Understanding structures of radicalisation

Speaker 1: Uncover the threat landscape with assured and interconnected threat intelligence from Janes. Covering military capabilities, terrorism and insurgency, country risk, and CBRN. Support your threat and capability assessments and enhance your situational awareness with Janes threat intelligence solutions. Find out more at Janes.com/threat.

Mark: Hello, and welcome to the latest edition of Janes World of Intelligence Podcast. So today we're joined by Tim Clancy, who is the founder and CEO of Dialectic Simulations Consulting. So for now, welcome to the podcast.

Tim Clancy: Thanks for having me today, Mark.

Mark: Maybe you could give us more of an introduction to your work.

Tim Clancy: Sure. My focus of my work is on violence and instability of non-state actors. And really, I look at that from a perspective of system structure, trying to understand the underlying system structure with the purpose to identify policies we can use to reduce it.

Mark: Wow, okay. So this process of radicalization, it can be a pretty complex process at the best of times, I know in the course of your work, you've developed a structure of radicalization, I guess you'd call it. And maybe you could give us a bit of an overview.

Tim Clancy: Sure. So our structure that we looked at, we went over all the expert theories we could, but we were trying to look for something that was agnostic to violent ideology. So the structure I'm about to describe, it's not just for white supremacy, it's not just for radical Islam, Salafi Takfiri, it's really a generic structure that we feel can represent a wide variety. And really in the governed space, the key process that an individual goes through, begins with a grievance. There's a grievance, they perceive a grievance or some form of moral outrage. And this, I want to be clear here. These could be legitimate grievances.

Mark: Sure.

Tim Clancy: These are not necessarily nefarious themselves, but the key is then they advanced to this narrative and they gain the narrative. And the way they gain it is through cultural scripts. You can describe these as the memes, the information, it's coming throughout their networks. The narrative explains why the grievance occurs. And here it's key that when we talk about violent radicalization, it's almost always a conspiracy narrative that says," Here's why the grievance exists, who is to blame, and what needs to be done." And this is really the roots of a violent ideology. Then they go through a fixed station phase where their lives get consumed around this narrative. They can't get enough of it. They're always looking for more information. They're beginning to go out and explore and find areas where they can get more. And then if they continue advancing where it gets dangerous, is they sort of adopt this identification and activation and the identity that they adopt, the personal identity is sort of a pseudo commando, or warrior identity. I've got to go do something on behalf of this grievance. I've got to take it into my hands. They sort of rationalize a violent response as part of this identity. And it's this identity that really gets to the next stage pathway to violence. And that ends up in a terrorist acts. So you can think of this loop for lack of a better word. And then when they commit terrorism, that provokes a societal response that can lead to grievance, the cycle continues. Now influencing this gear is all sorts of domestic network effects. We're going to talk a little bit later I think about the difference of influence from your local connections versus a non- state actor.

Mark: Sure.

Tim Clancy: But these network effects are going on whether people who are at risk to being susceptible to this are being sort of moderating influences or radicalizing influences, but this is all happening in the governed space. What's key to understand though, is in the ungoverned space, there are non- state actors and these could be organized hierarchical groups that run websites, or it could just be the sort of a crowd of folks like you might see in a Reddit or 4chan, where they're sending cultural scripts into the near space. There's two main forms. There's broadcasting scripts, which is what gives the narrative. This is how people find a narrative for a grievance. This broadcasting, it's un- targeted, it's a unspecific, it's just out there. And it's spreading that conspiracy narrative often aligned with a violent ideology. But there's a second kind, which is narrow casting, which is much more targeted and personal. And this could be following a Twitter of someone you think is speaking to you. It could be these videos that are directed to followers. It could be very tailored and that's what drives the fixation and the identification. And this is what really can accelerate this radicalization process. So this structure I've kept it fairly simple, but this structure we think actually is explanatory for a wide variety of radicalizations. Whether you talk about school shooters replicating the Columbine incident in the US, or those who may have been, for example, followers of ISIS, or more recently, you have the concern with QAnon and radicalized groups on the right wing extremism in the US and UK and other areas.

