Federated Conferences
A Possible Future for the Academic Conference?
For several years now, I’ve been toying with the seed of an idea about the future of academic conferences based on distributing them into regional subconferences. I don’t claim that this idea is very novel, or even that it is my own idea in the first place; I have heard so many variants from other people over the years that I can’t even trace its lineage. In fact, earlier this year at the SIGCHI Futures Summit 2024 in Milan, there was a fair bit of discussion along these lines. But the other day, I had a flash of insight for how to think about this concept that I think is worth developing further: the federated conference.
Let me first say that these are my own thoughts and not representative of the opinions of the SIGCHI Executive Committee or the IEEE Visualization Steering Committee (both which I happen to be part of). The SIGCHI and VIS committees are hard at work thinking about the future of conferences, but this post is not a reflection of either body.
To discuss this topic properly, I need to establish a few things. What is an academic conference, anyway? Quite simply, it is a meeting of academics for discussing a specific research area. As far as I can tell, most academic fields have them. In the field of computer science, which I am a part of, conferences also happen to be the main archival publication outlet, ostensibly because the review process for an annual conference moves faster than any academic journal. This last point bears on the topic of hand, but it is not central to my argument.
More important is to consider why people go to conferences in the first place. After all, most academics advance their careers by publishing papers in academic journals, which by definition are asynchronous communications that are geographically distributed to its readers (formerly in print, now mostly online) rather than require people to gather in the same physical place. Why do we even need an annual physical gathering when these periodical letters serve to keep everyone informed of new results?
The answer, of course, is that a scientific community, regardless of its lofty goals of advancing knowledge, is still a community of people, and people are social beings. They thrive on meeting physically to discuss new ideas and present old ones (as well as make new friends and hang out with old ones). There are many tangible and intangible benefits to regular professional meetings that academic conferences currently fill.
Of course, the COVID-19 pandemic showed us that when physically meeting up in the same place is impossible, we can still use the Internet to achieve more or less the same effect. For one or two years after that fraught spring of 2020, several of the conferences in my community held a virtual conference entirely online facilitated by video streaming (Zoom and Youtube), synchronous chat servers (Slack and Discord), as well as various online networking services (such as Gather and Shindig). Even after the worst of the pandemic has faded into not-so-distant memory, many of these conferences still support hybrid participation — where some participants are remote and some are in-person — with varying success.
Better minds than I have spent a lot of time analyzing the relative merits of virtual and hybrid conferences, but I can only report on my personal experiences of remote conference participation: it does not work. It seems my fleeting attention span can only be fully captured by going somewhere and physically meeting with my colleagues. If I am forced to live the conference experience through my Zoom and browser window, the distractions of email, Internet, and Overleaf are too readily at hand to be ignored. And this is not even starting on the myriad distractions of domestic life still present when attending a conference from the comfort of your own home: laundry and cleaning, childcare and after-school pickups, dinners and faculty meetings. I am sure there are people that do remote conferences better than me, but I find them poor substitutes for the real thing.
Yet, obviously there are many positive benefits of a virtual — or at least hybrid — conference. For one thing, reducing travel is key to combating climate change. It does not really make good sense to have thousands of researchers from around the world converge on a single geographic location several times a year. It is becoming increasingly difficult to motivate a net gain in scientific knowledge from this massive carbon dioxide footprint when we already have perfectly adequate academic journals to facilitate scientific conversation. Furthermore, for some people, travel is just not possible: perhaps the costs are too high, perhaps they have people at home they have to care for, or perhaps they don’t have the physical ability or endurance for long travel. It is pretty clear that promoting remote participation in conferences is the right thing to do.
Let’s summarize where we are: remote conferences kinda suck (at least for me), but sooner or later we’re going to have to make them work. So, what are we to do?
Enter the federated conference (or the fedirence, fediconf, or distributed conference, if you will). Note that the term “federated conference” traditionally has referred to a joint conference combining many different conferences happening in the same time and place; the ACM Federated Computing Research Conference is one example. Such “federated conferences” provide basic infrastructure for registration, websites, and venues that help the organization of many smaller conferences. However, this is not the meaning of the term “federated” I am going for.
Rather, here I am using the term in the same way that federated social networks (or the “fediverse”) uses it: as a collection of separate communities that communicate with each other in real-time.
Federation is a computer science concept signifying a group of platforms that follow the same operating principles and act as a collective. The concept has become quite the thing in recent years. Part of the reason is that it prevents fragmentation, where different platforms, each with their idiosyncratic protocols and practices, splits users across isolated “islands” with no mutual communication. As such, it is an important topic in computing in general, and while the idea has been around since the early 2000s, it was mostly the decline of the microblogging platform Twitter and the subsequent rise of its federated equivalent Mastodon that popularized it (at least for me).
In other words, what I am saying is that a conference could be federated as a set of smaller conferences happening in different places and connected by the Internet rather than as a single centralized one in a single location.
A federated conference is a physical event happening at the same time but not the same place. Instead of a single conference location, the conference is split into several regional sites. These regional sites would be carefully selected to require minimal travel (preferably via rail or at least ground transportation) for the majority of attendees. The sites would correspond to federated servers in a federated social network such as Mastodon; they serve a local community with all the benefits of a physical in-person conference, and they connect and communicate with each other using the Internet. The communication should be designed so that there is minimal difference between communicating with someone on your own regional site as with someone at a different site.
Exactly what this communication looks like is where the magic has to happen for a federated conference to be successful. Regional sites should likely have their programs based on their own time zones (another downside of a traditional conference for long distance travelers), but would still need to have some synchronized events for plenaries and keynote speakers. There should likely be some kind of duplication of sessions so that people at different regional sites can still catch a talk at a convenient time. Viewing parties of recorded content from other sites should likely be a prominent schedule item.
As it happens, the federation infrastructure would also enable those individuals who can’t travel at all, not even to regional sites, to participate. This is similar to how you can run your very own Mastodon server with a single user that can still federate with the rest of the fediverse.
At the same time, the reliance on internet technologies for conference activities that don’t translate well to online delivery is mostly removed. Networking and social events such as coffee breaks, receptions, and banquets, for example, can be done locally at each regional site and don’t have to be federated. Yes, it’s true that splitting the community between geographical regions will necessarily reduce socializing and networking across continents. Perhaps it might cause scientific communities to become a little more provincial and less international. Perhaps some ideas and collaborations will never happen because the right set of people never meet in person.
On the other hand, perhaps we need to reduce the cult of personalities in science anyway; maybe it should be more about someone discussing a topic rather than a specific set of people discussing it? Perhaps we want each region to cultivate the specific strengths and individualities they possess? Besides, there is nothing stopping individuals that have the capacity to travel to a more distant regional site than the one closest to them. We just want to eliminate travel for all participants.
There are many technological and social hurdles we have to overcome to make this idea a reality and I haven’t even begun to tackle them in this thought experiment. However, there are examples to look to that can be inspirational. For example, the ACM Creativity & Cognition conference is held as a virtual event every other year, and an in-person one during other years. The virtual versions are held in Gather Town, a game-like virtual space with an 8-bit visual aesthetic. Some conferences, such as ACM CHI, have experimented with remotely operated telepresence robots. Some conferences have made hybrid participation a key part of their identity. There are many models, but as with any federated service, the key is to settle on a common and interoperable set of standards to facilitate each regional site communicating with the others in a seamless and well-designed fashion.
What do you think is the future of the academic conference? I would love to hear any feedback on this idea, or how it can be improved. Federation can perhaps be part of the solution, but it is a niche technical term, and I am eager to hear if there are better alternatives out there.
