QAnon followers flocked to a major four-day conference in Dallas, Texas over Memorial Day weekend to hear speeches from prominent QAnon supporter Michael Flynn and the biggest names spreading disinformation about the 2020 election, including former Trump attorney Sidney Powell. 

The event was a stark reminder that QAnon has not disappeared since Donald Trump’s election defeat. Its followers have simply adapted, changing tactics to avoid getting kicked off mainstream social media platforms. 

However, the language usually employed by QAnon followers, who believe that the levers of power are controlled by a shadowy cabal of Satanist pedophiles, is becoming less common online, according to new research from the Atlantic Council’s Digital Forensic Research Lab (DFR Lab).

Jared Holt is a resident fellow at the DFR Lab, where he oversees research on domestic extremism. He’s the co-author of a new study, which analyzed more than 40 million appearances of QAnon catchphrases online from January 1, 2020 to April 1, 2021.

Holt spoke with me about what this change in QAnon’s language indicates about the conspiracy theory’s progression and the effectiveness of deplatforming its followers.

This conversation was also featured on Coda Currents, our weekly podcast.

Coda Story: It used to be easy to see who was a QAnon follower on Facebook or Twitter because they had Q-related hashtags in their bios. Often these hashtags came from messages from Q, the conspiracy’s anonymous leader or leaders who first started posting about the theory in 2017. Has that changed?

Jared Holt: There were slogans that you could pick out as a dead ringer that this is a QAnon follower. Phrases like “Where we go one, we go all” or “The storm is upon us.” That has changed in wake of deplatforming and the Q author — or authors — explicitly encouraging people to mask their language in response to moderation efforts. We’re starting to see the language of QAnon shift. It’s morphing into almost a second-gen QAnon.

Coda Story: Does it indicate to you that the conspiracy theory has changed substantially? 

JH: What it doesn’t indicate is that QAnon’s dead. Communities built around QAnon are still very active. 

One thing that the data could indicate is that there’s less discussion of hallmark postings from Q. A lot of these communities, since the January 6 riot at the Capitol, have moved on to include a whole umbrella of other topics. Extremist worldviews like QAnon require constant disinformation to keep themselves alive. In the absence of material from the poster writing as Q, there is likely a motivation to scour the rest of the information landscape for things that can fill that void. 

Coda Story: Facebook, YouTube and Twitter all cracked down on QAnon. What’s the impact of that deplatforming?

JH: As these moderation actions started to kick in and communities became fearful that they were going to be banned.

Even just the threat of moderation is a very big risk for these communities. Movements like QAnon were able to get as big as they did because they were able to operate in plain sight on mainstream platforms and leverage the nooks and crannies of the algorithm to get their message in front of a lot of people. Taking away those tools makes it a lot harder to keep the thing going, whether it’s by virtue of getting ejected or it’s the fact that an alternative platform like Gab or Parler has a small fraction of the user base that a huge platform like Facebook would.  

Coda Story: Let’s talk about Parler and Gab, because there was this concern that, as QAnon was kicked off Facebook and Twitter, they would just go to more fringe platforms. Did QAnon ever reach that level of chatter there? 

JH: I have to make the caveat here that we didn’t have a way to assess Telegram widely, and Telegram is a huge venue for QAnon. But you have Parler and Gab as places where these communities continued to post. What we found on those platforms is that there wasn’t that translation over.

Even a slow day on Twitter would outpace all the alternative sites we were able to analyze, like 4chan, 8Kun, Parler, Gab. It would outpace them like 40 to one, and that was after moderation efforts. It really did seem to contradict this idea that if you deplatform people from these mainstream sites, they’ll just go to a new place. It doesn’t seem like that really happened at a scale that supports this notion.

Coda Story: So does deplatforming work?  

JH: Deplatforming works, but with that asterisk there of making sure that we’re clear about what we think platforming does. Removing content does not remove the social conditions that caused people to create that content to begin with. By kicking QAnon to the curb on a mainstream social media platform, you’re not undermining the social, economic, personal conditions that might have led people to be susceptible to those kinds of beliefs. 

It does make it harder for that ideology to spread to an audience that is not already plugged into that universe. And it disrupts any organizing or community building that might be happening attached to a movement. So, destroying these Facebook groups, where there were tens of thousands of QAnon believers, that goes a long way, because the community is the backbone. 

Coda Story: Let’s talk about the four-day long QAnon conference in Texas over Memorial Day weekend, which brought together Trump supporters, election conspiracy theorists, and some big names within QAnon circles, including Michael Flynn. What does that tell you about QAnon’s ability to gather offline? 

JH: The QAnon movement has had real-life events before. A big part of that was this desire to put faces to screen names and build something more solid than a group chat on a social media platform. 

I couldn’t help but think, “This is very expensive.” You’ve got camera work, professional sound, all this graphic design. This was a lot of money and influence, all coming together into one big event. It hit a scale that I haven’t seen really play out before. My gut fear is that this might be an indication of what’s yet to come.