Online postings and social networks offer one of the few access points for researchers studying secretive, violent, armed groups. Groups can use online platforms to recruit members, rally public support, and broadcast propaganda. But not all online spaces are created equal—some are more private than others—and online data is not always the most useful place to learn about violent acts planned by an armed group. The Oath Keepers in the United States and the Islamic State, with its base in the Levant and its global aspirations, operate in vastly different contexts, and with very different goals. But researchers benefit enormously from comparing the research techniques they apply to the online activities of armed groups. How can we most effectively study violent armed groups online, and with what limits?

Sam Jackson: Digital media provides a great source of data for studying hate groups or militias (or just about any other social or political actor), depending on what we want to learn about those groups. Anyone who uses the Internet uses different digital spaces for different purposes. Some social media platforms are particularly useful for finding new people who share your thoughts, concerns, or offline location. Chat platforms and messenger apps let people talk with existing contacts (or contacts of contacts) in a less public space. File-sharing sites can help people share content that they wish to be more durable rather than transient (for example, a manifesto versus commentary on breaking news).

Scholars of digital media point out that the affordances of different digital platforms shape and constrain what users do on those platforms. The fact that Telegram is encrypted, for example, shapes how it is typically used: it’s safer from surveillance (and research) but it’s not terribly effective for reaching new audiences.

But it’s important to recognize that we might not learn about what we want to learn about if we look in the wrong place. For example, if we want to know if hate groups are planning criminal activity, we’re not likely to find that activity on Facebook, but we might be more likely to find it on Telegram or Discord or one of the “chans.” On the other hand, if we want to understand the public presentation that groups share to build their public reputation, Facebook might be a good site for research (assuming that the group hasn’t been banned from the platform).

Propaganda versus Planning

Aron Lund: Great points, Sam. As for me, I’m a bit of a Luddite and I have long since given up on keeping pace with the Internet, which means I’ll probably never learn what a Discord is. I have nevertheless done quite a lot of online research when studying transnational jihadists and militias in the Middle East, looking at Twitter, Facebook, and a few encrypted apps like Telegram. Although my Middle Eastern militants are different from the people you follow, much of what you say about online research feels familiar.

There also seem to be similarities in how the online environment—social media dynamics, the competition for visibility, and all that—affect their own behavior and relationships. Social media produces ostentatious narcissism, self-radicalization, and in-group bubbles, and nowhere more so than among people who are already on the crazy fringe to begin with. Look at the foreign fighters who joined the Islamic State: they’re basically al-Qaeda’s YouTube generation.

Your point about open communications being for public relations, not planning, brings to mind the freelance work I used to do for risk consultancy companies about a decade ago. This was before jihadism migrated to Twitter and Telegram, and they wanted me to read al-Qaeda-linked Arabic online forums and search for chatter about specific individuals or companies. There was never any of that. These forums were incredibly boring. A typical discussion thread would start with a statement by al-Qaeda or al-Shabaab or some other group, and then you’d have several pages of reply guys with handles like “mujahed___999” writing “thank you brother” or “allahu akbar.”

So, while that kind of online research was great for keeping up to date with jihadist propaganda, it was almost useless for learning secrets. The real inner workings of these groups are offline or in encrypted environments that we’re not going to have easy access to as researchers or journalists.

Sam: Absolutely! So much online activity is brief or mundane, especially when you start thinking about rank-and-file members and supporters of these types of groups. If what we care about is the juicy stuff—or more formally, if we care about online data that can tell us something about either the ideologies or activities of these groups—we have to be really thoughtful and deliberate about what kind of online data might be useful.

It’s interesting to think about some of the consequences of social media that you identify for people in these groups and movements. It seems plausible that some of those dynamics also happen in offline “total institutions” or other offline communities where participants overwhelmingly interact with like-minded people. In both these online spaces and offline total institutions that are characterized by militancy or extremism (following J.M. Berger’s definition of the term), participants might find themselves in “echo chambers of awful” where they are surrounded by dangerous speech about some enemy or other that dehumanizes them and rhetorically establishes a reality characterized by a battle between good and evil. Perhaps this should draw our attention to parallels between certain online spaces and certain offline spaces. Some online spaces are, in practice, more public than others, just as in the physical, offline world, different public spaces offer differing degrees of privacy. And the type of behavior that happens in public online spaces might have more in common with behavior in public offline spaces than private online spaces.

Aron: I’m straying from the topic of online research now, but that is an interesting point that I think goes far beyond extremist or militant groups.

As far as the public–private distinction: isn’t the thing that an online environment can function both as a public and a private space, simultaneously? An open and abundant information environment will organically lead to the creation of self-contained clusters of people with weird, overwrought ideas and paranoid mindsets. Information is now so easy to access that you’ll quickly find whatever you’re looking for. Self-selection goes haywire.

If you read the morning paper, on paper, it will have been curated by editors who weren’t just thinking of pleasing you. They will have had some kind of broad audience in mind, which means you’ll run into a lot of news and views that you wouldn’t otherwise have sought out. But if you get all your news online, your entire information intake will be structured by your own preferences, your network, and your interests. Your favorite political stimuli will constantly be available at the click of a button, and you’ll be cheered on by the like-minded. So echo chambers start to form, despite the objectively open and diverse nature of the online space.

