Building an Archive of Our Own: The Ephemerality of the Digital Fanfiction Archive


For decades, fanfiction has been a beloved practice among fans of just about any media fandom. Whether it is novel-length stories about members of One Direction romancing ordinary girls or short, rambling anecdotes about which character from the Harry Potter universe is most likely to order a pumpkin spice latte, fanfiction offers readers a chance to live out their fantastic and mundane fantasies in a fan community that is relatively free of judgment.

Compared to the days when paper copies of fanfiction were circulated around conventions and mailing lists, the Internet has enabled today’s fanfiction can reach a wider audience at a faster pace. In 2017, the bulk of new fanfiction online can be found on Archive of Our Own, also known as AO3. Founded in 2007 by the Organization for Transformative Works (OTW), AO3 offers a noncommercial and nonprofit hosting place for fanworks, including videos, art, stories, and audio content. As of December 20, 2017, AO3 has more than 3,509,000 works from over 1,344,000 users. Though the digitization of fanfiction has dramatically increased its accessibility, it also has brought up questions about the effectiveness of the digital archive.

In this essay, I will reflect on how the relationship between fanfiction and the archive. This relationship, I argue, has shaped fanfiction creation as both a cultural practice and a form of labor that has the potential to be exploited for capital gains. I will also delve into the nature of the archive as infinitely expanding, yet still susceptible to loss, and what this potential loss could mean for digital archives like Archive of Our Own. In acknowledging the ephemerality of the fanfiction archive, we can think on the ways in which it manages to endure.

Digital Fanfiction and the Papyrus Scroll

Though one might think of printed fanfiction as the direct precursor to digital fanfiction, Lev Manovich offers an alternative perspective of the shift from print to digital. Manovich observes that text is unique among media types, in part because it “is a metalanguage of computer media, a code in which all other media are represented” (74). Text, he argues, is the primary means of communication between a user and a computer. And if text is the metalanguage of computers, then cultural interfaces are derived from the principles of text organization, such as the page. Manovich uses the term ‘cultural interface’ to describe “a human-computer-culture-interface – the ways in which computers present and allow us to interact with cultural data” (70). New media cultural objects like online museums and multimedia encyclopedias are examples of cultural interfaces.

According to Manovich, cultural interfaces stretch the definition of the page interface to include new concepts made by the computer. For instance, overlapping windows on a computer monitor function like a set of book pages that the user can scroll through. As technology has progressed, the virtual page has extended even further to include elements like HTML. In its transition from clay tablets to paper to HTML, the page, Manovich suggests, has become more unstable as it mixes together different historical forms. The scrolling feature of a web page shares more similarities with a papyrus roll than it does with the sheet pages of a book, and the integration of photographs and sound on a web page is conceptually similar to a newspaper page.

AO3 web pages operate much like the web pages that Manovich describes. Fanfictions are broken into chapters (at the author’s discretion) and each chapter can be opened in a new tab or window, allowing users to jump through chapters at their own pace. At the same time, authors who choose to write long chapters, or forego putting their stories into

A general tutorial for HTML text-formatting on AO3.

A general tutorial for HTML text-formatting on AO3. (AO3, 2015)

chapters in favor of page breaks, require their readers to scroll through a web page that can go on for thousands of words Tech-savvy users can use HTML to format their text and pages or add tables, lists, links, or images into their stories, creating a multimedia experience. Although these processes are happening online, fanfiction writers are drawing on the traditions of pre-digital books, scrolls, and periodicals, revealing the complex nature of online cultural interfaces.

On the “Archive”

The development of the fanfiction archive has had a profound impact on how fanfiction is written, read, and discovered. Abigail Derecho calls for an analysis of fanfiction not as a cultural phenomenon, but as a cultural practice. Derecho defines fanfiction as a subgenre of archontic literature. Archontic literature, Derecho explains, refers to Jacques Derrida’s definition of archives as always expanding. Looking at archontic literature from the seventeenth century to the present, Derecho evaluates how people of color and women have used archontic writing to critique social and cultural expectations. Influenced by poststructuralists like Gilles Deleuze and Edouard Glissant, Derecho suggests that fanfiction, as a form of archontic literature, challenges outdated notions of hierarchy and property.

