From Codes of Conduct to the Digital Covenant

Hello! I'm Robert W. Gehl, speaking to you from Toronto in Canada, where I am a professor in the department of Communication and Media Studies at York University. I'm delighted to have this presentation be part of the Behind The Scenes conversations on platform governance research.

This presentation is based on research I conducted for my forthcoming book, Move Slowly and Build Bridges: Mastodon, the Fediverse, and the Struggle for Democratic Social Media, which is set to be published from Oxford University Press later this year. The book focuses on alternative social media -- specifically, "the fediverse," a network of many small social media systems. The most notable fediverse system is Mastodon, a Twitter-like microblogging system.

My plan for today is to give a quick overview of the fediverse.

Then I will present materials from my book, predominantly from chapters 3 and 4, which deal with codes of conduct and defederation decisions.

The final third of the presentation will be some "behind the scenes" looks at how I think about consent in my research as well as how I publicly share my research process.

Let's start with some background on the fediverse. To understand the fediverse, let's start with what seems to be an ancient digital technology...

...email. Email is a federated system.

Email resources appear on multiple computer servers, which can communicate with one another. For example, let's say a friend of mine uses Gmail, another friend uses Microsoft Outlook, and I use Protonmail. Despite the fact that we're each using the services of very different companies, we can still communicate. This is because all email servers agree to abide by a set of technical rules, a protocol called SMTP (Send Mail Transfer Protocol.) Today, there is a means for social media data to be treated like email.

In 2018, five technologists working with the World Wide Web Consortium published an open standard called ActivityPub. ActivityPub provides a protocol for social media data, much like SMTP provides a protocol for interoperable email servers. It standardizes typical social media activites: making a profile, posting something, liking or favoriting that post, boosting a post, and commenting on posts. Let me walk through this by considering three distinct fediverse systems.

The top logo is that of PeerTube, which is server software that allows for people to post videos, comment on them, and share them. The logo to the left is for Pixelfed, which allows people to post images. And the logo to the right is Mastodon, which is a microblogging system. All of these servers can communicate with one another via ActivityPub. That means my friend could post a video on a PeerTube server, another friend could see it in Pixelfed and comment on it, and I can like and boost that post in Mastodon.

It's hard to overstate how radical a departure this is. You CANNOT do that with corporate social media. You cannot post to Facebook and expect a friend on X to comment on it.

You cannot even post to Instagram and expect a response from Facebook, even though they are both owned by the same company, Meta.

The result of ActivityPub has been a global network of heterogenous social media servers. The average size of each server is around 500 accounts, but the network itself is comprised of 10s of thousands of servers with somewhere between 8 to 14 million users.

So we have a network of interconnected social media servers. A major question is: what social or cultural factors play a role in how these servers are interconnected? After all, we're talking about platform governance. That brings me to a major focus of my book -- the concept of "the digital covenant." A way to understand the digital covenant is to start with what it is not.

Ok, so here's a tiny slice of a 7000 word document that few of us want to read: it's a Terms of Service agreement. When you see this, how many of you instinctively reach for the scroll on your mouse, or want to swipe the screen down, down, anything to get past this legalese? We scroll past these all the time.

Terms of Service agreements are a contractual relationship between you, the lonely you, and a vast Corporation, attempting to specify in advance every possible interaction you might have with that Corporation. This is not the practice on the fediverse.

On many fediverse instances, in contrast, a very different relationship is imagined. Consider server rules, such as these from Mastodon.art, a Mastodon instance. Compared to Terms of Service Agreements, server rules are easier to read. Instead of specifying every possible interaction, they provide a relatively simple set of deontological rules to abide by.

Server rules are often part of Codes of Conduct, used on many Mastodon instances. Codes of conducte are often developed by the communities themselves. As an example, I can point to AOIR.social, an instance I helped to set up, which has a code of conduct developed by myself, the othe admin, and the moderation team. In addition, the AoIR code of conduct is open to modification by AOIR members.

