Self-Insert Fanfiction as Digital Technology of the Self

Self-insert fanfiction is a long-established but still controversial mode of writing, even within the already marginalized genre of fanfiction. Moreover, many of the specific terms and practices used to describe this kind of writing have not been formally explored or theorized. We maintain that self-insert fanfiction can be understood as a digital technology of the self, building upon Foucauldian roots and extending into digital platforms and their affordances. We begin by making connections to the precedents established by “Mary Sue” characters, then continue by tracing the shifts from those conversations to more explicitly self-insert subgenres of the present day. Then, drawing on a survey of self-insert fanfiction conducted across four platforms (Ao3, FF.net, Tumblr, and Wattpad), we explore how such works can be discovered, read, and engaged with, and we offer specific observations about self-insert subgenres, as drawn from a selection of these works. Ultimately, we maintain, self-insert fanfiction expands the possibilities offered by other digital technologies of the self (avatars, blogging, etc.) by attempting to create a self that can be open to any reader who encounters it, although this expansion is not without its own limitations and drawbacks. We conclude by offering potential directions for further work in this area that fall beyond the scope of this initial exploration.


Introduction
Although fanfiction tends to be shared among like-minded communities, reading and writing work in this tradition is often a highly personal experience at its heart. Part of the inclination toward the personal in fanfiction is certainly historical and stems from certain long-standing stigmas: despite greater recognition in contemporary public discourse, fanfiction is often still associated with criticisms of "bad" (i.e., amateurish or non-normative) writing, as well as an overt focus on sexual or other "wish fulfillment" content over actual storytelling and the supposedly self-stifling tendencies of its authors, who dedicate financially uncompensated time and craft to existing media properties instead of producing their own "original" work. Engaging with fanfiction on a personal basis-whether entirely independently, alongside trusted peers, or in spaces comprised of like-minded community members-can become one way of minimizing encounters with those continuing stigmas. However, the very decision to read and/or write fanfiction is also a personal choice that entails devoting one's own resources-time, labor, and energy-to the pursuit of personal tastes and enjoyment in non-commercial works: desires that are seen as unruly and unproductive in multiple ways by the standards of kyriarchal late-stage capitalism.
In this project, we turn to yet another, and often even more divisive, facet of the personal in such writing: as actualized in self-insert fanfiction. Here, we trace connections from the often-derided "Mary Sue" concept to more recent forms of fanfiction, such as "imagines", "Y/N", and "xReader" works. While the arguably avatar-like nature of Mary Sue characters and the self-serving focus of Mary Sue stories have long been sticking points for their detractors, we observe how these same features are technically necessary and entirely fundamental to the appeal of these newer forms, which center fan-participants within the form of the text, not simply as part of the writing or even the reading of that text. In certain ways, we argue, self-insert fanfiction can function somewhat akin to what Foucault (1988) terms "technologies of the self": that is, as a means of expression and knowing that hinges upon "the interaction between oneself and others, particularly as mediated by writing of one's self" (p. 19).
We turn to this concept because we find self-insert fanfiction hinging upon a comparable move-using "self" as a device to facilitate particular, affectively-driven kinds of interaction-despite key differences in who that self is and what it is meant to accomplish. For one thing, the "self" in self-insert fanfiction is meant to be more of a palimpsest, open to readers too rather than signifying merely the author, as in Foucault's examples. In addition, the interactions that this device mediates are mainly those between that palimpsestuous, created self and the fan-author's versions of characters or personages from the original object of their focus, rather than primarily between the text's writer and its reader, as is again the case in Foucault's examples. Such key differences, and the types of narrative and writing that stem from them, enable the fan-authors and fan-readers of self-insert fanfiction to indulge in their own fondness, desire(s), and other affective experiences, in what might be seen as a particularly personal way.
In many ways, we envision this work as building upon Kristina Busse's (2016) assertation that Mary Sue characters and fanfiction are often positioned as "effectively oversharing the writer's personal [interests] without mediating and coding them properly" (p. 162). We are particularly interested in this idea of "mediating and coding" one's interests "properly." To us, Busse's framing of fandom communities' common issues with Mary Sues also suggests that there could be means of classifying such highly personal interests and presenting them to others "properly"-i.e., in some manner that seems correct or satisfactory to the community encountering them. We draw upon this possibility to suggest that self-insert fanfiction is one such means of "mediating and coding" (Busse 2016, p. 162) particular kinds of personal interest in a favorite text or character, such as the desire to interact with these characters directly or the desire to imagine participating in sexual encounters with them. Through its formal conventions, its characteristic narratives, and its communities' uses of specific "fantagging" practices (Price and Robinson 2021), we might say that selfinsert fanfiction "codes" personal interests by forewarning audiences about what they will encounter in the work, and "mediates" such interests by offering an inhabitable narrative self that makes such encounters available to fan-readers, rather than limiting them to the fan-author.
In this work, we trace certain shifts-although they are not a teleological "evolution"from Mary Sue characters to contemporary forms of self-insert fanfiction. Of course, we can only offer a partial perspective on a dynamic phenomenon that is still changing at the time of writing this paper. We also use the term "self-insert fanfiction" very broadly in this work, focusing more on the presence of *a* self within fanfiction texts, rather than focusing only on forms that feature either fan-authors or fan-readers more specifically. That is, while the term "reader-insert" certainly describes a particular writing practice, and "self-insert" can be used to distinguish between that practice and others, we envision the present work as providing more of a starting point from which further such theorizations can be built later.
Taking Harry Potter (HP) fanfiction as our primary example, given its relative visibility, extensive corpus, presence in multiple online spaces, and ongoing popularity, we investigate how self-insert subgenres such as "imagines", "Y/N" stories, and "xReader" fanfiction, facilitate interactions between a "self" and fictive others through the kinds of "mediating and coding" we have described above, building upon the problem spot that Busse (2016) observes (p. 162). By considering the formal, intentional presence of *a* self in such texts, we argue that-rather than fully blurring the boundaries between text and reader, or between personal desire and source canon-these subgenres first demarcate and then deliberately traverse such boundaries in ways that go beyond merely allowing the reader to add subjectivity to the text. Instead, these highly personal kinds of fanfiction writing actively encourage and require it.

The Original Self-Insert? Remembering Mary Sue
However, discussing self-insert fanfiction also requires stepping back for a moment to consider "Mary Sues". Historically, this term has been understood to denote a new female character created by the fan-author: a character who is then made the central focus or driving force of a particular fanfiction text and whose interactions with the canon setting and characters tend to exemplify the fan-author's own personal interests, rather than fulfilling the narrative's actual need for an engaging story. Later, the term also came to denote particular kinds of fanfiction, a development that we will revisit shortly. Most importantly for our purposes in this paper, the following should be made clear: while "Mary Sue" is not fully synonymous with "self-insert", it has often been treated as if it were. To put this another way: while Mary Sue characters can be self-inserts, they are not always so, or are not so by default. Still, the two terms do evidence certain additional overlaps: both are particularly subjective approaches to writing fanfiction, and both have drawn criticism for their supposedly overly personal nature, even from others who also read and write fanfiction. Revisiting the concept of Mary Sues, then, offers both a useful starting point, as well as a basis of comparison for later discussions of self-insert fanfiction.
As early fan studies scholar Bacon-Smith (1991) documents, the term "Mary Sue" stems from a short, parodic work by Star Trek fan Paula Smith. Smith's ten-paragraph story responded to certain trends that she had observed in earlier Star Trek fanfiction by recounting the adventures of one Lieutenant Mary Sue, who is desired and admired by the male officers of the Enterprise, saves everyone during a mission gone awry, and eventually meets a dramatic demise to universal mourning (Bacon-Smith 1991, pp. 94-96). However, both Pat Pflieger (1999) and Ashley Barner (2011) also trace comparable traditions of selfinvested writing all the way back to women's popular literature of the late eighteenth century. Barner in particular identifies the "close relationships between writer, character, and reader" (Barner 2011, p. 1) that characterized much of this writing tradition and its successors today; she notes that the largely female audience's ability to become "absorbed" in reading such works discomfited critics, who advocated against these texts and their associated reading practices in ways often mirrored by contemporary criticism. Still, from Bacon-Smith's (1991) foundational work through to the anti-Mary Sue fan initiatives, the Mary Sue "litmus tests", and even some scholarship still visible in the present day, Mary Sue characters are frequently described as being personal fantasies or wish-fulfillment devices on the part of the fan-authors who create them-particularly because these original female characters are often overly powerful, exaggeratedly knowledgeable, or "too" impactful within a story, compared to the canon characters with whom they feature alongside. Barner (2011) sums up this line of thought thus: Mary Sue characters are often seen as "encroaching authorial self-insertions" that enter and "twist" the source canon to suit a particular fan-author's own ends (p. 11). This is a particularly strange criticism, Busse (2016) points out, when this kind of transformative reworking is really just one "variant of what defines most fan fiction, if not fiction: The ability of the writer to translate their own fears and hopes, disgusts, and desires" (p. 16) into fictional and shareable forms. Still, the virulence directed against Mary Sue characters, and the many value judgments that get caught up in this animus, mean that, at least in Busse's (2016) view, the term "has become all but useless as a descriptive or critical tool" (p. 160), given how it can refer to "a character, trope, developmental stage, writing style, and all too easy dismissal of female characters" (p. 160), depending on who is using the term and within what contexts this use occurs.
As Bonnstetter and Ott (2011) see it, this is essentially the objective of Mary Sue characters: "not to extend the meaning of the original author, but to author meaningful extensions of one's self" (p. 361) into a beloved narrative or a story-world. However, what actually constitutes such an extension of self might differ widely according to a particular fan-author's specific interests, experiences, and preferences. Bonnstetter and Ott (2011), for instance, maintain that a Mary Sue character affords the fan-author who creates her a means of being "accepted and acknowledged, celebrated and loved" by fictional characters, to whom the fan-author extends this same sort of effect (p. 353). For Ika Willis (2006), however, this subjective extension of self-presence means acknowledging queer desires as well as the very existence of queerness (p. 155); meanwhile, Chander and Sunder (2007) contend that for fan-authors of color, idealizing or flattering self-inserts can offer a "partial antidote to a media that neglects or marginalizes certain groups" (p. 608). Elsewhere, Dreisinger (2017) also identifies disability as another axis of identity and lived experience that is often sidelined or erased by popular media but that can be revisited through Mary Sue characters, who thus become "powerful disruptive agents that challenge and resist narratives of compulsory able-bodiedness" (p. 2). As these critical readings of Mary Sue characters demonstrate, self-insertion into recognizable story-worlds alongside existing characters and narratives becomes "a powerful form of literally 'inhabiting texts' to respond to what can sometimes be painful subject-text relations" (Cho 2008, p. 106), and this potential offers a significantly different perspective from much of the early criticism of Mary Sue characters.
The other side of this coin, however, is that fan-readers bring their own stakes to these same works of fanfiction, particularly in the case of highly personalized examples such as Mary Sues, where these incoming interests and values may clash with those of the initial fan-author. Internalized misogyny and ableism, latent queerphobia, and overt racism can each play a part as well. Dawn Walls-Thumma (2019), for instance, documents the vitriol and "blanket condemnation" of all women characters as being Mary Sues that was endemic in early Tolkien fanfiction communities, leading to the founding of the Protectors of the Plot Continuum (PPC) initiative in 2002 and instigating the ongoing harassment of fan-authors across multiple fanfiction archives (p. 25). Indira Neill Hoch (2020) notes that today, similar accusations of Mary Sue-ness become particularly pointed regarding fan-authors writing original female characters (OFCs) of color, where any perceived move away from a default whiteness is often seen and treated as a threat (p. 109)-this even occurs in fanfiction for source texts such as videogames, where the character is already an avatar that is female-bodied and fully customizable prior to the fanfiction written about her. Here, Neill Hoch (2020) points out, the accusations of "bad" writing and Mary Sue characterization are in fact driven by a preference for whiteness. Multiple scholars also document how fan-readers have historically relied on their own genre knowledge and perceptions of the source text in order to identify supposed Mary Sue characters, whether "by physical description and by certain elements of plot" (Pflieger 1999, para. 6) or by "the exaggeration of [positive] traits to an unattainable ideal" (Chaney and Liebler 2006, p. 54). It is also worth noting here that fan-readers can perceive such characteristics in a fanfiction text where the fan-author merely writes a female character-either an original or a pre-existing one-in ways that those particular fan-readers did not expect or enjoy. These various disconnects between the fan-author and the fan-reader of Mary Sue fanfiction ultimately lead Beck and Herrling (2009) to compare Mary Sue characters with RPG avatars. In this kind of fanfiction and with these kinds of characters, Beck and Herrling (2009) maintain, the "game" of simulating one's own adventures within a beloved story-world is the main point (para. 2.5); however, most readers simply do not have the same investment in that highly personal and individualized "playthrough" of the source canon that the fan-author had when writing this new character.
Looking forward, though, it must be noted-as both Indira Neill Hoch (2020) and Elizabeth Minkel (2017) have also done-that the strict distinctions between different kinds of Mary Sues, or even between Mary Sues and other types of characters (particularly original characters [OCs] and original female characters [OFCs]) have broken down significantly since Bacon-Smith (1991), Pflieger (1999), and other early scholars first began documenting them. Likewise, as an increasing number of examples from popular culture writ large demonstrate, "Mary Sue" has also become what culture writer and acafan Stitch (2021) describes as a more generally used "pejorative to negatively describe a female character that [the speaker or writer is] currently insulting. Despite attempts to reclaim the word in some fandom spaces, chances are that if someone calls a character a "Mary Sue", they really don't like the character they're talking about" (para. 3). In such conversations, this dislike may stem from the fact that "the character in question is not white" (Stitch 2021, para. 9), or the way that criticism using this term is often "chock-full of internalized misogyny" (Stitch 2021, para. 8)-or, particularly outside fandom spaces, the fact that the character in question may simply exhibit a power, skill, or interiority that certain audiences expect to see in male characters (Coggan 2016). In addition, as we observed while working on this project, contemporary fan-authors' own use of "Mary Sue" in tags and summaries is often connected to works that are more meta-commentary or sardonic and parodic fanfiction: that is, works more akin to a modern update of Paula Smith's short story featuring Lieutenant Mary Sue, rather than actual female characters created by fan-authors for traditional narrative purposes. This being said, we also concur with Busse's (2016) cautionary note about analyses that depend on assuming what fan-authors "intend" by writing what others may consider a Mary Sue character: analyses such as these are necessarily "fraught" with potential, and often highly gendered, misunderstandings and oversimplifications (p. 160).
Again, our intention in this section is not to say that all Mary Sues are self-inserts, whether intentional or otherwise: instead, we have revisited some of the existing literature and ongoing debates on this topic more to demonstrate how common that perception is, and to reveal how much it impacts related conversations. Mary Sue characters and stories have often been taken to reveal something highly personal about the individual interests and disruptive desires of the fan-authors who create them, often without further evidence save fan-readers' own perceptions of the text, and by extension, the fan-author creating it. Likewise, critics of Mary Sue characters and the stories featuring them tend to hold the view that "publishing such a 'private fantasy' is transgressive" (Barner 2011, p. 12), even in the already transgressive, highly personal, and at least partially private spaces of fanfiction. In reality, other factors-such as the unacknowledged expectation of characters' whiteness that Neill Hoch observes-also underlie and complicate that dislike, but still, the supposed and "unseemly" revelation of the fan-author's self via a Mary Sue character provides both a reason for such critiques, as well as a metaphorical coat-hook on which to hang them.

