Further thoughts on the Rolling Stone article: “Everything You Know About Radicalization is Wrong”
PART 1 of 2
By John Horgan
There have been lots of reactions to the @rollingstone article on radicalization and terrorism. I am grateful to @johnknefel for approaching me to contribute thoughts to those of @jamiejbartlett. For the record, I was quoted accurately and additional thanks to John Knefel for both writing the piece and helping kick off a spirited discussion. Just don’t confuse me with his title 😛
Jamie Bartlett has since retired to the countryside, and while I await my record deal, I’ll take an opportunity to do two things:
a) Reflect on reactions to the ongoing debate (with 1-2 clarifications on my position)
b) In part 2, to follow, offer a modest attempt to make progress with reference to some key concepts in psychology, some of which will include reference to MacCorquodale and Meehl’s classic 1948 paper: “On the Distinction Between Hypothetical Constructs and Intervening Variables”. Or put another way, ensuring that “Radicalization debate” doesn’t become the new “terrorism definitional debate” with 30+ years ahead of us to figure out what it means. Yes, maybe it might be good for “terrorism studies”, but still, nobody wants that right?
My approach to (a) I can couch as completely subjective, but nevertheless important to clarify some thoughts and not alienate fellow radicalization researchers (of which I actually am one – see my Walking Away from Terrorism for how I conceptualize radicalization) and so I’ll indulge with some extra latitude on my personal thoughts on this past week.
As someone who is very new to blogging, what follows can be rightfully considered a ‘draft’ at best and stream of consciousness at worst.
With that in mind – here goes.
There have been some good exchanges this week, ironically, not long after Dr. Marc Sageman feels that terrorism research has become stagnant. In a co-authored response with Dr. Jessica Stern, we respectfully chose to disagree. In fact I think there has never been a better time to get involved. Despite the obstacles, studying terrorism is not an academic career killer, but does require careful navigation. Dr. Sageman is correct in a broader sense (though this did not emerge in his Chronicle article) that there are structural problems limiting the field’s development – for research to be effective, it has to be conducted by, and amongst, equals and those ‘equals’ have to tolerate challenge and debate – Dr. Sageman is absolutely right to point to the weaknesses of government investment, but we as academics have to rise to the challenges that may bring. There’s far more going on here than either Dr. Sageman’s original piece or my response with Dr. Stern captured, so more on that another day, and before too long.
In the days that followed the initial article, Mr. Daveed Gartenstein-Ross provided a detailed response here, and Dr. Jamie Bartlett’s further thoughts on the debate arrived here. But it only took off from there. Mr. J.M. Berger (aka @intelwire) provided critically incisive observations here, and just when you thought it couldn’t get any better, @ThoughtsonYemen’s anonymous blogger not only summarized the debate to date, but offered insightful observations here.
So after all that, and at the risk of sapping the momentum built up by a week of spirited debate, I will offer some further thoughts.
Do I believe “radicalization” exists? Seriously? Hard to believe I have to actually ask this, but well, it’s been an odd week and so let’s leave no assumption unaddressed here: yes I believe it exists. Yes, I recognize it is a process of immense social and political significance. Yes, we should continue to unravel its complexity – in particular its social psychology-centered questions of how attitudes and behavior relate to one another (which, really is at the heart of this problem). Yes, we should understand radicalization and its implications for the development of policy (and vice versa) and yes we should continue to examine how radicalization and terrorism relate to one another. I understand and accept that radicalization can refer to both an attitudinal process and a behavioral process alike (though it might appear that it is commonly represented via a type of attitudinal process with some behavioral outcomes), but I contend that however it is conceptualized, its relationship to terrorism remains grossly misunderstood.
It is on this last issue that I long for clarification because the implications for getting it wrong far outweigh academic discussions about whether one causes the other, stems from it, or, in some cases, isn’t even relevant. And no, I’m not talking about the Tsarnaevs, about whom I refuse to speculate because it’s still too early to say anything meaningful about their “motivations”.
That current thinking is dominated by the assumption that terrorism must be understood with reference to radicalization is the heart of my concerns.
