1. Introduction
[1.1] Who reads and writes fan fiction—and why—has long been a central concern of fan studies. Many of the foundational works in the field of fan studies aim to answer these questions, including Bacon-Smith's (1992) Enterprising Women, Jenkins's Textual Poachers ([1992] 2013), Penley's "Brownian Motion" (1991), and Russ's "Pornography by Women, for Women, with Love" ([1985] 2014). These early works depicted fan fiction to be the domain of women—women who until recently have most commonly been depicted as white, straight, cisgender, middle-class, adult, and Anglophone (see, e.g., Busse and Lothian 2018; Hellekson and Busse 2006; Scott 2013; Lothian, Busse, and Reid 2007; Stanfill 2011) (note 1). This understanding of who produces and consumes fan fiction has fundamentally influenced how we conceive of fan fiction itself, including its genres, structures, contexts, functions, and aims.
[1.2] Nonetheless, there has been increasing acknowledgement that this stereotypical female fan figure is, perhaps, outdated. As reading and writing fan fiction have become more mainstream, fandoms' demographics have shifted (Barnes 2015; Coppa 2006; Hellekson and Busse 2006), yet we know very little about who is reading and writing fan fiction online today. Researchers like Click et al. (2018) suggest that fan studies "has much to gain by looking at the vast amount of materials circulating freely online" (442), while Barnes (2015) argues that "there is a need for broader research that investigates the cultural demographics of fan fiction writers" (75). Indeed, fans' identities are of interest not only in and of themselves but also because fan fiction's content, modes of circulation, structures, communities, and sociopolitical functions are inflected by fans' locations and identities. The work that fan fiction does with popular texts, the networks surrounding it, and the systems of representation bound up in and bounding it are all inflected by who produces and consumes it, just as individuals, in turn, are affected by the fan fiction they read.
[1.3] This article presents the results of a wide-scale qualitative study of fans' demographic information, as shared in the paratexts of 1,939 Harry Potter fan fiction works posted to Archive of Our Own (AO3). While the study is limited to one fandom and one archive, it nonetheless broadens our understanding of who engages with fan fiction. The findings suggest, in particular, that the queerness of the Harry Potter fan community is more widespread and diverse than previously estimated and thus ought to be more widely acknowledged in future research.
2. Who reads and writes fan fiction? An overview of existing research
[2.1] Early American (note 2) studies of fan fiction focused on zine culture (e.g., Bacon-Smith 1992; Jenkins [1992] 2013; Penley 1991; Russ [1985] 2014; see also Coppa 2006) and "overwhelmingly focused on female fans, often suggesting that fan culture functioned as both a feminine and a feminist space" (Click and Scott 2018, 1) (note 3). Such studies implicitly or explicitly characterized fan fiction as composed and shared within communities of adult, cisgender, straight, "middle-class, educated, liberal, English-speaking, white North American women" (Lothian, Busse, and Reid 2007, 104). The similarities between the women in these communities were implied to be perpetuated by the communities' invitation-only status (Hellekson and Busse 2006; Bacon-Smith 1992).
[2.2] However, fandom's shift online has challenged this stereotype. Not only have scholars admitted that "the initial focus on women as fans resulted in some hasty and easy generalizations" (Scott 2013, ixx) but reading and writing fan fiction have also become more mainstream activities (Barnes 2015, 74). As Hellekson and Busse (2006) have suggested, the digitalization of fan fiction has diversified and democratized fan fiction–centered communities, making them more accessible to all fans despite their ages, financial means, ethnicities, nationalities, locations, linguistic knowledge, sexualities, and genders. Nonetheless, fan communities' preference for anonymity (Brennan 2014; Hellekson and Busse 2006), combined with the popularity of qualitative, self-referential methods in fan studies (Barnes 2015; Click et al. 2018; Hills 2012), have ensured that we know very little about who actually produces and consumes fan fiction today.
[2.3] Additionally, fandom's move online and the concomitant diversification of the fans participating in fan fiction communities have resulted in some shifts in these communities' priorities. These have included bottom-up moves to make slash fan fiction—fan fiction depicting homosexual relationships—less objectifying of homosexual men (Busse and Lothian 2014) and to make people of color more visible (Fowler 2019; Wanzo 2015). However, practices of gatekeeping have not entirely disappeared and can quickly "slip into toxicity…[and] transform communities into hostile spaces, especially for fans or characters that are traditionally Othered or marginalized" (Walton 2018, 239), leading to arguments, ostracism, and siloization (Hills 2002, 1). As Brennan (2014) and Walton (2018) have argued, marginalization within fandoms is very often a matter of degree—groups marginalized outside a given fan community may nonetheless be a majority within that community, and they may themselves exclude or malign others. As such, while fan fiction–centered online communities at times act as a space for cisgender women to explore a female heterosexuality that objectifies (gay) men (Brennan 2014; Zubernis and Larsen 2012), for others, it is a space in which fans can explore nonmajority perspectives and experiences (Brennan 2014; Brough and Shrestova 2012; Busse and Lothian 2018; Driscoll 2006; Duggan 2017a; Hampton 2015; Russo 2013, 2018; Lothian, Busse, and Reid 2007; Willis 2006).
[2.4] Resultantly, there has been a marked increase in qualitative work examining the diversification of fandom since its move online, including work discussing who has been marginalized both within online fandoms and by fan studies. Hellekson and Busse (2006, 13), for example, have drawn attention to "ever-younger fans who previously would not have had access to fannish culture" and to those whose financial resources would not have stretched to allow them to take part in conventions or purchase zines. Along with a number of other scholars, they have also drawn attention to the increasing transnationalization of fandom (Chin and Morimoto 2013; Hellekson and Busse 2006; Jenkins, Ford, and Green 2013; Morimoto 2018; Williams 2012) and, with it, increasing linguistic diversity within and across fan communities (Chin and Morimoto 2013; Duggan and Dahl 2019; Morimoto 2018). Scholars have also drawn limited attention to the varied sexualities and gender identities of fans (Brennan 2014; Busse and Lothian 2018; Driscoll 2006; MacDonald 2006; Russo 2013, 2018; Willis 2006), as well as their varied ethnic and racial identities (De Kosnik and carrington 2019; Fowler 2019; Gatson and Reid 2012; Stanfill 2011, 2018; Thomas 2019; Thomas and Stornaiuolo 2016; Wanzo 2015). This research has brought the normative figure of the fan into question, but further research examining what traits can be considered (non)normative or majority/minority within various fan communities is required. More than anything else, these recent qualitative studies have made it clear that the idea that fans who produce and consume fan fiction are mostly ciswomen, white, straight, English-speaking and living in Anglophone-majority countries, middle-class, and higher educated should be questioned.