(Digital) Technologies of the Self
When theorizing the kinds of writing that he would come to call "technologies of the self", Foucault (1988) articulates his driving interest as: "How had the subject been compelled to decipher himself in regard to what was forbidden?" (p. 17). That is, Foucault conceptualizes technologies of the self as describing the writing practices that result when subjects attempt to articulate personal ideas and information concerning impulses of the body, sexual desire, and similar taboos against the conventions of religious dogma and societal mores. For Foucault, then, technologies of the self are self-governing measures undertaken by individuals in order to effect certain "operations on their own bodies and souls, thoughts, conduct, and way of being" (p. 17), typically through writing about themselves in confessional terms to trusted others. However, Foucault also stresses that-in its earliest forms, at least-this type of writing about one's own self also served as a means of knowing and caring for that self; scholars who have since picked up on the term tend to highlight this aspect more than that of self-governance.
This tendency to prioritize knowing or recognizing one's own self becomes particularly evident in theorizing regarding digital technologies of the self. Broadly speaking, this term encompasses the scope of the means and practices by which individuals construct and communicate details about their selves using digital technologies. As Abbas and Dervin (2009) note, for instance, technological offerings such as blogging, game avatars, and virtual worlds evidence important similarities to the confessional letters that Foucault focused on: in the same way, these technologies reveal a self that is not the actual subject, but instead, a curated version of that person, intended to interface with other people at one remove from the actual body (pp. [1][2][3]. However, for Abbas and Dervin (2009), one major difference that is made possible by digital technologies, as opposed to journaling or letter-writing, is that they "enable the individual's self/selves to emerge . . . [and] be worked upon" publicly (p. 2). That is, Abbas and Dervin point to the increased scope and visibility that is made possible by digital technologies of the self: such texts or depictions reach much larger, and often less pre-determined, audiences; they are also capable of undergoing change while still being visible, rather than being presented to only a select audience in a fully finalized, static form. Likewise, between personal devices, shared social platforms, and the affordances of each one, "the opportunities for staging and transforming the self/selves have become nearly limitless" (Abbas and Dervin 2009, p. 2); this, in turn, reconfigures certain forms of digital writing as technologies themselves, creating a cycle in which the "artifacts and practices of the self mutually shape each other" (Siles 2012, p. 409).
As mentioned earlier, we maintain that self-insert fanfiction functions as one such digitally-mobilized technology of the self. In the most foundational sense, of course, this claim is tied to the fact that this entire genre of fanfiction focuses on articulating and exploring highly personal desires that are explicitly attributable to the fan-author, the fan-reader, and/or both. However, we are also struck by the formal and structural ways that "self" actually becomes a means of accomplishing this expression.

Some Types of Self-Insert Fanfiction
As Elizabeth Minkel (2017) accurately observes, self-insert fanfiction encompasses a variety of types that often "work very differently in form and function" (para. 20). Minkel also discusses several of the same subgenres that we have focused on here, ranging from "fleshed-out second-person narrators to 'x Reader' stories that eschew identifying details, to 'imagines', short prompts that exist in a murky space between fiction and daydream fodder" (para. 20). However, at the time of writing this paper, we are not aware of any extant scholarship on these specific subgenres: The existing literature focuses specifically on Mary Sue characters and fanfiction or else touches on practices of self-insertion in fiction more generally. Part of our project here, then, also requires providing brief, limited, and almost-certain-to-change definitions of these various subgenres.

Imagines
First, "imagines" are short works of fanfiction, often consisting of mere descriptive passages that place the reader in proximity to a character or person associated with the fan-object, to the effect of "imagine meeting..." or "imagine dating..." that character or personage. Such passages might tell full stories, or they may simply describe scenarios without a traditional plot progression to structure them. As the community-authored resource Fanlore (n.d.) documents, "imagines" are more common on platforms such as Twitter and Tumblr, which support shorter forms of writing and enable fan-authors to pair visual elements with text, "such as a gif of the singer being named or an image associated with the fictional character: usually not a screenshot, but an edit or a manip that calls them or the scenario to mind" (para. 3). If using a gif, the fan-author may either create it or locate an existing one: if using an edited image or a "manip"-which is a fan-created "photo manipulation . . . depicting TV and movie characters in situations different from canon" by repurposing screenshots or promotional material (Fanlore n.d., para. 1)-then the fan-author will often make this, as well as writing the text.

xReader
Next, "xReader" describes a work of fanfiction where the protagonist is the reader, often being sexually or romantically paired with a canon character (this pairing being signified by the "x"). These "xReader" fanfiction texts are often written in the second person (i.e., "you"), and may vary in length, often depending on the platform of origin. While we use the "xReader" name here for clarity and ease of reference, this actually represents an entire constellation of slightly different forms, some of which are unique to specific platforms, given different affordances and practices such as tagging. Other versions of "xReader" include "ReaderX", "Reader", "Canon X Reader", and "CanonXReader" (spelled without the spaces).

Y/N
Finally, "Y/N" is an acronym for "your name"; the fan-reader who sees this phrase is meant to mentally replace it with their own name as they read, thus actively working with the fanfiction text's own features to position themselves within that narrative. In this scenario, "Y/N waits outside" becomes, say, "Effie waits outside" or "Maria waits outside". As with "imagines", these "Y/N" fanfiction texts also tend to be shorter, and they are also found more often on micro-blogging or image-based platforms; they may also focus more on descriptions of scenarios than on providing traditional, plot-driven narratives.

A Brief Codicil
Broadly speaking, many of these subgenres are colloquially termed "reader-insert" fanfiction: that is, they are often understood to be highlighting the fan-reader as that self who is being positioned within their narratives, rather than the fan-author as that "self" (i.e., the common criticism that we have seen being levied against Mary Sues). Despite fan communities' use of such terms, though, in this project we purposefully turn to the even broader-and often interchangeable-term "self-insert fanfiction", and we do this for a number of reasons.
For one thing, there is slippage between the concepts of "reader-insert" and "selfinsert" as described above (some of which we will explore later); for another, the term "selfinsert" can foreground the self more generally, and also enables us to trace the connections between these subgenres and the Mary Sues of older tradition with greater ease. Future work, however, could certainly delve into exploring how the terms "reader-insert" and "self-insert" are used differently, as well as the specific types of transformative fanwork that each one describes and the overlap(s) between them.