Much of this can be summarized by saying that whilst radicalization (whatever it is – and see below) may be a necessary factor in the development of the terrorist for some (and that is an empirical issue) for those same people it is not a sufficient account – logically it cannot be.
Radicalization is all around us. Terrorism is not.
The signs of radicalization are easy to spot.
The signs of a ‘terrorist in development’ are not.
Spotting radicalization as a means of identifying terrorists in the making? Well, however we define radicalization (whether wholly in terms of attitudes/opinions or also in terms of a kind of action that is in some way related to the former), that seems to spell serious trouble.
Lots of myths have been prevalent in terrorism research over the years. The idea that we can profile terrorists is probably the most famous of these (we still can’t get a terrorist profile, but we can explore interesting ways of differentiating terrorist behavior with studies like this – and no, differentiating different kinds of terrorists via analyses of their behavior isn’t really ‘profiling’ under another name. It’s behavioral analysis.).
The new myth, in my view, is that to prevent or in some way counter the development of the terrorist, we must address radicalization.
The importance of knowing radicalization
Radicalization has many definitions, and many detailed and thoughtful examinations of this subject. Believe it or not, it was even something that people were talking about before the “T” word came into focus.
Thought it might seem invidious to highlight individual contributions, a notable recent example is that of “Friction: How Radicalization Happens to Them and Us”. That book (and its revisions through several subsequent papers – all of which are critical to read), by scholars Clark McCauley and Sofia Moskalenko identifies twelve mechanisms contributing to radicalization. A major feature, the authors warn, is the need to understand that radicalization is not just a process, but one that feeds from the reaction to it. The distinction here is between a linear process, and a more complex, interactive process: an important quality of McCauley and Moskalenko’s framework – but while we are used to understanding complex feedback and multiple operating systems in say, biological processes, it seems odd that we seem to be blind to understanding this in the context of terrorism.
In attempting to respond to radicalization, we change it, sometimes for worse rather than for the better. Furthermore, radicalization, McCauley and Moskalenko, suggest, offers a way to understand the development of terrorism (more on that below).
And further still, a critically unexplored facet of radicalization in the words of J.M. Berger, is how it remains ‘underweighted’ outside the context of terrorism. The broader and deeper significance of “Friction” is that the mechanisms that give rise to radicalization (e.g. love, fear, grievances etc.) are far more common outside the terrorism process. That we seem to primarily discuss radicalization today in the context of terrorism is itself a problem, and one that has barely been recognized.
The issue for me is not, nor has it ever been, about radicalization per se. I am not suggesting radicalization is a myth, nor am I attempting in any way to devalue the work of far smarter people than I who research radicalization. My concern, and one I have long communicated is that in our attempt to prevent terrorism, we have focused on identifying and pre-empting radical views, attitudes and opinions. We certainly seem to have become so intolerant of expressing radical alternative views of society.
We can find plenty of exceptions where radicalization (both attitudinal as well as actionable) does not appear to precede engagement in terrorism, but the overriding narrative that has taken hold in discussions about national security is that to understand the development of the terrorist, we must understand that person(s)’s radicalization process as if it were a means of preventing terrorism.
The questions that arose in the wake of the Boston bombings were typically: “what was their radicalization process? How did they become radicalized? Where did they become radicalized? Were they self-radicalized? Did someone else radicalize them?” and so on.
Detecting radicalization has implicitly become a proxy for detecting the development of terrorist behavior.
As a result, I think there has been a loss of focus in prioritizing the problem set.
I see this in terrorism research, academic journal articles, books, reports, popular media, policy discussions, law enforcement and intelligence discussions. It seems to be an idea that has taken root. And why not? Addressing radicalization is plausible, logical and reduces the complexity of the problem of terrorism into something we believe we can a) identify and consequently b) root out, displace, prevent, counter, or whatever “intervention” word is in fashion.
Radicalization might be seen as a particular kind of “motivating state”, which with other things might be an element in, say, the Boston story, but it is not “the” story. When it exists at all, it is always only a part of an explanatory account.
In my 2005 Psychology of Terrorism, I proposed we could understand the development of the terrorist first and foremost by focusing on three stages:
IED – easy to remember, right? There are MANY models of radicalization out there. Some are about radicalization – a few of these focus only on ideas, opinions and attitudes. Others focus on behaviors. Others still attempt to examine the relationship between radical action and radical behavior. Some claim to be about both radicalization and terrorism. Some use radicalization as a proxy for understanding involvement in terrorism, viewing these as essentially synonymous. Some of these models are comprehensive, some actually draw on data, while others are little more than neat metaphors – they are memorable, but don’t stand up to serious scrutiny.
There are quite a few, though I don’t have the time or space to address all of them here (I do this in my forthcoming 2nd ed.), but the reason my research currently focuses primarily on those three IED phases is that if our goal is to prevent terrorism, I don’t see a convincing reason for focusing on “radicalization” to help achieve that.
Let me unpack this a little. On those three phases of involvement, engagement and disengagement (conceptual distinctions long recognized in criminology by the way and by terrorism researcher Max Taylor in the early 1980s) a critical starting assumption I embraced is that the reasons that help us understand why someone becomes involved in terrorism do not necessarily help us understand how, why and when someone engages in an actual act of terrorism. They might, but not necessarily.
From those three stages, we can identify three potential phases for prevention:
a) Preventing initial involvement in terrorism
b) Disrupting or pre-empting the development and execution of terrorist attacks
c) Facilitating or promoting disengagement.
Where does “radicalization” fit into this?
Well – here’s the rub. To be fair – it might depend on what we think radicalization is, but one common view (and despite slight differences in emphases) is that “radicalization” generally refers to a process of increasing social and psychological commitment to ideas/ideological content that in and of itself is viewed as conducive/supportive of violence. The broad logic being – if you become “radicalized” by embracing ideas, whether through person to person contact, the acquisition of material deemed (at least in some jurisdictions) illegal or problematic (e.g. Inspire) that even the acquisition of said material puts one at greater risk of engagement in terrorism. So “radicalization” seems to refer to the process by which one becomes “radicalized”. If one is in a state of being “radicalized”, the logic goes, one is at risk of engagement in terrorism.
There are other definitions of radicalization (I will consider another below) and these give rise to different implications. But let’s first acknowledge and accept this one view that the solution to the debate about the relationship between radicalization and terrorism might be solved with greater clarification around what is meant by ‘radicalization’. @ThoughtsonYemen reached the same conclusion.
Let’s consider a different position. Let’s accept for the moment that actually, what is meant by “radicalization” is far broader than just “holding particular kinds of opinions” and instead it is something accepted as “the process by which people become involved in terrorism”. Now let’s imagine that such a view of radicalization was broadly held. Does that mean that these problems are solved?
There is widespread acceptance that there are many more people that are ‘radical’ than there are those who engage in terrorism. The numbers of those who engage (or even attempt to engage) in terrorism remains relatively tiny. So there is already a tacit acknowledgement of the key distinctions between these terms. What is not understood is the implication of that distinction.
Additionally, thinking about this a little closer begs the further question as to whether radicals who do not participate in terrorism represent some kind of meaningful control group? It brings with it a deeply worrying assumption that all radicals are therefore potential terrorists. And that is an assumption that takes us way beyond issues of nomenclature.
These are just some of dozens of conceptual questions raised by many, but let me try to move the debate on by anchoring this in more specific questions. Back to the above – “where does radicalization fit into discussions about terrorism?”
Well, it fits into at least two places.
1. Radicalization might be related to the reasons for initially becoming involved in terrorism.
2. Radicalization might also be relevant for understanding how and why someone might choose to disengage from terrorism.
Yet, this is not so clear as it might first seem.
As above, the overwhelming majority of those who are radicalized do not engage in terrorism.
There is plenty of evidence that those who become involved in terrorism either a) are not ‘radical’ to begin with and/or b) become ‘radical’ as a consequence of increased engagement in terrorism (either through activity, or through increased social involvement with others who themselves display commitment to a group, cause or movement). If we look beyond the narrow lens of involvement and engagement in terrorism, plenty of examples of these kinds of cases abound.