[2.5] One of the fandoms that has been most noted by scholars to be concerned with social justice issues is the Harry Potter fandom. While the books themselves, despite their seeming celebration of difference and championing of diversity, problematically champion white, straight, and able bodies (Duggan 2019; Horne 2010; Pugh 2011; Pugh and Wallace 2006, 2008; Rana 2011; Thomas 2019), a majority of acafans have argued that the online fandom is a space in which reparative readings are emphasized and social justice championed (Duggan 2017a, 2019; Fowler 2019; Tosenberger 2008a, 2008b; Willis 2006)—and sometimes translated into real-world activism (Brough and Shresthova 2012; Hinck 2012; Jenkins 2012a; Kligler-Vilenchik et al. 2012). Of course, groups do exist who insist upon the series' whiteness and heteronormativity, and discrimination against minority fans certainly occurs (Fowler 2019; Thomas 2019; Walton 2018); however, the Harry Potter fandom is nonetheless one in which queer subtexts are actualized (Duggan 2017b, 2019; Fowler 2019; MacDonald 2006; Tosenberger 2008a, 2008b; Willis 2006), in which critical race concerns are increasingly central (Gilliland 2016; Seymour 2018), and in which intersectional concerns are addressed (Fowler 2019). Due to the tensions that exist within the fandom and between the commercially produced texts and their fannish interpretations, the makeup of this fandom is particularly intriguing, but while it has been made clear through various studies that minorities are a real presence within the fandom (MacDonald 2006; Thomas 2019; Willis 2006), how Harry Potter fan fiction writers and readers identify remains a mystery.
3. Zooming out
[3.1] One theorist in particular shaped our understanding of how fan scholars should do fan studies: Henry Jenkins. He is credited with coining the term acafan to refer to academics who are themselves fans and who use their "subcultural knowledge" to inform their academic work (Jenkins, Rand, and Hellekson 2011). The term prioritizes the first-person, the immersive, the experiential, and the close because these are "presumed to combine scholarly practice with depth, detail and rigor of fan knowledge" (Hills 2012, 15; see also Geertz 1973; Love 2015) and to allow us to write in sympathy with our research subjects (Click et al. 2018; Jenkins [1992] 2013; Scott 2013). As such, (auto)ethnographic studies relying on "qualitative analysis and cultural immersion" (Click et al. 2018, 441) and small-scale case studies, in which the experiences of a group of fans are closely examined, are favored methodologies (Barnes 2015) (note 4).
[3.2] However, overreliance on such approaches has garnered some criticism. Notably, Hills (2002), Scott (2013), and Duffett (2013) have expressed concern that we acafans use the "imagined subjectivity of fandom" to romanticize fans' "'affect,' 'love,' or 'excessive positioning'" (Hills 2002, 15). This has resulted in a "rather dismaying short-sightedness" (Hills 2002, 15), making us "blind…to certain aspects of the community" (Click et al. 2018, 440). Critiques emphasize that normative tendencies in the field perpetuate our finding of our own "mirrors" (Bishop 1990) in fan cultures; as Wanzo (2015) notes, the citationality of the concept of acafandom often conceals the practices and methods subsumed by the concept. She and a number of other scholars (Click and Scott 2018; Hills 2002, 2012; Sandvoss 2005) have noted that we tend to focus on fans who are like ourselves and on the aspects of fandom we find most exciting. This means that who researches fandom and who fans are become conflated and coconstitutive in research, misrepresenting the fannish landscape. As Hills (2012, 21) convincingly argues,
[3.3] The mirroring of specific fan identities in scholar-fandom is…a skewed, distorting mirror which threatens to render specific fandoms automatically canonical…whilst also marginalizing a massive range of [fans,] media fandoms, and material cultures.
[3.4] The experiential weight of attachment is the price paid for our widespread acceptance of the tenets of acafandom. As Love (2015) has argued, "stabilizing methods—with their suggestion of a neutral or unbiased view—have been seen as a form of violence [from the point of view of late-twentieth- and early twenty-first-century humanities]; however, we might also understand these methods as implying respect for our objects, since their aim is to keep open the possibility that one might be surprised or proven wrong" (84). Because most fan scholarship to date has relied heavily "on subjective analysis, personal experience, small sample sizes, and analysis of specific fandoms[,]…there is a paucity of research that provides quantitative…analysis (Barnes 2015, 80) and wider-scale considerations of fandom. For Barnes (2015) and Click et al. (2018), mixed-methods approaches that bring together quantitative methods with "discourse and/or (para)textual analysis" (442) could minimize bias in future research.
[3.5] There are precedents in critical theory that allow us to reconsider the value of distance. Foucault (1980, 81–82) argues that finding and bringing attention to subjugated knowledges is the key task of criticism. Haraway (1988) suggests that we must not only acknowledge our own "limited location and situated knowledge" (583) but also champion what she terms passionate detachment from our objects of study, that is, the seeking of "perspective from those points of view which can never be known in advance" (583–85). For her, this "requires more than acknowledged and self-critical partiality" (585)—it requires our considering the ways in which we provisionally map boundaries onto our objects of study as we come to know them over time and our acknowledgment that these mapping practices can encourage us to maintain a static understanding of the object of study (595). Similarly, Love (2015) argues that "the humanities view of social scientific objectification fails to recognize…the important role of objectification in cultivating self-reflexivity" (87). Distanciation allows us "access to the object as something other than a reflection of our own values" (84).