Derecho observes three arguments among fan scholars regarding fanfiction and its origins. The first argument is that fanfiction originated several millennia ago, with myth stories, and today encompasses works from authors who identify as fans and those who do not write from within fandoms. The second argument is that fanfiction should be understood as a product of fan cultures. Star Trek, Jane Austen, and Sherlock Holmes fan societies are just some examples of twentieth century fan cultures. The third argument, which Derecho declares is the most accurate line of thinking, is that fanfiction neither encompasses most of literature nor emerges out a relatively recent trend in audience response.

Derecho’s use of the term “archontic” comes from Jacques Derrida’s 1995 work Archive Fever, in which he asserts that no archive is ever complete or closed. Although Derrida imagines the archive as ever-growing, Alexis Lothian notes that Derrida also recognizes that a complete archive would be an impossible concept. Loss is an inevitable part of any archive, including the digital archive. The archontic principle, Derecho elaborates, “is the drive within an archive that seeks to always produce more archive, to enlarge itself” (64). Derecho argues that unlike the terms “derivative” or “appropriative,” which imply hierarchy and ownership, “archontic” best describes how fanfiction is archival in nature. She explains that “all texts that build on a previously existing text are not lesser than the source text, and they do not violate the boundaries of the source text; rather, they only add to that text’s archive, becoming a part of the archive and expanding it” (65). The archontic text has the potential for infinite expansion, which calls into question whether all texts can be considered archontic. Derecho asserts that for her analysis, archontic refers only to “works that generate variations that explicitly announce themselves as variations” (65). Since works on AO3 are clearly delineated as variations (listing the ‘source’ fandom(s) is mandatory), the fanfictions on the archive would fall under this classification of archontic literature.

Derecho is also interested in how archontic writing incorporates repetition. Citing Gilles Deleuze, who argues that a repeated material can in fact contain differences from the first iteration, Derecho claims that “when one reads a work of archontic writing, in other words, one is really reading two texts at once” (73). For the fanfiction reader, the source material remains in their mind while they read fanfiction, and the fanfiction subsequently impacts how they view the source material. Consequently, the fanfiction reader might come to prefer their own version of a story to what is happening in the canonical text, possibly abandoning the show altogether in favor of the online community.

Wrangling the Tags

To keep the archive organized, AO3 relies on a tagging system, but this system is not without issues. In her examination of tagging practices on AO3, Kelly Lynn Dalton argues that AO3’s hybrid of freetagging and some vocabulary control and hierarchy works well. According to Dalton, AO3 employs a free tagging system by allowing users who post their own fanworks to create their own tags. Once a user posts their tags, volunteers known as “tag wranglers” step in behind-the-scenes and link the tags to AO3’s canonical tags. These canonical tags are set up by the wranglers for the character names, relationships, and the source material. For instance, Dalton explains, the tag “Romeo and Juliet” would fall under the canonical tag “Juliet Capulet/Romeo Montague.”

Despite this creative freedom with tagging, AO3 still requires certain types of tags, mostly to prevent readers from stumbling upon inappropriate or upsetting content. According to Dalton, AO3 requires users to include a rating, such as “general audiences” or “teen and up audiences.” There are also options to put a content warning for violence, death, rape, and underage sexual content. Users can somewhat work around these required tags by labeling their works as “not rated” and choosing “author chose not to use archive warnings” in lieu of content descriptors.

Dalton notes that AO3 maintains some vocabulary control by pushing autocomplete suggestions at taggers. Once an author has assigned a story to a specific fandom, the system will start suggesting canonical names from that fandom based on the first letters the user types into the system. Autocomplete can help prevent typing errors, which streamlines the tag-wrangling process. Users are not required to use the autocomplete suggestions, but they are encouraged, suggesting that AO3 is influencing the language of fandom despite its claim that it is giving users full creative control.

AO3 is not the only fanfiction website on the Internet, but its tagging system is arguably the most complex. Dalton offers some insights into the divide between AO3 and, another popular fanfiction-hosting website. Based on a quantitative and qualitative survey of AO3 users, Dalton finds that most respondents prefer AO3’s complex search system compared to’s bare-bones system that limits tagging for up to four characters, two genres, and two fandoms. Although she does not address the issue in her survey, Dalton suggests that some of this preference for AO3 may stem from’s refusal to allow explicit fanfiction and the automatic deletion of explicit works from the website.

While AO3’s search system seems to be preferred by fanfiction readers, Dalton reports that users are not completely satisfied by the website’s filters. Many respondents indicated that they would like to frame their searches around the exclusion of certain characters, pairings, and fandoms. Rating and warnings make it easy to filter out graphic violence and sexual content, but filtering out other tags is much more challenging unless one is intimately familiar with both AO3’s tagging system and advanced search commands like (-) and (+).

