By J.M. Berger
This article is an adapted and lightly edited excerpt from Online Extremism, a chapter in the forthcoming Routledge Handbook of Online Violent Extremism.
While the Internet can be and is used as a simple broadcasting tool for spreading ideas and ideologies to new audiences, its most powerful innovation was participation. Unlike radio or television, the Internet quickly established itself as a fully participatory medium. But not all participatory platforms are alike, and structural affordances vary wildly. Differing styles of participation create varied challenges for those seeking to combat extremism online. In this article, I will examine the major styles of online participation.
Closed Private Communities
Closed private communities are online spaces where audiences are strictly limited, such as direct messaging or small invitation-only chat groups. They can be found on platforms and protocols such as e-mail, SMS, Signal, or Slack, and in chat-oriented apps with wider functionality such as Discord.
Closed communities generally do not empower network discovery—meeting new people who share an interest—but rather serve as online spaces where previously established connections are deepened. In other words, you don’t typically wander into a closed private community; you agree to enter a controlled space comprised of people you already know and friends of friends.
Extremists use closed private communities for community building and to organise collective action. For example, the White nationalist Patriot Front has organised postering and flyer drops on Discord, while ISIS notoriously uses a number of private chat services for communication, including WhatsApp, Telegram, and other, more obscure apps.
Gated Semi-Public Communities
Gated communities, such as dedicated web forums and member-only Telegram channels, require users to register and sometimes pass additional vetting in order to be admitted and permitted to post. Unlike closed communities, gated communities often include public-facing content (generally posts) that can be viewed by non-members. However, they also include members-only areas or invitation-only subforums not visible to the public. Other functions, such as search or viewing user profiles, may be limited to members only. Examples of gated communities include threaded forums like Reddit, or dedicated extremist web forums such as Stormfront, al-Ekhlass, or incels.co. Extremist forums tend to attract users who are usually at least part of the way down the road to radicalisation, rather than serving as a magnet for the uninitiated.
Public Content-First Communities
Content-first communities focus primarily on sharing media or selling products, becoming social due to the secondary inclusion of engagement features such as likes, shares, and, most significantly, comments. Content communities include YouTube and TikTok—where users share videos for other users to like, dislike, or comment on—and online retailers such as Amazon.com, where users can review and discuss products, including sharing images and video. Most user social activity on these networks is public.
Newspapers, newsletters, magazines, and blogs also put content first, while allowing social interaction in the comments section. Most modern content communities include at least some ability to thread topics, for instance, by replying to a specific user’s post.
Content-first communities often have robust content moderation practices. The broadcasting component of content-first platforms may be especially useful for movements that have complex or convoluted beliefs that require long-form explanations, such as sovereign citizens, QAnon, and the antisemitic Black Hebrew Israelite movement, all of which found homes on YouTube for many years.
Structured Public Social Networks
Structured public social networks include platforms such as Facebook, LinkedIn, and Threads, which strongly encourage users to engage with algorithmic feeds and/or algorithmic network discovery and offer multiple classes of post and interaction.
While users must typically sign up for an account, structured public communities create relatively little friction for entry. Unlike closed or gated communities, structured public communities are not usually organised by specific interest, although some extremist or extremist-adjacent iterations such as Gab, Parler, and Truth Social are more focused. But these are the exception. Most structured public social networks present themselves as general interest communities. Once a user joins, however, they are steered into subnetworks that (in theory) reflect their interests.
In structured communities, users are guided to find both content and other users, often amplifying pre-existing interests. Content that receives high engagement is pushed out to more users, so artificially boosting engagement (through astroturf, sock puppets, or community organising) is a reliable way to spread content. All these tactics can be implemented at low cost, making them accessible to a wide range of bad actors, including extremists, conspiracists, and authoritarians.
By design, algorithmic recommendations also help create communities around topics of interest, which can be problematic when the topic is extremism, or when conscious and unconscious biases shape results, for example by recommending accounts that reflect a user’s subconscious racial, religious, or partisan biases.
Extremists actively seek to exploit structured public social networks, since algorithmic manipulation is key to reaching far larger audiences by going viral. Al-Shabaab was one of the first major violent extremist movements to establish a significant official presence on structured social media, with its @HSMPress Twitter account and related accounts on Facebook and elsewhere. ISIS perfected its social media approach in 2014 and 2015, building out a massive and influential network on Twitter that necessitated a dramatic escalation in content moderation.
Progress on community standards was set back as White nationalists and other extremists made political gains in the late 2010s and 2020s, and structured platforms—especially larger companies such as Facebook—have largely abandoned efforts to moderate extremists who hold political power.
Less-Structured Public Communities
While structured communities offer significant advantages to extremist groups and movements, less-structured online domains have also proven useful. Perhaps most notoriously, less-moderated online forums such as 4chan, 8chan, and 8kun have allowed extremists to circulate toxic content that would otherwise be moderated, including open and virulent racism and, at various points, extremist manifestos. Chan-type forums tend to be loosely threaded but are otherwise highly unstructured, allowing users to post anonymously with comparatively few safeguards.
Other less-structured communities, such as federated social networks, feature some content moderation and no mandatory algorithmic direction. Among the best known federated systems are Mastodon, whose most popular interfaces feature a chronological feed with very limited discovery tools, and Bluesky, which offers a chronological feed but allows users to create custom algorithmic feeds.
Mastodon, the most established federated social network, illustrates the trade-offs that come with federation. The far-right social network, Gab, which transferred its infrastructure to Mastodon’s protocol in 2019, faced limits on its reach thanks to aggressive blocking by some of the largest Mastodon communities. Federated protocols are open to all users, but specific communities can and do impose moderation, many forms of which are still early in their evolutionary journey.
Conclusions
Styles of online participation are constantly evolving, and new forms are already taking shape on the horizon, such as private and public user interactions with generative artificial intelligence products like chatbots. In addition, new features or policy changes can shift a platform’s affordances for better or for worse.
Nevertheless, this overview of online participation styles is intended to provide a strong opening argument for considering and classifying online communities, especially in the context of how those communities are exploited by bad actors. This taxonomy also aspires to inform comparative work by helping scholars identify apples-to-apples platform groupings.
Dr. J.M. Berger is the author of Extremism (MIT Press, 2018). He is a senior research fellow with VOX-Pol and with the Center on Terrorism, Extremism, and Counterterrorism (CTEC) at the Middlebury Institute of International Studies.
Image credit: Zulfugar Karimov on Unsplash