Mark: What do you think makes people more vulnerable to accepting those grievances that you've outlined there? And how would you fit that into your model of radicalization?

Tim Clancy: We rely on the research here a researcher named Malloy. He has something called the TRAP- 18. You may hear this as dimensions of vulnerability. There's been a lot of research into characteristics into what we would describe as an at- risk population. And so these characteristics in the TRAP- 18 that we use they're divided into sort of distal characteristics, things that are way in the past, and this could be they're dependent on a virtual community. They may have had stymied career progress or thwarting of their goals. They've had a trouble bonding with people. They may have had a past mental disorder. They may have had a history of criminal behavior. These are things in the past that, it's not saying if you have one of these things, you're at risk, it's a collection. And another thing we found that was really key for this at- risk is they're looking for self- similar people. So this is a part that was novel to our research. And I think comes from looking at other ways contagion spread, is an at- risk population, you can create a profile for them, but they're not out there necessarily looking across all terrorism and all violent ideologies. They're looking for people who look like themselves and taking cues and signals from them and are therefore more receptive, which is one reason you start to see channels create in the way radicalization not only forms, but how it's expressed in violence that is very much ideological specific is because of this, not just these background characteristics that may be shared, but they've got self- similar behavior. They're looking to someone who looks like them. They see themselves in them, they're inspired by them. They may take signals from them. And that's the pathway that lends more weight to the radicalizing scripts.

Mark: Got it. Yeah. I mean, I think you've touched on a really important point there, it's when these different combinations of vulnerabilities are brought together, sometimes it almost creates that, for want of a better phrase, that perfect storm in an individual where maybe their beliefs that they once held have been shaken and then that makes them somehow more vulnerable, more susceptible to taking on maybe the ideas of an extremist propagates or whoever else. But I mean, I know your focus is, it's not just on understanding why radicalization happens. I mean, I know you're also interested in how do we would reduce violence like you say. How do we reduce violence and instability in society? And I guess with the topic that we're talking about today, radicalization, I guess we're asking ourselves, look, what could we do to stop this thing once it's started. Or, hey how do we even prevent this thing of the mountain in the first place? So to your mind. Are there any factors out there that you think may help put the brakes perhaps on an individual's process towards violent radicalization?

