A little less conversation,a little more action: Schools and the prevention of violent extremism through reasoned engagement

Abstract

Many liberal democratic countries have enacted policy programmes aimed at preventing support for violent extremism. In countries such as the United Kingdom, part of this responsibility falls upon state-maintained schools. Teachers are charged with – among other things – challenging the claims, arguments and ideological narratives supportive of terrorism as well as promoting liberal democratic values as a counterweight. Insofar as the aim is to rationally dissuade those in danger of being ‘radicalised’ in this way, this article argues that this approach is unlikely to succeed in all but a narrow range of cases. It argues that the claims that radicalisation occurs through lack of knowledge or poor critical reasoning skills, and that improving these things is the antidote, is at odds with the consensus of social scientific research on partisanship and knowledge, and the phenomenon of motivated reasoning. Challenging beliefs and ideological narratives – whether directly or through more indirect mechanisms – will likely serve to entrench and harden those beliefs in those being drawn towards them. Contrary to this, the article argues that adoption of beliefs supervenes on membership in groups or affiliation to social movements and causes. Individuals receive the benefits of belonging, self-esteem and friendship from such affiliations. It is the receipt of these goods that drives commitment to cherished, identity-forming beliefs. Reasoned engagement therefore misses its target because those beliefs are not held on an epistemic basis. This has important consequences for education. Reasoned engagement is only likely to work where the person seeking to persuade, and the target of that persuasion, have fundamental beliefs and commitments in common. Educational resources are better focused on providing the contexts for students to receive and enjoy the feelings of belonging, self-esteem and friendship, rather than focusing those resources on defeating arguments.

We need to stop just pulling people out of the river. We need to go upstream and find out why they are falling in.

-Attributed to Desmond Tutu.

Introduction

Policies for preventing ‘radicalisation’ are now a familiar component of counterterrorism strategies in many liberal democratic societies. The United Kingdom’s ‘Preventing Violent Extremism’ framework is one example (HMSO, 2023). Part of the aim of Prevent is the ‘disruption’ of the ideological ‘narratives’ and processes by which individuals might come to adopt attitudes and form beliefs supportive of violent extremism (HMSO, 2011: 7). In the United Kingdom, part of this responsibility falls to schools.

One aspect of the discharging of this duty in schools is the direct challenging of the ideological positions and apologetics that underpin violent extremism (HMSO, 2023: 27, 36). This process of challenging can take place through classroom engagement with such views, as well as the promotion of the ‘fundamental British Values’ of liberty, toleration, democracy, and the rule of law (Department of Education, 2014: 5). This is intended to create resilience against the pull of pernicious ideas.

Insofar as the aim is to persuade those in danger of being radicalised that they should reject such beliefs, I argue in this article that this approach is unlikely to succeed in all but a narrow set of cases. This is because the assumption on which it is based – that those vulnerable to extremist ideas are vulnerable (in part) because they lack knowledge or understanding – and which gives rise to the idea that better knowledge is the solution, is mistaken. The available evidence strongly indicates that the reverse is true: higher levels of knowledge are the corollary of stronger beliefs and deeper attachments. Adding more knowledge or increasing critical reasoning capacities does not, moreover, tend to weaken existing beliefs, but to strengthen them. To the extent that educators are expected to engage in this way, the Prevent-style policy risks being counterproductive.

I argue instead for a view of radicalisation that is rooted in the idea that attachment to extremist ideas and beliefs tends to occur because of non-epistemic reasons. That is, social and material benefits, such as belonging, and self-esteem accrue to individuals through their membership in various social groupings. Based on the receipt of these benefits, individuals are motivated to adopt the beliefs and behavioural norms which define the group. These identity-forming beliefs are, therefore, the result, rather than the cause. This makes them particularly difficult to dislodge through reasoned argument. This has consequences for education. The reasoned-engagement approach is destined to largely miss its target because the beliefs it aims to destabilise and undermine are not held on the basis of epistemic reasons. Instead, preventing radicalisation would be better served by schools focusing resources on addressing issues of relationships, inclusion, discrimination, and so forth that lead to individuals seeking a range of social goods in tight-knit, radical ideological groups.

