One of the most frequently asked questions about terrorism is also the most intractable. Why? Why do they do it? Why do people join terrorist groups and participate in acts of terrorism?

There are as many answers to this question as there are terrorist groups, and everyone from clerics to caustic cab drivers seems to have a confident opinion on the subject, as though the interior world of terrorists can be easily mined and mapped. But this confidence is often misplaced, given how little scholars actually know about terrorism and the people who are involved in it. It also betrays an epic obliviousness about just how difficult it is to access the internal, subjective desires and emotions that shape the outer world. Instead of asking why people join terrorist groups and commit terrorist atrocities, a more worthwhile starting point for explanation is to ask how.

One culturally prevalent answer to the why question is that terrorists are “driven” or “pushed” to do it, and that the decisive driving or pushing agent is pathology. This answer has evolved in recent years in line with advances in knowledge and moral sensibilities. In terrorism studies in the late 1960s, it was not uncommon for scholars to conceive of pathology as a psychological abnormality or affliction rooted inside the individual. Since the 1980s, this idea has fallen into disrepute, and the scholarly consensus now holds that the roots of terrorism lie not in the individual, but in the wider circumstances in which terrorists live and act.

This reflects a broader consensus in the social sciences about violence: namely, that it is “socially determined,” a product of deeper historical, economic, or cultural forces over and above the individual. It is perhaps best summarized by the renowned social psychologist Albert Bandura. Drawing on studies of violence from across the human sciences, Bandura concluded that “it requires conducive social conditions rather than monstrous people to produce atrocious deeds. Given appropriate social conditions, decent, ordinary people can be led to do extraordinarily cruel things.” Social scientists argue about the nature and impact of the “social conditions” in question, but few would question the essential point that violence, however personalized or idiosyncratic its expression, is primarily rooted in historical structures or social relationships, not individuals, still less their “pathological” mindsets.

This consensus is also reflected in much liberal-left commentary about terrorism, especially of the jihadist variant. For example, in some quarters of the “radical” left it is asserted that the roots of jihadist terrorism lie not in Islam but in the myriad historical crimes and injustices of Western, and specifically U.S.-driven, imperialism—most notably, in the post-9/11 era, the 2003 invasion of Iraq. Jihadist violence, from this perspective, is an inevitable reaction fueled by Muslim anger and vengeance; and Westernized jihadists, far from rejecting the civilized norms and ideals proclaimed by the West, are in fact alienated from a West that excludes, demeans, and harasses Muslims.

The scholarly consensus on violence has a lot going for it. It humanizes the perpetrators of violence by insisting on their ordinariness and contextualizing their actions. It obliges people to reflect on their own possible shortcomings and vulnerabilities, and how, in different circumstances, they too could do monstrous deeds. And it compels people to recognize that they do not act in a social vacuum, and that what they think, feel, and do is powerfully shaped by the broader historical circumstances in which they are compelled to live and act. Moreover, Westernized jihadists, as a recent report cogently suggested, assuredly are alienated and feel that they do not belong in a secular world that often mocks and challenges their religion and identity as Muslims.

But the consensus can’t divest itself of the idea of pathology. Rather, it simply relocates the notion, tracing the causes of violence to pathological “background factors” operating on the violent. No doubt this is a more illuminating and edifying narrative than that sketched out in earlier psychological accounts. But its explanatory power is limited, because, as the eminent sociologist Jack Katz has convincingly argued, “whatever the validity of the hereditary, psychological, and social-ecological conditions of crime, many of those in the supposedly causal categories do not commit the crime at issue, ... many who do commit the crime do not fit the causal categories, and ... many who do fit the background categories and later commit the predicted crime go for long stretches without committing the crimes to which theory directs them.” Or as the British writer David Aaronovitch once joked, “Why don’t black lesbians blow up buses? Aren’t they alienated enough?”

One of the most sensitive and profound explorations of terrorism in recent years comes not from a scholar, but from a novelist. Philip Roth’s American Pastoral is a murder mystery in which the focal point is not the who, but the why. The protagonist, Seymour Levov, is a successful businessman whose 16-year-old daughter Meredith (“Merry”) blows up a post office to protest the Vietnam War, killing a bystander. All Seymour can think about is why Merry did it. She was an adored only child who grew up in a privileged and decent family in the idyllic hamlet of Old Rimrock, New Jersey. Seymour is desperate to locate “the wound” that caused Merry’s violence. Was it her stutter? Was it that anomalous kiss on the mouth he gave her one summer when she was 11 and he 36? Or was it the mysterious firebrand Rita Cohen who radicalized her?

As the novel progresses, Seymour’s disbelief gives way to clarity. But it is a negative clarity. “He had learned the worst lesson that life can teach—that it makes no sense,” Roth writes. He had learned that his daughter “was unknowable,” and that “there are no reasons,” that “reasons are in books.” In capturing Seymour’s efforts to understand the disaster that befalls his family, Roth holds up to scrutiny conventional efforts to explain terrorism—and exposes just how imaginatively cramped and simplifying they can be.  

Do terrorists have their reasons for committing atrocities? They certainly reel off any number of reasons in their pronouncements, but, as law professor Stephen Holmes has observed, “private motivations cannot always be gleaned from public justifications.” Sometimes people do what they do for the reasons they profess. Sometimes not, because what they do is motivated by reasons that are too dark, shameful, or bizarre to be openly acknowledged. Sometimes people do things that are so morally contentious that when called to account they are liable to excuse or justify, rather than to explain, their actions. Terrorists unquestionably fall into this category.

And sometimes people do what they do without the slightest sense of knowing why. I once met someone who robbed a liquor store in his teens. He was caught and did jail time for it. This person is now an accomplished writer. Doing that stick-up was a hinge moment in his life and today, some 30 years later, he still cannot make sense of it. The motive simply eludes him.

Terrorism scholar John Horgan has made a similar point. “The most valuable interviews I’ve conducted [with former terrorists] have been ones in which the interviewees conceded, ‘To be honest, I don’t really know,’” he writes. “Motivation is a very complicated issue. To explain why any of us does anything is a challenge.” It’s a challenge further compounded by the fact that some actions are informed by multiple motives, and even if these can be reliably identified it is often difficult to disentangle them and calculate their respective causal weight.

As Horgan suggests, a more manageable and useful question to ask about terrorism is not why, but how—and when and where? How did this specific person come to join this specific organization? What networks helped facilitate the act of joining, and where and how were these networks accessed or sought out?

Because these questions are about the circumstances of terrorism, and not the interior world of terrorists, they are not only more intellectually tractable for scholars, but also more directly relevant to efforts to prevent or stop terrorist recruitment. Law-enforcement agents can’t disrupt a motive, but with the right intelligence and skills they may be able to disrupt a network of terrorist recruiters. Marc Sageman’s work on Western “leaderless” jihadists demonstrates the promise of this kind of approach. Although Sageman has some interesting things to say about the why question, the strength of his research lies in showing just how decisive social and kinship networks are in the radicalization process.

This isn’t to suggest that the why question should be abandoned, but rather that those who ask it better appreciate the magnitude of the question and acknowledge the possibility that some momentous life decisions will remain forever opaque and mysterious—not only to outside observers, but also to the people who take them and must live with the consequences.