[3.6] Recent critical work has reconsidered the value of distanced reading. For example, Sedgwick (2002) argues against a hermeneutics of suspicion (cf. Ricoeur 1970), which she terms paranoid reading, in favor of reparative readings, which undertake "a different range of affects, ambitions, and risks" and seek to "extract…sustenance from the objects of culture" (150–51). Warner (2012) highlights the value of uncritical reading—aesthetic rather than political readings. Best and Marcus (2009) term most critical reading practices undertaken today symptomatic reading—readings that look for "what is hidden, deep, and in need of detection and disclosure by an interpreter" (1)—and argue that we should instead consider the surfaces of texts, including the surface "as materiality" (9), "as the intricate verbal structure of any…text" (10), "as the location of patterns that exist within and across texts" (11), and "as literal meaning" (12). Moretti (2000) advocates for distant reading, a move to a wider scale of analysis that allows us to observe the mechanics of cultural systems and to pinpoint what might require close considerations in future. Finally, Love (2010) supports reading that is "close but not deep," suggesting that we should experiment with methods from the sciences and social sciences, "including mapping, systems theory, the statistical analysis of genres, [and] evolutionary modeling" (373). She argues that the humanities would benefit from the widespread use of methods made visible through, for example, digital humanities, statistics, and "data mining" (382), as distance allows us to consider the larger picture. All of these critics emphasize how dominant methodologies championing the deep and close consideration of texts through a particular theoretical lens, while valid, have limited the scope of scholarship. They advocate for a renewed interest in experimentation, transdisciplinary borrowing, and distanciation in order to refresh and update disciplinary discourses while nonetheless drawing attention to the limitations and affordances of varied methodological approaches.
[3.7] These critiques impel us to consider the ways in which dominant ways of seeing in fan studies as a field may have limited our consideration of other perspectives in the past. We do not have adequate knowledge of the groups who exist in fan fiction–centered communities because we have been so focused on looking from a particular perspective and with a given set of scholarly tools, assuming that these tools are adequate (Barnes 2015; Hills 2012). Let me be perfectly clear: my arguments for distanciation, passionate detachment, and objectification are not arguments against immersion, experience, and closeness. The two are not mutually exclusive. Indeed, fan studies as a field can only benefit from including both. Close reading, small-scale case studies, and (auto)ethnography are valuable, but they are not enough on their own. Wider-scale, distanced studies complement these more entrenched approaches, allowing us to be surprised or proven wrong in our assumptions.
[3.8] In fan studies, recent critiques call for scholars to "push past comfortable topics and techniques to places that feel less familiar and less certain" (Click and Scott 2018, 3). This does not mean that closeness must be altogether abandoned; however, we must expand our approaches. As regards representational politics, distanced approaches will also allow us to "expand the range of familiar identity categories explored in the field" (Click and Scott 2018, 4), becoming more aware of our own biases as regards race, gender, sexuality, class, and age (Click et al. 2018, 439). As we move forward, we must also question the essentialism bound up in identity categories and, particularly in our qualitative work, emphasize intersectional analyses that consider the ways in which different aspects of identity intertwine (Crenshaw 1989, 1994; Columbia Law School 2017). Furthermore, when considering subjects such as gender, sexuality, age, location, ethnicity, and (trans)nationality, we must also consider local differences, including differences within and between (counter)publics, in expressing belonging and identity. Categories that are used in one locale may not translate well to others. Identity categories, "as cultural forms and lived experiences, flow in ways that are historically restricted both legally and practically" (Gatson and Reid 2012, ¶ 1.2).
4. Method
[4.1] The present study aims to identify, analyze, and discuss the identity markers used by Harry Potter fans on AO3 and linked sites, and thus to shed light on the makeup of the fandom, in response to repeated calls for wide-scale data from the fan studies community (Barnes 2015; Busse and Lothian 2018; Click et al. 2018; Hills 2012). AO3 was chosen as an ideal source of data because it has become an archive for Harry Potter fan fiction originally posted on various fan fiction sites that have since been closed (e.g., Silver Snitch, Unknowable Room) and profiles that have since been deleted (e.g., from LiveJournal, Fanfiction.net, Skyehawke), as well as hosting stories that were only ever posted on AO3. As such, AO3 represents a cross-section of Harry Potter fan fiction produced and shared by various factions within the fandom, and the data can be considered to represent wider trends in the fandom.
[4.2] The data recorded include fans' self-articulated sexes/genders, sexualities, ages, locations, nationalities, ethnicities, and linguistic competencies. These data were coded by inductively determined emergent themes and were analyzed using a mixed-methods approach. Data were only gathered from publicly available sources, including authors' notes, AO3 profiles, and any linked paratextual profiles such as Twitter, Tumblr, LiveJournal, Dreamwidth, and Fanfiction.net profiles.
[4.3] While broad categories emerging from previous research on fans were used to determine which data to gather, the present study did not precategorize data. Instead, data were coded according to inductively determined emergent themes. Further qualitative information is included when it provides additional insight. Codes that emerged as significant within the data set were as follows:
- [4.4]Gender: female, male, trans, (gender)queer/-fluid/nonbinary, genderless
- Pronouns: she, he, they, she/they, he/they, she/he, she/they/he, pronounless
- Sexuality: queer/gay, lesbian, pan-/bisexual, polyamorous, asexual, demisexual, aromantic
- Location: Great Britain; Mainland Europe; Scandinavia, incl. Iceland and Greenland; Asia; South America and the Caribbean; North America; Oceania; Middle East
- Age/life stage: teenager/school-aged, studying at university/20s, middle-aged/working, retired/"old"
[4.5] Data were coded in this manner to avoid eliding differences within the studied population and to respect how fans chose to describe themselves. Nonetheless, it is important that I acknowledge here that any attempt to codify large groups of fans will to some degree flatten differences within those groups (¶ 3.8, above). Moreover, although the data are presented numerically, it is important to note that they do not lend themselves to statistical analysis. This is both because some profiles are used by multiple individuals and because many fan fiction paratexts remained blank. As such, these data can only be considered to represent wider trends of identity and self-representation in the fandom.