Another issue with AO3’s tagging system is the weighting, or lack thereof, of tags. Dalton observes that users struggle to determine who the “main” character or pairing is for a story because AO3 gives equal weight in the tags to each mentioned character or pairing. Dalton suggests that authors who tag every single character who appears in their

On a similar note, this AO3 tag generator generates a random AO3 tag, such as “#potentially deadly greco-roman OC’s” or “#cringeworthy breakfast love story.” The relative popularity of this generator on AO3 suggests that users are self-aware of their elaborate tagging styles.

On a similar note, this AO3 tag generator generates a random AO3 tag, such as “#potentially deadly greco-roman OC’s” or “#cringeworthy breakfast love story.” The relative popularity of this generator on AO3 suggests that users are self-aware of their elaborate tagging styles.

fanfiction can create a sort of “tag fog” for readers looking for stories focused on their favorite characters (85). Oftentimes, tag fog is the result of users practicing a sort of stream-of-consciousness style tagging. Tag fog has become such an issue in the AO3 community that there are entire blogs dedicated to showcasing some of the more colorful tags. Highlights from the “AO3 Tag of the Day” Tumblr blog include “#honestly this fic could very much be my horcrux #i poured my entire soul into it” and “#innaccurate descriptions of jobs and other things i got off google #or just made up tbh #the important stuff is accurate #i swear.”

In allowing freeform tagging, AO3 expands creative freedom at the expense of efficiency and accessibility, demonstrating one of the inadequacies of their archive model that values users over moderators. While users may find’s limited tagging features to be too stifling, they also find AO3 to be at times, too relaxed.

The Digital Archive Culture

The reason for this user-centric archive model stems from the origins of the organization behind AO3, the Organization for Transformative Works. Alexis Lothian explains that the name “Organization for Transformative Works,” or OTW, comes from the provision in US copyright law that ‘transformative’ uses of copyright material qualify as ‘fair use’. From the OTW’s perspective, fanfiction and similar types of fanworks like videos and drawings, are transformative works and therefore do not infringe on intellectual property laws. As a supporter and member of the OTW (and having served on the founding editorial team of the OTW’s open-access academic journal, Transformative Works and Cultures), Lothian seeks to complicate the discussion around what fannish and scholarly archivable content is and should be.

Drawing on Jacques Derrida’s 1996 work, Archive Fever, Lothian investigates how contemporary fandoms form digital archive cultures. Lothian suggests that online fandom developed an archival formation in order to share amateur art and writing. By becoming digital archive cultures, Lothian writes, “online fan communities have transformed from marginal worlds, of interest only to members and scholarly admirers, into discursive landscapes at the heart of debates about digital media and intellectual property” (544). These fan communities are larger, more exposed, and more interconnected with other fandoms than ever before. When AO3 users search for specific tags like “hurt/comfort” or “alternate universe: coffee shop” without specifying which fandom they want to read about, they are put into conversation with potentially hundreds of fandoms and thousands of users who are all talking about the same tag. These newfound relationships consequently expand into other online spaces like Twitter and Tumblr, incorporating an even larger audience. As AO3’s userbase continues to expand, so does its potential to be recognized by potential to be recognized by “official” media producers.

Fanworks and Fans Working

Lothian is concerned that fans’ archontic production has been viewed by corporations as a resource to be exploited. She elaborates that fans’ willingness to do free labor is profitable to producers because it increases the presence of copyrighted characters and narratives. However, Lothian asserts, fans are conscious of some of the ways in which media companies attempt to exploit fan labor. Back in 2007, users on, a fanfiction-hosting website that was sponsored by media companies, complained that the people behind FanLib were profiting off other people’s intellectual property. Calling for an archive of their own, some of these fans drafted a vision of a centralized archive that would encourage fans to participate in a subcultural and non-corporate fannish archive culture. This vision turned into the umbrella nonprofit Organization for Transformative Works, which went live in October 2008 and bought its own servers in September 2009. Running without commercial advertisements or official corporate sponsorships, the OTW continues to thrive today because of donations from its users. OTW is also supported by its staff of volunteers, who serve as tag wranglers, testers, coders, support staff, and abuse report staff.