The case is advanced in the following way. Section “Policy background and reasoned engagement” sets out the basics of the challenge model as reasoned engagement. Section “Challenging the causal assumption” undermines the assumption of ignorance that underpins the reasoned engagement approach. Section “Challenging knowledge as the solution” challenges the idea that providing stronger arguments, better information, improving reasoning skills, or teaching directively towards the adoption of liberal values will successfully dissuade or divert people away from extremist ideas. Section “Indirect engagement – reasoning from conjecture” considers an indirect form of challenging that schools and teachers may undertake – conjectural reasoning – and which may not trigger the defensive mechanism of motivated reasoning. Section “A non-epistemic-based account of radicalisation” sketches an alternative explanation of why some come to hold extremist beliefs. Section “Some possible contributions by schools to preventing violent extremism” offers some areas of focus for schools based on the preceding arguments. Section “Conclusion” concludes the article.

Policy background and reasoned engagement

I begin by setting out the idea of reasoned engagement as part of the process of challenging extremist narratives. The most well-developed policy example in this respect is the United Kingdom’s Prevent framework. Prevent sits within a wider policy framework for detecting, stopping and apprehending terrorists. Prevent focuses on ‘radicalisation’ – the process by which persons are said to come to hold beliefs supportive of violent extremism (HMSO, 2023). ¹ Its aim is to stop individuals from becoming terrorists or supporters of terrorism – the prevention of radicalisation – and to provide for the rehabilitation of those already involved in terrorism – the ‘undoing’ of their beliefs. The Prevent remit includes the tackling of ‘the ideological causes of terrorism’ by ‘disrupting’ and ‘reducing the influence of radicalisers’ as well as reducing ‘permissive environments’, and the availability of ‘terrorist content’ (HMSO, 2023: 9).

The genesis of the Prevent framework is well documented (see, for example, Lundie, 2023; Stevens, 2011; Thomas, 2009). Its earliest manifestation was based on the perceived need to respond in a multi-dimensional and multi-agency way to the increased threat posed by Islamist terrorism in the wake of the events of 9/11 and the July 2007 bombings in London. Its more recent manifestations have been broadened to include extremist threats from the Far-Right, Eco-terrorism and other political ideologies. ² Prevent has been subject to severe criticism, such as singling out the Islamic community, unnecessarily aggravating societal divisions, for fuelling feelings of exclusion and for increasing discriminatory attitudes (see, for instance, Jarvis and Lister, 2013; Thomas, 2009). I will not revisit these criticisms here. Nevertheless, Prevent has served – and continues to serve as a model of similar policies throughout the world, including Australia, Canada and several European countries (see Lundie, 2023). This includes attempts to use educational institutions and other bodies to intervene in, and disrupt, the radicalisation process with the aim of creating resilience against, and dissuading individuals from adopting, extremist beliefs. Prevent is used in what follows as illustrative of a set of policies and approaches for dealing with extremist beliefs that has – despite criticism – continued to gain traction world-wide.

Schools in the United Kingdom are one environment where Prevent is delivered. This includes duties of safeguarding and referring pupils deemed in danger of radicalisation to the authorities for further support and interventions by external agencies. The task of challenging extremist ideologies and narratives is also devolved to schools. ³ Classrooms are to be places for the discussion and understanding of sensitive issues, ‘including terrorism and the extremist ideas that are part of terrorist ideology, and learn[ing] how to challenge these ideas’ (HMSO, 2023: §5.2.6.1). This is to include challenging ‘more broadly harmful ideas, such as misogyny and antisemitism, or the concept of blasphemy to justify or condone violence’ (HMSO, 2023: 20).

