Announcing the D/ARC!

By PS Berge

🐦 @theiceberge       📧 hello@psberge.com

The D/ARC Discord server is now live! If you want to join, and/or help us grow, use the https://darcmode.org/invite link to access the server, and don’t forget to share with your colleagues who might be interested!) 💖

Over the past year, I’ve been lucky enough to connect with a handful of Discord researchers. Each time, without fail, we have the same exchange—one of us shakes our head and says “yeah… Discord is critically understudied.”

Goodness, what an understatement!

As of this writing, Discord is arguably one of the most rapidly expanding social media platforms. It’s increasingly becoming a go-to tool for remote teaching and academic spaces (hell, I’ve had to join almost 10 different conference and academic servers in this last year alone!). While some recent studies have evaluating Discord’s pedagogical success as a classroom management tool (e.g. Arifianto & Izzudin, 2021), relatively little scholarship has tackled the networked implications of Discord’s ecosystem, the platform’s culture, or the practices of users.

Last month, my collaborative publication with Daniel Heslep, "Mapping Discord’s Darkside," came out in New Media and Society. Our study examined the metadata of 2,741 Discord communities as hosted on server bulletin site Disboard. We are especially glad to finally be sharing this work because this study is, to our knowledge, one of the first longitudinal studies of hate networks on Discord. But while our study is, we hope, a useful model for discussing hate networks on the platform—it is only a small glimpse at the larger breadth of Discord communities. There remains so much work to be done in examining Discord through the lens of technology and culture. But we can’t do much without established methods, usable tools, or the support of community.

And so it’s in response to this urgent gap in knowledge and resources that we’re thrilled to announce the Discord Academic Research Community, an inclusive and supportive network for anyone researching Discord and Discord-related communities, platforms, or culture. We’re taking inspiration from many of the incredible research networks already out there (a big 💜 to the TikTok Cultures Research Network and the Meme Studies Research Network) and hoping to build a meaningful space for community, resource-sharing, and tool-making.

What is the D/ARC?

First and foremost, it’s a Discord server for Discord researchers. When you join the server, you’ll gain access to our community resources. If you complete a short verification (just so we know you’re not a bot) you can also participate in discussions. We’re only getting started, but already our community maintains a number of resources, including:

• A curated Zotero library of all extant research we have gathered on Discord as well as a comprehensive list of resources (links to important blogs, scrapers and repositories, videos, historical wikis, and tools for Internet research).
• Multiple channels for discussing ongoing changes to Discord (new technical features, headlines and drama, etc.), relevant CFPs, and new research. We’re also coordinating some live workshops and events led by our moderators and guests!
• A Google Colab notebook which contains the modified scraper code (with some improvements) that Dan and I used to collect Discord server data in our NMS study as well as a handful of analysis tools. I have also uploaded the full GitHub repository and a post to the D/ARC Blog about working with Disboard data!

We are sharing these resources as a starting place for current and aspiring Discord researchers in the hope that people will take, adapt, and improve our methods-and ultimately begin to close the research gap on Discord as a platform. We encourage you to use, share, teach, and remix our tools and resources!

Who’s Flying This Thing?!

The D/ARC is run by a handful of dedicated graduate students and junior scholars who believe in building community and open-source tools for Internet research. Our current Moderator team includes:

• PS Berge, Doctoral Student @ University of Central Florida
• Rachel Berryman, Doctoral Candidate @ Curtin University
• Hibby Thach, Graduate Student @ University of Illinois Chicago
• Nick-Brie Guarriello, University of Minnesota and currently on the Job Market! (Please hire them 💖)!
• Daniel Heslep, Graduate Student @ University of Alabama

The mod team has been hard at work setting up the server for launch, and we’ll hope you’ll join us and share the invitation (https://darcmode.org/invite)! The moderators are also working to coordinate events both in-server and at forthcoming conferences. If you need to reach the mod team, you can email us at mods.darc@gmail.com!