Though the OTW says that the two ways to support the organization are donating and volunteering, creating fanworks is ultimately what keeps the archive alive, bringing in new visitors and keeping current users engaged. Mel Stanfill and Megan Condis argue that fan activities, including fanfiction writing, have been framed as participatory and voluntary, which calls into question how labor functions in fandom. According to Stanfill and Condis, industry recognition and encouragement of fan value creation is a relatively new phenomenon that can be attributed to the increased visibility of fan practices online.

Stanfill and Condis emphasize the importance of thinking of fan practices as a form of labor. They note that fandom has typically been thought of as a gift economy in which a gift is not a freely offered expression of affection but is instead obligatory. This line of thinking, Stanfill and Condis argue, overlooks how fan labor, however pleasurable it may be, is still labor. Fan labor needs to be recognized as such because it shapes online fandom in a significant way. Fan labor is what built the Organization for Transformative Works and maintains AO3. Though there are issues of exploitation to investigate, Stanfill and Condis assert that framing fan action as work provides a way to appreciate what fans are doing.

An Archive of “Our” Own?

The ‘our’ in ‘Archive of Our Own’ calls into question the relationship between the community and the archive. The OTW professes a rhetoric of community with the motto “by fans, for fans,” but Lothian, drawing on Derrida and Miranda Joseph, insists that “all communities, like all archives, are constituted through violence and erasure” (548). The conflicts and exclusions happening in the community and the archive are central to their formation.

One of these conflicts concerns the OTW’s relationship to law and capital. To protect fans from infringement accusations, the OTW claims that fans are loyal customers who will ultimately allow others (media companies) to profit from fans’ free labor. While the OTW is a nonprofit organization, Lothian explains that their drive for legal legitimacy rests on this strategic claim. Lothian emphasizes the importance of asking who and what will be left behind as the OTW evolves into a public-facing organization that represents fan cultures’ interests to academics and journalists.

Lothian addresses the concern from some fans that fan studies is invasive. She notes that these fans “would prefer not to be legitimated into a scholarly archive, not to be a source for articles like this one, perhaps not to be archived at all” (549). Lothian proposes honoring fans’ discomfort with scholarly archiving by looking at the ways that fan cultures exceed the OTW’s archival model. When the OTW was developed, LiveJournal was one of the more popular social networking services for subcultural fans. According to Lothian, LiveJournal’s ephemeral style of engagement emphasized user interactions rather than archival features. Links to past postings enabled users to store and share fanworks in the manner of an archive, but the website did not differentiate structurally between art, fiction, and photographs. Lothian presents LiveJournal as an example of how fandom’s relationship to the archive is neither static nor stable. By thinking of online participation as an ephemeral performance, she proposes that scholars can approach the digital with attention to the affective elements that exceed the recorded archive.

Remembering the Archive

Wendy Chun offers a closer examination of digital media and memory. Chun argues that “the major characteristic of digital media is memory” (154). She observes that the technological shift from calculator to computer depended on regenerative memory. If the calculator was a machine, then the computer was a human-emulator. The Internet’s content is similarly based on memory. Chun notes that many websites and digital media projects focus on preservation, suggesting that “memory allegedly makes digital media an ever-increasing archive in which no piece of data is lost” (154). This idea that no piece of data is ever lost has led some to see the Internet as threatening. Regardless of this threat of exposure, people continue to share compromising information about themselves online.

Chun refers to Vannevar Bush’s “As We May Think” to reflect on the conflation of the memex and the Internet. According to Chun, Bush’s memex is “linked to a mechanical, analog future that has not and arguably may not come to pass” (155). Thinking of the memex as a precursor to the Internet conceals the ephemerality of digital media. Bush envisioned the memex as a mechanical solution for researchers struggling to access scientific records. Ideally, the memex would save people from repetitive thought and repetitions in thought. Being able to pull up any piece of knowledge with a machine would allow one the privilege of forgetting.

The problem with Bush’s vision of the memex is that Bush does not differentiate between access to and understanding the record. According to Chun, Bush assumes that human records make possible the construction of an archive that has no gaps. By Bush’s logic, discontinuities in human knowledge are due to a failure to consult the human record. Unlike humans, machines are somehow accident-free and permanent.  The reality of machines, however, is that they are susceptible to failure. Microfilm fades, paper burns, and hard drives are corrupted. In the world of online fanfiction, stories can be deleted by users, erased due to a website crash, or removed by moderators for violating the website’s terms of use.