[4.6] At the time this study was undertaken, there were between 178,976 (October 18, 2018) and 189,890 (January 23, 2019) Harry Potter fan fiction works available on AO3. These stories vary in length, ranging from 1,500 to over 3 million words, as well as in genre and focus. A cross-section of all types of Harry Potter–tagged fan fiction works was used. To ensure that the data were randomly sampled, length was the sole organizing factor used. Stories between 1,500 and 5,000 words in length were randomly included in the analysis.
[4.7] The study analyzed the paratexts of 1,939 stories. This represents approximately 1 percent of the Harry Potter stories available on AO3 during the data collection period. As I only have reading competence in Danish, English, French, German, Norwegian, and Swedish, 136 stories written in languages including but not limited to Chinese, Greek, Indonesian, Japanese, Korean, Polish, Russian, and Turkish were excluded from the final data set. A further three texts were excluded as irrelevant, as they were not fan fiction but were instead other types of text, such as film reviews. In total, 1,800 stories and 1,959 AO3 profiles were analyzed.
[4.8] Because the data were collected on a story-by-story rather than a profile-by-profile basis, some user profiles were included more than once in the data set. These have not been collated but have rather been treated as separate entries, as the information provided in authors' notes, including which external profiles were linked, differed between stories.
5. Gender and sexuality results (n = 265)
[5.1] Of the 1,800 fan fiction items in the data set, 265 included data relating to gender and/or sexuality. Some of these items included information for more than one person, for example, if there were multiple authors or if the story was gifted to a named fellow fan whose AO3 profile was linked. Coded data are grouped into three main categories: gender, pronouns, and sexuality.
[5.2] In the theme gender, items were coded as female, male, trans, genderqueer/nonbinary, and genderless. Those coded female include those who referred to themselves using the nouns "female," "girl," "woman," or similar. Those coded male include those who referred to themselves as "guy," "boy," and similar. Those coded trans include those who referred to themselves as "trans," "transsexual," "transgender," "ftm," "mtf," and similar. Those coded nonbinary include those who referred to themselves as "queer" or "LGBT" in general (note 5), as well as those using terms such as "genderfluid," "genderqueer," "nonbinary," "intersex," "2s," and so forth. Finally, those coded genderless used terms like "genderless" or "cassgender." A total of 127 individuals provided information regarding gender. The results are presented in table 1.
| Gender | Number | (%) of total items re: gender |
|---|---|---|
| Female | 64 | 50.39 |
| Nonbinary | 27 | 21.26 |
| Male | 17 | 13.39 |
| Trans | 14 | 11.02 |
| Genderless | 5 | 3.94 |
[5.3] Pronouns were considered separately from gender, as gender identity does not always directly correlate with preferred pronouns. The total number of items coded as pronouns was 127. Results are presented in table 2.
| Preferred pronoun(s) | Number | (%) of total items re: pronouns |
|---|---|---|
| She | 94 | 74.02 |
| He | 17 | 13.39 |
| They | 6 | 4.72 |
| She/They | 4 | 3.15 |
| She/He/They | 2 | 1.57 |
| "Pronounless" | 2 | 1.57 |
| He/They | 1 | 0.79 |
| She/He | 1 | 0.79 |
[5.4] Items included in the theme sexuality were divided into seven categories: queer/gay, including identities expressed as "queer," "gay," "LGBT," or a rainbow flag emoji; lesbian, including "lesbian" or "wlw"; pan-/bisexual; polyamorous; asexual, including "purposefully celibate"; demisexual; and aromantic. The total number of items coded under sexuality is 124. Nobody identified as "straight" or "heterosexual."
| Sexuality | Number | (%) of total items re: sexuality |
|---|---|---|
| Queer/Gay | 43 | 34.68 |
| Pan-/Bisexual | 36 | 29.03 |
| Asexual | 17 | 13.71 |
| Lesbian | 11 | 8.87 |
| Polyamorous | 10 | 8.06 |
| Demisexual | 4 | 3.23 |
| Aromantic | 3 | 2.42 |
6. Location results (n = 496)
[6.1] The vast majority of fans' profiles focused on location rather than nationality. A small number of fans listed prior locations or dual nationalities. The locations listed below are fans' most recently provided locations.
| Location | Number | (%) of total items re: location |
|---|---|---|
| North America | 296 | 59.7 |
| Great Britain | 84 | 16.9 |
| Mainland Europe | 50 | 10.1 |
| Oceania | 31 | 6.3 |
| Scandinavia, incl. Iceland, Greenland | 14 | 2.8 |
| Asia | 11 | 2.2 |
| South America & Caribbean | 9 | 1.8 |
| Middle East | 1 | 0.2 |
[6.2] Although such information was not given by those in other locations, seven North American fans of color specified their ethnicity: two African Americans, one Asian Canadian, two Asian Americans, and two Latinx Americans. A further five fans explicitly identified themselves as white, four of whom were North American.
7. Age/life stage results (n = 277)
[7.1] The category age/life stage included four subcategories: teenager/school-aged, studying at university/20s, middle-aged/working, and retired/old. Some profiles listed ages or decades outright (e.g., 30s), while others made mention of school life or deadlines (e.g., finals), professions (e.g., librarian), or kinship-related age markers (e.g., grandparent). Fans were allocated to the most likely age category based on qualitative information included in their profiles.
| Age/life stage reported | Number | (%) of total items re: age/life stage |
|---|---|---|
| Teenager/school-aged | 55 | 19.8 |
| Studying at university/20s | 157 | 56.7 |
| Middle-aged/working | 59 | 21.3 |
| Retired/old | 6 | 0.2 |
8. Discussion
[8.1] While only a small proportion of fan fiction paratexts provided information regarding the themes analyzed, the data, viewed as a whole, allow us to infer a great deal about Harry Potter fans' identities. They also suggest a number of areas in which future research would be valuable. Nonetheless, the overwhelming trend was secrecy, and this is a limitation to the study.