Some Conventions of Self-Insert Fanfiction
While the specific types of self-insert fanfiction named above each have their own unique genre conventions-in fact, the most prominent ones are often signaled by a subgenre's name, e.g., "Imagines" and "Y/N"-one particular convention that most types hold in common is an interpellative use of second-person narration. The "you" might be implied, as in the case of "Imagines", or might be directly stated, as in the case of some "xReader" works, but the general practice builds upon existing traditions from both postmodern fiction and ergodic, or "playable", texts.
For its part, postmodern fiction may use second-person narration to extend what Irene Kacandes (1993), early on, calls an "irresistible invitation" to readers: one that attempts to "put the reader in the text" through the use of "you" as a signifier that only functions in the instant that it is uttered to interpellate the reader (pp. 139-40). In this light, the narrative use of the second person "produces an ontological hesitation between the virtual and the actual by constantly repositioning readers" (Herman 2002, p. 378) between the act(s) of reading the text and inhabiting it or, alternately, of experiencing the narrative and feeling that one is actually driving it. The English language's own idiosyncrasies further enrich this array of possibilities, since "you" as a grammatical form "homonymically references male and female, singular and plural addressees, but can also be used as a generalized pronoun replacing 'one'" (Bell and Ensslin 2018, p. 312). Likewise, ergodic texts, such as interactive fiction (IF) and hypertext fiction, encourage readers to participate in "constructing" a story by selecting from pre-set choices that will send them down various narrative paths. Using "you" to present and frame these choices helps "create the illusion of being present in a story-world . . . constructed by the reader in creative collaboration with the programmed text" (Bell and Ensslin 2018, p. 312).
However, despite the array of differences between these two approaches, it is also worth noting that both postmodern texts and most of their ergodic counterparts proceed along fully formed, predetermined routes once their readers have chosen to participate in them. Thus, the self being signified by the "you" of these genres is not fully synonymous with the highly personal and personalizable "you" of self-insert fanfiction.
Still, this is not to claim that self-insert fanfiction is some utopian ideal or fully democratic reading experience. Many Fanlore pages, archive forums, and other fandom spaces feature long discussions of how and why this particular form of fanfiction is particularly strange, upsetting, or disliked, and in particular spaces, this disquiet has become codified in specific rules. In its content guidelines, for instance, the fanfiction archive Fanfiction.net (FF.net) disallows "any form of interactive entry: choose your own adventure, second person/you-based, Q&As, etc." (Fanfiction.net (FF.net) [2005] 2008, para. 4). While selfinsert fanfiction is not specifically designated here, terms such as "interactive" and "second person/you-based" can certainly describe that type of writing, even if that was not the initial intent of the rule. As a result, many users take this guideline as effectively a ban on publishing self-insert fanfiction to FF.net.
However, the slipperiness of defining what constitutes self-insert fanfiction continues, even here in the form of FF.net's content guidelines. For instance, archived conversations from an FF.net forum discussing the rule mentioned above point out many such ambiguities, and some even present fans' own versions of Busse's (2016) cautionary note about assuming writerly intent. We were particularly struck by one user asking whether the rule is banning "interactive fanfiction that put the reader as themselves into the story (which is weird and I can see why it would be banned on a family-friendly site), not the less-used second-person narrative style using 'you' for the main character in the narrative" (Rtarara 2013). While this user is comparing formal conventions (i.e., the interactivity and second-person language that is actually mentioned in the FF.net guidelines), their question ultimately frames each example in terms of whom it places "in" the story and how. For interactive fanfiction, what this user finds "weird" and worth banning about this type of fanfiction is the way that it lets a non-fictional self-here, the reader-intrude upon the fictional narrative. However, in terms of second-person language, what makes "you" potentially worth using-and perhaps risking or contesting FF.net's unevenly enforced guidelines-is the fact that. alternately, this "you" can signify a fictional self: "the main character in the narrative" presumably from the original media text, rather than the situation of the fan-reader intruding where their presence is not welcome on a "family-friendly site" (Rtarara 2013), a platform that also banned NC-17-rated adult content in 2002.
With these examples and observations in mind, we assert that subgenres of selfinsert fanfiction are not always "performing a sense of self" in the ways that Bonnstetter and Ott (2011) consider to be happening with Mary Sue characters. However, we find that Bonnstetter and Ott's framing already echoes key ideas present in Foucault's (1988) "technologies of the self", particularly regarding how the author-self that is present in such a text is a conscious construction working to acknowledge, care for, and sometimes even govern the author-self creating it. Bonnstetter and Ott (2011), for instance, maintain that Mary Sue characters tend to mirror the fan-authors who create them and enable those fan-authors to seek validation, comfort, and consolation by inserting themselves among fictional characters and into fictional settings. (Issues with this framing, of course, include the way that Bonnstetter and Ott's argument assumes that Mary Sue characters in general are deliberate self-inserts, which we have already seen is not necessarily the case.) With the more contemporaneous self-insert fanfiction that we examine, however, Bonnstetter and Ott's almost-Foucauldian notion of "performing a sense of self" could also describe how these texts function, albeit with addendums.
With self-insert fanfiction, we find that "performing" a sense of self differs on three axes: first, intent (i.e., self-insert fanfiction is written, specifically and unambiguously, to let the fan-author and/or fan-reader "put . . . themselves into the story" (Rtarara 2013)); then, what we follow Busse's (2016) lead in calling "mediating and coding . . . properly" (p. 162). "Coding" or categorizing "properly", we observe, entails using platform affordances and community "affordances-in-practice" (Costa 2018) to inform fan-readers that the text is a specific kind of self-insert fanfiction so that those who enjoy the subgenre can choose to engage, while those who do not have enough information to avoid it. This is often the purpose of using tags, summaries, formatting, and other paratextual appendages, and while it is not dissimilar from tagging in other fandom contexts, it becomes particularly important when the work being "coded" in this way caters to niche and widely criticized tastes.
Then, "mediating" or facilitating "properly" entails using formal conventions-such as those we have discussed in our brief summaries above of the "imagines", "Y/N", and "xReader" subgenres-to place this self within the narrative in specific, circumscribed ways, unlike the ways in which earlier Mary Sue characters and stories were often "identified" more by fan-readers' personal preferences, perceptions, and even biases.
Thus, with these important differences regarding intent and "mediating and coding . . . properly" (Busse 2016, p. 162) in mind, we now turn to a sample of self-insert fanfiction.

Insert Self, Part II: Imagine Going to Hogwarts
Given that this project is-we hope!-one early venture of many into the topic of selfinsert fanfiction, we have focused here on fairly broad strokes that we believe future forays could adjust or rework as needed. Thus, we formulate our argument for the intentionality, coding, and mediating of self-insert fanfiction following an investigation of such texts, drawn from four highly popular sites: The Archive of our Own (Ao3), Fanfiction.net (FF.net), Wattpad, and Tumblr. The first three are fanfiction-focused reading platforms, while Tumblr is a social media site; however, we turn to these specific spaces regardless of that difference because, as fanfiction readers ourselves, we know that fan-authors and fanreaders engage with self-insert fanfiction substantially across all four platforms. Likewise, we envision this project more as setting out the starting points for continued dialog than as a truly deep delve into any one platform or subgenre-as has been done concerning Mary Sues on Wattpad (Binike 2018), for example-and we strive to avoid forcing any teleology onto the present co-existence of Mary Sues and self-insert fanfiction, despite their many areas of overlap and similarity.

A Brief Note on Our Methodology
Our approach to selecting and analyzing specific works of HP self-insert fanfiction has been heavily influenced by Milena Popova (2018) in their work, "Rewriting the Romance", a study of arranged marriage stories in slash fanfiction. After noting the prevalence of auto-ethnographic approaches in fanfiction scholarship, since most researchers in this area began as fans, Popova argues for certain merits to drawing from one's fannish experience: for one thing, searching for fanfiction using the tools and approaches that fans employ can help researchers to locate texts that might be particularly visible or popular (p. 6). In their case study, Popova reports using Ao3 s own features to "search, sort and filter" fanfiction works of potential interest, before then "immers[ing] myself in [these texts]" using "a range of auto-ethnographic insights . . . my understanding of the technical features of the site . . . of the community's usage practices, and of dynamics and trends within the particular fandom, pairing and trope of interest" (p. 6). By using this combination of archival/technological savvy, fandom familiarity, and the application of fannish reading practices, Popova asserts, the stories that they ultimately selected for analysis in greater depth could be trusted as fairly accurate representations of popularity, impact, trends, and themes touching upon a specific trope and character pairing (pp. 6-7).
Popova also avoids providing complete URLs for the two specific fanfiction texts they analyze, correctly citing a need to protect individual fan-authors' privacy (p. 6). We concur with this need, as well as the observation elsewhere that fanworks, while technically accessible to anyone with an internet connection, do exist in "layered publics" (Busse and Hellekson 2012, pp. 38-39): that is, fans and fan communities tend to consider "a shared fan space to be private, even when it is [technically] accessible and thus public" Hellekson 2012, quoted in Busse 2018, p. 12). Thus, given the particularly personal and often polarizing nature of self-insert fanfiction, we have taken Popova's precautionary measure a step further and omitted fanfiction authors' pseudonyms from the following analysis, as well as the works' URLs.
Appendix A of this work outlines our approach to locating Harry Potter self-insert fanfiction in greater detail, but as the following subsections also demonstrate, we both drew from and somewhat modified Popova's approach here. Following their example, we also began with exploration and immersion, as informed by our own personal experience with certain fanfiction platforms and fanfiction-searching practices; later, we also then turned to the anonymization of the specific works that we will consider in more detail. However, similar to the community distinctions between particular flavors of "reader-insert" and "self-insert" texts, further research in this area will also need to grapple further with questions such as how to cite specific works of self-insert fanfiction, given their particularly personal nature. For now, though, given our focus on the aggregate or larger picture of this phenomenon, we are confident that the measures we employ here can provide at least a starting point.

Why Harry Potter, Given Ongoing Issues
As previously mentioned, we look to Harry Potter (HP) self-insert fanfiction for this project. This decision was made for four main reasons. First, there is an extensive corpus of HP fanfiction, available across multiple different platforms: as of writing this article, this corpus included 361,000 fanfiction texts on Ao3; 840,000 on FF.net; and tallies estimable in the millions for both Tumblr and Wattpad. These numbers offer rich possibilities for sampling, even among supposedly more niche genres such as self-insert fanfiction.
Second, there is the breadth of scholarship already available on both Harry Potter more generally and on HP fanfiction specifically, which provides an established foundation from which to build newer observations about specific subgenres. For example, in terms of Mary Sues and self-inserts alone, we might begin from work on queerness and narrative gaps (Willis 2006), readers' engagement with particular characters (Alderton 2014), or even just observations of how commonplace Mary Sue characters are in this particular body of fanfiction (Almagor 2006;Mackey and McClay 2008;Lehtonnen 2015).
Third, the historical arc of the HP fandom tends to coincide with certain well-mapped phenomena within the English-speaking fandoms of popular culture texts from the Global North, such as the Mary Sue characters discussed earlier. That is, the production of HP fanfiction spans both well-established and burgeoning trends in fanfiction, making comparisons between such practices possible within a single fandom.
Fourth and finally, HP fanfiction is usually easily accessible, both for us as researchers and also for anyone who might wish to build upon our work here. In the first place, both of us have read HP fanfiction fairly extensively, and Effie in particular has been researching it for years. Then in the second place, HP fanfiction's "canon", or definitive body of texts, is relatively more straightforward than, say, Star Trek or Star Wars, despite a growing number of transmedia add-ons to the 1997-2007 books by J.K. Rowling and the 2001-2011 HP films: this lowers certain barriers to entry for researchers.
Looking forward, however, it will also be interesting to see how the historical position of HP fanfiction continues to change, particularly as the books' own issues (entrenched racism, anti-Semitism, etc.), various transmedia properties' reiterations of such issues, and the author's consistent transphobia on social media alienate more and more readers with each year.

