Poster

Click to view at full size.

Text Description of Poster

Poster Title: (Re)Archive: The Rise and Fall (and Rebound?) of Independent Fanfiction Archives

By Dawn M. Walls-Thumma
@DawnFelagund on Twitter and Tumblr
dawnfelagund.com

The left sidebar includes information on my research, and the center section includes the data itself.

Sidebar:

An independent archive is a website for sharing fanworks, run by fans but not affiliated with any for-profit or nonprofit corporations or organizations. Independent archives have been a significant part of online fandom, but in recent years, dozens of these sites have disappeared.

I wondered what factors predict these sites’ creation, survival, and disappearance?

Methodology

I included archives listed on Fanlore from the Tolkien and Harry Potter fandoms that were

• single-fandom,
• multi-author, and
• English-language.

To be included, archives must have the following data available, if applicable:

• year of opening
• final year of activity
• year of closure
• year ported to the OTW’s Open Doors project

This data was often discoverable via Wayback Machine captures, news posts on- and off-site, and update pages. All dates on Fanlore were verified.  An important part of my process includes updating Fanlore with new information and documentation, when it is discovered so that my research extends the historical record.

Findings

87% of archives opened in the years 2001-2005. These years saw major film releases and media attention toward both fandoms studied. Home internet access (in the U.S.) passed the halfway mark, and Web 2.0, which allowed internet users to interact with content and each other, arrived.  Community and conflict are sometimes given as reasons why fans build archives: as a place to preserve the fanworks of a community (De Kosnik, 2016) or in response to conflicts, such as content purges or fan disagreements. Certainly, these are true in individual instances, but archives that arose from communities didn’t survive longer than archives with affiliated communities. Nor did major fandom conflicts predict widespread archive building.  Instead, the biggest predictor of an archive’s survival was automation, which shifted labor off of archive volunteers.

In order for independent archives to exist and thrive, automated scripts must be available to fans.

AO3 and Tumblr moved the fandom toward consoli- dation, and the last major eFiction update occurred in 2015, leaving potential archivists without an easy site- building solution.

Why Independent Archives Matter

With the existence of the Archive of Our Own (AO3), do we even need independent archives? We do! Independent archives

• establish and preserve unique fan cultures
• encourage community-building and collaboration
• grant autonomy to communities to manage issues like tagging, moderation, and inclusiveness     differently than AO3, creating spaces that are safer     and more accessible to diverse users
• integrate features and resources unique to their     community or fandom
• ensure fanworks are preserved in more than one     location
• cost little in labor and money to maintain, if (when?)     fandom moves beyond internet-hosted fanworks
• empower communities through self-governance and     self-sufficiency
• serve as fanworks in their own right

(Re)Archive? The Future

Recent years have seen several independent archives close or go inactive. At the same time, there is growing interest in creating software and platforms that can be used for independent fanworks archives and educating fans in how to building and maintain archives.

• Fandom Coders
• BobaBoard 
• Ourchive
• Neocities
• Umpteen site builders, blogging platforms, and     content management systems that could be     repurposed as archives

Data:

The data is represented as a series of stacked, color-coded timelines, one for each archive in the study. Color-coding shows when an archive was active, inactive (but still online), offline, and available via the Organization for Transformative Works' Open Doors Project.

At the bottom of the timeline are a series of events:

Panfandom

• 2002: Fanfiction.net NC17 purge
• 2004: eFiction released
• 2007: LiveJournal Strikethrough
• 2008: AO3 opens
• 2009: Geocities shutdown
• 2012: Fanfiction.net M-rated purge
• 2017: LiveJournal ToS change
• 2019: Yahoo! Groups shutdown

Tolkien Fandom

• 2001-2003: Lord of the Rings films
• 2012-2014: Hobbit films

Harry Potter Fandom

• 1997-2006: Harry Potter books
• 2001-2011: Harry Potter films

U.S. Home Internet Access

• 2001: over 50%
• 2007: over 60%
• 2010: over 70%
• 2016: over 80%

The data are divided into two sets. The first set is hand-coded archives. These archives were maintained entirely by hand. HTML (and sometimes CSS), graphics, and updates were hand-coded by volunteers. The immense amount of tedious labor involved meant that these archives tended not to survive long. The arrival of eFiction and other automated software marked a shift in how archives post-2004 were built.

Hand-coded archives were

• 5 median years online
• 2 median years active
• 61% (or more) had affiliated communities

The color-coded timelines show that, for the majority of hand-coded archives' existence, they are inactive or offline. These archives were early in the history of the two fandoms studied; there were none eligible for the study that were built after 2004. There really are no discernible patterns that explain why these archives go offline. The reason, based on looking at many of them, seems to be the amount of labor involved. On many, the final post of the webmaster was an apology for a lack of updates or a promise to update soon.

The second set is automated archives. These archives used scripts to automate many or most aspects of the archival process. Labor transferred from site admins/moderators to creators and users. Community features, such as comments, became possible. Some sites were custom-coded but most used open-source scripts, with eFiction by far the most popular.

Automated archives were

• 12 median years online
• 10 median years active
• 57% (or more) had affiliated communities

There are many more automated than hand-coded archives. The peak of building automated archives was 2002-2005.

The color-coded timelines show much more red, denoting that these archives are not just online but active much more than hand-coded archives are. Going offline or inactive after just a few years is relatively rare; only six of the sixty-eight automated archives included were active for two years or less.

There was a wave of archives going inactive or closing in the mid-2000s, before AO3 opened. Archives sputtered out across the fifteen years after, though the last two years have seen nine (of sixty-eight) archives going inactive or closing.

Presentation Audio

Presentation Text

A year ago at this conference, I raised the question of small, independent fanworks archives, whether they were worth preserving, and what was needed to do so. I have spent the last year looking closer into the latter question especially: What conditions foster the establishment and survival of independent archives, and what causes their inactivity and closure? For the purposes of my study, I defined an "independent archive" as a site that was built and run by members of the community it serves and unaffiliated with any large for- or nonprofit corporations or organizations.

I began lurking in the Tolkien fandom in 2004 and participating actively in 2005, an era when most Tolkien fanworks were being posted on small single-fandom archives. In 2007, I opened my own Tolkien fanworks archive, the Silmarillion Writers' Guild (or SWG), which remains online and active today. During the past two decades, I watched independent archives flourish, then fade, going from essential institutions within the Tolkien fandom to nearly nonexistent and (for new fans) hard to even imagine once existing. What happened? What caused the rise and demise of these sites and—in keeping with the conference theme of reengaging and reinventing fandom spaces—what do we need to rearchive, that is rebuild these sites for fans who want and need them?

To begin to understand the bigger-picture history of independent archives, I began about a year ago collecting data on these archives. These data form the crux of my poster for today's presentation. Primarily using Fanlore and the Wayback Machine, I collected data about when an archive opened, how long it remained active—defined as fanworks being posted to it—and when it went offline. If all of these data points were available, I included an archive if it was single-fandom, multi-creator, and English-language. For today's presentation, I was able to review all archives listed on Fanlore for the Tolkien and Harry Potter fandoms. Using these data, I created a series of stacked timelines, one for each archive, showing the status of the archive across the years and aligned with historical moments within fandom.

The best way to interpret the data is to begin zoomed out, noticing what trends you observe in how the different colors are clustered and aligned. The top timeline shows hand-coded archives, meaning that all processes of preparing a fanwork for archiving were done manually by volunteers. The bottom timeline shows automated archives, which use scripts to perform some of the steps of archiving. You will notice right away that the second set of timelines is much more red, meaning that automated archives show a much higher activity level than hand-coded archives do.

You might also notice that both timelines show that the bulk of archive creation occurred across rather limited timespans. For hand-coded archives, this was 2001-2003; for automated archives, this was 2002-2005. In fact, there is a noticeable shift away from hand-coded archives, with the last of them opening in 2004, at least among those eligible for the study. The likely explanation here was the availability, in 2004, of the eFiction open-source script, which allowed fans to easily set up an automated archive on their own web server.

But what led to the mass creation in those five years of independent archives? These archives are typically seen as arising from a fan community; however, the data show this is not always the case. As the fandom timeline shows, some can be explained by conflict within the fandom. Actions by large sites like Fanfiction.net or LiveJournal to censor fanworks provoke fans to build archives, as can conflicts within the fandom that lead groups of fans to split into their own communities. But this, too, is not the full picture. Instead, archive creation in these two fandoms appears to be motivated by high levels of excitement in the fandom and availability and empowerment to use technical tools. Archive-building occurred at a moment in history when home internet access was expanding rapidly and Web 2.0 made it possible to interact online without a high level of technical skill. Furthermore, the single best predictor of an archive's survival is whether or not it is automated, shifting labor more equitably from one or a few volunteers to the community members who use the archive.

Likewise, the flip side of the coin—archive inactivity and closure—points to the same trends. 15% of the archives included on the automated archive timeline have closed or gone inactive just in the past two years.

If the rise of automated archives can be explained by the availability of eFiction, so can their demise. The eFiction script last received a major update in 2015, and while several developers have since tried to take over and reignite the project, none have so far been successful in doing so. As PHP, the coding language eFiction is written in, evolves, the scripts have become deprecated, and the site stops working. This happened to us on the SWG, and while we had the community backing to undertake a labor-intensive migration to a new software platform, changes to Yahoo! Groups and LiveJournal occurring at the same time cut off that community support from many sites. It was easier to simply close.

So what does it take to re-archive? It perhaps seems oversimplified to say that the tools to do so must simply exist—scripts that can be easily and inexpensively used by fans with minimal technical training—but my data certainly points in that direction. While trends around consolidation onto large sites like the Archive of Our Own and Tumblr certainly play their part, small fan communities have never ceased to exist. In many cases, they simply lack the tools and platforms needed to exist.

Further Reading

Abigail De Kosnik, Rogue Archives: Digital Cultural Memory and Media Fandom, MIT Press, 2016.

Dawn Walls-Thumma,  A Fanworks Ecumenopolis: Tolkien Fanfiction Archives and the Implications of Consolidation, FSNNA 2022 presentation.

Dawn Walls-Thumma, Independent Archive Survey, 2023. (Note that this was an informal survey and did not undergo IRB review for human subjects research. Hence, I do not cite it or the data in my poster or presentation, but some of the data is interesting in the context of this topic.)

Synonymous (Dreamwidth) documents the issues with the OTW/AO3 that have led to the interest in 2023 in re-archiving.

Play with the Data!

I encourage other fans and researchers to build on my data! Credit me as Dawn Walls-Thumma if you use what I have started. I'd also love if you'd let me know what you make and discover.

View the Data | Play with the Data