[8.2] There are myriad possible reasons that fans may elide personal information from paratexts. First, fans are known to be private (Brennan 2014; Hellekson and Busse 2006) and thus may be disinclined to share information about themselves in publicly accessible forums. Second, AO3 may not be a space in which fans choose to share personal information. More, the spaces in which fans share personal information may differ between diasporas, some of which may be less accessible to scholars working out of the global west due to linguistic or cultural barriers, as is true in this instance. Finally, and perhaps most importantly, fans may use silence to protect themselves from hate speech, threats, and violence from both within and without fandom.
9. Gender and sexuality discussion
[9.1] This study suggests that gender identity in the Harry Potter fandom is much more diverse that previously acknowledged. First and foremost, fans who considered themselves to be female were only a majority by a very slim margin: just 50.39 percent of those who reported their genders considered themselves to be female. Nevertheless, 74.02 percent preferred she/her(s) pronouns. We can therefore infer that between half and three-quarters of these fans consider themselves to be female. Genderless, nonbinary, and trans individuals made up the second-largest group in the data set, with 36.22 percent of those who reported their genders falling into these categories. Noncis individuals can therefore be inferred to be a much larger proportion of the fandom than previously acknowledged. However, it must be noted that those who reported themselves as being queer may have intended only to signal sexual nonnormativity. Finally, 13.39 percent of those reporting their gender identified as male.
[9.2] Gender is the most discussed aspect of identity in media fan studies (Click et al. 2018), particularly as regards fan fiction (Bauer 2012; Busse 2013; Hellekson and Busse 2006; Jenkins [1992] 2013). This is in part because, as Busse (2013) has argued, "gender discrimination occurs on the level of the fan, the fan activity, and the fannish investment…Female fan interests are much more readily mocked…as is fan fiction, an activity more commonly ascribed to females" (Busse 2013, 75). As such, the passionate defense of fan fiction and its authors and readers has become central to the feminist project of media fan studies (Click and Scott 2018). The drawback to this is that the stereotypical figure of the female author of fan fiction has also become entrenched.
[9.3] Because gender and sexuality are intimately interwoven categories (Butler 2004, 2011), they are here presented together. A key gender norm that exposes the interweaving of gender and sexuality is heteronormativity—our assumed attraction to the opposite gender within a rigid gender binary (Butler 2004, 2011). While fan studies has commonly considered fan fiction and its surrounding communities queer-positive, it has nonetheless often done so while assuming that most producers and consumers of slash are both female and heterosexual (Brennan 2014; Busse and Lothian 2018; Hellekson and Busse 2006; Jenkins [1992] 2013; MacDonald 2006; Zubernis and Larsen 2012).
[9.4] Dissenting voices within the field have recently drawn attention to the lack of attention paid to lesbians (Russo 2013, 2018), gay men (Brennan 2014), and trans and nonbinary individuals (Busse and Lothian 2018), who have received very little scholarly attention (note 6). Some therefore argue that scholars' emphasis of straight, cisgender female fantasy in studies of fan fiction shuts down the very real sociopolitical importance of slash to many of its readers and writers (see, e.g., Brennan 2014; Driscoll 2006; MacDonald 2006; Woledge 2005). These scholars have emphasized the importance of slash for making visible desires and identities elided from commercially published texts (MacDonald 2006; Duggan 2017a; Willis 2006), while others have argued that slash actualizes ever-present queer subtexts (Duggan 2017b, 2019; Fowler 2019; Tosenberger 2008a, 2008b; Willis 2006), "is inextricably linked to meaning[, desire,] and identity," and acts as "a place where lived identities are forged" (Brennan 2014, 376; see also MacDonald 2006; Russo 2013, 2018; Willis 2006).
[9.5] While the majority of fan scholars argue that the fandom not only gives in to but actively celebrates the queerness that is stifled in the commercially published Harry Potter texts (Duggan 2019; Tosenberger 2008a, 2008b; Willis 2006), there has been limited discussion of the queerness of the fans themselves, with focus primarily given to queer reading and writing practices (Duggan 2017a, 2017b; Tosenberger 2008a, 2008b) and only a few scholars discussing the importance of these queer communities to queer people (Willis 2006). This study reveals that more attention ought to be paid to queer fans.
[9.6] The data presented here reveal a number of interesting points of tension. For example, while only half of those who reported their genders identified as female, nearly three-quarters used she/her(s) pronouns. There are several possible reasons for the difference between reported gender and reported pronoun. The first is that those who are cisgender and report their preferred pronouns may take it as given that these pronouns clearly imply a gender, that is, that she/her(s) clearly indicates ciswomanhood. The second is that nonbinary individuals may continue to use the pronoun associated with their assigned sex for a variety of reasons—one fan included in this study indicated that she used female pronouns "for my mother," for example, while another indicated that it was "for simplicity." It must be noted that such statements were most common in non-North American western contexts, such as Mainland Europe, and may be associated with cultural differences in the discourses surrounding gender and in the use of gendered pronouns between languages (e.g., French versus English), as a gender-neutral pronoun and its use are not necessarily as widely accepted elsewhere as they are in Anglophone contexts, and indeed, the acceptance, use, and frequency of use of pronouns such as "xy" and "they" varies across and between communities, generations, and countries even where English is spoken. More, use of pronouns may also be politically motivated, especially in the many contexts globally—and increasingly in many European countries—where nonheteronormativity is suppressed, policed, or even punishable by law (note 7). It may also be that there are discrepancies between fans' online profiles: a fan may not have updated "her" AO3 profile as recently as "their" Tumblr profile, for example. A fan's gender identity may have shifted over time, or there may have been changes in the norms of reporting gender within the fan community. Finally, the predominance of the pronoun "she" as compared to the self-reported female gender may relate to earlier paratextual norms within the fandom. A great number of the texts on AO3 were copied verbatim from fan communities active in the early 2000s, such as LiveJournal, when the genders of fellow fans appear to have been more commonly assumed to be female. As such, a great number of these fan fiction works' paratexts refer to fellow fans as "she."