To enable this challenging of extremist ideas, teaching is to provide relevant ‘knowledge, skills and values’ (HMSO, 2023: 41). Improving knowledge and critical reasoning skills includes classroom discussion aided by specific government-approved teaching resources with information and reasoning tasks ‘related to different forms of extremism, including Islamist extremism and Extreme Right-Wing ideology’ (HMSO, 2023: 42). The non-statutory guidance to schools on promoting the ‘fundamental British Values’ of individual liberty, democracy, toleration of those with different faiths and beliefs, and the rule of law, also fits within the Prevent framework (Department of Education, 2014: 5; HMSO, 2023: 41). Broadly speaking, the strategy is one of reasoned engagement with extremist ideas and arguments for the purpose of dissuading pupils from coming to hold beliefs supportive of violent extremism (HMSO, 2023: 8). By creating attachments to liberal values, providing accurate information, developing critical reasoning skills, and directly countering apologetics, young people are to be equipped with the resources necessary for resisting the allure of pernicious ideas. Call this the knowledge-as-antidote claim. The knowledge-as-antidote claim is necessarily grounded upon the assumption that part of what places individuals in danger of radicalisation is a lack of relevant knowledge, understanding, critical reasoning skills and appropriate moral values. Outside of specific vulnerabilities that influencers exploit, the general susceptibility to radicalisation is taken to be one of ignorance (HMSO, 2023: 12). Creating resilience requires removing or mitigating this ignorance. Call this the ignorance-as-cause assumption.

Challenging the causal assumption

Are those at risk of radicalisation, or those who are in the process of coming to hold extremist beliefs, ignorant in the way the ignorance-as-cause assumption implies? While it is tempting to assume that those who are the least knowledgeable are the most easily seduced by extremist ideas, this runs contrary to a large body of social scientific evidence regarding relative levels of knowledge and strength of ideological affiliation and political partisanship. Instead, the reverse is true. Those who hold the strongest beliefs, or who identify most strongly with an ideological view tend to be the most knowledgeable about relevant matters (Kinder and Kalmoe, 2017: 63). ⁴

On reflection, this is unsurprising. It would be odd if someone who claimed to be deeply passionate about something turned out to know little about it. We would expect the avid football fan to know more about the game, its history, their team, its players and results, than the non-football fan (Hannon, 2022: 31). Those with little interest in something tend to know the least, and those with more interest tend to know more relative to their degree of interest (Kinder and Kalmoe, 2017: 40–42; Michaud et al., 2009: 32). Consequently, those in danger of being radicalised are not likely to lack knowledge or be the least informed. This is the case not only for adults, but for children too – and across a range of beliefs including political and religious beliefs, as well as beliefs in conspiracy theories (Jolley et al., 2021; Van Deth et al., 2011). As an individual’s interest in an area develops, their knowledge levels also increase relative to the depth of the interest. That many terrorists have turned out not to be ignorant, naive dupes, but intelligent, well-educated, informed or high-achieving individuals, bears out this fact (Stevens, 2011). Similarly, those who affirm conspiracy theories are often highly educated, including possessing well-developed critical thinking skills (Dyer and Hall, 2019).

Perhaps, rather than lack of knowledge, susceptibility to radicalisation turns on poor reasoning capabilities or the kinds of critical thinking skills that Prevent charges schools with developing. Absent such capabilities, young minds might be more susceptible to manipulation by apologists and less able to see through falsehoods and faulty reasoning. Indeed, adopting outlandish, radical and extreme ideas often strikes us as necessarily irrational. How could any right-minded or clear-thinking person come to accept pernicious, harmful, ideas without irrationality, deception or manipulation playing some part.

Sociological studies of a range of parallel cases where seemingly bizarre choices are initially assumed to be the result of poor reasoning, manipulation and deception cast significant doubt on this claim. One of the most researched set of cases are those of the New Age cults of the 1970s, such as the ‘Moonies’ and Hare Krishnas. Although these cults were largely peaceable and often explicitly abjured violence, their organisational features, recruitment practices, the kinds of operational problems they encountered and the solutions they implemented are sufficiently similar to extremist groups in this regard to allow for meaningful comparison. As per contemporary worries over radicalisation, much was made at the time of stories of coercion and ‘brainwashing’ to explain how otherwise normal young people could take up such strange ideas, practices and affiliations.