Locating and Reading HP Self-Insert Fanfiction
As previously mentioned, our approach to locating and selecting HP self-insert fanfiction for this analysis was adapted from the approach outlined by Popova (2018), who reports that they drew from fannish experience and "followed the path any fan new to a fandom, trope or pairing may follow to find stories that are considered good or impactful by the community at large" (p. 6). On Ao3, the platform Popova's study focuses on, this approach entailed using the site's features "to search, sort and filter stories of interest" (p. 6). Popova describes using tags and sorting the results by the number of "kudos", which is the Ao3 platform's one-click, one-use feature intended "to let a creator know that you like their work" (Archive of Our Own (Ao3) n.d., "What are kudos?").
Informed by our own experience in locating fanfiction of different genres on different platforms, plus our own observations of self-insert fanfiction encountered during personal browsing, we decided to begin with specific keywords: "Y/N", "xReader", "Reader-Insert", "self-insert", "imagines", and "Mary Sue". (In some cases, we also queried slight variations, such as "reader" [i.e., instead of "reader-insert"] or "imagine" [i.e., without the pluralizing "s"] when the initial search did not yield the volume of results that we were expecting.) In many cases, we treated these terms like site tags-i.e., as navigational, classification, and indexical tools-and adapted our search according to the slightly different ways that each of our four target sites uses them. On Tumblr, for instance, tags are often "put to expressive rather than organizational purposes" (Stein 2018, p. 87) and are used "to create poetry, analysis, conversation, and even fiction" (p. 89). Meanwhile, on Ao3, tags become part of a "curated folksonomy" (Bullard 2014, as quoted in Price 2019, p. 2) that consists of "a combined self-tagging and automanual system" (Price 2019, p. 2) where users can choose from predefined site tags and/or create their own, which are later checked and possibly ratified by volunteer "tag wranglers." By contrast, other platforms often have simpler, more rigid, and less folksonomic tag systems. FF.net relies on more traditional querying methods and a far more limited, less expressive, and less fan-curated tagging system, while Wattpad offers sorting options/filters based on length, recent updates, WIPs versus completed stories, and ratings, among others.
We mention these differences in order to highlight how even an apparently straightforward functionality, such as tags and tagging, will often work differently on various platforms and, in the case of Tumblr, may also be leveraged by "fan-users" (Alberto 2020, para. 2.6) in ways that differ from the applications of those that developers consider a general userbase. Thus, for a project such as ours, locating self-insert fanfiction may sometimes be as simple as searching specific tags-if one knows the tags du jour to search for in the first place-but more often, also entails knowledge of what Elisabetta Costa (2018) calls "affordances-in-practices": that is, "the enactment of platform properties by specific users within [particular] social and cultural contexts" (p. 3651). Because they can be used in such vastly different ways according to users' different needs, Costa (2018) argues that a platform's features should be understood as relational and contextual, as well as technological; we found micro versions of this phenomenon at play in our searches across all four sites. Thus, we would add to Popova's (2018) methodology that a familiarity with fannish affordances-in-practice, as well as with a site's purely technical features, can be another important step in locating relevant fanfiction.
Adjusting and running these searches on each of our four selected platforms netted a wildly varying range of results (see Table 1). We provide a broad overview of our search process, organized by platform, in Appendix A; there, we also provide a listing of the self-insert HP fanfiction stories that received a close reading. Before we move on to our discussion(s), however, we must reiterate here that this data collection and analysis are by no means comprehensive. With such a wide scope regarding both fanfiction types and platforms, it would be difficult to make definitive, water-tight statements rather than broad observations-nor do we believe that we should aim for the latter anyway, as fanfiction and fan communities are constantly shifting and transforming. Instead, with these observations, we aim more to provide a more general, working overview, different aspects of which can then be taken up and expanded upon in future research. Tumblr n/a n/a n/a n/a n/a n/a Wattpad 7600+ 496 1400+ 126,000+ 12,500+ 12,300+

Insert Self, Part III: An Overview
Following the exploration and data collection stages outlined above, we now move into a discussion of certain traits that we found characterizing these particular subgenres of self-insert fanfiction, in a sample of HP fanfiction texts located across four large and largely fandom-populated platforms. Given that we have focused primarily on the fanfiction written for a single fandom-even if it is a very large and highly active one-we doubt that all the findings noted here will map exactly onto self-insert fanfiction produced within other fandoms but, again, we hope that this can provide at least a starting point for ongoing conversation.
We might summarize our thoughts here as follows: we found fan-authors "mediating and coding" (Busse 2016, p. 162) their writing and publishing of self-insert fanfiction in ways that exhibited cross-platform similarities of subgenre, as well as identifying some key, and often platform-specific, differences within these same shared subgenres.

First Primary Search: Imagines
As previously discussed, "imagines" are short fanfiction texts-often comprising more descriptive passages than a traditional narrative-that situate the reader in proximity to a character or real person associated with the fan-object, to the effect of "imagine [your] meeting..." or "imagine [yourself] dating..." that character or person. And, much as in the other subgenres of self-insert fanfiction that we outline here, the specifics of "imagines" seem to vary depending on the platform. On Ao3, the keywords "Imagine" and "Imagines" yielded totally different results. We observed that the keyword "Imagines" (i.e., with the -s) was most often appended to collections of "one-shot" stories, many of which were originally posted on the authors' Tumblr accounts and were then "cross-posted" to Ao3 afterward. (Conversely, our searches on Ao3 using the keyword "Imagine" (i.e., without the -s) yielded stories that were more in line with the "Reader" keyword results on this platform; thus, we focused more on our "Imagines" results.) Some of these works were written in response to prompts received from other Tumblr users; others stemmed from prompt lists (see Section 6.1 for more on this topic). Many of these stories use first-or second-person narration, but as we observed with other self-insert subgenres, the majority featured a female character in a heterosexual relationship. The "Imagines" from Ao3 that we read in more depth for this project all fell into this pattern: collections of individual, sexually explicit texts, all focused on heterosexual relationships between a canon male character and a female self-insert character. Some of these collections are themed around a specific canon male character, while others explore scenarios with a variety of characters. Several of these collections also feature shared "kinks" (beyond fandom, these are non-normative sexual practices or desires often associated with BDSM; within fandom, the same applies and/or the term also denotes an "elaborate fannish framework . . . within which fans have engaged [in] sexual experiences, fantasies, and even politics" (Busse and Lothian 2017, p. 124)). Regarding kinks tagged and chosen to describe these collections, we noticed a predominance of "daddy kink", "innocence kink", and dom/sub (i.e., dominant/submissive) relationships.
The results on FF.net were inconclusive, given that this platform's search queries function more in the form of a traditional BOOLEAN search; that is, our query selected all results with the word "imagine" in either the title or story summary, which brings up works that use the term creatively or narratively, rather than only those that use it categorically. However, when we searched "imagine collection", this yielded far more relevant results, although we also realized that this would exclude stand-alone "imagines" or even collections that might not be described as such by the author. Most of the results obtained in this way specified that they are collections of "drabbles" (i.e., a short piece of fanfiction; today usually of fewer than 1000 words, but originally, of exactly 100 words) or "one-shots" (i.e., a fanfiction story that has only one chapter in its completed form, rather than having multiple chapters); we also noted that these results did not use the term "imagines" very often. We hypothesize that this is because "imagines" are a fairly recent fandom development, while FF.net is the oldest of the four platforms examined here, with the fewest updated search functions and the lowest level of multimedia support. In fact, of the few "collection" results that mentioned they were of "imagines", the author also stated that these stories had stemmed from Tumblr prompts and requests.
Tumblr and Wattpad both yielded significantly more results using the search keyword "imagines." On Tumblr, we found, "imagines" are predominantly written in the second person, using "Y/N" in place of a name for the self-insert character even when these texts are not specifically described as "Y/N" fanfiction; they are also largely focused on female characters in heterosexual relationships. Of the top ten results emerging from this search on Tumblr in March 2022, six were written in response to user-submitted prompts or "asks"; thus, the fan-author is writing at more of a distance from their story, as opposed to a story emerging from an idea that the fan-author themselves had. About 90% of the top results were also sexually explicit stories, interspersed with collections of pornographic gifs sans text. Additionally, pornographic gifs or static images are often attached to the "imagine" fanfiction texts themselves. The gifs typically did not include faces but were predominantly of white heterosexual couples engaged in sexual acts. While the static images were less explicit, they also seemed chosen specifically to enrich the fan-reader's experience, visually. In "subspace", for instance, the author has included a static image of a woman kneeling-the image only depicts the woman's body from the upper thigh to the knee-while in "Just Ours", the author includes a series of images depicting a female body in pink clothing. All of the top five stories paired the main character with one or more of the Marauders. Additionally, in terms of kink/fantasy, the majority of these stories featured similar interests (e.g., "daddy kink", "sub kink", "innocence kink", etc.); we even noticed a specific trend placing the male canon character in the therapist profession and the main female character as his patient, as seen in the story "Feels too right." On Wattpad, "imagines" seem most commonly to be written in the first or second person; they are rarely in the third person. They are also much less explicit than those found on Tumblr and Ao3; we hypothesize that this is likely due to the younger age range of Wattpad users. For example, the story "imagines Humanities 2022, 11, x FOR PEER REVIEW 14 of 32 On Wattpad, "imagines" seem most commonly to be written in the first or second person; they are rarely in the third person. They are also much less explicit than those found on Tumblr and Ao3; we hypothesize that this is likely due to the younger age range of Wattpad users. For example, the story " ϟ " includes four very short stories, each with a female main character; three out of four stories pair this character with a canon male character, but the fourth one pairs the reader-character with Hermione Granger in a same-sex relationship. Similarly, in the collection titled "IMAGINES|HARRY POTTER CAST X READER", there are several stories pairing the self-insert reader-character with a female canon character. Throughout this whole search process, same-sex relationships between the self-insert character and the canon character did not emerge often in the top results on any other platform, whereas Wattpad was the exception to this rule. Also interestingly, some of the results on Wattpad are actually merely lists of prompts with matching gifs pulled from the Harry Potter films; others are bullet-pointed "headcanons". This is similar to some trends we also noticed on Tumblr; that is, both platforms offer multimedia integration in ways that Ao3 and FF.net do not.