Instead of a relationship between lonely you and a corporation, you are part of -- dare I say this most dangerous term -- a *community* of people on a particular server. Moreover, your instance is only a little part of a big network of other servers or "instances," each with their own rules. The current, ActivityPub-powered fediverse emphasizes the Instance as the organizing principle. The federation functions largely as an instance-to-instance relationship as opposed to the user-to-service relationship that we see in corporate social media.

Codes of conduct at an instance level help shape connections between instances. The result is what Diana Zulli and I have called the "digital covenant" -- an agreement between instances who abide by similar rules.

In fact, the Mastodon project uses this language. The Mastodon server covenant is a key document here -- it's a requirement for any Mastodon instances that wish to be listed on the project homepage.

I want to answer two questions. First, how did we get such a covenant? There's actually a very long history of federated relationships in political philosophy. What I will talk about today however is the shorter history of covenants in free and open source technology groups.

A second question is about breaking connections. I keep talking about how instances might connect through shared values. But what about when they disagree? In fact, there are very good arguments for breaking connections -- I'll talk about them in a bit. In what follows, I'll draw on research I conducted for my forthcoming book.

Let's address the first question -- how did we get the covenantal fediverse? A key resource to understand this is Christina Dunbar-Hester's book,_ Hacking Diversity,_ which has documented diversity advocacy in the FOSS world.

And a key activist we can consider is Coraline Ada Ehmke, whom I've interviewed for this project. The struggles of activists such as Ehmke are documented in Dunbar-Hester's book.

Starting in the early 2010s, Ehmke and allies called for Codes of Conduct at tech conventions. This was in reaction to documented harassment of women and marginalized people at tech conventions, and observations that FOSS development was dominated by white men.

Ehmke and allies faced death threats, but succeeded in their fight. While it's not a tech convention, the Association of Internet Researchers ratified their own conference code of conduct not long ago as part of a broader wave of organizations doing so.

Ehmke's activism also led to the practice of online Codes of Conduct. This is important because FOSS projects are often developed across the internet, not just in-person at events or within organizations. There is now a ubiquitous document found on many FOSS projects, contribute.md -- Contributor Covenant.

This connects to Mastodon: Since it started in 2016, Mastodon emerges in this milieu.

It initially attracted many queer and trans developers who were invested in having codes of conduct in projects.

Not only did Mastodon adopt a Code for the development of the software in 2017. It also inaugurated a radical practice: instances themselves, where people would sign up and engage in social media, adopted Codes of Conduct. The earliest servers, including Mastodon.social and Awoo.space, helped establish this practice in 2017.

As instances with like-minded codes of conduct federate, it leads to the digital covenant -- ethical values shared across many small communities.

What emerges is what feminist philosopher Iris Marion Young might call “decentred diverse democratic federalism” -- a network of instances sharing a “thin set of global values”

Now, most of this analysis is Mastodon-centric, which makes sense since Mastodon is central to the story. But the fediverse is bigger than Mastodon. What about other projects?

Well, the Reddit-style fediverse software system Lemmy now regularly adopts codes of conduct. Bookwyrm, which is a federated alternative to GoodReads, goes even further, adopting not only codes of conduct, but an ethical source-style license called the anti-capitalist license, attempting to bake its values into its license. The use of codes of conduct have spread from Mastodon to many (though not all) other federated social media systems.

Recall my two questions from earlier. What about non-federation -- that is, denying federation? So far I've only focused on positive connections -- what about the negative, the dis-connection?

Well, codes of conduct matter here, too. Consider Awoo.space, one of the earliest Mastodon instances.

They use approve-list style federating, vetting every single connection made to their server. I interviewed the admin of Awoo, who told me that codes of conduct are central for that process. The allowlist approach is a case where disconnection is assumed from the start -- connections are opened slowly and purposely, rather than easily and then maintained over time.

