In September 2014, when the Islamic State (ISIS) was at the height of its power, Director of U.S. National Intelligence James Clapper acknowledged that the United States had underestimated the terrorist group’s will to fight. “We underestimated the Viet Cong and the North Vietnamese and overestimated the will of the South Vietnamese,” he told The Washington Post. “In this case, we underestimated ISI[S] and overestimated the fighting capability of the Iraqi army … It boils down to predicting the will to fight, which is an imponderable.”

Scholars and policymakers have long sought to determine what drives people to keep fighting when the chips are down, and, if need be, to give their lives to a cause. Traditional explanations, based on rational choice theory or focused on mental abnormalities, have largely failed to explain what motivates the members of extremist insurgent movements. But Clapper was wrong to suggest that the will to fight is imponderable. In fact, it is possible to predict who is willing to fight and die, based on a combination of cultural and psychosocial factors. Research on the human brain suggests that people fight when their sacred values—that is, the values that define their identity and therefore can’t be compromised—are under threat.

In a series of behavioral studies of frontline combatants in Iraq and brain imaging studies of self-identified extremists in Spain, research teams of which I am a part at Artis International, Oxford University, and other partner universities found that those most willing to make costly sacrifices, including fighting and dying, were motivated by sacred values and shunned deliberative reasoning. Even more consequential for policymakers working to prevent and counter radicalization, we found that social exclusion and political marginalization heightened the importance of sacred values—and even caused non-sacred values to mimic sacred ones—increasing people’s willingness to fight and die for those values.


Despite what the term suggests, sacred values are not necessarily those enshrined in holy books. They can stem from country, religion, family—any personal or collective value that defines identity. Previous research focusing on Native North American, Mayan, Palestinian, and Israeli populations indicates that sacred values are strongly resistant to negotiation. Costs, consequences, risks, and rewards—all the motivators on which classic economic and political theory rely—don’t seem to matter when sacred values are involved. Most people hold sacred values but become aware of them only when the values come under threat. People see their sacred values as immutable. They find it easy to cooperate unconditionally with people who share them and just as easy to fall into conflict with those who don’t.

To test the relationship between sacred values and violence, we interviewed and surveyed captured ISIS fighters as well as fighters from the Kurdistan Workers’ Party (PKK), the Iraqi army, and Sunni Arab militias, among others, in 2015 and 2016. We asked these fighters to identify their important values, and we rated the relative strength of those values by asking if the fighters would exchange them for material benefits. We treated an absolute refusal to contemplate a tradeoff as indicative of a sacred value. We found that ISIS fighters and Kurdish fighters from the PKK, which the U.S. government considers a terrorist organization, were more likely to consider the cause they were fighting for (the caliphate, for ISIS, and a Kurdish homeland, for the PKK) sacred than were the soldiers from the Iraqi army or Sunni Arab militias. Moreover, the ISIS and Kurdish fighters expressed a greater willingness to fight than the other militants did. Some even said they would leave their families in harm’s way in order to defend their sacred causes—an excruciating sacrifice that we observed some of them make during the course of the study.

ISIS and Kurdish fighters expressed a greater willingness to fight than the other militants did.

For the ISIS and Kurdish fighters we interviewed, spiritual strength was more important than material strength, such as manpower or firepower. The fighters expressed willingness to march into battle knowing they were outgunned, because they believed in the sacredness of their cause. This calculus wasn’t rational but, rather, born of an ingrained sense of duty that resembled instinct. As one Kurdish fighter remarked of ISIS, “They are weak now because they have used up their resources, but their fighters don’t retreat even if the battle is lost.”


Having established a connection between sacred values and willingness to fight, we sought to determine if sacredness itself was ever malleable: Could a value be non-sacred one day and become sacred the next, or vice versa? Given the link between sacred values and violence, the ability to shift the boundaries of sacredness could be extremely useful. Previous studies had indicated that social exclusion intensifies extreme behavior and increases the risk of radicalization, so we designed a study to test if such exclusion could strengthen people’s attachment to sacred values or even cause them to assign sacredness to previously non-sacred values. It was a two-stage social experiment in which the participants—38 men living in Barcelona, Spain, who expressed willingness to perform violent acts associated with jihadism—agreed to answer questions, play a virtual ball game, and have their brains scanned.