Second Primary Search: xReader
As noted earlier, xReader fanfiction (also known as ReaderX or even just Reader) features a self-insert protagonist who is sexually and/or romantically paired with a character from the source text. Our preliminary observations across these four platforms suggest that in some fanfiction communities, these terms are quite similar, while in other communities, they are vastly different. To test this, and in keeping with our earlier observation that "reader insert" often signifies a particular subset too, we looked for fanfiction texts classified as "xReader" as well as for those categorized as "Reader-Insert", hatty potter" includes four very short stories, each with a female main character; three out of four stories pair this character with a canon male character, but the fourth one pairs the reader-character with Hermione Granger in a same-sex relationship. Similarly, in the collection titled "IMAGINES|HARRY POTTER CAST X READER", there are several stories pairing the self-insert reader-character with a female canon character. Throughout this whole search process, same-sex relationships between the self-insert character and the canon character did not emerge often in the top results on any other platform, whereas Wattpad was the exception to this rule. Also interestingly, some of the results on Wattpad are actually merely lists of prompts with matching gifs pulled from the Harry Potter films; others are bullet-pointed "headcanons". This is similar to some trends we also noticed on Tumblr; that is, both platforms offer multimedia integration in ways that Ao3 and FF.net do not.

Second Primary Search: xReader
As noted earlier, xReader fanfiction (also known as ReaderX or even just Reader) features a self-insert protagonist who is sexually and/or romantically paired with a character from the source text. Our preliminary observations across these four platforms suggest that in some fanfiction communities, these terms are quite similar, while in other communities, they are vastly different. To test this, and in keeping with our earlier observation that "reader insert" often signifies a particular subset too, we looked for fanfiction texts classified as "xReader" as well as for those categorized as "Reader-Insert", in order to explore potential overlaps or differences.
On Ao3, the search keyword "xReader" yielded very few results; thus, we turned to the keyword "Reader" because it provided a much wider array of results. "Reader" fanfiction on Ao3 seems to function in a similar way to the "Y/N" fanfiction tag on FF.net and Wattpad; that is, these texts offer more descriptive detail of the "reader" character and, sometimes, there is even a complicated background contextualizing them. For example, in "The Potions Master's Apprentice", the "reader" character is described as a Muggleborn, Gryffindor, female character; however, within the text, the author uses the self-insert language of "(your name)", "(hair length)", and "(hair color)", to encourage the reader to place their own descriptors in these parts of the story. Meanwhile, in "Broken Silence", the author's summary provides context, including the fact that the character has been "expelled from an American school [and] relocated to Hogwarts." It is also worth noting that these two texts feature a relationship between the self-insert character and Severus Snape, as do the stories "Black Sheets" and "The Potions Master." Of the top five search results for "Reader" on Ao3, the only one that does not feature a relationship with Severus Snape is "I Hate You", which features a relationship with Draco Malfoy instead. This story also differs from the other four in that it appears to be written in response to a request that the fan-author received, rather than being of the fan-author's sole creation. All five of these stories are written in the second person, which is mostly consistent across all the "Reader" stories that we found on Ao3.
Stories tagged as "Reader-Insert" on Ao3 overlap quite widely with "Reader" stories on the same platform. Of the top five "Reader-Insert" stories, three are also found in the top five "Reader" stories (e.g., "I Hate You", "Black Sheets", and "The Potions Master's Apprentice"). However, the very top search result is a story entitled "the salt and the sea", which features an OFC (i.e., an original female character) but does not use any of the self-insert language that we have seen elsewhere; moreover, the author does not provide any insight as to why they have tagged this story as "Reader-Insert." The other new result appearing in the top five here is "Dear Professor-Snape x student!reader." Similar to the "Reader" stories discussed above, the author provides contextual information in the story summary and indicates that this was written in memory of Alan Rickman, who played the character Severus Snape in the HP films. The chapters alternate between the third-person point of view (POV) of Severus Snape and the second-person POV of the reader-character, who is not given a name: instead, the placeholder (. . . ) is used to indicate that the reader should insert their own name for this character.
On FF.net, the "xReader" stories are quite similar to the other insert genres that we encountered on this platform; that is, they allow for some ambiguity so that the reader can place themselves in the story, with relatively little need to adjust for their own lived experience or bodily reality. Our results included a story where the reader is in a relationship with Fred and George Weasley, a student-professor relationship between the reader and Newt Scamander, and quite a few stories pairing the reader with Draco Malfoy. It is also worth noting that-despite the FF.net guideline banning second-person fanfiction, as discussed earlier-the more we explored this keyword, the more second-person fanfiction stories we found ("Newt Scamander X Reader", "Ravenclaw (Draco × Daughter-Reader × Father-Snape)", and more). This keyword also elicited a fair number of first-person stories. In "The both of us", the author seems to be attempting a gender-neutral character but slips into using she/her pronouns near the end of the story. In "I've Been a Bad Girl, Professor Scamander", the gender of the character is indicated immediately, in the title. Another interesting text that emerged in these results is titled "From the Library of June Williams", although it actually features an OFC in a Marauders-era story. This story opens with a long author's note, in place of the first chapter, where they explain that this story was written when they were fifteen and that it reflected their desire to live out the life that they have given their OFC. Nearly a decade later, mid-pandemic, "stuck inside a childhood home with childhood books", they have decided to revisit and rewrite the original story in this form; although nothing immediately explains why this particular text came up when searching "xReader" on FF.net, it is possible that something unseen in the metadata is attributing it to this genre. In terms of "Reader-Insert", only one result appeared on FF.net, and this is not exactly the type of self-insert that we have seen across other genres. In "Power of the Elements", the first-person narrator is an eleven-year-old male child who is familiar with the HP story-world; after being admitted to Hogwarts (which he had previously thought was fictional), he is declared Morgan Le Fay's heir. In comparison to our other results here, this story functions more in the manner of a self-insert we might find on Ao3.
A significant number of the Tumblr "xReader" fanfiction results that we found originated from requests sent to the fan-authors who eventually wrote them; these requests are often quite specific, which does not allow for as much flexibility on the fan-author's part when writing the corresponding text. For example, we saw one Tumblr author being sent an anonymous ask that requested: "Maybe James or Sirius is the reader's older (by year) brother and they are really protective because of something that happened in the past. They could walk in on Remus and their sister kissing or something and they go all big brother on him. Maybe they both do because the other feels like a big brother to her as well." The subsequent story, titled "Remus gets caught Kissing James's Sister [R.L]" follows the precise plot described by this anonymous user. In other instances, though, there are fewer details included in the ask/request, which means the fan-author can include more of their own creative ideas or preferences. For instance, one fan-author received a request stating: "I was wondering if you could please write a fic with Draco and a gn!Crush? Maybe Draco wants to ask the reader to the yule ball and gets jealous when someone else tries to do the same(?)" The fan-author writes them a story in reply, titled "Are you Jealous?" but mentions in an author's note that they hope the story is "accurate to what [the anonymous user] wanted." Given that authors do not know any personal information about the fan-readers requesting such stories, we note that these fan-authors often tried to maintain a certain neutrality in the self-insert character and that they often did so quite effectively.
The Reader-Insert stories that we found on Tumblr were split evenly between those that emerged solely from the fan-author's imagination and those that resulted from fanauthors answering anonymous "asks", as described above. The story "tricks and charms", for example, was written for a writing event rather than in response to a request/prompt. In the prefacing material, the fan-author notes that the fan-reader should either pretend that the timeline of the story is contemporary (despite it featuring the Marauders) or else "pretend youtube [sic] existed back then". Similarly, in "My Dream Girl", the request specifies that the story should take place at a particular canon location, but the fan-author states: "I didn't want to write about [location . . . ] so just pretend . . . " and situates the story in another place. Utilizing this keyword in our search here also brought up one of the only explicitly same-sex self-insert stories that we encountered. "Love Languages" is a story where the character is a male Ravenclaw student starting a relationship with James Potter. This story was written as a reply to the request: "James Potter falling for his sweet gay Ravenclaw tutor?" In almost all the other stories that we read, the self-insert character is assumed to be female or else is explicitly described as gender-neutral.
On Wattpad, there was a significant overlap between the search results for "xReader" and those for "Y/N". In fact, of the first five results listed for "xReader", two were also in the top five results of the "Y/N" search ("The Boy Who Loved" and "Her"). The top result for this keyword, however, is a text titled "Potter?||Draco Malfoy x Reader", which has 23.5 million read counts and 808,000 votes. It comprises 157 chapters, including an epilogue, and tells the story of Harry Potter's sister, who has attended other schools before now transferring to Hogwarts. Elsewhere, with "Her", another text that comes up as a result of both "xReader" and "Y/N", the author's note states the reader will "go through the stages of like, love, lust, and loss with Harry and y/n as they venture through their years at Hogwarts." This statement seems to imply that "Y/N" is a character and not a holding space into which the fan-reader or fan-author inserts themselves; however, the story is told in the first person. Other search results that emerged here included story collections rather than single texts. These could be written in either the first person or the second person, but the top collections all feature "collections" of ficlets (i.e., very short fanfiction texts) with a heterosexual romance between the female self-insert character and an HP canon male character.
When searching "Reader-Insert" on Wattpad, the bulk of the top search results were collections of stories, similar to the aforementioned collections found with the "xReader" search term. The story titled "Different|dracomalfoy|" comes up as the second result and is one of the only ones that is not a story collection. Instead, this is a story that begins in the third year of school, and the character, y/n Potter, has transferred to Hogwarts from the American magic school, Ilvermorny. It is a story wherein the self-insert female character enters into a relationship with Draco Malfoy after being sorted into Slytherin, similar to the previously mentioned story, "Potter?". In another story, titled "Harry Potter x Reader-More Than Best Friends (First Year)", the author explicitly states that this is the "first part of a seven-part book of you and Harry". The story is written in the first person and alternates between the "Reader's Point of View" and "Harry's Point of View" (also written in the first person). Although the use of a first-person narrative in the reader's point-of-view sections allows for self-insertion, remaining within this perspective to explore Harry Potter's point of view could certainly create obstacles for continued self-insertion.