The other aspect of an open, abundant information environment is that the political enthusiasts and entrepreneurs who want you to engage with them—whether for the ad clicks or to shape your views—need to stand out against a very busy background. They need to grab your attention. And you will also want to be noticed, because otherwise no one will share your posts and giggle at your dumb memes. In political terms, the way to do that is to lean to the extreme, pick fights, and court controversy. So everyone starts competing for attention, and it creates spirals of radicalization.

I guess none of that is news to anyone. That’s how social media works for all of us. But it seems like the same dynamic is playing out inside ideological movements online, and in the constituencies they court for recruitment. Ideological purity, action, and brutality pay by drawing attention and giving you online reach. Pragmatism and gradualism are boring, and therefore ineffective.

Learn More About Century International

Online Ramps to Offline “Total Institutions”

Aron: I really think this allure of the extreme in online spaces was key in determining the political and communication culture of the Islamic State. When it broke off from al-Qaeda, most of the young Western generation of jihadists went with the Islamic State, which was basically a gory clickbait version of al-Qaeda. Its political posture related to old school Ayman al-Zawahiri-type jihadism like Buzzfeed relates to The New York Times.

That dynamic seems like a huge part of why extremism thrives today—be it jihadism, far right or left, or QAnon conspiracism. The Internet offers a social environment perfectly tailored for polarization and radicalization, leading individuals into extremist networks and then continuing among and inside the networks themselves. I suspect that structural effect is much more important for extremist recruitment than the easy availability of encrypted communications and private spaces. But of course, they have that, too.

Still, the Islamic State example also shows the limits of looking at something from an online angle. Most of its foot soldiers weren’t recruited online or self-brainwashed by social media. They came to the group through the civil wars in Syria and Iraq. It’s a story of rebellion, conflict trauma, warlordism, tribal politics, and all manner of hyperlocal issues, and we can learn very little about it online. As for the Islamic State’s senior leaders, they emerged from precisely the type of closed offline “total institutions” you were talking about: from clandestine work in the Iraqi resistance, and from Abu Ghraib, Camp Bucca, and Sednaya—prisons for jihadist captives, which turned into hotbeds for their production of ideology and networking.

Sam: I think there’s a really important point that’s implicit in what you’re describing here. Perhaps we should expect to see some online activity that is more about feeding an appetite for outrage and some that is more about getting people to make dramatic changes in their lives to join an illicit organization. Perhaps the feeding of outrage can set the stage for successfully encouraging someone to join an illicit group, but it’s unlikely that most people who feed their outrage—or even help feed other people’s outrage—will get involved beyond propaganda. (I hope that doesn’t come across as suggesting that propaganda doesn’t matter—it absolutely does, even beyond direct connections between propaganda and illicit offline activity.)

In a sense, the propaganda and recruitment side of online activity is a numbers game. The analyst J. M. Berger developed a semi-facetious Berger’s Rule of Radicalization: “On average, 15 percent of any given population are assholes, but considerably less than 1 percent are violent assholes.” If the end goal for a militant group is to increase the number of participants in illicit activity (in other words, make more violent assholes), one way to do that might be to just increase the number of people who believe the ideas promoted by the militant group (in other words, make more assholes).

“I do think some types of extremists are more than just users of social media—they’re products of it. QAnon did not exist before social media, and could not have.”

This also reminds me of a satirical piece that Seamus Hughes wrote a few years ago about fax machine radicalization. Hughes’ fictional analysis describes fax machines under children’s beds as warning signs of radicalization. The joke cautions against treating technology itself as the driving force behind extremism. As important as it is to study the mediums used by militant groups, we need to be careful to avoid fixating on that medium. Social media can be an important tool for these actors, but these types of groups existed before social media and will exist after it.

Aron: Ha! Well, I agree, and I don’t. The Hughes article is funny, and it makes a good point. Officialdom has a way of being baffled by ostensibly new threats that does lend itself to parody. But here’s the disagreement: I do think some types of extremists are more than just users of social media—they’re products of it. Even if the ideological vocabulary stays the same, there’s a difference in the way these movements operate and who joins them. I made a related point about the Islamic State versus al-Qaeda earlier. A clearer example, although it’s a very different one, would be the QAnon cult. It did not exist before social media, and could not have.

The Berger rule seems accurate! But increasing the number of believers, or assholes, is a big task, especially if you’re a small group. Let’s also recall that many extremist groups aren’t just interested in swelling the pool of the like-minded. They also want to be the organization leading it all. Whether in fragmented insurgencies and civil wars, like in Syria, or among extremists online, be they jihadists or left- or right-wingers, there’s often competition among groups that should, logically speaking, be allies.

Online politics seems to be naturally competitive. Alternatives are easy to find and switch between, so it devolves into a race for attention. I think that might help explain why modern organized terrorism so often seems introverted and like it is catering to like-minded communities, with very little effort to engage wider society. The Islamic State’s gore videos, for example, are a type of performative hyper-radicalism that plays well with sectarian extremists but is terrible if you want to appeal to normal Muslims. But the fact that this brutal posturing repels the normies is not a problem if your target audience is the sectarian fringe, and it is often an advantage: the shock value creates controversy, gets you the media coverage you’re looking for, and builds the brand.

This dialogue is part of “Transnational Trends in Citizenship: Authoritarianism and the Emerging Global Culture of Resistance,” a TCF project supported by the Carnegie Corporation of New York and the Open Society Foundations.