In the first stage, we asked the participants about the sacredness of their values, again ranking these values according to how willing the participants were to trade them for material gain. The value propositions we ranked ranged from forbidding images of the Prophet Muhammad to preventing gay marriage to requiring women to wear the veil. While most of the participants considered forbidding images of the Prophet and preventing gay marriage to be sacred values, few of them felt the same way about requiring women to wear the veil.

Social exclusion appeared to cause non-sacred values to mimic sacred values, both in terms of brain activity and willingness to fight and die to protect them.

In the second stage of the study, the participants played “Cyberball,” a virtual ball-tossing game, against other players with Spanish names. During the game, half of the study’s participants were abruptly excluded from passing the ball. After the game, all of the participants answered questions about their willingness to fight for the previously discussed values while their brains were scanned. When asked about sacred values, all participants showed greater activity of the left inferior frontal gyrus, an area of the brain associated with non-rational thinking. Included and excluded participants alike expressed greater willingness to die for sacred values, compared to non-sacred values, indicating that traditional cost-benefit analyses were not driving their reactions. When the participants were asked about non-sacred values, the scans revealed less rapid, more deliberative responses.

But clear differences emerged between the excluded group and the one that had been allowed to keep playing. When excluded participants were asked about sacred values, their levels of left inferior frontal gyrus activity surged even higher than those who had been allowed to continue playing. Even more strikingly, non-sacred values generated levels of left inferior frontal gyrus activity similar to those generated by sacred values. Excluded participants also expressed greater willingness to fight and die for those values. In other words, social exclusion appeared to have caused non-sacred values to mimic sacred values, both in terms of brain activity and willingness to fight and die to protect them.


By demystifying the will to fight, our research should help prevent the United States from underestimating highly motivated insurgencies in the future. But the biggest practical implications of our findings concern efforts to prevent and counter radicalization. Most governments in Europe and North America have adopted policies grounded in rational-choice theory, deploying costs and penalties as means to dissuade radicals from acting on their values. But as our survey and brain imaging studies revealed, those motivated by sacred values are likely to resist rational cost-benefit analyses. Instead of preventing radicalism, “rational” policies might even be perceived as morally outrageous, backfiring and increasing radicals’ commitment to sacred values.

Our research also helps to explain why attempts to change or replace violent ideologies have generally had little effect. Efforts that “emphasize facts over propaganda”—as the Belfer Center at Harvard University suggests—or prioritize “evidence-based and data-driven” arguments, as the U.S. State Department recommends, are unlikely to sway people’s thinking on values that tend to be immune to a rational or logical calculus.   

Rather than challenging sacred values, those seeking to prevent or counter radicalization would be better off acknowledging these values while at the same time offering alternative interpretations of their meaning. Salafi preachers have had some success dissuading would-be suicide bombers this way, embedding the Quran’s arguments against violence into interpretations of how to defend Islam. This strategy requires deep engagement with actual social networks, not just in the realm of ideas. But if one person can be convinced to follow a less violent path in support of the group’s sacred values, others may follow.

But the best counterterrorism approach would be to prevent new and potentially dangerous values from becoming sacred in the first place. That, in turn, would require combating the kind of widespread social exclusion that currently afflicts vulnerable populations in Europe and the United States. If would-be radicals of all stripes regard as sacred fewer tenets of their worldviews, they will have fewer things to fight and die for.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • SCOTT ATRAN is Research Professor of Public Policy and Psychology at the University of Michigan, Ann Arbor and Co-Founder of Artis International and the Centre for the Resolution of Intractable Conflict at Oxford University.
  • More By Scott Atran