Radicalization and Political Violence

Rolling Stone has published a new article entitled “Everything You’ve Been Told About Radicalization is Wrong.” It is primarily an attack on the NYPD’s study Radicalization in the West: The Homegrown Threat, written by Mitchell D. Silber and Arvin Bhatt, but it more broadly makes the bold claim that there is no causal connection between radicalization and violence. (I will define some of these terms subsequently — something that Rolling Stone failed to do, and that certainly contributed to the piece’s lack of clarity.)

In studying what drives people to undertake terrorist violence, it is extremely important to be open to new ideas, and new ways of thinking about the issue. Further, I have my own criticisms of the NYPD’s model, which I have already outlined at Al-Wasat. But the major problem with Rolling Stone‘s argument is that it is a broadside against entire lines of inquiry without actually presenting evidence that these ways of thinking about the problem set are flawed.


It is worth beginning any discussion of radicalization by exploring what the concept refers to. I’ll offer two definitions. Since Rolling Stone is attacking the NYPD’s study, we might start with how that study conceptualizes the process of radicalization.

The NYPD’s study views salafi jihadist ideology as “the driver that motivates young men and women, born or living in the West, to carry out ‘autonomous jihad’ via acts of terrorism against their host countries.” Radicalization, according to that study, is a four-step process (sequential though not necessarily linear) that terminates in a final step that it refers to as jihadization. NYPD defines this phase:

Jihadization is the phase in which members of the cluster accept their individual duty to participate in jihad and self-designate themselves as holy warriors or mujahedeen. Ultimately, the group will begin operational planning for the jihad or a terrorist attack. These “acts in furtherance” will include planning, preparation and execution.

NYPD’s study notes that “individuals who do pass through this entire process,” to the jihadization phase, “are quite likely to be involved in the planning or implementation of a terrorist act.”

Now, if I were critiquing the NYPD’s study, I would argue that the definition of jihadization makes the study’s conclusion almost tautological. Is it any surprise that people who “accept their individual duty to participate in jihad and self-designate themselves as holy warriors,” and then take the extra step to “begin operational planning” for a terrorist attack are “quite likely to be involved in the planning or implementation of a terrorist act”? But, oddly, Rolling Stone takes the opposite tack. It actually argues that there is no connection between completing the steps in NYPD’s study and the propensity to undertake violence.

Moving away from the NYPD’s definition of radicalization — which, as I said, seems to contain a tautology — let’s also employ a more generalized definition of radicalization. I find the definition offered by Peter Neumann and Brooke Rogers useful: “radicalisation describes the changes in attitude that lead towards sanctioning and, ultimately, the involvement in the use of violence for a political aim.” In other words, under this definition radicalization refers to various changes in attitude that culminate in the idea that using violence for political aims is acceptable. So if we’re querying whether there is a causal connection between radicalization and political violence, the question is if embracing an ideology that holds political violence to be acceptable or even required makes one more likely to engage in political violence, or if it makes no difference.

Finally, I should mention that there is no single pathway to terrorism: this discussion is not about whether ideological radicalization is the only cause of terrorist violence. As I wrote on Al-Wasat earlier, sometimes “political anger, group dynamics, even sense of adventure” may be the dominant factor driving people to violence. Rather, this discussion is about whether ideological radicalization is one such pathway to terrorism, or whether that causal connection is a myth.

John Horgan: “The Greatest Myth Alive”

“The idea that radicalization causes terrorism is perhaps the greatest myth alive today in terrorism research.”–John Horgan

I recognize Horgan as a talented academic, albeit one whose past work has sometimes had a blind spot with respect to the role of religious ideas. So I was interested to see the evidence for such a bold claim. But none of the points he made to Rolling Stone establish that the connection between radicalization and terrorism is a myth. I engaged Horgan on Twitter, and he provided two different sources that he said grounded his argument in far deeper research than is reflected in the Rolling Stone piece. But neither of those sources support his strong conclusion either.