Third Primary Search: Y/N (Your Name)
As previously mentioned, Y/N fanfiction enables readers to place themselves directly within the story-world. The subgenre's name stems from the use of "Y/N" in places where the reader is meant to insert their own name, while specific Y/N fanfiction texts may also make other rhetorical or structural moves to open up places where readers can insert their own selves.
On Ao3, the use of "Y/N" is relatively consistent in several ways. The first is that most of the stories we came across were written in the first or second person, further encouraging readers to place themselves in the narrative-even when many of these texts also indicate that the author might be self-inserting as well. Of the five Y/N stories we ultimately read from Ao3 following our search query, one is in the first person, three are in the second person, and one is a list of bullet-pointed "headcanons" (i.e., a fan's own personal ideas or preferences regarding the original canon).
In "Wonderful Mischief", the author indicates in the story summary that it is a "very self-fulfilling" story, thus gesturing to the idea that they have written this story with themselves in mind. Written in the first person, this story's opening leans heavily on the sort of life the author might have known; the character's parents are Muggles (non-magical people) and the character references their headphones, a "house in the suburbs", and their parents' car. Similarly, "A Million Times Over" is a second-person-narrated, sexually explicit story where the Y/N character is in a heterosexual relationship with the secondary character Bill Weasley. In the author's notes, the author indicates that they have "been in love with Bill since [they were] 14/15"; we can assume, then, that there might be an overlap between the author's self and the self in the story. On the other hand, in "A Potions Storeroom Story-One-shot", the author has chosen to write in the second person and indicates in the summary that they are replying to an ask from Tumblr emerging from a list of dialog prompts that the author created; this story, then, is intended for readers to self-insert themselves, rather than the author. Another work, "NSFW/SFW Harry Potter Headcanons", consists of three chapters of bullet-pointed headcanons (i.e., what the fan imagines these characters would do), all listed in the second person. A chapter apiece is dedicated to headcanons about Draco Malfoy, Fred Weasley, and the original character Lorenzo Berkshire (for more on this, see Section 6.2); even though other male characters are also tagged, and even though the author attempts to keep the Y/N character genderneutral, slip-ups seem to reveal that this is a female self-insert in a heterosexual relationship with the male character featured in each chapter. The Fred Weasley chapter also switches from a second-person narrator to a first (i.e., "PDA (slaps your ass in public but only if you're okay with it, puts his hand on my thigh all the time, kisses me; mouth, cheek, forehead, or hand(s)"), thus perhaps unintentionally revealing that the author has been inserting themselves all along. Conversely, the final story we read in this subgenre on Ao3 was a "neutral!reader" story called "Sweet Dreams", which avoids identifying a specific gender or physical descriptors much more successfully; readers can mainly intimate that the Y/N character is of school age. The story also avoids any physical intimacy between the characters, aside from a touch on the face and a hug, perhaps signaling an attempt to leave "neutrality" open for asexual and aromantic readers as well. Looking at the range of "self" open for readers to engage with in these five stories, we find perhaps the most complete instantiation of that neutrality in this last text.
On Tumblr, "Y/N" fanfiction seems written primarily in the third person. These texts also tend to be significantly shorter, often only around one thousand words in length, and feature predominantly female Y/N characters. Aside from this gender identification, often implied via pronoun usage, Tumblr Y/N fanfiction texts are more careful than Ao3 Y/N fanfiction about keeping physical descriptors vague or nonexistent. We speculate that this is more common on Tumblr for two reasons: (1) the shortness of story length, and (2) the fact that many stories are written in response to reader requests, and, thus, are less likely to be influenced by the author. It is also worth noting that, for the most part, these Tumblr Y/N stories tend to be sexually explicit, and most of the ones that we located paired the Y/N character with a member of the Marauders (i.e., the Hogwarts generation before the HP narrative present: Sirius Black, Remus Lupin, and James Potter), with Draco Malfoy, or with Severus Snape. Authors on Tumblr tended to provide either an abundance of information prior to the beginning of the story (i.e., content tags, summaries, and background information) or else no information at all. Many of these Y/N fanfiction texts also incorporated or linked away from the site to pornographic gifs depicting male bodies (that are visually similar to various HP characters) engaged in sexual acts with female bodies. Such gifs rarely included the adult performers' faces, which left some room for the readers' own imagination to fill in that blank, while also still specifying the participation of white, able-bodied figures (for more on this topic, see Sections 6.4 and 6.5 below).
Broadly speaking, then, we notice that Ao3 and Tumblr Y/N stories tend to use self-insert language for the character's name and then place more emphasis and effort on making the character's physicality neutral. On FF.net and Wattpad, however, we found that the authors tended to use self-insert language more broadly, creating a textual "self" that may seem like an even more explicitly blank slate for the readers to fill in themselves. For example, in "Look at Me: A Draco Malfoy x Reader Fanfiction" on FF.net, the author uses h/c (hair color) and e/c (eye color) throughout this third-person story to allow the reader to mentally input their own defining characteristics. Similarly, in "The Boy who Loved" on Wattpad, the Y/N character's mother is identified as being Y/Na (your nationality) and the story goes on to state that the Y/N character looks just like their mother, without specifying that physical appearance. However, Y/N fanfiction authors on both FF.net and Wattpad were also more likely to set up an existing background for the Y/N character at the beginning of the story; this practice often functions in ways that foreclose some of the possibility that the in-text characteristics leave open. In the aforementioned story, "Look at Me", for example, the character is explicitly identified as a sixth-year Gryffindor, half-blooded, female character. Likewise, on Wattpad, the aforementioned "The Boy Who Loved" is the story of Y/N Scamander, the grandchild of Newt and Tina Scamander (introduced in Fantastic Beasts), who is also identified as a third-year, Gryffindor, pureblooded, female character. Thus, while the trend of a female main character is common across all four of the platforms we examined, Wattpad and FF.net authors often provided so much additional detail before the actual text that-if these preliminaries are taken into account-these stories could actually exclude many readers from inserting themselves into the otherwise blank slate of the self-insert character right from the beginning of the text.
This leads to an interesting overlap between the original characters (i.e., OCs) and blank-slate characters on these platforms. For example, in "Medusa's Heir Book I: The Sorcerer's Stone", on FF.net, the author uses Y/N instead of a character name for the main character, but this character is explicitly identified as a "girl"; the author also shares that this character's "right eye was completely liquefied gold color while her left eye was a soul-piercing blue [ . . . ] her hair was also two different colors, her left side was a pastel pink while the right side was a pastel blue." Despite attempts to create a Y/N character into which readers can insert themselves, the stories on Wattpad and FF.net often end up featuring an OC rather than a self-insert, in ways that often mirror the characteristics that we have seen attributed to Mary Sue characters.

First Supplementary Search: Self-Insert
While our overall project has focused on "self-insert" as a sort of umbrella term for an entire subgenre of fanfiction, we have also shown how fanfiction communities often use that term to indicate a specific type: i.e., texts in which fan-authors envision themselves as the subject being inserted into the text, as opposed to "reader-insert" texts that welcome the fan-reader in to do so. While we have not delineated this distinction very sharply in the present work, given our focus on theorizing the presence of that broader umbrella first, we still ran keyword searches on "self-insert", to explore what stories would be found. When we ran the search, we immediately noticed the afore-mentioned distinction: unlike the "xReader", "Y/N", or "imagine" fanfiction texts that we have explored thus far, the works tagged "self-insert" tended to imply or even state explicitly that the self-insert character represents the work's author, and that they are not necessarily open to self-insertion by the work's reader. It is also fascinating that the results for this keyword were the most dissimilar across the four platforms. On Ao3, the "self-insert" results revealed a trend wherein the story's author inserts themselves into the story as someone reborn into the HP universe; they are either reborn into an existing character or are a new character. In many of these texts, the story is told in the first person and the "reborn" main character is already intimately familiar with the Wizarding World as a fictional universe. In "Rose Petal Red", the main character dies in a car crash and is reborn as the twin sister of Ginny Weasley. She identifies herself, prior to her rebirth, as an American woman from the 21st century, identifiable by mentions of Google and iPhones; this main character also states that she has been to college, which makes her pre-rebirth age older than the school-age characters of the canon. From the author's notes and tags (e.g., the tag "EXTREMELY self-indulgent"), we can also infer that there are similarities between the main character and the author. This same trend (of a contemporary American college-aged woman) is seen in "Harry Potter and the Reluctant Rebirth." Interestingly, this story is the first in a series of five (so far), of which three are ranked in the top results for the "self-insert" keyword. Likewise, in "The Mudblood of Slytherin", the main character is once more an American college-aged woman, although, this time, the character is reborn as the younger sister of the canon character, Ted Tonks. Unlike the other stories, this one does not begin with the character's death but rather with her Sorting into Slytherin House in the 1970s. In the opening notes of "The Ghost of Privet Drive", the author states: "this is a grossly self-indulgent story [ . . . ] while this is a selfinsert fix, I've changed a few modest details of my life for privacy purposes." This is a clear admission from the author that the self-insert main character is meant to be themselves. This is a trend that does not appear in our results for other self-insertion keywords and, from this finding, we see further evidence for the way that, in many fandom communities, "self-insert" implies the insertion of the fan-author, rather than the fan-reader; thus, out of this distinction, there emerges a whole new trope.
On FF.net, the "self-insert" stories that we analyzed were also written in the first person and featured similar plots to those in the Ao3 results: somehow, the author ends up in the HP universe. On this platform, however, the authors often explicitly share that they are the self-inserted character in the story. For example, in one story's author's note, the author states, "this story is just basically about Sirius, with me thrown in as an OC", and that is the exact plot of the story, wherein Sirius Black develops a relationship with an original female character, who is a stand-in for the author. In the majority of these top search results, the authors insert themselves into the Wizarding World in an attempt to "fix" it. They often explain that they are dissatisfied with the decisions made by the author and want to remedy these "mistakes." For example, one author shares: "This story will go through what I would have done differently, knowing what I know from the books and from fanfiction". Ultimately, fanfiction is about rewriting an existing story-world to fix it, or engage in its fantasy, or explore unseen parts of that world. These authors, however, feel that there is a need for them to enter the story-world themselves, in order to effect such changes. In "self-insert with attitude", the author does not explain how they can enter the fictional universe but intimates that they have done so before. This is similar to the trend that we noticed on Ao3; in "What I Would Have Done", the main character dies in a train crash and wakes to find themself reborn into the fictional world as a new character at the moment when the book series begins; they have two sets of memories, their old ones from their non-magic life and their new ones, and they are already familiar with the story-world. Likewise, in "Cedric's New Life", the author shares that this is an attempt at writing a self-insert story in media res. This impulse of writing a modern character who is reborn into a bygone and sometimes fictionalized world would greatly benefit from further exploration and analysis, perhaps in comparison to practices such as the transmigration stories that are popular in Chinese web-based fiction. On Tumblr, the "self-insert" keyword offered a mix of results, many of which we had seen in other search results (namely, the "xReader" results), as well as posts that were not fully narrative or were not in story form at all. Common examples of this second category included bulleted lists outlining specific scenarios, and how the different characters would react (i.e., "when you're on your period (as a trans guy)", "you watching a movie with . . . ", "a day with Draco Malfoy on Valentine's", etc.), as well as fanart where the fan artist has included themselves (and the fictional character(s)) in the piece. Of the story results, most include a self-insert main character and are either written in the first or second person. Surprisingly, given our results for the other keywords queried in this analysis, these Tumblr results were less sexually explicit, and, instead, tended to be more comedic or friendship-centric in nature. For example, in "A Special Friend, Part Two [Fred Weasley, George Weasley X Reader]", the plot is centered around a friendship between Fred Weasley, George Weasley, and the reader. In "Oh No . . . R.W. x F!reader", the main character's hickeys, a product of her relationship with Ron Weasley, are seen by his mother and the consequences are humorous.
On Wattpad, neither the keyword "self-insert" nor any variation of it yielded many results; this tag seems to be used sporadically and inconsistently on this platform. In fact, the bulk of the results on the first page was clearly marked as Drarry fanfiction (i.e., stories depicting Draco Malfoy and Harry Potter in a sexual and/or romantic relationship), and, similar to our Mary Sue results on this platform, there does not seem to be any indication as to why these stories would be selected when this keyword was searched. Some of the results were self-insert stories-many of which we have seen in searches with other keywords-but these are in the minority compared to the number of Drarry fanfiction results.