[9.7] Although it has been theorized by the fan studies community that there are more men and nonbinary individuals reading and writing fan fiction than has been acknowledged (Busse and Lothian 2018), most nonetheless assume that women are still the clear majority. This study suggests that fans' gender diversity has been underestimated and underacknowledged in scholarly work (note 8). It must be acknowledged that it is possible the high proportion of noncis/trans [36.22 percent] fans may be due to a desire within these minority groups to make themselves visible, particularly in light of the limited number of fans who clearly designated their genders in their fan fictions' paratexts. However, it can also be argued that noncis/trans fans and men are more likely to avoid reporting their genders for fear of ostracization and discrimination (cf. Brennan 2014; Hills 2002; Walton 2018). Nonetheless, the proportion is large enough to suggest that much more attention ought to be paid to these fans in future. Male fans [13.39 percent] also deserve more attention. Indeed, many of those who identified themselves as men did so by defining themselves in opposition to the normative figure of the middle-aged female fan. That they felt the need to make this discursive move demonstrates their clear understanding of themselves as a minority group in this community.
[9.8] Notably, not one person who provided information regarding sexuality described themselves as straight. There are several possible reasons for this. One is that straight womanhood may be the assumed default, as it is in much of the literature cited above and as it is clearly understood to be by the fans discussed above. While references to marriage and parenthood were not considered in this study to be indicative of a given gender or sexuality, fans who mentioned being married or having children may have expected to be read as straight women. Secondly, minority individuals may be more inclined to share personal information in order to make their population visible to the larger fannish community as part of a sociopolitical project. However, the opposite is also true: minority individuals may be less likely to articulate their identities for fear of ostracization or abuse (cf. Brennan 2014; Hills 2002; Walton 2018). Finally, given the queerness of fan fiction communities (e.g., Busse and Lothian 2018; Lothian, Busse, and Reid 2007), it may not be as acceptable to label oneself "straight" as, for example, "bisexual," even if one is predominantly attracted to the opposite gender. It may, in fact, be the norm in these spaces to imagine oneself to be on a queer spectrum rather than heterosexual (cf. centreoftheselights 2013). All these possibilities merit further scholarly attention.
[9.9] It is also important to note that some fans referred to themselves differently on different platforms—for example, as lesbian women in one space and as queer and genderfluid in another. Where this occurred, the most recently updated label was used. Possible reasons for variations include that different platforms encourage different labelling practices; for example, Hellekson (2009) argues that the digital platforms used by media fans allow the "performance of gendered, alternative, queered identity" (116). These discrepancies also emphasize that identity categories shift over time and may be politically or socially influenced. Moreover, they emphasize the role that fannish "cultures of anonymity" (Brennan 2014, 363; Hellekson and Busse 2006) can play in identity experimentation. There is no guarantee that fans are who they present themselves to be. Indeed, fan spaces may constitute a key space in which fans are able to experiment with identities they do not or cannot embody in other spaces, to fulfill a fantasy, to experiment with cross-identification, to try on imagined future identities, or to experience aspects of their imagined selves that are foreclosed or impossible in their real lives. Nonetheless, the ways in which fans describe themselves may gesture toward concerns central to their fan communities.
[9.10] Overall, the findings of this study suggest that communities producing and consuming Harry Potter fan fiction are queerer than previously theorized, with between 20 percent and 40 percent of fans included in the data identifying themselves as noncis/trans and the clear majority identifying themselves as nonstraight, although it must be acknowledged that heteronormativity is often marked by silence, as is likely the case here. Nonetheless, as various scholars have suggested (Lothian, Busse, and Reid 2007; Willis 2006), fan fiction–related fan communities are queer spaces. I here think of queer in the larger sense of the word, as an unstable term defined mainly by its "resistance to regimes of the normal" (Warner 1993, xxvi) or "antinormative project" (Sommerville 2007, 187) rather than simply as an umbrella term encompassing all nonheteronormative/nonbinary individuals (187). The Harry Potter fandom is a clearly antinormative space—or queerly "oriented" and queerly "orienting" (Ahmed 2006)—in which nonheteronormative fan fiction is extremely popular (Duggan 2017a, 2017b, 2019; Tosenberger 2008a, 2008b; Willis 2006) and which has championed the rights of the queer community, for example, through the Protego campaign of the Harry Potter Alliance, which aimed to make trans lives safer (Click and Scott 2018, 3). As such, this particular fandom may not only appeal to queer individuals (i.e., LGBTQ+) but also may also encourage those who participate in it to consider their orientations both inside and outside fan spaces and to be more accepting of difference.
10. Location, nationality, and ethnicity discussion
[10.1] Location was the most openly discussed identity marker among Harry Potter fans on AO3, with almost one-third of the fan fiction paratexts analyzed addressing this theme. The data suggest that the majority of the Harry Potter fans using AO3 are western: 95.8 percent of those who reported a location came from Mainland Europe, North America, Oceania, Scandinavia, or Great Britain. Few fans outside of the so-called global west provided information about themselves. However, my linguistic abilities limited my access to nonwestern profiles, which likely skewed these data.
[10.2] Fans living in certain countries may elide location information because AO3 is a notably queer space not only due to the large number of stories posted there portraying queer relationships but also because it seems to be a space frequented by a number of people who identify as queer (see section 9). We must remember that in many countries, individuals face severe social and legal penalties—up to and including death—for displaying nonheteronormative desires or identities. Even in the west, homophobic hate violence continues to be a problem (see note 7). This may discourage some fans from identifying their locations. Moreover, we must keep in mind the concomitance of location—both spatial and temporal—and identity articulation: our locations shape the identities available to us, politically, imaginatively, and linguistically, through discourse.