Chun argues that this belief that machines are more stable than human memory is at odds with the material transience of the Internet. She emphasizes the ephemerality of digital media, writing that “digital media is degenerative, forgetful, erasable” (160). Computer memory is akin to erasable writing because the computer’s memory can constantly be rewritten, and the age of a computer memory seldom corresponds with the age of the memory it holds. Because fanfiction on AO3 can be updated and edited, a user can easily rewrite a story, making the outdated version essentially “disappear” from the archive.

Chun is concerned with the conflation of storage with memory. For Chun, memory is an active process that looks backward, while storage looks toward the future. In other words, memory is an act of remembering that requires constant regeneration to keep it from fading or moving. Echoing the argument of Wolfgang Ernst, Chun asserts that digital media is a time-based medium. She elaborates that digital media “depends on a degeneration actively denied and repressed” (167). While film stock and paper are recognized as non-durable, the digital computer is still perceived as a permanent memory machine. The computer appears permanent in part because it is constantly refreshed so that its ephemerality survives.

This “enduring ephemeral,” which Chun classifies as a “battle of diligence between the passing and the repetitive,” is also applicable to content on the Internet (167). Though the Internet may be available around the clock, specific content may disappear and reappear at any time, which can prove frustrating for those trying to erase content. The Internet Wayback Machine (IWM), a webservice that makes archives of deleted or changed websites available, is one avenue for recovering lost content. The webpages that the IWM preserves are “not quite dead, but not quite alive either” because there remain gaps in the preserved pages (such as broken links or missing images) but they visualize how constant regeneration affects what is regenerated (169). A proper commemoration would require greater effort than IWM’s technology has to offer. AO3 is also available on the IWM, though the OTW advocates for users to download their favorite stories in case AO3 vanishes from the IWM, as it did in 2016 for seven months. For all that AO3 and the IWM can do to archive fanworks, it too is susceptible to the loss that afflicts all archives. If the digital archive is ephemeral, as Chun suggests, then AO3 users must accept the possibility that one day their stories could vanish.

Conclusion: Why Do We Archive Fanfiction?

Though some might argue that the loss of fanfiction would have a significantly smaller impact than say, the loss of a scientific record, digital fanfiction is vital for online fandoms because of its liberating potential. Derecho describes fanfiction as the “literature of the subordinate, because most fanfic authors are women responding to media productions that, for the most part, are characterized by an underrepresentation of women” (71). Although fanfiction, Derecho argues, tends to reinforce traditional gender roles and social norms, it still functions as a form of artistic resistance. The virtual realm plays a significant role in expanding the possibilities of fanfiction. According to Derecho, Deleuze sees the virtual as a set of potentialities that could happen at any given time. While print culture involved the exploration of only a handful of possibilities within texts, Derecho asserts that the Internet has allowed thousands of potentialities to be actualized and circulated online. Reflecting on the importance of fanfiction and archontic literature, Derecho suggests that these works open up possibilities for a different perspective on the institutional and the social.

The Archive of Our Own might be an archive of wacky Dragon Ball/Sons of Anarchy crossover fanfictions, but it is also a commemoration of fan resistance to a world that constantly undervalues women, people of color, people in the LGBTQ community, and other disadvantaged groups. That this archive somehow emerged out of the traditions of the papyrus scroll and the magazine leaflet demonstrates the interconnectedness of old and new media in a what has been thought to be an exclusively digital age. As scholars think about how such an archive can be preserved, I echo Chun’s sentiment that we must embrace the ephemerality of the Internet and appreciate how, despite its constant regenerations, the digital archive survives.

Works Cited

Chun, Wendy Hui Kyong. “The Enduring Ephemeral, or the Future Is a Memory.” Critical Inquiry, Volume 35, Autumn, 2008, 148-171.

Dalton, Kelly L. Searching the Archive of Our Own: The Usefulness of the Tagging Strucutre. Thesis. University of Wisconsin-Milwaukee, 2012.

Derecho, Abigail. “Archontic Literature: A Definition, A History, and Several Theories of Fan Fiction.” Fan Fiction and Fan Communities in the Age of the Internet: New Essays, edited by Kristina Busse, Karen Hellekson, McFarland & Company, 2006, 61-78.

Lothian, Alexis. “Archival anarchies: Online fandom, subcultural conservation, and the transformative work of digital ephemera.” International Journal of Cultural Studies, Volume 16, Issue 6, 2012, 541-556.

Manovich, Lev. The Language of New Media. MIT Press, 2001.

Stanfill, Mel and Condis, Megan. “Editorial: Fandom and/as labor.” Transformative Works and Cultures, Volume 15, 2015. Web.