphor of debt. For seme, "it's like you owe me five million. " She would keep a poker face or smile with only one side of her mouth moving upward, while applying some force on her eyelid. For uke, in contrast, it could be either "[you] owe me three million or I owe you three million. " Behind this debt metaphor is a matter of power difference and righteousness for exercising domination. Facial expressions for cosplaying seme, such as a scornful smile, provoke a sense of guilt in the viewer, positing them as subordinate to the character/cosplayer. While facial expressions for cosplaying some uke characters do so to a lesser extent (owing three rather than five million), those for other uke characters flip this power dynamic (I owe you, rather than you owe me), inviting the audience to take a more authoritative position in the visual encounter. Luoyu also noted differences among uke characters. While cosplaying tsundere uke, who are cold on the outside and warm on the inside, involves such poses of denial as "turning your head away" or "pushing the seme away, " when it comes to "lamb-like uke, " poses and facial expressions would be more similar to those for cosplaying girls.
Personality difference also plays out in body modification. Ruoqian, for example, distinguished makeup for seme from makeup for other softer male characters, like shōta and uke. For male characters with a softer aura, she tries to make eyes larger by adopting such tactics as using circle lenses, thickening the middle portion of upper eyeliners, and leaving a small gap between the eyeliner and her lower eyelid at the end of her eyes. For seme, she aims to make eyes smaller and slenderer with smaller-sized cosmetic lenses and by applying eyelashes closer to her eyes. Such eyes, Ruoqian suggested, look more forceful and aggressive.
The interplay between personality category and sex category thus creates rich intragender variation in character presentation, making it important to distinguish the nuanced details of each performance, such as how one sits or how eyebrows are drawn, against the general templates described in the previous section. Gender play in cosplay relies on collective body work. Regardless of the personality or gender of the source character, both body movement (through poses and facial expressions) and body modification (through makeup, post-processing, and other means) are crucial for the success of gender play in cosplay, as they infuse gendered sensibilities into cosplayers' body. Yet the importance of authenticity in the ACG fandom means that the cosplay team needs to draw upon their comprehension of the source character's personality and carefully decide what specific information to convey with the cosplayer's body movement/feature, and what kind of masculinity/femininity to present. Blindly following the general template to perform hyper-masculinity for male characters, or hyper-femininity for female characters (Leng 2013), deprives the character of its soul, reveals the team's lack of understanding of the character, and renders the performance inauthentic.

Conclusion and implications
Despite the global popularity of cosplay, scholars have mainly focused on why fans participate in cosplay and whether it subverts gender norms. Missing from these discussions are the processes of how gender is performed during cosplay and how these processes can inform our understanding of gender. Integrating theoretical insights from sociology of gender (Goffman 1976;West and Zimmerman 1987) and sociology of culture (Becker 2008;Peterson and Anand 2004), this article not only shows that gender is embodied via conventionalized performances and that gender embodiment in cosplay is a collective effort, but goes a step further to document the minutiae details of body work, the communicative effects of bodily cues, and complications that can arise in the quest for authenticity.
Broadly speaking, whereas cosplayers convey gendered messages with body movements, their supporting crew infuse these messages into bodily features. Whereas uninhibited body movements and angular, prominent, and slender facial features can convey gendered sensibilities of aggressiveness, toughness, and dominance, constrained body movements and curvy/round facial features can convey softness, passivity, and submissiveness. These collectively produced bodily cues can not only bolster membership claims to a sex category (West and Zimmerman 1987), but also invoke certain feelings in the audience, leave them with a gendered impression of the character/cosplayer, and render the perceived power difference as essential and biological.
In the art worlds of cosplay, this collective process is further complicated by the importance of authenticity and the salience of personality category. Instead of simply donning hyper-masculinity or hyper-femininity (Leng 2013), cosplay participants also need to attend to personality category and stage different versions of masculinities/femininities as consistent with this additional categorization. In addition to signaling participants' deep comprehension and sincere appreciation of characters, such orientation can also introduce further sets of conventions for designing and evaluating cosplay performance. Accordingly, as much as imposing constraints and drawing boundaries, the performative nature of authenticity (Duffy 2017;Grazian 2010;Peterson 2005) can direct participants to resources that might otherwise be overlooked.
While my interview-based design privileges participant's understanding of, and rationale for, using certain bodily cues in the making of cosplay photos, the study has other limitations. For instance, the spatiotemporal separation of the interview site and the cosplay site may encourage respondents to typify their tactics and give reconstructed accounts. They may thus be less inclined to talk specifically about how they exercise artistic creativity and how their team deal with disagreements that might occur as a cosplay shooting session unfolds. Hence, this article focused more on conventions and collaboration than creativity and conflicts. Future work may benefit from more ethnographic approaches that can better capture how participants find rooms for artistic creativity, document what convention goes without saying and what can be verbally contested, and unpack whether these moments of negotiation can make the collective product of cosplay more than the sum of its parts-not just additive but multiplicative.

Cosplay as a new heuristic?
For social theorists, the point of studying a community is finding out something beyond the community we study. Just as Garfinkel's ([1967] 1991) analysis of Agnes revealed not only passing strategies of trans people, but the seen yet unnoticed aspects of gender, just as Goffman's (1979) analysis of advertisements unpacked not only gender images in cultural production, but the orchestration of gender displays in everyday interaction, just as Butler's (1990) analysis of drag exposed not only subversive practices in queer communities, but the illusory naturalization of an interior gendered self, a study of cosplay may generate theoretical insights that can be extended beyond the narrow (albeit now global) confines of the ACG fandom (see also Silvio 2010). Theory, as Abend (2008) reminds us, includes not only propositions and explanations but also hermeneutics and ways of looking at the social world. Passing, display, and drag have all been useful heuristic tools that helped us theorize gender. But what if we take cosplay as a new heuristic? Having looked at how gendered images are manufactured in the Chinese cosplay scene, what insights can we gain by flipping our "case" and "theory, " by looking at gender as a cosplay? To end on a constructive note, I argue that using cosplay as a heuristic helps us bring together different lines of inquiry to sketch out the art worlds of gender performance.
First, using cosplay as a heuristic requires us to attend to the variegated ways our body figures into our accomplishment of gender. As Messerschmidt (2009, p. 86) notes, "most writing on 'doing gender' ignores the body. " Under the demarcation of sex (biology) versus gender (society), the body is often excluded from the social realm of gender. But my respondents reveal that it is also a useful resource for enacting femininities and masculinities. In Wacquant's (1995, p. 67) words, the body is both our "means of production" and the "raw materials" for gender presentation: We "work with" the body as we interact with each other in the social world, but we also "work on" the body by imbuing cultural messages into embodied physical features, thus disguising these messages as natural elements of the material world (see also Hoang 2015).
Second, even as beauty filters relieved the headache of photoshop, the labor of gender embodiment is not always done by ourselves. Nor is collaborative embodiment necessarily limited to artistic realms like cosplay or fashion modeling (Mears 2010). When even feminist scholars are tempted by neoliberalism to see gender as an individual achievement based on personal choice (Risman, Myers, and Sin 2018), cosplay as a heuristic urges us to take what is made explicit in subcultural spaces to other arenas of everyday life, where it may be obscured in different forms. That is, as sociologists, we can examine if there are "ghost" co-authors behind seemingly solo-authored gender presentation of "self, " and if its collective production varies by contexts of interaction and means of mediation. These co-authors may be parents or partners who enforce gender-appropriate conducts at home (K. A. Martin 2005; Ashwin and Isupova 2014) or salon workers who help us invest in our appearance (Barber 2016); They may be co-participants of interaction who enhance our gender subjectivities with their "gender labor" (Ward 2010;Xiao 2011;Hoang 2015;Mears 2020), or non-participating audiences who look at our performance and conclude that it is indeed masculinity/femininity (Messner 2000). Just as the gender play in cosplay would be compromised if a role of the cosplay team is unfilled, our accomplishment of gender may have a different result if these "collaborators" are absent.
Third, using cosplay as a heuristic garners attention to the construction of "authentically" gendered selves. Like fictional anime characters, qualities like "hegemonic masculinity" are abstract ideals against which our conducts are assessed (Connell 2005). Although they have no ontological status within us, 8 we often strive to embody them in such a way that our gender displays can be read as expressions of our innate essence (Goffman 1976;Butler 1990). Yet as West and Zimmerman (1987, p. 134) note, if one tries to be "120 percent female, " she "could have given herself away by overdoing her performance. " For cultural sociologists, the issue here is "authenticity to constructed self " (Peterson 2005(Peterson , p. 1089. Such authenticity is also achieved in interaction, involving impression management and authentication (Grazian 2010;Hughes 2000;Peterson 2005). If the arbitration of authenticity has the discursive effect of materializing constructed distinctions, examining the construction of "authentically" gendered self can help us better unpack the naturalization process behind the "doing" of gender.
Fourth, in the age of intersectionality, the quest for authenticity can be further complicated by the multiplicity of our identities (Collins 1986;West and Fenstermaker 1995;. As Duffy (2013; shows, while ideals of femininity had historically been infused with white, middle-class, heteronormative assumptions, these ideals are increasingly challenged by "real woman" narratives: with the rise of social media and fashion blogging, celebrating expressive individualism, body diversity, and racial/ethnic/sexual identities has become integral to the quest for authentically gendered self. These intersecting identities can be seen as resulting from jointly invoked membership categorization devices (Sacks 1972). While the relevance of personality categories may not be directly transposable to other contexts, using cosplay as a heuristic can attune us to the grounds on which multiple categories are invoked and the interactional resources they introduce. In some situations and for certain interactional goals, one sex category alone can be adequate (Sacks 1972;Schegloff 2007), but in other situations and for other ends, such as for the quest for authenticity in cosplay, they may not be sufficient. It is the job of sociologists to examine when, and for whom, claims to authentically gendered self can rest solely upon sex-category-based narratives, and why sex category is jointly invoked with a category from one membership categorization device rather than another such device.
In short, to the extent that femininities and masculinities are cultural objects produced and embodied through everyday performances, gender can be said to have its own "art worlds" (Newton-Francis and Young 2015). Cosplay as a heuristic allows us to reconceptualize gender embodiment as a collectively manufactured artwork that requires a division of labor for its production. It also offers an exciting opportunity for further theoretical integration between sociology of gender and sociology of culture. For sure, my specific findings (e.g., the use of Photoshop) may be limited to the ACG subculture, and in themselves they cannot "prove" gender is a collective accomplishment. By definition, reconceptualization relies on extrapolation, as it implies a new way of looking at the world. Gender production in the internet-mediated ACG fandom may well differ from gender production in face-to-face interaction. Understanding the specific ways such difference plays out requires further empirical analysis. But when we take cosplay as a new heuristic to examine dissimilar cases like these, we may find new momentum for further theoretical advancement. And this is indeed one of the many beauties of social theory.