Under the United Kingdom’s Terrorism Act 2000, it is illegal to invite support for proscribed organisations. For instance, if you deliver public lectures or sermons encouraging people to support a prohibited group you can be convicted in a British court and given a lengthy prison sentence. But are the actions of terrorists really a consequence of merely listening to speeches? A growing body of research suggests that what motivates suicide bombers and other forms of self-sacrifice is something much deeper: the sharing of transformative personal experiences with others, together with perceptions of outgroup threat. These basic conditions for extreme pro-group action seem to owe little to religious dogma. In fact, it is quite possible that doctrines of hate are little more than hot air, a way of rationalising ‘after the fact’ why one is committed to carrying out acts of terrorism. If that is indeed the case, then arresting religious and political activists whose speeches are offensive to the liberal majority may be quite ineffective in the battle against terrorism.

For many years, my colleagues and I have been studying extreme pro-group behaviour – that is, actions in support of a group that carry a heavy cost for the actor. The heavier the cost, the more ‘extreme’ we would say the action is. A particularly extreme form of pro-social action would be to blow yourself up in order to harm a perceived enemy. This isn’t always obvious to people. Seen from the perspective of the victim, or indeed the wider society, suicide attacks are a supremely anti-social form of behaviour. But from the perspective of the terrorist, suicide missions are acts of extraordinary pro-sociality, directed to furthering the interests of the in-group.

My personal interest in this topic began at a remote location in the mountains of Papua New Guinea in the late 1980s, where I went to conduct two years of field research as part of my doctoral project in anthropology at the University of Cambridge. The group I lived with took me in as a member of their tribe, helping me to learn their unwritten language and the intricacies of their ancient culture. Many of my friends in the rainforest had joined a newly established cargo cult, repudiating all acts of violence in the belief that this new way of life would bring their ancestors back from the dead, ushering in a time of great peace and plenty. But prior to the spread of these ideas, or the arrival of missionaries, the region had a long history of tribal warfare. And many of the cultural institutions I ended up studying served to bind together military units through the intensely painful rituals of all-male warrior cults.

Some of the older men in my tribe described homicidal exploits with obvious enthusiasm. But these actions did not seem to be motivated or even justified by a hard-line religious or fundamentalist ideology. As I grew more intimately acquainted with their lives and personal histories, I came to realise that what really drove their warlike behaviour were not doctrines of hate but a very intense kind of love: love of their fellow warriors, love of the in-group.

The love that bonded together male raiding parties in the jungles of Papua New Guinea was forged through pain and suffering – especially the agonies of male initiations in which, for example, they were forced to wear heavy masks supported on their backs by a cassowary bone driven through the skin at the base of their spines and daubed in blood extracted from their tongues. These huge masks, many with grotesquely enlarged eyes and gaping mouths, were strictly taboo to women and children who hid or ran away whenever they came near, and who were strictly forbidden to enter the areas of the forest when men and initiated boys met in secret to construct them and plan their rituals. The shared ordeals of the male cult produced such powerful bonds of solidarity that every initiated man knew he could bank on his fellows to stand by him on the battlefield, come what may. Warriors could depend on each other never to run away no matter how overwhelming the odds stacked against them and instead to continue fighting side-by-side or to die trying. When I first learned about this form of intense group bonding, I didn’t know what to call it – still less how to measure it. But later I met group psychologist William B Swann at the University of Texas and learned about a construct he and his colleagues had developed known as ‘identity fusion’ – sometimes also described as a ‘visceral sense of oneness’ with the group. I suspected that identity fusion was exactly what I had observed out in the field.

To understand identity fusion, it is important to appreciate that we all have basically two kinds of identities. On the one hand, we all have a personal self, qualities that make us unique as individuals. But on the other hand, we are also members of groups – sometimes we wear many hats as members of lots of different kinds of groups. Most people, most of the time, see their personal and group identities as quite distinct and separate. So, if you make people’s group identities salient, their personal identities become less accessible. Partly for this reason, social identity theorists often describe group identification as depersonalising. It’s a bit like losing oneself in the crowd. But with fusion it’s different. If you are fused with a group, then the boundary between your personal and group identities becomes more porous. So, when your personal self is made salient, your group identity is made salient at the same time and actually fosters an enhanced sense of inner strength and invulnerability. And it works the other way around as well: if you make the group salient to a fused individual, this taps directly into a sense of personal agency. People who are highly fused with a group say that they will do almost anything to protect their fellow group members, risking or even laying down their lives to save them.