And there is also the use of blocklist style defederation. This is where instances are listed as bad actors, and the list is imported by instance admins to sever connections between servers. A key example is Ro and The Bad Space. I interviewed Ro for my book, and he talked to me about the Bad Space as being born out of experiences of anti-blackness on the fediverse. Codes of conduct are important here, too -- but the emphasis might be more on Conduct, which is to say that being put on the Bad Space is largely a function of acting unethically on the fediverse.

And here we see another covenantal network emerging: The Bad Space isn't just run by Ro, but is based on the actions of a network of admins who have banded together. Through their individual blocking decisions, a collectively-developed list (or rather, series of gradated lists) has emerged. Those lists enable new instance admins to preemptively block unethical parts of the fediverse.

So let me conclude part 1 this way: Reading corporate social media terms of service is hard.

But agreeing to them is easy. Corporate terms of service are designed to be clicked past and forgotten, even though they hold incredible weight.

By contrast, Reading a Code of Conduct is easy. They tend to be short and very readable.

Building a covenantal network comprised of instance/communities who abide by a thin set of global values found across multiple Codes of Conduct is hard. But that's the point. Building societies is messy. Self-determination is messy. Ethical interactions are messy. Federated social media is messy and hard because it compels people to self-govern in ways that corporate social media simply do not demand. (Credit to Jim Brown for this contrast -- I saw him make a similar case at a conference.)

With that quick overview of my book behind us, let's talk about the "behind the scenes" aspect of my work.

As I've been indicating throughout part 1, much of this book is based on interviews. Here is a list of just some of them.

And here are a few more. I'm grateful to them, and frankly I'm proud that I was able to interview them.

The book is also based on participant observation. I've been on the fediverse since 2017 (arguably longer, but I will save that point for Q and A). I've spent years being a member of instances, but now I am helping to run an instance -- all of this gives me insight into how it all works. And the central lesson I've learned is:

Trust and consent have to be at the core of the project.

This is because of corporate social media history. We have seen unethical social media research and use of data -- there are many examples. Two notable ones being Cambridge Analytica and the Facebook emotional contagion study of 2015.

We've also seen FOSSBro and gamer harassment of marginalized people across the history of corporate social media.

So what I have tried to do with my work is to be as open as possible. My first, most dedicated start to the project began in 2020, with a blog called Fossacademic.tech. I have been blogging about my process in the open as a signal to people on the fediverse of who I am, my intentions, and how I'm thinking about my project.

If you look at my blog, I would draw your attention to key posts -- posts describing my interview process, how I use my funding to send money to fediverse artists, my critiques of unethical fediverse research, and posts that discuss the contents of my book. Everything I am writing and doing in this space -- including today's presentation -- is about a key audience:

A key audience in my mind is the people of the fediverse.

For example, recently I was asked to give a keynote at a research symposium at the University of Warwick. Based on my participant observation on the fediverse, I can say that this symposium was viewed with some suspicion by people on the fedi. The worry was about terms like "research" and "tool": * would the workshop take the fediverse as an object, rather than an equal participant in, research? * Would the outcome be tools that have no benefit to the fediverse -- that only benefit the researchers? * Would my giving the keynote be seen as endorsing such practices?

In reaction to these concerns, I drafted a presentation called "The Road to AoIR.social", and I posted the draft online ahead of time, sharing it with the fediverse. I also posted notes about the conference and shared them with the fediverse. And I offered to answer direct questions from fedi members about the Warwick symposium. I'm glad to say that the Warwick conference was, in my view, a wonderful exploration of ethical research practices.

In addition, I've begun sharing the full text of my fediverse-related presentations on my website. This one is included!

Whenever possible, I make sure my fediverse journal article research is open access.

And now I am part of a network of scholars who are thinking about ethical approaches to alternative social media research.

I am running out of time, so I will conclude here by saying I hope my book is not just about the fediverse, but also is FOR the fediverse. This is not to say it won't be critical, but then again, people on the fediverse are very critical, too! I hope the work reflects some of their perspectives.

Thank you!

Default Statcounter code for Homepage2 http://www.robertwgehl.org

End of Statcounter Code