After “Friction”, McCauley and Moskalenko moved beyond a definition of radicalization that as only a matter of opinion/attitude. What I would like to do is to offer some additional thoughts, drawing also on how McCauley and Moskalenko conceptualize and differentiate actions and attitudes, on how we might begin to resolve this conceptual and empirical confusion. Though these issues might seem purely academic, they are absolutely not without immense policy significance. More on that later.
In a forthcoming paper by Dr Lily Cushenbery, Mr Casey Hilland and myself, we identify and explore four broad ‘types’ of case commonly found in terrorism.
- The person who is radicalized (meaning here the acquisition, embracing and expression of particular ideas, opinions and attitudes) and THEN becomes engaged in terrorism
- The person who engages in terrorism, and THEN becomes radicalized
- The person who engages in terrorism WITHOUT being radicalized (prior to, or after engagement)
- The person who is radicalized but does NOT become involved in terrorism.
These distinctions offer only a starting point, and we can achieve far greater granularity if we explored this further, but a critical framing question would be to examine the representativeness of each of those ‘types’ in even a specific sample or context.
McCauley and Moskalenko acknowledge the differences here by differentiating between what they call their “narrative pyramid” and their “action pyramid”, stressing in a 2010 article produced for the Dutch National Coordinator for Counterterrorism (NCTb) that “acceptance of narrative elements is correlated with levels of action, such that accepting a personal moral obligation for jihad…is most likely among the terrorists and least likely among the inert. Similarly, belief in none of the aspects of the Global Jihad narrative…is most likely among the inert and least likely among the terrorists…But the correlation is only probabilistic, not deterministic. A few individual jihadist terrorists may accept no part of the Global Jihad narrative – for instance an individual who joined a terrorist group for the thrill of guns and fighting. And there may be a few politically inert individuals who feel a personal moral obligation for jihad – for instance an individual who does not want to hurt his parents by leaving for jihad…it is neither obvious nor known what parts of the Global Jihad narrative appear with what frequency in different levels of the action pyramid. Mechanisms of radicalisation that do not depend on ideology or narrative imply that the Global Jihad narrative is not necessary for radicalisation in action.(p.65) [emphasis mine] (see Leuprecht, C., Hataley, T., Moskalenko, S., and McCauley, C. ‘Narratives and Counter-Narratives for Global Jihad: Opinion versus Action’, Countering Violent Extremist Narratives, National Coordinator for Counterterrorism).
To illustrate, a recent example from the literature of when ideology is mostly learned “after activism” is conveyed in Ziad Munson’s “The Making of Pro-Life Activists: How Social Mobilization Works” (Univ. Chicago Press, 2008). In Clark McCauley’s own review of Munson’s work, he cautions that “The Making of…”: “issues a major challenge to the idea that radical ideology is the center of gravity of jihadist terrorism…militants are not distinguished by their radical goals but by their radical means” (C. McCauley, ‘Does political radicalization depend on ideology?’ Dynamics of Asymmetric Conflict, 2008, 2:3, p. 215).
Let me be clear. I am not attempting to ‘win’ an argument by setting up radicalization as a straw man – that it is only about attitudes or opinions. But both for radicalization as well as for terrorism, we lack any meaningful (i.e. reliable and valid) behavioral referents for the terminal state – i.e. “terrorist behavior”. Lots of different kinds of behavior seem to fall within the category of terrorist behavior, and in the absence of any kind of behavioral differentiation, how can any causal mechanism (i.e. terrorism is caused by radicalization) be meaningful for multiple behaviors? Terrorist behavior, at least legally, encompasses a broad variety of different activities, as does, it seems, radicalization.
The broader question becomes whether radicalization is a meaningful explanatory mechanism for understanding the development of the terrorist. And it is at this point I see greater value in making far more explicit not just what we think “radicalization” should be, but how we operationalize radicalization in studies of terrorist behavior.
The real danger that follows closely from this is unless we demand greater specificity, we run the risk of allowing radicalization to become an ‘explanatory fiction.’ And furthermore, I think we can find some clarification here via a distinction that has long recognized in the social and behavioral sciences, and elegantly discussed in Psychological Record’s 1948 now classic paper “On the Distinction Between Hypothetical Constructs and Intervening Variables” by Kenneth MacCorquodale and Paul E. Meehl.
More to follow in PART 2.