Second Supplementary Search: Mary Sue
Our analysis of the "Mary Sue" keyword results on these four platforms will be brief; we ran the search for comparison purposes, but then found that the bulk of the results was either extremely similar or irrelevant. On Ao3, most of the search results were identified by their authors as crack!fic (i.e., fanfiction that is intentionally silly or ridiculous), 1 parody, and/or satire. This is a trend that we saw across all four platforms where this keyword led to relevant search results. However, on Ao3, we made several interesting observations. First, in several of the top search results, the authors stated that their story was originally published on FF.net and that they had either reposted to Ao3 "for ease of find-y-ness", or else that they were rewriting the FF.net version of the story chapter-by-chapter and reposting it on Ao3. Second, in one of the top search results, the fan-author shared that they "wrote this fic as part of [their] Master's thesis on Mary Sues." These findings add additional layers of complexity to the questions we posed earlier on Mary Sues in particular, and on self-insert fanfiction more broadly.
On FF.net, the "Mary Sue" keyword also yielded search results that were mainly crack!fic, parody, or satire. In some cases, the authors presented their stories as attempts to "write a Mary Sue", while in others, the authors indicated that the original character in their story was a Mary Sue. On Tumblr, the Mary Sue keyword led mostly to fans' meta-commentaries on whether certain characters can be considered as Mary Sues in canon or whether certain characters were written as Mary Sues in fanfiction. Finally, on Wattpad, there were no commonalities or consistencies within the search results, aside from the stories being HP fanfiction. In fact, we could not find any clear indication as to why these results were selected using this keyword; none are tagged "Mary Sue", nor do they mention Mary Sues in the comments, notes, summaries, etc. However, many of these stories would likely fall under the Mary Sue umbrella (i.e., those with characters named Ember White or Ophelia Noble, for example), which leads us to suspect that, somewhere in the metadata, they have been tagged as Mary Sue fanfiction by their authors.

Observations and Discussion of Self-Insert Fanfiction
As outlined above, for this project, we located and read a sampling of Harry Potter self-insert fanfiction from multiple subgenres, across four platforms that are widely used for fannish writing: Ao3, FF.net, Tumblr, and Wattpad. While this analysis reveals a wide variety of ways that a "self" was made evident in such texts, we also did notice several broad similarities and recurring themes of interest.
The first is that self-insert fanfiction was "coded . . . properly" (Busse 2016, p. 162) by its fan-authors using various tags, "fantagging" practices (Price and Robinson 2021), and paratextual appendages that are already widely recognized across the platform in question in order to let readers know, immediately, precisely what kind of highly personal material they would encounter in the subsequent text. Likewise, the actual texts then "mediated" such highly personal interests "properly" (Busse 2016, p. 162) through their use of conventions specific to various types of self-insert fanfiction, including the implied second-person narrative of "imagine" fanfiction and the "insert your name here" constructions of "Y/N" fanfiction; these conventions outline specific, formalized means for the fan-author and/or fan-reader to "put . . . themselves into the story" (Rtarara 2013), rather than leaving such interests up for every fan-author and fan-reader to renegotiate in every new text. Here, we return to Busse's formulation, which was initially used to describe common fan gripes with Mary Sue characters, in order to stress how formality and convention constitute a major shift in self-insert fanfiction. Put differently, self-insert fanfiction tends to be more overt and explicit about its purpose(s), even when those purpose(s) may differ in terms of text, type of self-insert, and/or platform.
Other macro-observations that are worth noting include the fact that each of the platforms we surveyed had its own practices regarding textual perspective (i.e., whether the story was written in the first, second, or third person), paratextual material(s), such as author's notes, and so on. We also observed a significant amount of overlap between various subgenres, as well as with works being tagged as both "xReader" and then also appearing in the search for "Imagines".
From these observations, we maintain again that any description of self-insert fanfiction is describing a form of writing that will rarely remain uniform, contained, or rigidly bounded. Likewise, the roles played by specific platforms cannot be overstated, and, much as we have seen Elisabetta Costa (2018) argue, understanding platform affordances-such as tags, in our case-is much improved by considering "the practices of usage within situated environments" (p. 3643). In other words, we reiterate that site tags, metadata, search results, etc., do not definitively define self-insert fanfiction or a specific subgenre of it, although they can be used for location purposes in the type of broad overview that we have offered here.
We also observed more specific trends during our skimming of the top results of each subgenre-upward of 100 fanfiction texts altogether-then, again, in our selection and close reading of five samples from each for the comparative survey above. In the following subsections, we expand upon some of these broad observations.

Community: Requests, Prompts, and Events
Across multiple platforms, we noticed that self-insert fanfiction works are often requested by a fan-author's followers and readers, or else are written for story exchanges and similar events: likewise, information identifying these points of origin is often included in paratexts, such as story summaries, author's notes, etc. This occurrence caught our notice because it highlights a certain community aspect to self-insert fanfiction: that is, despite the decidedly personal nature of this type of writing, its creation is often highly purposeful and such texts are produced on the understanding that specific people apart from the fan-author want to read them.
Requests, prompts, fanfiction-writing events, and fan-authors' paratexts regarding them were far more common on some platforms than others. For instance, Tumblr's selfinsert fanfiction frequently seems to be written in response to "asks" (i.e., a user question that is submitted directly to a specific blog) and requests made by other Tumblr users. Likewise, many fan-authors writing self-insert fanfiction on this platform run their blogs as sites where anonymous users of the Tumblr community can submit their story idea or prompt through Tumblr features such as asks. FF.net had few such notes, while their counterparts on Ao3 are often identified as cross-platform postings and are attributed to requests received on Tumblr (also, see Section 6.2, Conversations and Movement Among Platforms). Meanwhile, on Wattpad, self-insert fanfiction often seems to be published in story collections; many of these collections accumulate works written in response to reader requests that are made to fan-authors, either by private messages or via a post on their profile's "Conversations" page. On Wattpad, many fan-authors also publish an opening "chapter" for their story collections that outlines how fan-readers can request stories (i.e., make sure to list the character's school year, House, blood status, etc.). Cross-posting is also common, as some Wattpad fan-authors also seem to be receiving requests from Tumblr and cross-posting their short works on both platforms (see Section 6.2 for more).
Beyond reader requests and formalized events, such as fanfiction exchanges, another element of community that we noted regarding self-insert fanfiction is the circulation of prompts, either as rebloggable lists (on Tumblr) or static lists (on Wattpad). There are a range of ways to create and use lists for writing prompts: fan-authors can create their own, fan-authors may poll blog followers for prompts of interest to add to such a list, which is then open for readers to request stories from, or a fanfiction-writing event can create its own list, with a prompt for each day of the week or month, to which event participants can then respond. (It is also of note here that writing prompt lists of this sort are not unique to self-insert fanfiction-both Effie and Maria, the authors, have seen them used extensively among broader writing and fandom communities on Tumblr.) Some of the community elements that we observed here also dovetail with some of our later observations about the portrayal of self (see Sections 6.4 and 6.5, Slipping Into and Out of Self-Insert and Whiteness as Common Default). That is, we noticed that some of these prompt options give authors a little more structure to follow (i.e., receiving a request for a specific, tailor-made story creates the need to try and match that particular reader's desires for the text) while others are more open-ended (i.e., receiving a prompt from a prompt list does not come with additional, preferred information). Likewise, since many of these community requests remain anonymous, the fan-author may not have much information about the requester, and so may try to write a self-insert fanfiction that is even broader than usual, in order to accommodate them: These circumstances or concerns may also be brought up in paratextual, introductory material. The preponderance of requests for fan-authors to write self-insert fanfiction for their readers and followers also indicates the presence of a specific fandom gift economy (Turk 2014), which we find particularly interesting because both the labor and the text being produced in such instances are so highly personal.

Conversations and Movement between Platforms
Our discussion thus far has noted that many subgenres of self-insert fanfiction take specific forms when written for and distributed on particular platforms, and also that the community element of writing and consuming these subgenres may entail circulating prompt lists and "cross-posting" the results. However, we also noted that several strands of self-insert fanfiction, its fan-authors, and its fan-readers navigate between various platforms while engaging with this kind of writing. For instance, readers might have sent fan-authors prompts on one platform, then the fan-authors have posted the finished work on another; or the same work might be cross-posted on two different platforms, to take advantage of the different reader communities and distribution practices on each one. Some of the cross-platform conversations and movements that we observed included fan-readers and fan-users utilizing Tumblr's messages and other functions to circulate prompts, fanfiction texts referring the reader to one of the other three platforms (Ao3, FF.net, and Wattpad) as a distribution or archival space, or else transferring the work from its original home on FF.net to Ao3 as a "cleaned-up" and updated version. The three fanfiction reading platforms also often included notes indicating which platform where the text originated. Another movement that we observed affecting self-insert fanfiction in particular was between Wattpad and other social media platforms with younger user bases, most often TikTok, 2 which often had different practices altogether. 3