Breaking Ground: The Obstacles of Studying Discord

As we launch the D/ARC, and begin building new tools and methods for critical Discord research, it’s going to be increasingly important that we know what we’re up against. There are no shortage of answers as to why Discord remains an understudied platform. It is undoubtedly a combination of factors, some unique to the platform and some shared by other social networks. Below, I characterize a few of these most significant obstacles that Discord scholars are facing. As important as it is to share tools and resources, we also require wary attention to the challenges ahead:

Novelty

Discord is a relatively nascent platform (founded in 2015)—and pivoted quite quickly from being a niche, gaming-centric service to a widespread tool used by educators, artists, entertainers, public figures, and companies. Much of the earliest research on Discord (such as Brown Jr. & Hennis, 2019) identifies Discord as a platform exclusively for gamers. Yet, as Discord’s marketing shows, the platform is in the midst of an extensive rebranding to broaden its userbase and this has been accompanied by an expansion of Discord’s corporate structures. This undertaking has won them a lot of attention—interviews on NPR, numerous headlines, and a several-billion-dollar valuation from Microsoft (in an offer which Discord turned down). For many, Discord has only recently emerged as worthy of examination as a mainstream social media platform. While this means that more eyes are on Discord and new attention is being placed on the platform, we have also noted that Discord’s legacy as a gaming-centric space has alienated some researchers.

Along with this comes the difficulty of studying a still-emerging platform. For example, in the time it took our study to finish peer review and editing, several of Discord’s services had already changed significantly: the 7000 member requirement for Discovery had been lowered to 1000, and the Live Stages feature had been launched and sunsetted before we even had time to address it! As Discord shuffles its public rhetoric, its audience, and adapts its features, it’s going to be important that researchers are equipped to keep up with the rapid changes of Discord’s evolutions. For this reason, one of the key functions of D/ARC is to maintain collective attention and note the ever-shifting dynamics of the platform.

Few Extant Tools

The better the tools to study a platform, the more that platform will be researched. It’s a common issue in internet studies, and a sentiment often used to point out how sites like Twitter are ‘over-researched’ compared to other social media platforms. While social science and digital humanist tools integrate easily with Twitter, Instagram, YouTube, other mainstream social networks—chat-based applications are less often supported. Discord, as a strange blend between social platform and chat application, has few publicly known tools for analysis and a limited canon for established methods or approaches.

It’s our hope that D/ARC can begin to alleviate this—building, modifying, and circulating new tools for analysis. As a small part of this, I am providing an open source, guided and code-free version of my own Disboard scraper. But I look forward to new collaborations and further tools operating within Discord’s ecosystem.

Closed Systems

Perhaps the most significant obstacle to Discord research is the closed nature of servers themselves. According to Discord’s usage statistics, there are over 19 million active communities on the platform, yet a vast majority of these are not public-facing. Most are private, ‘invite only’ servers that can only be accessed via a direct link. As such, it is difficult to analyze behavior and culture on Discord at scale. Although Discord servers can designate themselves as ‘community servers’ with a permanent, public invite link—there are some additional complications:

• An invite link is still required, even to join ‘public’ community servers (blurring the line between public and private spaces).
• Community servers must have, as of January 2022, at least 1000 members in order to qualify for Discord’s ‘Discovery’ feature (which makes them searchable within the platform).

This creates a number of difficulties for internet scholars. Because these communities require the user to be logged in to Discord and receive an ‘invite link’ (even if that invite is publicly available), many researchers will need to get approval from a research board to conduct studies within any Discord server. For broader studies, this often means getting IRB approval for every community that the researcher intends to investigate. Public servers (such as those in the Discovery menu) and community servers with links online remain a grey area in academic study, and I’ve heard from colleagues at several institutions who received entirely different guidance on whether to treat such Discord servers as private or public for the purposes of collecting data. Arguably, this is a space where we need more specific clarification from the academy, internet studies, and review boards in how we approach semi-private networks like Discord.