[10.3] There are also cultural differences in sharing information. Even between the groups who shared their locations, there were notable differences. For example, Anglophones seemed not only more likely to share their home country but to share their home region or city, with the clear assumption that the other fans using AO3 would know where the Bronx, Northumberland, or the Pacific Northwest are, while those from other western countries were more likely to suggest their location by identifying a wider geographical area, such as Mediterranean Europe or Scandinavia. It is clear, then, that there are differences in how fans conceive their geographical or cultural belonging as it is mapped onto space, in addition to which different fannish diasporas may have different privacy norms relating to, among other things, government oversight or societal conventions regarding information sharing. Furthermore, some fans who do not have English as a first language may not wish to share their locations or nationalities due to fear of linguicism or xenophobia in fandoms, like this one, that center on English-language cultural objects. And, of course, that fans of color receive abuse in fan fiction communities and that fan fiction can incite racial conflict is well documented (e.g., Fazekas 2014; Fowler 2019; Thomas 2019), while whiteness is often considered to be transparent (e.g., Dyer [1997] 2017), and may be a factor in the relative silence of fans regarding race/ethnicity.
[10.4] As regards ethnicity, it is of note that only seven fans of color identified their ethnicities and all of them lived in North America. Of these, six also identified themselves as queer. A further five fans identified themselves as white, four of whom also identified that they lived in North America, and four of whom identified as queer. Explicit reference to ethnicity thus appears to be more expected in the North American queer milieu than in the other communities using AO3. Further research into this phenomenon would be welcome.
[10.5] While a number of those who were living in North America, Oceania, and Great Britain (Anglophone majority countries) indicated that they were immigrants or refugees, this was very rarely the case for those who lived in other areas. This is not only intriguing in terms of the assumption of fans' whiteness (Gatson and Reid 2012; Stanfill 2011, 2018; Wanzo 2015) but also in terms of thinking through the complex relationships between ethnicity, location, and nationality and how (un)expected it may be in various places and communities to explicitly mark one's place of origin, current location, refugee/immigrant status, or ethnicity.
[10.6] Readers are likely aware that fans' whiteness has been assumed in fan studies and popular culture until quite recently. Indeed, Stanfill (2011) calls whiteness "the unmarked category (marking others), the unexamined category (subjecting others to examination), and the norm (making others insufficient), the cumulative effect of which is privilege (and disadvantage for others)" (¶ 2.4). Walton (2018) argues that many of the largest Harry Potter fan websites "marginalize or exclude…people of color or members of the LGBTQ community" (233). Discrimination within the fandom may be one reason ethnicity is rarely referenced. More, it is clear that only some people of color in specific locations feel it important or safe to identify themselves as nonwhite. Again, location-specific sociopolitical concerns and cultural variance are likely behind such differences, but further research is required.
[10.7] The complicated relationships between location, nationality, and ethnicity, as well as the varying norms of articulating these aspects of identity, underscore the importance of intersectional (Crenshaw 1989, 1994; Columbia Law School 2017) approaches to the experience of identity categories. The data make clear that fans' experiences of and with ethnicities, including their discursive relationships to ethnicity, race, and nationality, differ depending on fans' own locations, nationalities, and ethnicities and are likely inflected by their own politics, education, social circles, and intersecting identity categories, among other factors.
11. Age/life stage discussion
[11.1] This study found that the majority of Harry Potter fans on AO3 are young: over half of those who listed an age or life stage (56.7 percent) were at university or in their twenties, while one-fifth (19.8 percent) were school-aged, that is, children or teenagers. This is consistent with the data collected by centreoftheselights (2013), who found that 63 percent of fans using both AO3 and Tumblr were twenty-four or under. Moreover, a great number of those coded in the over-nineteen age categories of this analysis indicated that they had first entered organized fandom as children or teenagers under the age of eighteen, with one fan reporting that she began reading and writing fan fiction online at eight years old and another indicating that she and a friend first started sharing fan fiction at age six. This mirrors trends in acafans' self-reporting (e.g., Jenkins 2012b).
[11.2] Most fan scholars appear to feel that fan fiction is written and consumed by adults (Bacon-Smith 1992; MacDonald 2006), with insistence on fans' adulthood often seeking to undermine discourses that seek to define the fan as immature or "infantile" (Busse 2013). To the contrary, scholars who work in education and children's literature tend to consider readers and writers of fan fiction as adolescents (Black 2008; Bond and Michelson 2008; Tosenberger 2008a; Wikström and Olin-Scheller 2011). Problematizing the matter of age even further, Walton (2018) argues that "what scholarship does exist regarding Potter [fan fiction] readership is generally more focused on teens than children…If anything, children seem to be assumed but understudied presences within these online spaces" (235; see also Hunting 2019). Indeed, age is one of the most openly policed identity categories in fan fiction–centered communities, with many sites barring those under eighteen from entering (although it must be noted that such age policies are easily circumvented). AO3's general age policy is that anyone thirteen and older can open an account; however, this is mitigated by local factors, such as the recently passed GDPR legislation in Europe, which prevents most EEA citizens under sixteen from opening an account (https://archiveofourown.org/tos). Moreover, research focusing on younger participants is ethically fraught and could cause legal and social difficulties for younger fans and the sites on which they are active; getting the required informed consents from both child fans and their parents is nigh impossible, as child fans often participate in fan spaces in secret. Hunting (2019) therefore accurately describes children as "often overlooked" in fan studies research (94).
[11.3] A further factor in sites' age policies is discomfort with child sexuality. Fan fiction is often openly sexual, and many researchers are keen to consider readers and writers of these stories as at least adolescent, if not adult (Duggan 2017b). Western societies tend to be uncomfortable with child sexuality and seek to police it through laws and restricted access to information (Kincaid 1992, 1998; Levine 2002). Commercially published texts for young people rarely present desire "in a manner designed to titillate," and while erotic fiction may be written about adolescents, "such works are usually not considered to be for teenagers" (Kokkola 2017, 93–94). Queer sex, in particular, is traditionally associated with pain and shame in commercially published texts for young people (Crisp 2009; Duggan 2017a; Flanagan 2010; Kokkola 2013; Trites 1998). It is thus likely due not only to laws such as ages of consent but also to adult discomfort regarding child sexuality, a sense of societal duty to safeguard children against sex, and the fear of being associated with pedophilia (Duggan 2017b) that the ages of fans are so policed. As Kokkola (2017, 93) argues,
[11.4] Investments in maintaining the idea that teenagers [and children] are not sexual beings seem to be part of the larger cultural project aimed at preserving the notion of childhood innocence. This notion, with its onus on adults to care for and protect children, has had positive social outcomes…[but] children can also be disempowered by the notion that they are innocent.