Explicit Content and NSFW
Aside from works that were tagged or described using the appellation "Mary Sue", sexually explicit self-insert fanfiction was found predominantly on Ao3 and Tumblr: in fact, on Tumblr, around 90% of the works under the subgenres of "imagines" and "Y/N" fell into this category. On Ao3, these stories are also predominantly written in the second person and tended to feature a heterosexual female character paired with a canon male character. If the character was gender-neutral, as occurred in only about 10% of the stories that we surveyed, the work tended not to be explicit. On Tumblr, these explicit self-insert fanfiction works were also written predominantly in the second person across the subgenres; they, too, also tended to feature a heterosexual female character, although in this case, she is often paired with a/multiple Marauder(s), Draco Malfoy, or a(/both) Weasley twin(s). Given the nature of Tumblr as a micro-blogging platform, designed to support particular types of content, these self-insert examples were shorter, often under 2000 words in length. However, even on Ao3, much of the fanfiction tagged as belonging to the "Y/N" genre was short and consisted of one-shots or comprised shorter stories that are told in single installments; there are only a few exceptions where the story is multi-chaptered.
Additionally, on Tumblr, due to its nature as an image-based platform, many of these explicit stories were enhanced through the use of gifs, edits (edited images), and "manips" (photo-manipulated images, often superimposing the character's face onto another's body or in a new situation), all of which tended to evoke or match the story itself in some way.
While, at times, such images were taken from the Harry Potter films, more often they were not, and were, instead, reworked and re-contextualized from some other form of media: photoshoots, stock images, and even pornography. For those images that did stem from the films, those that featured the actors Daniel Radcliffe and/or Tom Felton (who portray Harry Potter and Draco Malfoy, respectively) seemed the most common, probably attesting to the popularity of both these characters and the fan-favorite (relation)ship, "Drarry". However, it seemed equally likely that these images should come from other sources, although presented in ways that either "fandom iconography" (Nielsen 2021, p. 208) and/or story context made it clear whom they were meant to be depicting. For instance, a significant number of these HP self-insert stories are set during the decades prior to the HP books' narrative present (i.e., the "Marauders era"), and so there is no film footage for the love-interest characters from this time period: thus, fan-authors may offer gifs, edits, and manips of models or actors from other films.
Beyond the settings, such differences may also be broken down according to the subgenre of self-insert stories. More specifically, "Imagines" seem predominantly to use gifs from the HP films; "xReader" stories are less likely to use images, but, when they do, they are often from the HP films; and reader-insert fanfiction offered a balance between using no images at all, or else using images from the films. As an obvious outlier, Y/N stories seemed to use explicit, body-focused gifs (i.e., the faces are not shown) pulled from live-action pornographic material; these images tend to feature white individuals in various explicit positions, although still being mindful of Tumblr's 2018 crackdown on pornographic images. Additionally, however, interspersed among the top results for Y/N on Tumblr are links to pornographic gifs/video clips on Twitter that feature male bodies similar to the HP characters; the faces of the individuals in the gifs are rarely shown, heightening the readers' ability to superimpose these visuals and their surface similarities to HP characters over the explicit narrative itself.

Slipping into and Out of Self-Insert
As we have already seen, many contemporary forms of self-insert fanfiction utilize formal conventions that leave a literal space for the reader within their narratives, such as the implied subject position that results from inserting one's own name to replace the "Y/N", or envisioning one's own self participating in the narrative, along the lines of "imagine [you were]. . . " However, our analysis of a select sample here also uncovered how, often, the fan-authors of these works also include physical descriptors that likely allude either to themselves or, alternatively, to their ideal selves. Often, these slippages occurred when the author described the Y/N character's hair ("long hair, "braided hair", "thick, tousled hair", or hair that another character can "[run] their fingers through"), or when providing the Y/N character's physique ("small frame", "bikini", "slim waist"). These descriptions of the character's hair and body immediately exclude those potential readers who do not have these hair types or body types. Similarly, as seen in the previous sections, the great majority of the top stories emerging from our keyword searches featured a cisgender, neurotypical, non-disabled, female main character engaged in a heterosexual relationship with a cisgender, neurotypical, non-disabled, male canon character. Thus, despite the effort to allow for fan-readers to fully insert themselves into the story, these characterizing details immediately exclude certain readers, if they are looking for truly blank-slate self-insert characters.
This impulse leads us to question whether self-insert fanfiction is driven by fantasy or by representation; our analysis has certainly indicated that it could be a mixture of both, depending on the genre, platform, and individual inclination. However, if even some of these stories do indeed emerge out of a type of wish-fulfillment, then, future research should consider the implications of heterosexual, cisgender, neurotypical, non-disabled, female, and white (see Section 6.5) characters being adopted as (or being assumed to be) the desired identity/fantasy by fan-authors. This could be an especially interesting facet to explore on a platform like Wattpad, thanks to its distinctly younger user demographic.
Simultaneously, we also observed even more distinct and deliberate attempts to keep self-insert characters neutral in works that the fan-author has written directly in response to an ask or request, i.e., on Tumblr (see Section 6.2, Conversations and Movements Between Platforms, above). In these cases, fan-authors were much less likely to include physical descriptors of the self-insert character when the ask or request did not specify additional details; however, in some cases, if the ask or request specified a detail (i.e., "nerdy Ravenclaw with glasses"), then the fan-author ensured that this detail was included in the subsequent story. From this finding, we speculate that fan-authors may be more able to delineate between own-self and story-self when responding to anonymous requests, where the impetus for the story has come from a subject beyond the author's own creativity. Conversely, we wonder if, when fan-authors are writing a story based on an idea that they themselves had or would like to experience, they are more likely to let parts of themselves slip into the story.

Whiteness as a Common Default
A final key observation for the present project builds on the previous point regarding the accidental insertion of specific selves, whether those of the fan-author or of an ideal(ized) self. We noted that one of the most common and recurring instances of slippage between potential selves concerns the self-insert character's implied whiteness. That is, we found that most fan-authors writing across these various subgenres seemed to try and keep their self-insert characters fairly neutral, steering away from overt descriptions of physical characteristics, so that fan-readers could then envision themselves in these scenarios without having to negotiate a position regarding whether the bodies or appearances being explicitly described were unlike their own. As we noted above, however, these attempts were not always fully realized, and fan-authors may unconsciously have included descriptions of specific hair types, body types, and more.
Whiteness becomes another such descriptor, the placement of which, within the formal structures of a self-insert subgenre (i.e., implicit or explicit second-person narration, the injunction to "imagine" a scenario, an unstated understanding that the reader's own name replaces the acronym "Y/N", etc.) can exclude particular readers, thus negating the main point of self-insert fanfiction entirely. We noted multiple examples of implied or explicit whiteness occurring across the sample of stories read for this project, including Y/N characters who have, respectively, a "white chest", a "face white as a sheet", "bright red cheeks", and certain hair types, among other identifying characteristics.
We highlight these instances, not to claim that they are always self-reflexive on the part of their fan-authors but, instead, to emphasize that self-insert fanfiction and its subgenres are not immune to the all-too-common, fandom-wide settings in which "the logics of whiteness structure the assumptions of both fans and scholars in these spaces" (Pande 2020, p. 4), exacerbating mainstream Western media's and society's assumptions of whiteness as a default (Dyer 1997) within smaller subcultural or even grassroots spaces. The long history of this trend, its overlooking of entire fan communities, and the various types of violence that it perpetuates (including those weaponized along the lines of self-insert characters; see Neill Hoch 2020; Stitch 2021) all make their recurrence in the subgenres we are studying here even more important to highlight, particularly when every single one of the stories that we found, slipping into racialized descriptions in this way, highlighted whiteness.

Some Concluding Thoughts and Future Directions
Our goal with this project has been to trace certain shifts in the development of selfinsert fanfiction, including common threads from Mary Sue characters-particularly the way in which these were often understood and judged as authorial self-inserts, often based on fan-readers' own biases and preferences-to contemporary forms, including "imagines", "Y/N", and "xReader" fanfiction. To this end, we have offered a brief overview of Mary Sues, some definitions of specific self-insert types, and a discussion of findings drawn from a survey of HP self-insert fanfiction. With these analyses, we hope to provide a starting point for further discussions of this subgenre.
We have also taken a Foucauldian approach to this topic, viewing self-insert fanfiction along the lines of fiction-mediated "technologies of the self" (Foucault 1988). However, rather than adopting Foucault's framework for personal, confessional, self-governing writing in its entirety, we also look to the possibilities inherent in digital technologies of the self, as theorized by Abbas andDervin (2009), Siles (2012), and others, which often tend to stress knowledge of the self over the governance of the same. While we do find common threads between this concept and self-insert fanfiction, we also demonstrate how the umbrella term of self-insert fanfiction at least attempts to open up the "self" that is being created and is performing for the enjoyment of others besides its author. Moreover, this subgenre of fanfiction also avoids some of the criticisms often levied against earlier Mary Sue characters, by "mediating and coding" certain forms of particularly transgressive desire "properly" (Busse 2016, p. 162)-that is, in ways that can be recognized immediately and either engaged with or avoided, as desired.
Likewise, our preliminary survey of HP self-insert fanfiction across four platforms has led us to ask further questions that future work in this area could continue exploring and expanding upon. For instance, in HP fanfiction, at least, around 85% of the fanfiction texts we encountered in this subgenre involved a "het" (i.e., heterosexual or opposite-sex) pairing of a female self-insert character and male canon character, whether or not the story itself was sexually explicit: we found this to be a fascinating departure from the preponderance of slash (i.e., male/male, same-sex) pairings, particularly in sexually explicit fanfiction. Likewise, we imagine that future work could further explore whether the origin points of different self-insert fanfiction types also influenced their current characteristics. For instance, Fanlore (n.d.) documents how both "Y/N" and "imagines" seem to have begun on Twitter, even if they are now more common elsewhere. So, while platforms such as Wattpad and their multi-million-hit stories demonstrate an increase in the prevalence and acceptance of certain forms of self-insert fanfiction, these genres rarely began in these highly visible spaces, even though they may have developed specific conventions there. Another area of great potential for future study can be found in the multi-media compositions of self-insert fanfiction, which, in the case of HP fanfiction at least, may use green-screen effects and scenes spliced from the films to place the fan-creator's own body into the story-world (Sapuridis 2021).
We might articulate some of our other remaining questions as follows: • What can we make of the slippage between self-insert writing as being primarily a reader-oriented fantasy and then as (potentially) a representation of the reader in what are often fantasy settings? • What does a deeper delve into the "fantagging" (Price and Robinson 2021) of self-insert fanfiction reveal about this overarching practice, the subgenres within it, and those who engage with it? • Why, as our preliminary sample here suggests, do the majority of self-insert fanfiction across subgenres, such as "imagines" and "Y/N" stories, feature female-identified characters in heterosexual relationships, as contrasted with the predominance of slash and queer relationships in other types of fanfiction? • If this is true, how might this tendency dovetail with the nature of self-insert writing?
However, as mentioned in our introduction, our intent with this project is not to set out this recent shift toward purposeful, participatory self-insert fanfiction as an "evolution" of the Mary Sue: There are far too many shortcomings to that particular approach, most notably its risk of becoming the teleological argument that contemporary forms are better, more progressive, or more self-aware than their predecessors. This is both disingenuous and also risks reifying the particular stigmas surrounding certain forms of authorship, reading practices, and encodings of reader desire (Morrissey 2014). There is also the possibility of inadvertently holding fan-authored works to commercial or cultural standards that they were simply never created to meet, and/or of equating standard writing conventions with value or meaning.
Considering self-insert fanfiction in terms of being a digital technology of the self, however, can also offer a means of understanding some of the backlash against it. Unlike the private confessional letters between only two people that Foucault deems the technologies of the self, digital technologies of the self are explicitly designed and are expected to reach much wider audiences, the majority of whom the author will not actually know personally and who, thus, lack certain frames of reference that Foucault's (1988) formulation presumes. To reapply Busse's (2016) terms here, we might point out that what such a text is "mediating and coding" (p. 162) and also how it is doing so become key aspects of these digital technologies of the self-here is where many divides originate, as those who author the texts and those who encounter them.