Before addressing Horgan’s arguments, let me point out that the idea there is some causal connection between radicalization and terrorism is one of the more intuitive connections in all of terrorism research. One would expect that someone who believes in an ideology that justifies the use of violence for political ends would be more likely to engage in politically-motivated violence than someone who does not. Of course, sometimes the intuitive answer is not the right one — so let’s look at Horgan’s evidence.

Arguments to Rolling Stone. Horgan provides two arguments to Rolling Stone about why the idea that radicalization causes terrorism is a myth. First, he says that “the overwhelming majority of people who hold radical beliefs do not engage in violence.” I agree with this, and in fact made a similar point in a recent CNN interview. (When I said this, I was defining “radical beliefs” in a different way than the NYPD does in its jihadization phase, and Horgan is almost certainly employing a different definition as well.) But the point that most people who hold radical beliefs don’t engage in violence doesn’t disprove a causal connection. I find that, because people tend to be uncomfortable discussing religion, they have a far lower threshold for accepting evidence discounting the connection between religious ideas and terrorist violence than they might in other contexts; I noted this tendency in a recent review of a book by Clark McCauley and Sophia Moskalenko. So let’s take this out of the context of religion and examine parallel claims and refutations:

  • Poverty is a causal factor in crime. “That’s not true. The overwhelming majority of poor people do not commit crimes.”
  • The widespread availability of firearms is a causal factor in gun violence. “That’s not true. The overwhelming majority of gun owners do not commit gun violence.”
  • Government repression is a causal factor in revolutions. “That’s not true. The overwhelming majority of repressive governments do not experience revolutions.”

In these cases, the refutations do not disprove the initial claims. Similarly, Horgan’s statement makes the point that there is hardly a one-to-one relationship between radical ideas and terrorism. But that doesn’t mean there is no causal connection.

Horgan’s second argument is that “there is increasing evidence that people who engage in terrorism don’t necessarily hold radical beliefs.” I asked him about this on Twitter, and he explained that “it’s not that they don’t hold extreme beliefs. It’s that the beliefs don’t always precede involvement.” This is obviously a different claim than his quotation in Rolling Stone. Since I don’t have access to the data he’s referring to, I can’t really speak to it, except to make one point: the fact that in some cases involvement in a violent extremist movement doesn’t precede beliefs does not disprove a causal connection between radicalization and terrorism.

Horgan also makes a third point later in the Rolling Stone article: “There are the bigger social, political and religious reasons people give for becoming involved… Hidden behind these bigger reasons, there are also hosts of littler reasons – personal fantasy, seeking adventure, camaraderie, purpose, identity. These lures can be very powerful, especially when you don’t necessarily have a lot else going on in your life, but terrorists rarely talk about them.” But as Adam Elkus pointed out on Twitter, multicausality is part of social science. While Horgan’s point is, again, true, it doesn’t substantiate his claim that the causal connection between radicalization and violence is a myth.

Terrorism and Political Violence. Horgan pointed me to his article in Terrorism and Political Violence (Jan. 2007) entitled “A Conceptual Framework for Addressing Psychological Process in the Development of the Terrorist.” It was an interesting and worthwhile read, but did not demonstrate that the causal connection between radicalization and terrorism is a myth. Here is what that article had to say about ideology and terrorist violence:

The influence of ideology in a psychological sense has been little explored with respect to terrorism. As noted earlier, Hall describes ideology as ‘‘the framework for thinking and calculations about the world—the ‘ideas’ that people use to figure out how the social world works, what their place is in it and what they ought to do.’’ In a psychological sense, this implies influences on an individual’s cognitive state, a point noted by Aaron T. Beck in his paper on terrorism: ‘‘Ideology concentrates their thinking and controls their actions.’’ For Beck, ideology implies not only cognitive influences, but that it operates as a process, changing or ‘‘controlling’’ behaviour; ideology therefore might be expressed in forms of cognitive structure, but it also has a sense of content that may exercise significance influence over behaviour. From a different context, as noted earlier, Louis Althusser also emphasises the significance of meaning and representation expressed through language and social action and practice. Taylor and Taylor and Horgan explored the role of rule governance as one way of understanding the behavioural process that might be involved in ideological control over behaviour, and it may be that further empirical exploration of this would yield valuable results.