[11.5] As a result of the complexities of adults' relationships with child sexuality as well as of operating websites accessible in countries in which ages of consent may differ, many sites require that participants be above the legal age of consent in their local contexts. But what this means can be contradictory and confusing, not least because consent is a fraught subject and different age regulations often apply to the act of sex than to accessing explicitly sexual media. Moreover, legal and social categories of age rarely function as desired—in the United States, for example, only one in five teenagers remains a virgin by eighteen (Kokkola 2017, 93).
[11.6] Furthermore, while this study appears to suggest that most Harry Potter fans are young, these results do not necessarily indicate the true proportions of age groups within the fandom; older fans may simply be less inclined to share personal information openly online, as they may be more enculturated in the fannish norm of anonymity (Brennan 2014; Hellekson and Busse 2006). Further studies regarding possible correlations between fans' ages and fandoms and their willingness to share information online would be beneficial. Moreover, we may need to consider new, less essentializing approaches to discussions of age in fandom, bringing fan studies into direct dialogue with age studies to highlight "the social constructedness of age norms, to fight discrimination on the basis of age, and to foster intergenerational understanding and dialogue" (Joosen 2017, 80). We must consider the "various personal differences [that] affect the way in which people experience a given age" (80), including intersectional considerations, such as how race, class, gender, and sexuality inflect children's perceived innocence (Beauvais 2017; Bernstein 2011; Dyer 2017; Joosen 2017). Fandom is a space in which children defy age norms by making them apparent through "deviations" (Joosen 2017, 82), and is thus a fruitful space for the exploration of child sexuality as well as the ways in which age norms are performed within and across the different spaces of our lives. We ought also to consider "children's voices and texts not in comparison with adult-authored texts per se, but rather in conversation with other texts that share generic or thematic traits" (Beauvais 2017, 270), as well as the intergenerational relationships that reading and writing fan fiction may foster (Tosenberger 2014).
12. Conclusions
[12.1] It is difficult to categorize fans by their profiles, as profiles are fictions of a single moment and changeable. Online fandom is not only ephemeral but various sites' policies as well as local political and legal considerations limit the freedom with which fans can describe themselves. Ways of constructing the self vacillate, fashions and possibilities may vary between platforms and local contexts, and understandings of the self shift. Nonetheless, and although I must emphasize that the majority of fans on AO3 do not provide demographic information in their profiles and authors' notes, the findings presented here indicate trends in Harry Potter fandom and have important implications for future research.
[12.2] The present study reminds us that we must use theory and method conscientiously, borrowing varied and transdisciplinary approaches, in order to ensure balance and limit bias in our depictions of media fandom—and varied approaches have varied affordances and limitations. Acafans' intimacy with specific fan communities has afforded us in-depth understandings of fandom, but this intimacy comes with the price of self-selection and a limited perspective. The experiential weight of attachment we carry blinkers us, and continually close considerations may prevent us from being pleasantly surprised by fandom. As Love (2015) argues, objectification and distanciation cultivate self-reflexivity. This study shows how varied approaches to data collection and analysis can complement each other and, in particular, how wide-scale studies can make apparent biases or stereotypes that go unremarked within the field. Nonetheless, further studies exploring various fandoms' demographics or using different data collection methods would be welcome complements to the present study, as there are notable gaps and silences in the profile data here analyzed.
[12.3] The study nevertheless demonstrates why thinking about who writes fan fiction is important; it undermines stereotypes and challenges us to reconsider our conceptions of fandom, and it highlights that the content and functions of fan fiction are anchored to fans' locations, identity categories, and sociopolitical concerns. The work fan fiction does with canonical texts in community networks using systems of representation is inflected by who is writing and reading the fan fiction. This study suggests that we can learn from this and other fandoms about how to define "community" in transnational, digital spaces; about child and adolescent sexuality; about how children and adults interact; about the balance between expression, identity, and secrecy in public depictions of the self; about writing as a tool for self-fashioning and for expressing nonnormative and intersectional identities; and about the different norms of expression between locations and (counter)publics. Both close and distanced follow-up studies will shed further light on the gaps and silences in our own research, highlighting the communities and aspects of fandom that have so far remained underresearched (cf. Hills 2017).
[12.4] This paper suggests that nonbinary and trans fans make up a much a larger proportion of Harry Potter fan fiction–centered communities than previously realized, although precise proportions are impossible to glean from these data. It emphasizes that this fandom may be queerer than previously acknowledged, as no fans in the data set explicitly identified as straight. While in some instances this may be due to fans' default assumption that they will be read as female and straight, we cannot assume this to be the case. As I argue above, there are multiple reasons that nonheteronormative individuals might also be silent (see also Duggan n.d.).
[12.5] The present study also indicates that fans in Harry Potter fandom vary in age. One key finding is that most fans appear to enter fandom as children or young adults. As children are often overlooked in discussions of fandom (cf. Hunting 2019; Walton 2018), further examinations of age and its relation to fandom, as well as how age affects fannish relationships and activities, would benefit the field.
[12.6] While the location data in this study were skewed by my limited linguistic competencies, there are some intriguing differences in fans' articulation of place. Notably, fans in Anglophone countries were more likely to provide precise location information, including the suburb, city, or region in which they live, while fans in other locales used larger geographic areas, such as "Mediterranean region," to express place-based identity. Moreover, expressions of ethnicity appear to be made mainly by queer North American fans, with slightly more fans of color than white fans providing this information. Further research into why would be welcome.
13. Acknowledgments
I wish to acknowledge the kind support and generous feedback of my supervisory team, Hanna Musiol, Catherine Driscoll, and Anna Krulatz, as well as of the team at Transformative Works and Cultures.