A crucial point to make about identity fusion is that it doesn’t necessarily lead to violence. Most people who are fused with a group are perfectly peaceful, law-abiding individuals. And the most common group to be fused with is one’s family, which is hardly the first group that comes to mind when you think about inter-group violence. But if you threaten a fused group, it is an entirely different matter. The Liam Neeson movie Taken, about a father’s attempts to rescue his daughter from kidnappers, plays upon the intuitive urge to defend one’s family even if that means leaving a trail of carnage in your wake. Among fusion researchers this has sometimes been called ‘the mother bear effect’ because bears are not on the whole very interested in humans unless they have cubs to defend – and then you had better keep your distance! The same idea applies to fused individuals. They aren’t particularly interested in out-groups, unless they pose a threat – and then you should really watch out because fused people will stop at nothing to defend their in-group.

The earliest measures of identity fusion were pictorial. Respondents viewed a series of pictures. Each picture contained a small circle (that’s you) and a big circle (that’s your group), intersecting to varying degrees. At one end of the spectrum the two circles were completely separate, but at the other the little circle (your personal identity) was shown to be completely contained within the larger circle (your group identity). And people were asked to say which of these representations best characterised their relationship with the group. Those who chose the one in which the small circle was entirely encompassed by the big circle were said to be ‘fused’ with the group. Later, a set of verbal measures was developed based on interviewee descriptions of their thoughts while responding to the pictorial measure. Respondents were asked to indicate the extent to which they agreed with various statements, using a Likert scale: ‘I am one with my group; I feel immersed in my group; I have a deep emotional bond with my group; my group is me; I’ll do more for my group than any other group members would do; I am strong because of my group; I make my group strong.’ With both the pictorial and fusion measures, the term ‘group’ could be supplanted by a more specific term, such as country, church, family, or even an individual such as brother or spouse. This made it possible to measure levels of fusion with a wide variety of group categories, relational networks, and dyads.

Armed with these fusion measures, my colleagues and I have been studying the causes of extreme group bonding among a wide range of groups – ranging from tribal warriors to revolutionary insurgents and modern armies, from university frats and sororities to football fans, and from adherents of world religions to participants in ritual tortures involving burning, piercing, whipping, and other ordeals. A common feature of all these different groups is that emotionally intense experiences endure in memory and become part of people’s unique identities as individuals. Sharing such experiences with others seems to be a potential cause of identity fusion. Since dysphoric events (ie painful, frightening, negative experiences) tend to be more intense than euphoric ones, we often find that traumatic ordeals (rather than joyful and pleasant ones) lie at the root of identify fusion. Among other things it means that the football teams that lose most frequently have the most tightly fused and loyal supporters. We call this the ‘shared dysphoria’ pathway to fusion. It involves going through a memorable, transformative experience that becomes part of your essential personal narrative, the things that uniquely make you, you. When such experiences define not only your own personal identity but also that of your group, it leads to fusion.

Our research has shown, however, that sharing dysphoric experiences is not the only way in which people can become fused with a group. Another common pathway to fusion is sharing biologically inherited traits or believing that you share common ancestors and family ties. In a large study of identical and fraternal twins we have shown that shared biology (as measured by zygosity) predicts fusion with twins independently of shared dysphoric experiences in their upbringing. And, not surprisingly, just as we have cultural institutions that cultivate shared suffering (such as painful initiation rituals), we also have institutions that emphasise shared ancestry such as beliefs about ethnic origins and tribal ancestor heroes. Military groups exploit these psychological tendencies in a range of ways by putting soldiers through excessively dysphoric forms of basic training or traditions of hazing, to create shared dysphoria. This is also achieved by requiring them to shave their heads and wear the same uniforms which gives the impression of their being more similar physically, and therefore more like clansmen who are descended from the same family heads or tribal elders. We also know from numerous studies that moving in synchrony tricks the brain into thinking that other people’s bodies are really extensions of one’s own body, further enhancing feelings of fusion between the personal self and the group. So, it is no accident that military groups so often engage in synchronous marching, drilling, parading, and dancing. These kinds of behaviours increase fusion with the group and, in turn, the willingness to sacrifice yourself when called upon to do so.

