An Archive of Their Own (CHI 2016)

For years, I’ve been bringing up the fan fiction site Archive of Our Own (AO3) to folks in the HCI community, as a cool example of two things: (1) an amazingly successful open source project designed and built mostly by women; and (2) thoughtful incorporation of existing community norms into design. That last one in particular came up a fair amount as I was doing the last batch of interviews for my dissertation. So last summer as I transitioned from PhD (under the advisement of Amy Bruckman at Georgia Tech) to my new job at CU, with the help of undergraduate research assistant Shannon Morrison (coming soon to a TBA graduate program!), we took a more purposeful look at that exact issue. How did these design decisions come to be, and what makes AO3 so successful? Are there lessons to be learned for how we can build social norms into technology design? And as we unpacked these issues, what we found was that an underlying commitment to core feminist values (like agency, inclusivity, diversity, empowerment) were part of this picture – it turns out that AO3 is an amazing case study of feminist HCI in action.

The TL;DR explanation for AO3 and its history is that it sprung up in response to other online communities (notably Livejournal and Fanlib) feeling unwelcoming or exploitative of the fan fiction community. (In the paper I briefly describe Fanlib as problematic because “a group of men created a website intended to monetize content that a community of mostly women had been sharing amongst themselves for free.”) The community galvanized to create the Organization for Transformative Works (disclaimer: I have been a part of the legal committee since 2009 but my research is independent of them), a nonprofit dedicated to preserving the rights and history of fan creators that would also be capable of supporting a fan-run fan fiction archive.

The blog post that sparked the whole thing points out: “I know we have project managers in our community – and coders and designers – can’t we do this? Seriously – we can come up with a site that would be miles better and more attractive to fanfic writers/readers than anything else out there, guys, because we actually USE the stuff.” The rallying cry became “we own the servers!” But this “archive of their own” also meant that they could design it to precisely fit the needs and values of their community. How successful has it been? Well, just in the time since this paper was submitted (August) and when I finalized the camera-ready version (January) I had to change “over 650k users and nearly 2 million works” to “nearly 750k users and over 2 million works.”

For this study, we interviewed 28 people – 6 developers/designers, 5 OTW staffers, and 17 archive users. Amazingly, my recruitment post for interview participants went slightly viral on Tumblr, shared over 700 times. I’ve never had to turn down interview volunteers before. And conducting interviews was a joy. I’ve been studying fan creators for over ten years now (since my masters program before law school!) and I continue to be amazed by their openness and insight. We talked to AO3 users about their history in fandom communities, their experiences with AO3, and how they compare and contrast, especially with respect to design features and values reflected. We were also lucky enough to talk to developers involved in the early days of the archive, including author (and fan, and computer scientist) Naomi Novik,* who was the major force in getting AO3 off the ground.

Feminist HCI (as articulated by Shaowen Bardzell in her 2010 CHI paper) as a framework emerged organically as we analyzed the interview data. We didn’t ask our participants about feminism, and yet there were all the values markers, as they talked about things like the importance of participation and ownership by the community, accessibility, inclusivity, advocacy, and nuanced handling of identity. You can read about all of these in more detail in the paper, but in sum, AO3 designers felt that integrating community values was critical to the design of the archive, and users named these same things as ways in which AO3 improves over some other online spaces. Both these values themselves and the careful way in which they were considered tracks well to the tenets of feminist HCI.

One of the most interesting things that emerged, though, were the tensions that exist in incorporating values into design. What happens when these values are at odds with each other?  We found that some of the design decisions were made to mitigate these tensions–for example, between a value of preserving history and a value of user control (resulting in the ability to “orphan” works), and a value of inclusivity versus safety (resulting in the content warning system). We also found examples of designing to influence values (particularly around remixing and permission), and how this can be tricky.

There is so much more here than could be written about in a 10-page paper (legitimate peripheral participation! norm migration! policy tensions!), and I’m really looking forward to delving even more into these online spaces in the future. In the meantime, I’m grateful to all of our interview participants, and to the community of fanworks creators generally for being awesome. If you’re a member of this community and want to know about future studies, please feel free to follow me on Tumblr or Twitter!

For more detail, see the full paper. I presented this work at the conference on computer-human interaction (CHI) in San Jose, CA in May 2016. I’m also pleased and honored that this paper received a Best Paper Honorable Mention!

Along with my other work on social norms and remix communities, this project was made possible by funding from the National Science Foundation, including allowing us to bring in Shannon as an undergraduate research assistant!

Citation
Fiesler, C., Morrison, S., and Bruckman, A.S. An Archive of Their Own: A Case Study of Feminist HCI and Values in Design. CHI ’16: Proceedings of the ACM Conference on Human Factors in Computing Systems, (2016).

* You’ll notice in the paper that we used real names of participants when requested (as appropriate). You can find some thoughts about and reasoning behind this practice in a chapter (written with Amy Bruckman and Kurt Luther) in the recently published Digital Research Confidential.