Tim Clancy: Yeah, I think there is, but I think it may be useful to take a step back and talk a little bit of how we see this behavior spreading. Because when we looked at these there's two general theories out there, which is that terrorism is being spread by groups of individuals that are loosely connected or perhaps a non- state actor. We actually found looking at the structure and we can talk about the evidence later that we reviewed, but we saw it's almost like in some ways a contagion. And I don't know if you or your listeners are familiar with the concept of a Werther contagion, but just to go through what that is, a Werther contagion is when a celebrity commits suicide and they do it in a specific way, the media picks it up because they're a celebrity, they spread that broadly. So everyone gets to see it. Now, people who see themselves in that celebrity and are susceptible, they're at a risky point in their life. This is a well- established phenomenon. They are more likely to ideate and perhaps even attempt suicide in a period of time after that, not just the attempt, but they'll use the actual methods that celebrity did, and Werther contagions are very well studied. They're very well understood. And it tends to taper off because any subsequent suicides themselves are not necessarily media worthy. So it's sort of a one and done. You have the initial spike, and then it falls. What we saw in our structure when we analyzed it was that there's a similar contagion effect that's almost inverted in that rather than an individual having celebrity, the mass violence itself creates the notoriety. And so if you have a successful event that kills a number of people, the media picks it up spreads it almost universally, and that then spreads. These scripts, this narrative, this modus operandi, and into that some set it will fall upon that at- risk populations, who sees themselves in that perpetrator and they'll begin following it. And the challenge here is that with a terrorism contagion, as opposed to a Werther contagion, the next act that succeeds doesn't need to be a celebrity, the act itself creates the notoriety. So what you have in these systems is the contingencies of what is a sufficiently large enough at- risk population to receive these scripts, and then a template that they can follow that is easy enough to access, or has a high enough success rate, such that the next violent act occurs soon enough that the media will pick it up and keep the cycle going. And this contagion theory provides a very powerful explanation for things that we would commonly call lone wolf actors. Columbine shootings, there's been 174 attempts to replicate Columbine, even by people who were born after the original perpetrators were dead, but they're still copying that contagion because they find themselves. It also, we think is explanatory for more traditional forms that people might be more familiar with, like Al Qaeda or ISIS inspired terrorism. Because in this case, you may have a very small at- risk population at the UK or in the United States, but they may have influences from abroad sending those scripts in. So you can really look at these, we look at it as all of these forms of radicalization that are out there as being determined by the contingencies of a system, in a given population, in a specific area, and that determines how it will manifest, and it can manifest in many ways. Now to get to your point about how you stop this, once you understand that feedback loop, some of the ways of stop... We already know how to stop or limit a Werther contagion. Like I said, there's a good amount of research of how you report on suicides to cut that feedback loop. And we think there's ideas in that to do it, but there's even some more specific things. One of the reasons we use simulations is so can test this. So if you think about that radicalization cycle I mentioned, obviously having alternative narratives out there is useful. So they're not latching on to the violent one, but that's hard to control in this day and age. We're trying to focus on what's practical and usable, but as they get that identification, they begin to identify. We tend to highlight the completed terrorists incidents with mass casualties, but there's a whole bunch of them that fail. And so one of the things that's very distinct is highlighting what we call failure, notoriety, which is the likelihood you're going to get arrested. You're going to get caught. It's not going to happen, you'll fail, and that can serve to dampen interest and increase abandonment. Now, again, this doesn't mean that there's not a risk, but we're trying to get them off ramped from a violent mass violent incident. Another way to do it is to if you can make it so that they don't see themselves in the perpetrator, there's not going to be as much replication. And I can go through a few examples of very high casualty terrorist incidents that were never replicated because the demographic and combination of factors with the perpetrator didn't appeal to a large enough subpopulation. And therefore there was no replication and this contagion effect kind of died out. And so I think there's some things we're looking at of how do you identify this at- risk population, where they're at, and do interventions to get them before they commit the acts themselves, because that cuts that feedback loop that then spreads the contagion. And if the contagion is underway, how do you use tools in the media to dampen or mitigate the contagion and keep it from replicating?

Mark: Yeah that is super interesting. I mean just going off your model there as well. I mean, you're talking about the ungoverned space, for example. Just thinking about what you just said there about contagion. I mean, obviously you have an attack, say for example, Columbine or whatever else, and then online, these kind of subcultures emerge that continually promote why this attack was a good thing in the eyes of the perpetrators or whatever else. So I guess that's a challenge as well. That contagion in the ungoverned space, and how does a government or wherever else, anyone trying to combat this thing, how do they actually contest that? Because it is in those spaces that exist for the very reason for promoting those attacks. So what do you think about that?

