On March 22nd, 60 Days of PVE interviewed Ryans Scrivens about his research on Stormfront.org, and the threat of right-wing extremism in Canada.
60 Days of PVE: What challenges did police officers express regarding right-wing extremism in your research interviews?
Ryan Scrivens: For them, a big thing was they just didn’t really have a sense of what was going on external to their own communities. Often they thought they were dealing with a localized problem, for example, maybe a group like Hammerskins. Police officers thought they had a problem working in community with a small group of Hammerskins when in fact there was a broader problem. I think the big thing for them was that they just didn’t really know, or perhaps pay attention to what was going on in a wider context.
They didn’t really talk a lot about the challenges of policing right-wing extremism. Our interest was in how they policed right-wing extremism in general. What we found was that they weren’t focusing on it at the time we did these interviews, which we started in 2012/2013. I would argue now they are a little more cognizant of what is going on in their communities with the regards to the far-right.
There are notable exceptions. Québec was doing a good job, as well as Alberta, and BC. There are others, however, that didn’t think far-right extremism was a threat in their communities or broader national context.
60 Days of PVE: So, would say their impressions of the current threat of right-wing extremism have changed?
Ryan Scrivens: I think it had to change.
We noticed after wrapping up our study in 2015 and publishing the first piece in 2016, that right-wing extremism was gaining media attention. I think that since the media have been focusing a lot more of their efforts on the far-right and right-wing extremism movement in Canada that police officers now see this threat as a public concern.
When we started doing this work right-wing extremism wasn’t on the agenda of policy makers, law enforcement officials, or academics. The focus was, and I still think it largely is, on violent Islamists. Due to this, perceptions of terrorism or extremism were equated to Muslims. For law enforcement officials, and not just them, there was a narrow conception of the threat of extremism in Canada. Particular groups that would come to mind for academics, such as Hammerskins or Blood & Honour, weren’t perceived to be anything more than a three man wrecking crew, or losers without a cause. But like I said before, there were some law enforcement officials that were doing a good job. The BC hate crime unit deserve an honourable mention.
60 Days of PVE: Can you tell us a bit about your research on Stormfront.org?
Ryan Scrivens: I’m working on Stormfront for my dissertation. What I am doing right now is more macro-level, not digging in from a qualitative micro-level perspective. We are developing machine learning tools to identify radical users or radical content, and we can use it on any web forum that we want. But right now, using it on Stormfront Canada. Amongst other things, we are trying to see what the trends are.
We use keyword analysis to try to identify topics of interests – we use a tool called Techware analysis to measure users’ opinions on these keywords- are their reactions positive or negative? Another thing we’re doing is tracking users’ trajectories over time, which on average, could imply to some extent radical trajectory i.e. to determine if peoples’ discussions are becoming more radical.
In terms of the discussions, it is very typical. They are very much targeted towards groups of people, particularly the Black, Muslim, and LGBTQ communities.
Really, in many ways, Stormfront is the hub of the movement. It is a virtual community in which like-minded individuals can come together and share their ideas, typically free from anti-extremists. It’s a place where people find a sense of identity around right-wing issues that perhaps they couldn’t discuss or have this particular identity in the “real world.” It is an interesting and important place for academics to continue to investigate.
I’ve heard some people suggest that extremists are no longer using online discussion forums. But I would argue, and some of my close colleagues who are very well-known in the field, would argue suggest otherwise. Fine, violent Islamists probably don’t use discussion forums as much, but right-wing extremists are still using them time in and time out. Their presence on Stormfront continues to flourish. It is one of the oldest websites, and not only one of the oldest white nationalist or white supremacist sites, but one of oldest websites since inception of the mainstream internet. And like I said, its membership has continued to grow.
Right now, there’s about 12 or 13 million messages, and the threads are continuously growing. It is an important avenue for academics to look at, whether it is an in-depth qualitative, or thematic content analysis, or work that we are doing using Stormfront as a test bed for machine learning.
We’re doing this on Stormfront, but I think this should be continuously looked at by academics. Considering how powerful it is, there really hasn’t been that much work done on it. Especially as of late, only a couple pieces have come out that are very anecdotal and only scratch the surface. We need to dig deeper into it.
60 Days of PVE: Some have suggested that allowing right-wing groups to post in forums alleviates tensions and could stop the commission of violent attacks. Do you think this is true?
Ryan Scrivens: I do.
You can’t suppress these people necessarily, that’s when they are going to continue to build resentment. I mean, I obviously don’t agree with what they say online. But they need an outlet to vent, they need somewhere where they can share their ideas with others.