Far from implying a lack of causal connection, this passage thus implies that further research into the connection between ideological radicalization and terrorism could “yield valuable results.” So did something change in the six years since his article came out to change his mind about that point?

START article. A second article Horgan pointed me to, published by the University of Maryland’s START center last year, speaks to that question. In it, Horgan argues that we should “end our preoccupation with radicalization so that we can effectively regain a focus on terrorist behavior.” This article is primarily concerned with problems with practically applying the concept of radicalization. Nowhere does it provide evidence that there is no causal connection between radical beliefs and violence. In fact, it addresses his above-discussed point that involvement may precede beliefs. He writes:

A lingering question in terrorism studies is whether violent beliefs precede violent action, and it seems to be the case that while they often do, it is not always the case. In fact, the emerging picture from empirical studies of terrorists (including over a hundred terrorists I have interviewed from multiple groups) is repeatedly one of people who became gradually involved with a terrorist network, largely through friends, family connections, and other informal social pathways but who only began to acquire and express radical beliefs as a consequence of deepening involvement with a network.

There are two things worth noting about this. First, it is useful that he quantifies that radical beliefs “often” precede violent action. Second, Horgan’s statement is entirely consistent with the conclusions in the NYPD study, which speaks of the role of social networks:

  • The key influences during this phase of conflict and ‘religious seeking’ includes trusted social networks made up of friends and family, religious leaders, literature and the Internet.” (p. 30)
  • “Clusters of like-minded individuals begin to form, usually around social circles that germinate within the extremist incubators.” (p. 31)
  • “These groups, or clusters of extremists … are not ‘name brand’ terrorists or part of any known terrorist group. For the most part, they have little or no links to known militant groups or actors. Rather they are like-minded individuals who spend time together in clusters organized, originally, by previously established social network links.” (p. 85)

Overall, Horgan’s paper makes some valuable points about practical applications of the concept of radicalization, and about some of the lack of clarity some scholars have when discussing radicalization. While I disagree with some points that Horgan makes in his paper, it is a) a valuable read, and b) one that does not support the very strong claim he made about the causal relationship between radicalization and terrorism being a myth.

Rolling Stone‘s Other Arguments

I examined Horgan’s ideas in depth because, as I said, I respect him as a scholar. Though none of the rationales he gave support his strong claim that “the idea that radicalization causes terrorism is perhaps the greatest myth alive today in terrorism research,” I wanted to examine them in detail and point out where they contribute valuable critiques to our discussion of radicalization as a concept. The rest of the article provides less meat, although it does contain one critique with which I agree.

The critique I agree with comes from Jamie Bartlett, who states: “I have found that many young home-grown al-Qaeda terrorists are not attracted by religion or ideology alone – often their knowledge of Islamist theology is wafer-thin and superficial – but also the glamour and excitement that al-Qaeda type groups purports to offer.” As I said in my last Al-Wasat post on radicalization, “an on-point criticism of the NYPD study … is that it assumes the primacy of ideology (religious or otherwise) in moving an individual toward the embrace of violence.” Ideology, as Bartlett notes, is not always the primary moving force. However, this doesn’t mean that the study and the concept of radicalization lack value: I find that the NYPD’s study provides a useful conceptual framework in cases where ideology is the predominant pathway toward undertaking terrorist violence. This seemingly includes the case of Tamerlan Tsarnaev, thus making this a strange time for Rolling Stone to launch a broadside against the concept of radicalization.

Other critiques that Rolling Stone offers:

If media accounts are to be believed, the accused Boston marathon bombers were “radicalized” by watching American-born Yemeni cleric Anwar al-Awlaki’s YouTube sermons and reading Inspire, the al Qaeda magazine.

False. My last Al-Wasat entry goes through media accounts carefully. This is not the impression one would come away with after reading my entry. If Rolling Stone‘s contention is that you’d get this impression by reading narrowly and selectively within the media accounts… well, okay, but not the best critique.