Tim Clancy: It's probably easier to say what not to do than what to do in those spaces. And that's the challenge is that the way I liken those spaces, is those spaces exist because of a grievance and the grievance is like water to a lawn, it's a carrying capacity that generates these things. So there's a lot of things that you see in the past, there might have been periodic raids in a physical area. Or they'll throw a few missiles at the information operations, try and physically destroy it. With the virtual space you may see these periodic, for example, after January sixth in the United States, Twitter did a big purge, Parler was shut down. The problem with these interventions is they're only affecting the casting capacity, the ability to send out cultural scripts, the grievance is like that water and that grass is going to grow again. And so you get a little dip and then it's sort of like an organic growth, the capability reconstitutes itself in one of these digital spaces and is right back at casting it. So one of the benefits of a simulation is we can model those types of interventions and show out exactly why they don't work or what combination. And you've really got to... It's kind of almost a long- term policy, but you've got to reduce the underlying grievance wile cutting out the organic capability. Because if you don't tackle that grievance, you're not going to do it. Now, some of these grievances, they may not be easily accessible to tackle, which is why we focus more on the feedback effects coming out after the incident itself, because we feel there's more tangible things, especially in the US where you have the First Amendment, and there may be complications of shutting down some of these sites, that's very hard to do. Now you can pollute the streams. You can use this concept called counter reification, where if you think about the way this contagion spreads, it's a very well cohered idea that communicates a narrative, that's tied to a violent ideology, that's tied to a template on how to conduct the attack. And I say tied as saying, it's not like there's one blueprint, but all of these things together connect to one idea. And you could call that idea inceldom or any of these things. Counter reification is simply diffusing that idea to mean many things and creating confusion. And that can be an effort, but again, you get into some legal and constitutional challenges there, but that's one way. If people can't see themselves or can't understand what the template is asking them to do, they can't replicate the attack and you can't get this feedback effect. And that's probably one of the better methods when you have this sort of swarm approach where there's no one actor that's generating these things, it's a collection and you're really going up against the hill there.

Mark: Yeah. Yeah, totally. Just want us to focus in for a minute on the topic of... You've kind of already mentioned it really, but the speed of an individual's radicalization process. Because this is something that when I look at various different cases, I'm always left a little bit bamboozled with the divergencies here. I mean, couple of UK focused examples on the white supremacist end of the spectrum here. So for example, we had one individual, guy called Thomas Mair. You've probably heard of him. He was the guy who perpetrated the murder of a British MP in 2016. Now, if you read some open- source reports on him, he had been consuming extremist or white supremacist propaganda for decades. Then suddenly one day in 2016, he decides he wants to go ahead and perpetrate his attack. That's on one end of the spectrum. Now on the other end of the spectrum, you've got another guy called Darren Osborne, who I think round about 2017, decides he wants to perpetrate a vehicle impact attack outside a mosque in inaudible. Now if you believe some open- source reports about that Osborne, self radicalized online, and some people say in the space of weeks. Now you've got massive diversions there, haven't you, in terms of the speed of an individual radicalization process. I just thought it'd be great to get your thoughts really on that divergence and maybe how perhaps you account for that in your model in any way. Because it does appear to just wildly diverge in some cases, what do you think?

Tim Clancy: Well that's a great point and I want to dig a little deeper in the Finsbury attack because that's a case study of our theory that we've used to illustrate some points. But to get to the broader question, I like to think of it as there's a fast gear and a slow gear. And I think part of this is to understand how humans express violent intent. Evolutionary speaking, we have two fundamental reactions. One is the effective reaction, it's an immediate reaction. It's very emotive, it's fast. It burns itself out very quickly and it's not very well planned. It's the threat that we developed evolutionary fo a bear entering the cave or somebody throwing a spear at you or something like that. It's very much in the moment. That's effective violence. Now, predatory violence or what some call instrumental violence, that's the evolutionary response of say going on a hunt where the threat is long- term abstract and it's a very cognitive effort. It's a very long- term sustained effort. It involves a lot of planning. It's very different than effective violence. And I think it's important to remember that both of these are in people, the effective response and the predatory response. And in most people there's barriers for us doing this against others. So to some extent, before you get violence on any other human, you have to drop those barriers. But this concept of slow burn, fast burn, I think gets into those contingencies. If someone has a slow burn accumulation, they may just be getting radicalized and if there's an opportunistic situation, they may go into an effective response that is hard to tell how long have they been at this? Where did it go at. Versus someone who is... A lot of these mass violence attacks they're planned for months. Someone radicalizes and then they're planning for between six to 18 months. And these are the contingencies that we look at. So if you think about, if you're going to spread a terror contagion, the attack itself needs to generate enough casualties to get the media attention on it, to spread it widely, to build. And that favors attacks that tend to be well- planned. So when we looked at the data, we actually looked at 4, 500 terror incidents over 20 year period, US and Western Europe. We classified them by the type of radicalization we want. And there was a whole bunch of them that failed that never made the list of most case studies. In our research, we tend to have survivor bias and I don't mean survivor of the incident, but we tend to focus on the incidents that completed and had high fatalities, but we looked at all of them and we found a lot of evidence, almost like a background noise of attacks that occurred, which seemed to be little planning, little effort, didn't do much and just kind of faded in the background. And that may be an example of an effective response to one of these radicalization. Now, a predatory response where you planned it, you get it in place and you generate a large amount of casualties in a completed attack, that tends to self elevate itself, spread itself wider, and so if you think of this almost like a cold virus, that virus is better adapted for replication, which is why things that cause many casualties tend to be what's replicated. And it gets to the accessibility of... The difficulty of replicating the attack. You don't see a lot of bombings as much these days, at least, especially in United States, because it's difficult, it's tricky. You're likely to get caught. Buying a firearm and going on a walking shooting spree is a lot easier to replicate. So you see these contingencies that are regional and location- based driving the channels within which terrorism and radicalization expresses itself. And so, yeah, I think it's very likely that this slow gear, fast gear are operating in all sorts of people. And the question is, was it that they really were radicalized quickly and then committed an attack, or were they radicalizing along and an effective opportunistic, I walk outside, I see something and I just decide to do it comes up. And I think that we have... Disentangling those because they can co- exist is very tricky.

Mark: Definitely, great response. Now I just wanted to explore with you finally, really about this debate that that has been going on, I guess, within the community that looks at terrorism and violent extremism, this so- called, the groups versus individuals debate when we're talking about violent extremism. How much influence does an extremist group have on radicalizing individuals to perpetrate violence these days?

Tim Clancy: And I think this gets... This is a fascinating debate. I'm very familiar with it in working with this and I often will hear it called the swarm, which is the groups of individuals, versus the fishermen, the non- state actor. I think it's been a debate, and when I look back, I think it's important to understand how this developed. In a lot of cases, this debate played out in the context of a global war of terror, where they were looking at a limited case set of completed terrorist attacks, largely within the Salafi Takfiri violent ideologies. We're talking Al Qaeda, ISIS, those types groups, and they were having this debate. Do we need to intervene to stop these non- state actors in the ungoverned space, because that will prevent it. But in our work, we actually took a big step back and said this whole question of whether it's individuals radicalizing one another or groups, are themselves just manifestations of this underlying system reacting to different contingencies. And so if this debate about root cause, that structure I talked about before that was kind of a blueprint. One of the things you can do with system science is you can do some interesting tests for causality in complex systems. And if you think about that structure and then turning it sideways and extending it vertically, so it's layers of a cake. Think of layers of a cake stacked on top of each other, level one would be the terror incidents themselves. These are from when the perpetrator leaves the door, goes out, conducts the attack, the attack either complete or fails. The terrorist is caught, captured, killed, whatever, just that pure discreet incident, that's level one. Level two are the individuals who comprise the at- risk population and how they move between radicalization and non radicalization within that. Three is where this contest plays out, which is the influence of near networks, your associates, the folks you're getting online and the influences of non- state actor. Four is these spaces. Four is the hierarchy where the non- state actor in a virtual space, like a 4chan or in a physical space like Afghanistan or Syria can spread these messages in. And level five is the symbolic or abstract. And so now we had a hierarchy and this whole debate is really focused I think, unfortunately on levels two, three, and four, they're not looking at the entire system. And what we found when we tested the causation, it was actually coming from the highest level down, the symbolic system of systems at the very, very top. And that those influences then created channels within which emerged manifestations that we then said, oh, that's a non- state actor one. That's evidence for this. Versus, oh, that's a swarm case. Without understanding those contingencies. And let me give you two examples. We are able to test with simulations, and this is synthetic data so there's a big caveat on this. But we're able to test with simulations, what are the values these contingencies need to have to spark a successful contagion? And one of the most basic elements is the size of the at- risk population. So take two different at- risk populations. And I can show you how this debate between swarm and fishermen emerges. In the UK, for example, or even in the US you might talk about, well, there's a much smaller at- risk population of men who follow the Muslim faith and have these at- risk factors in the past that they would then be susceptible when they see self similarity with a message that's coming from an Al Qaeda or ISIS. That's a small at- risk population. The population itself is so small that it may need something from the outside coming in those cultural scripts coming in from the far space, broadcasting it. And so it may appear that that benefits from a non- state actor group, an extremist organization. But take another example. We start with Columbine. Columbine was two white male teenagers, high school age men that had a sort of narcissistic aggressive angst perspective is how the media has portrayed them at least. And that's actually quite a lot of population in the US at any given time are white men in that age group, thereabouts who have those same things, huge at- risk population. So now you have a much larger base within the contagion to spread and self- replicate without necessarily needing outside groups intentionally directing it. So you have two contingencies that are different, and they manifest themselves in different ways. And I think there's been this kind of desire to say, well, it's one or the other. And we look at it as the contingencies of the system at any point in time will determine which one will manifest. In the United States we tend to go after organized groups pretty severely. We tend to clamp down on it, so you see, for example, when the FBI cracked down in white supremacy and sort of far- right militia groups in the eighties and nineties, you would see this concept called leaderless resistance, excuse me, emerge. Leaderless resistance. That is a reaction to the contingency within the United States of the successfulness of a group being able to survive long enough to conduct an attack. And you see, we had this Michigan militia group just last summer, try and make this plan to capture the governor of Michigan. And it was all tied up in COVID. That's an example of a small group that had been ostracized from the main group, because the main group's like, we don't need this heat, we don't need the FBI coming down on us. They pushed them away. This small group organized, they were penetrated by the FBI and broken up. Technology plays a role too. These days, it's not really, there's no differentiation in your access to media. If there's an incident anywhere, there's an incident everywhere. And it spreads in about five hours, but we can with the simulation replicate historical time periods, where that may not have been the case. And that's one of the interesting experiments we want to run or test is to say look, this theory means there has to be a transmission of the information fast enough, so that it's still in the public conscience to land on those at- risk and perpetrate. Now with the internet being what it is, you have this massive reservoir. So I'd say the contingencies favor the loose groups of individuals self radicalizing himself, but that's more of a contingency argument than a, which type. And as the contingencies change that will adapt. And here's the key with the adaptive system is the reason these are tricky is these systems are not static. And you know this, I mean this is obvious to anyone listening to this, I imagine, but as you go in and implement a policy, the policy itself will cause an adaptation and reaction in the space. And we simulated this by a law enforcement intervention that would, say go after a certain kind of right- wing extremist associated with the January sixth insurrection. They're going in heavy. They're moving it. Those actions themselves, if they're not done in a careful way, can provoke a backlash, which becomes a grievance. And the grievance then becomes the fertile ground by which other radically... Now this isn't to say, don't prosecute, don't go after them. But it's the same concept that you see in counterinsurgency that you have to be very thoughtful and credible and careful of how you go about these groups. And that's where we're trying to get into that really nuanced discussion, but in practical ways, so that we can advise policy makers, this is not just what you want to do, but to what strength and what timing window, and have that kind of sophisticated discussion, as opposed to, unfortunately, in public policy debate, it's kind of the single solution. And then people rally around that.

Mark: Yeah. Got it. Well Tim, I think that brings us to the end of the podcast, really. I just wanted to thank you for a really super interesting chat. I mean, I know this is a really complex topic, but I think you've broken it down and explained it super well. So thank you for that.

Tim Clancy: Thank you.

Speaker 4: Janes Capella interconnects millions of assured data points across Janes foundational intelligence with the ability to integrate and contextualize multiple sources, delivering the single source of truth. Janes Capella increases certainty and accelerates decision making for everyone in your organization. Find out more at Janes.com/ capella.