Discord also creates troubling precarity for researchers investigating hostile networks, predatory behavior, and hate groups: not only must the researcher join a server as a public member to view its content and activity but—as we note in our study—hate-oriented servers often set up security features including ID-checks, raid-requirements, and even skin-tone verification (yep, that really happens.) before allowing participants to view certain channels. This means that researchers must expose and potentially implicate themselves to gain access to these communities. Although there are a handful of scrapers that allow for gathering of text/image data from Discord text channels, in many cases, getting access to the text in the first place requires the researcher to expose themselves or engage the community directly. (Not to mention, scraping Discord data directly technically violates the site’s Terms of Service).

Our study circumnavigates many of these difficulties by examining the irrefutably public metadata of Discord servers. But while this is a useful angle (and one we hope more scholars will make use of) it is not a complete solution. We must find a way to approach Discord both from a longitudinal, networked perspective and through embedded and ethnographic approaches that recognize the nuance of activity within communities. Yet this assumes that researchers can even find these communities in the first place. The 1000 member minimum to qualify for Discord’s discovery feature (which must be applied for) means that what Discord provides is, at best, a cursory glimpse of only the most popular servers on the platform. Dan and I note this ‘iceberg effect’ in our study—Discord is able to curate an illusory public of partnered and verified servers while obscuring the activity of less-popular communities.

Although my approach to Discord has involved more structural and computational interventions, D/ARC is looking to expand this possibility space. A number of our moderators have interests and specializations in ethnographic methods, audience studies, and interview-based research. I must plug, for example, Nick-Brie Guarriello’s new article that interviews Discord users in videogame livestreaming communities. Together, we’re already developing new projects and approaches that can account for the networked realities of the platform.

Proceed With Caution

It’s important that I clarify here: I love Discord. I use it both professionally and personally. Originally, Dan and my research interests in Discord came from examining progressive activism and the way that social organizers could mobilize towards justice using the platform. Indeed, queer-support, indie gaming, sex work, activist, educational, and language-learning communities are all important functions of Discord’s ecology. I’m thrilled with (and hope we will see more of) the scholarly explorations of Discord’s value in these spaces. Yet it is important to note that, despite the cutesy-Wumpus-marketing, Discord is not removed from, and remains dangerously rooted in, toxic technoculture. This poses a double-edged sword for researchers. As we found in "Mapping Discord’s Darkside," marginalized communities exist under threat on Discord. We found complex networks of queerphobic and white supremacist communities whose emphasis on raids, attacks, and ‘purges’ indicated a chillingly familiar militarism harkening back to GamerGate.

To this end, I implore new Discord researchers to consider a twofold word of caution:

1. Take care of yourself. Don’t put your personal account in dangerous positions. Don’t risk being banned by the platform. Even if you are not joining hateful communities, be aware and be vigilant of raiders, scammers, and hate. If you are working with data from toxic communities, practice self care.

2. Take care of your research subjects. The semi-private nature of Discord can make it unfortunately easy to make someone a target. Don’t just look to protect individuals: protect their communities where possible. While Discord can be a wonderful refuge for marginalized people (especially POC, youth, sex workers, activists, queer, and trans folks) these people and their communities are not as protected as you may think by Discord’s semi-private structure. Be careful.

Community in D/ARC Times.

A big part of changing the landscape of Discord research is through community. We hope this can be a space to build and share together. Note that the D/ARC is open to all researchers, especially students, independent scholars, and junior faculty. To join the server, simply click the link below (or any of the other million links on this site, I’ve put it almost everywhere!!).

If you need additional help getting set up, don’t forget to check out the how to join page! Likewise, if you or your research team need additional support in setting up tools for studying Discord, please don’t hesitate to contact us (@darc_mode or mods.darc@gmail.com).

We hope to see you in the D/ARC!