So, we had a well-developed theory of what causes identity fusion and some evidence that fused individuals were more likely to express a willingness to fight and die for their groups. But would they really do so in practice? To find out we needed to find a population who not only talked the talk but were also willing to walk the walk – actually to put their lives on the line for their fellows where the chances of getting killed were extremely high. An opportunity to do this arose in 2011, during the Arab Spring, when one of my students managed to get into Libya in the midst of the revolution using his United Nations contacts. He was then able to get me into Misrata to design a survey with insurgent groups in a place where over 2,000 rebels and civilians had recently been killed in their efforts to repel Gaddafi’s forces. We interviewed 179 members of revolutionary battalions (katiba). Half our sample were frontline fighters and half provided logistical support (eg fixing vehicles, driving ambulances). Among the overwhelming majority of insurgents, we recorded ceiling levels of fusion with several groups, including family, friends in the katiba, and even people they’d never met from other battalions. But we found floor levels of fusion with fellow Libyans who supported the revolution but were not members of a katiba. In other words, it didn’t seem to matter that people shared your beliefs or were on the same side ideologically – what mattered was being a member of these groups that had suffered together.

Next, we said to participants in our survey: OK, we understand that you are fused with multiple groups but if you had to choose one, which would it be (ie a forced choice question)? And here we found a striking difference between frontline fighters and providers of logistical support. If you were a frontline fighter you were more likely to choose your friends in the katiba over your own family. This is a really striking finding because it meant that your fellow fighters were more like family than your actual family. But this preference was not as strong among providers of logistical support. This made sense in terms of our theory – that is, we expected those who suffered more shared dysphoria on the frontline to be more fused. But we couldn’t rule out an alternative possibility – what if fusion is what drove these fighters to the frontline in the first place? To adjudicate on that question, we ran a series of studies with US troops who had served in Vietnam, Afghanistan, and Iraq but who crucially had no control over their deployment. And we found that, here again, intensity of suffering on the frontline predicted fusion with unit.

If the perception of sharing self-defining experiences is one of the main ways in which extremists are fused to their groups, it may be possible to develop effective interventions to defuse them. This may be as simple as facilitating open discussion among those considered at risk of engaging in extreme pro-group action. Describing self-defining experiences to others in your group may have the effect of revealing fundamental differences, previously unacknowledged. Although this needs to be confirmed through additional empirical studies, the research reported above would lead us to predict that the process of discovering such differences would be naturally defusing. This may be why painful initiation rites are typically secret or why it is often taboo for soldiers to talk to each other about the horrors of frontline combat. Such discussion does not seem to happen naturally, so fostering it would no doubt require encouragement from trusted sources. In the case of would-be Islamist fighters, this might naturally involve buy-in from family members, close friends, teachers, or imams. But note that this kind of intervention would not involve challenging the person’s beliefs or resemble efforts to ‘deradicalise’ the individual. On the contrary, the focus would be exclusively on personal experiences, a topic that (on the face of it) would not be directly related to matters of religious doctrine.

So, to sum this all up, it looks like a key driver of extreme pro-group action (eg willingness to fight and die or carry out suicide missions) is a state of identity fusion, resulting from shared dysphoric experience. If we are right about this, then forbidding particular kinds of beliefs or curtailing freedom of speech is likely to be ineffective in combating terrorism. In fact, it could have the opposite effect, actually fuelling perceptions of out-group threat. Our research suggests that every time a hospital is bombed, every time a drone strike misses its intended target, and every time a religious leader is imprisoned, another potential cohort of terrorists is created.