Jamie Bartlett, head of the Violence and Extremism program at the think tank Demos, echoes these doubts. “The word ‘radicalization’ suggests a fairly simple linear path toward an ultimate violent conclusion,” he says. Studies suggest that although there may be stages in the evolution of a terrorist, placing them sequentially on a line, as the NYPD’s report literally does, is far too pat. The stages are fluid, not a simple trajectory, and it is virtually impossible to predict who will or won’t engage in violence based solely on their beliefs.

Ignores what the NYPD study actually says. That study clearly states (p. 19): “Each of these phases is unique and has specific signatures associated with it. All individuals who begin this process do not necessarily pass through all the stages and many, in fact, stop or abandon this process at different points. Moreover, although this model is sequential, individuals do not always follow a perfectly linear progression.”

As I flagged before, the idea that “it is virtually impossible to predict who will or won’t engage in violence based solely on their beliefs” is rather absurd if those beliefs are represented by the “jihadization” phase of the NYPD’s study. Once individuals “accept their individual duty to participate in jihad and self-designate themselves as holy warriors or mujahedeen,” it is probably a bit less difficult to predict who will or will not engage in violence. Further, while the ability to predict violence is relevant the utility of radicalization models, it does not really speak to the truth of the models. In other words, a radicalization model can simultaneously accurately describe the process leading one to commit violence while also doing little to help us distinguish between those who will drop out and those who will not.

“To be a radical means to reject the status quo, which in some cases propels society forward,” says Bartlett. “Equating radicalism with terrorism can produce a dampening effect on free expression – either by government or by self-censorship.”

Employs a different definition of radicalism than either the NYPD study or Neumann and Rogers. Salafi jihadism, the focus of the NYPD study, has never propelled any society forward.


The Rolling Stone piece also mixes in a great deal of criticism of the NYPD’s policing efforts. This is a fair area for discussion, but even if one concludes that the NYPD’s efforts were wrong in every way, the policies undertaken by NYPD do not invalidate the concept. Again, taking this outside the context of religion and political violence, let’s examine a few similar claims to Rolling Stone‘s contention that the connection between radicalization and terrorist violence is invalid because it led to what the magazine considers to be bad, discriminatory policies:

  • A carbon tax would be an economically disastrous, awful policy. Therefore, the idea that carbon dioxide causes global warming is invalid.
  • Gun control would put us on the road to dictatorship. Therefore, the idea that the availability of firearms is linked to violence is invalid.
  • Abortion is a positive evil, akin to murder. Therefore the idea that an abortion can ever be in the health interests of a mother is wrong.

Rolling Stone‘s readers would, I suspect, disagree with all three of these propositions, and see the logical flaws in them. The fact that you might oppose a carbon tax does not prove that global warming is a myth. Likewise, one’s views of the NYPD’s policing practices do not invalidate the connection between radicalization and terrorist violence. Overall, there may be a strong argument that, as Horgan suggests, there are serious problems with applying models of radicalization in the law enforcement context. But that is a far different argument than the contention that everything we’ve been told about radicalization is wrong.

Let me close by explaining, briefly, why the connection between extreme ideas and violence matter. Last month I did field research in Tunisia, where there has been an alarming amount of vigilante violence undertaken by hardline salafis against artists, activists, women, and religious minorities. If there were no connection between extremism and violence, we would expect vigilante violence to be evenly distributed within Tunisian society–undertaken sometimes by secularists, sometimes by Communists, sometimes by sufis. It is not. And if radical ideas and terrorism are not connected, then the growth of Ansar al-Sharia Tunisia, which openly shares al-Qaeda’s ideology, should not be seen as particularly problematic. After all, since it is not currently engaged in a terrorist campaign, its ideology should not be seen as predictive of future violence — right?

If we pretend that a connection between certain radical ideas and political violence does not exist, we will be taken by surprise, time and again, by future acts of non-state violence.

About Daveed Gartenstein-Ross
Senior fellow at the Foundation for Defense of Democracies; adjunct assistant professor in Georgetown University's security studies program.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: