A Warning About Twitter and Radicalization

Dave Troy
5 min readJul 11, 2020
Where we go one, we go all? (Illustration by the author)

I’d like to offer a warning about radicalization.

The debate currently raging about “cancel culture” is a byproduct of an ongoing process of radicalization and tearing of our social fabric which has been underway for some time. Some of this is intentional and being driven by bad actors; much of it is being driven by the design of technology platforms themselves.

When these platforms were designed, there was very little attention given to questions like, “What kind of society will this tool produce?,” or “How will this change the behavior of the people who use it?” Comparatively, much more attention has been given to how to make the products “sticky” (keeping you coming back again and again) and how to have them make money.

As someone who studies radicalization and online platforms together, I believe Twitter poses a special threat to society because its design emphasizes regular communication with people of like mind. If you think of it as a game (and it is a kind of game), the winning moves are to follow and be followed by as many like-minded people as possible, and for each tweet to score as many retweets and likes (hearts) as you can. This affects individual behavior, because it has people continuously thinking, “what would make a good tweet?”

Likewise, players will also develop a sixth sense for what kind of tweet might attract disapproval and scorn. They will start to consciously avoid those. They will also start to identify such tweets coming from other players, and pile on when they make a bad play. Getting “ratioed” is what happens when someone’s tweet gets more replies than retweets — a signal that the tweet was met with fierce disapproval or controversy.

The social capital (or goodwill) induced by players is self-reinforcing. Gradually over time, players find themselves interacting with people who are more and more like them, more and more often. This has the effect of making players more and more similar to each other. We all come to instinctively know the people we like most online and who are most likely to agree with us.

Because the player is not a static entity, but a dynamic one, this process constitutes a kind of thought reform, where certain positions become uniformly acceptable (i.e. tweets that would attract retweets and likes), while other positions become anathema (i.e. tweets that would attract negative attention and disapproval).

The more this process of self-reinforcing social capital continues, the more the network becomes a kind of distinct in-group with its own norms, codes, mores, and processes. These may or may not have any connection to broader societal norms, values or concerns. Generally, certain ideas are amplified and fetishized by the network, and may come in and out of fashion as the network churns.

A network with too much internal social capital is radicalized. Radicalization occurs when a sub-group places its own priorities and values ahead of those of society as a whole. Twitter, as the game is presently designed, acts as a machine for radicalizing parts of society.

Without regard to the debate around “cancel culture” itself (most reasonable people agree abuse and bad behavior deserve scorn, and sometimes rather a lot of it), what is notable here is the process of radicalization that produced this particular frame. Arguably, “cancel culture” would not exist as a phenomenon without platforms like Twitter. We might well have gone through a similar cultural reckoning through other mechanisms such as newspapers, gradually revising our cultural pantheon, and the like.

But for swift radicalization that affects all of society, Twitter is unmatched. While it has a relatively feeble 300 million users, its impact on society is outsized because of its heavy saturation with media figures, and the breathless punditry that accompanies every mindless word uttered by the President and other global figures.

Twitter is effectively a role playing game engaged in by a tiny slice of society, where the players get to alter our culture. This alone should be alarming enough, given how little attention has been paid to what kind of culture we want this game machine to build. But perhaps more alarming is what could come next.

While cancellers, cancellees, and critics sound off about the merits and dangers of “cancel culture”, we are meanwhile hurtling headlong into the next round of this game. Notably, the anti-cancellers have formed a pact (social group) to oppose the staunch group of pro-cancellers. We can expect that this will harden the resolve of the pro-cancel group even further, driving further radicalization and increases in internal social capital within that network.

And we may start to see hardening of the anti-cancellers into a larger, more unified bloc. That group may find allies (wanted or not) in the form of nationalist groups and other people for whom opposing the cancellers (whether materially or performatively) carries tactical or strategic value.

Increased social capital within these networks, and the accompanying hardening against other influences, is likely to lead to even further radicalization and violence. I am certain that some folks reading this will believe some violence is justified to achieve cultural change.

It will be a shame if people die because of mind control processes born on networks like Twitter. But indeed, that has already happened. To name just one instance in 2019, a 24 year-old man killed a reputed Gambino mob boss after being quickly radicalized by the QAnon conspiracy theory (whose motto is the ghoulish, cultish, “Where we go one, we go all.”) Some might say that QAnon is a far cry from “cancel culture,” and they are certainly not morally equivalent (one is nonsense, the other is rooted in legitimate concerns), but the difference in the networks that produced them is a matter of degree, not in quality.

While most of us like to think we are in full control of our senses and thought processes, science and the history of cults and marketing tell us that’s not quite true. While we debate the finer points of cancel culture, I’d encourage people to draw particular attention to group dynamics and how reinforcing social capital within groups drives radicalization.

Additionally, we have evidence that these group dynamics have come to the attention of people like Steve Bannon, Vladislav Surkov and their many protégés, who have deployed Active Measures style efforts to drive these radicalization processes faster and to greater extremes.

Twitter doesn’t yet seem to understand it’s a platform optimized for radicalization. There are changes it could make that might slow or reverse these processes. But until that happens, we should expect that radicalization will continue, and we must be prepared to guard against its worst effects in our society.

We will be increasingly faced with the question: how much violence born of online radicalization is too much, and when should we take action to curb it? Arguably, if we can act now (and we can), we should.



Dave Troy

Investigative journalist addressing threats to democracy. Public speaker, writer, podcaster. @davetroy on Twitter. See davetroy.com for contact info.