What happens if we try to shutdown these sites ? They just go underground again, to the dark web, or create another web forum. And I don’t think the idea of shutting down the website will deter any of their activity. It isn’t going to fix the root of the problem.
I think there are alternatives, Moonshot CVE and Jigsaw are trying to develop tools and technology that counter messages, websites, and web forums. For example, let’s say I am a right wing-extremist, and I search on the web for white power music. If I click on the link, instead of leading me to white power music, Jigsaw will open another webpage trying to counter the message of the music.
I think that those kinds of efforts are better than shutting down a site because that doesn’t deter people at all. What is interesting about Stormfront is you can’t really shut it down. The reason the site has been around so long is because they know where they sit, what free speech is and what it isn’t. They also have heavy, heavy moderators on the site.
Let’s say someone goes on Stormfront and posts, “I hate black people.” That isn’t likely to stay up there. A moderator edits it, or he or she deletes that message, or he or she bans the person from the forum. So what happens is that users are well aware of the rules, and that the moderators adhere to the rules. Otherwise the website gets shut down. Typically, you’ll find that although the sentiment is not nice at all, most times you don’t find the more hateful words. So, bottom line, I don’t think Stormfront can even legally be shut down.
60 Days of PVE: Does recruiting takes place on websites such as Stormfront.org, and how does it compare with recruitment offline?
Ryan Scrivens: It does happen, but research is showing that time and time again there isn’t a disconnection between the online and offline realm. I have worked very closely with former right-wing extremists whose job it was to recruit members, and it was and is a combination of both.
I don’t think a typical person will go online, and all of a sudden become radicalized. It doesn’t really work like that. They have to know somebody offline first, or maybe online. They are most likely going to connect with this person offline in some capacity as well.
Stormfront, is used to recruit, but it is not the only way in which they recruit. It is part of the process of recruitment. One of the formers that I worked with, for example, said that a very popular way to recruit people is to meet face-to-face, and to try and gain their trust in one way, shape, or form. People are brought into the far-right movement by people that they know, they’re not just coming out of nowhere. Often, skinheads for example, will get to know individuals offline and talk with them online, so that is how the recruitment process continues.
To answer your question in short, StormFront is used to recruit, for sure, 100%. It is not the only one. There are sites such as Vanguard News, the Hammerskins forum, and Blood & Honour forum to name a few. It should be noted though, that Stormfront is the main hub of those forums.
60 Days of PVE: Would you say there are any parallels between the radical right and radical Islamist activity on public domains and discussion forums?
Ryan Scrivens: It is hard to say, I have done some work on radical Islam as well. I would not call myself an expert by any means, but there are definitely parallels.
I think the difference between the far-right and radical Islam is that they are using different outlets. Radical Islamists typically, and I’m speaking in general terms, aren’t using forums anymore. They are using Telegram, Twitter, and/or Facebook. The far-right does too, but it seems like the far-right adhere to the old traditional forms of online communication, which are discussion forums. I think there are differences in terms of that.
But I can’t speak more to that, as I haven’t studied it enough in detail.
60 Days of PVE: Coming back to the tool you’ve been developing, what is its exact purpose? Will it be assisting law enforcement in the future, or is it purely academic?
Ryan Scrivens: It really is a combination of both. We do work with the RCMP, but it is not fully developed in terms of giving it over to them. I am not generally interested in that line of work.
It started off as a meaningful way to analyze big data. It started with what we called “The Challenge.” These groups of computer scientists had a challenge online for anyone who wanted to try it. They gave people 1 million messages to analyze and try to find the most infectious radicalized users, posts, sub forums, threads, etc. and try and develop new way of analyzing big data. It could be content analysis, social network analysis, or any other methods.
We found this pretty interesting since not much work done in this area, particularly at the time. We found no work using machine-learning tools to understand how extremists use the internet before this piece was published in 2015. They weren’t focusing on users specifically or finding radical users, so we found this is something we could potentially do. Let’s think of meaningful way to analyze a million messages, because what’s happening now, as I’m sure you are well aware, is information is growing online and manual analysis is no longer effective. So we have to think of new ways to analyze big data.
I think eventually it will get to the point where it is effective for law enforcement officials. And if we consider going down this route, we could frame it as a tool that officials can use to identify warning signs. So if it is tracking users activity over time, for example, and it’s seeing that a person’s sentiments are getting more negative on the discussions around Black or around Jewish communities, and law enforcement officials could have this information, and perhaps intervene if needed.