These explanations have been systematically overturned by numerous comprehensive studies in the intervening years. The results are telling. Cult converts were no more naïve, ignorant or unintelligent than the rest of the population (Robbins, 1984: 248–249). Those who joined the Moonies were generally healthy and intelligent, often with extensive education (Barker, 1984: 121–131, 207). This included recruits from college-age to older generations (Wilson and Dobbelaere, 1987: 186–187). Recruits also displayed lower levels of depression, anxiety and drug-use, and higher levels of well-being, than average for their age group (Koenig and Larson, 2001: 75). Claims of abductions, coercion and manipulation were quickly disproven (Barker, 1984; Robbins, 1984: 250–253). Whatever the power of the ideas being peddled or the charisma of those peddling them, they were not so seductive that they were incapable of being resisted by the vast majority that encountered them. Even the Moonies – the cult with the highest membership levels – was extremely poor at recruiting and retaining members. The famed two-day recruitment ‘retreat’ where participants were exposed to activities such as ‘love bombing’ was ineffectual. Less than 25% who attended joined for more than a week, less than 5% who joined were still members after a year, and less than 0.5% of those who joined were still members after 2 years (Barker, 1984: 149–172). A recruitment failure rate of over 95% of those who were sufficiently interested to spend a weekend finding out more is not an impressive record for the most successful group of its kind (see Iannaccone, 2006: 5–8).

If the reasoning capabilities of the vast majority were strong enough to reject the Moonie philosophy, it also played a role in those who did join. Recruits displayed standard cost–benefit reasoning about membership. The following from an interview with a young Moonie is, perhaps, only exceptional in its un-exceptionalness:

You know, sometimes at night I wonder what if none of this is true. What if Rev. Moon is not inspired by God? Maybe all of this is for nothing. But then I say to myself, if it’s true then I am in on the ground floor of the greatest event in history. And if it isn’t, so what? I was probably going to spend the rest of my life working in that plywood plant anyway. (Stark et al., 1998: 379)

Challenging knowledge as the solution

This section considers the knowledge-as-antidote claim. Recall, this is the idea of increasing student knowledge by providing more and better information, challenging extremist ideas and arguments, and developing critical reasoning capabilities. The purpose is to help students see through poor arguments and falsehoods, to develop resilience and to be moved by the force of the better reasons and evidence. Like the ignorance-as-cause assumption, the knowledge-as-antidote claim is to be found wanting.

Behavioural scientists have long recognised that human reasoning is beset with certain flaws. These include flaws such as confirmation bias, disconfirmation bias, prior attitude effect and framing effects. Confirmation bias is the tendency to seek out or only ‘notice’ reasons and evidence that is supportive of one’s pre-existing views or beliefs (Hannon, 2022: 31; Kahne and Bowyer, 2017: 6; Kalmoe and Mason, 2022: 75). Disconfirmation bias is the tendency to dismiss, fail to notice, or find reasons to object to information and evidence that contradict one’s pre-existing views and beliefs (Kahne and Bowyer, 2017: 6; Williams, 2023: 105). Framing effects is the tendency for individuals to agree or disagree with information and arguments depending not on their content but on whether the way it is presented aligns with one’s pre-existing views and beliefs (Hindmoor, 2004: 130–135). Prior attitude effect is the tendency to view arguments as weaker or stronger to the extent that they align with one’s pre-existing views or beliefs (Kunda, 1990; Taber and Lodge, 2006). In combination, these flaws severely restrict how much individuals learn from, and change their beliefs, through exposure to information or deliberation.

Taken together, these tendencies comprise the problem of directional motivated reasoning – ‘the tendency of people to conform assessments of information to some goal or end extrinsic to accuracy’ (Kahan, 2013: 408; see also Hannon, 2022: 31–32; Williams, 2023: 105). Motivated reasoning is especially prevalent and strong when the subject is important to the individual on an emotional level – ‘hot cognition’ – such as political and religious matters (Kahne and Bowyer, 2017: 6; Lodge and Taber, 2005).

This has two consequences relevant to the aims of preventing radicalisation via schools. First, the provision of more information or increasing knowledge about ‘hot’ subjects is unlikely to have the desired effect of undoing the views and beliefs of those being drawn towards extremism. Rather, such information will be used selectively to reinforce and embed those prior views further. More knowledge provides more resources to draw upon in support of preferred beliefs (Hannon, 2022: 32). Moreover, this does not vary with age. Young people are no less susceptible to motivated reasoning than adults. While we might expect the views of adults to be hardened and therefore ‘stickier’, this is not borne out by the available evidence. Studies of 15- to 27-year-olds demonstrate that young people exhibit the kinds of behavioural biases just as much as adults. They resort to motivated reasoning about political matters, judge the accuracy of information according to pre-existing beliefs, and employ motivated reasoning more as their level of knowledge increases (Kahne and Bowyer, 2017: 25). These tendencies are also evident in much younger children as well (Van Deth et al., 2011). ⁵

The second consequence for preventing radicalisation through schools is that direct challenges to pre-existing beliefs and views will risk triggering motivated reasoning. While we may hope that carefully crafted lessons, didactic methods of persuasion and so forth will create an environment for supportive and objective discussion that would serve to draw individuals away from false and harmful beliefs, the evidence suggests this is unlikely. Faced with competing reasons and evidence, our biases come to the defence of our favoured beliefs. That is, we ‘dig in’, we interpret evidence and twist data in ways supportive of our beliefs (Kalmoe and Mason, 2022: 75; Lord et al., 1979). Improving critical thinking and reasoning skills simply aids motivated reasoning by helping you ‘to find reasons to reject facts, figures, and arguments that conflict with your preferred views’ (Hannon, 2022: 32).

Rather than passive, badly informed and unwittingly manipulated individuals, those drawn towards extremist beliefs are more like Brennan’s (2016) ‘Hooligans’ than ‘Hobbits’ (p. 5). That is, they are the fanatical followers of sports teams, political parties, social causes and ideological movements (Brennan, 2016: 5). They are not merely passionate but docile; hooligans are actively prepared to go to battle in defence of their ‘side’. They are the cheerleaders – seeking out information that supports their position, cherry-picking facts, glossing over anything that does not fit. When challenged, they are more than willing to use their knowledge to rationalise away opposing arguments, to dismiss contrary evidence out of hand and to get aggressive and abusive (Iyengar et al., 2012: 406; Kalmoe and Mason, 2022: 75).

The same seems to hold for teaching liberal political values as a way of countering radicalisation as it does for directly challenging beliefs. The ‘Teaching Fundamental British Values’ guidance for schools (HMSO, 2014) is part of the wider Prevent framework. Class time must set aside for the promotion of the values of liberty, toleration, the rule of law and democracy. Promotion in this sense means directive teaching. ⁶ That is, teaching aimed at bringing children to endorse such attitudes and beliefs and for guiding their actions. Such values run contrary to the values of extremist ideology and is intended to provide a buttress against it.

Nevertheless, there may be good reasons to be concerned that teaching British values can also trigger motivated reasoning. First, many of those drawn to extremist positions are drawn to the view that liberal values are themselves the problem. Second, even where someone recognises the value of liberty, toleration, democracy or the rule of law, they may believe that other considerations, such as duties to a divine power, outweigh them. Reminding individuals of the benefits of living in a society that guarantees certain freedoms, or the worth of democratic processes in settling political disagreements is a worthy strategy, but it is precisely extremist beliefs often deny (Rawls, 1996: 386). Third, some extremist views, such as some Far-Right views, offer defences for their actions as motivated by the protection of freedom, or democracy, against those who they perceive as threatening it (immigrants, and so forth). Teaching British values can be understood as a call to arms, or as a challenge to that particular conception of what British values might mean, again possibly triggering motivated reasoning.

Indirect engagement – Reasoning from conjecture

If motivated reasoning is likely to be triggered when cherished beliefs are subject to direct challenges of the kind outlined, then it might be thought that more subtle, empathic, affective, less direct forms of reasoned engagement could be employed that would not trigger defensiveness on the part of the target of that engagement. Indeed, the Prevent guidance advises teachers to avoid directly denying claims or challenging beliefs. Instead, teachers are advised to begin conversations with questions that ask pupils to explain their views and beliefs more fully, and the authorities on which it is based. Challenging such beliefs is to come in the form of exploring that view together as a kind of mutual reasoning and learning (HMSO, 2023). ⁷

One example of this kind of reasoned engagement that has been developed by political philosophers in response to doctrines that deny one or more fundamental liberal values is conjectural reasoning or ‘reasoning from conjecture’: that is, we ‘reason from what we believe, or conjecture, may be other people’s basic doctrines, religious or philosophical, and seek to show them that, despite what they might think, they can still endorse a reasonable political conception of justice’ (Rawls, 1997: 591; see also Schwartzman, 2012: 6). Conjecture is a genuine attempt at persuasion. The conjecturer engages honestly and open-mindedly. In the context of schools, it would aim to place the student and their views and feelings at the centre of discussion. It is not instrumental reasoning. Someone who deliberately concealed their own beliefs or intentions or was not open to the possibility of being mistaken would not be a conjecturer because they would be engaging in reasoned discussion in bad faith. The aim is to support those in danger of being radicalised to come to support liberal political values and institutions for the right reasons, not through manipulation.

Note, that conjectural reasoning will be possible only for certain kinds of view. Some views that do not admit of the possibility of a more moderate variant could not be approached via such a method of engagement. Views that denied the fundamental equality of all human beings as a central and defining part of their doctrine, for example, would be impossible to engage with in a conjectural way. There is no version of Nazism or antisemitism consistent with a liberal interpretation.

That caveat aside, Wong (2019: 124) provides a useful education-based illustration of conjectural reasoning. Wong’s example is of presidential candidate Bernie Sanders speaking to students at Liberty University, the evangelical Christian college, in 2015:

Sanders, as a secularized Jew, claimed that ‘it is vitally important for those of us who hold different views to be able to engage in a civil discourse. . . . and argued that, from the Christian perspective, such vast inequality should be intolerable. During his speech, Sanders quoted several Bible verses, such as Matthew 7: 12 . . . and Amos 5:24 . . . to support a liberal claim that people should treat everyone ‘with respect and dignity’ and put themselves in the shoes of the poor. (Wong, 2019: 124. Footnote omitted)

Sanders is sincere – ‘Sanders confessed in his speech, “I am not a theologian, I am not an expert on the Bible”’ (Wong, 2019: 124). And his argument is conditional – he did not deny Christianity, but reasoned from Christian beliefs while acknowledging he was not himself a Christian. We might expect, in the case of schools, teachers and others engaging with students based on the beliefs, doctrines and views that those students endorse. Beginning from those premises, they could then argue that the resulting position is not one that endorses violent extremism.

There are, however, several reasons to doubt whether conjectural arguments will satisfactorily sidestep triggering motivated reasoning. ⁸ First, while the conjecturer may engage honestly and open-mindedly, the perception of bias or disingenuousness is likely to remain. The target of that engagement may view the attempt to persuade being instrumental. That is, because the conjecturer is still deliberately seeking to change the student’s mind, the latter may view the former as inevitably and unavoidably biased towards the liberal view (Schwartzman, 2012: 11–12).

Second, the conjecturer may be perceived as arrogant. When the conjecturer offers reasons for thinking that the person they are engaging with has misunderstood their own view, the conjecturer risks seeming arrogant or conceited. Think here of politicians uttering claims about the ‘true’ meaning of a particular religion or non-religious teachers pronouncing on interpretations of religious doctrine. ⁹ They appear condescending; telling those who have lived their lives within that faith that they know less about it than an outsider. Third, the conjecturer may be perceived as insincere. That is, they may be understood as deliberately refraining from saying what they really think about a person’s beliefs even while being honest that they do not share them. For example, consider saying:

I do not share your belief X, and I think there is good evidence to reject X as false, but if I did believe in X, then I would see it as supportive of liberal principles because of considerations C.

This rings hollow, even though it may be honest. The target of such persuasion may focus more on what is not said, rather than what is said. What is not said is that the conjecturer finds their view not merely false, but ridiculous, even contemptible. This may be taken as a signal of even less commitment to the worth of the view the conjecturer is engaging with.

Fourth, the conjecturer may be perceived as lacking credibility. When authority within a given group is conferred by such things as knowledge and understandings of traditions, doctrines or texts, outsiders who necessarily lack those things will have insufficient credibility to be heard, let alone taken seriously. Ignorant interventions, or arguments constructed on what the audience view as a simple misconstrual of an essential point may be taken as betokening a lack of understanding or impartiality. Even where the conjecturer is in fact knowledgeable, this can be the case. Some religions, for instance, contain content-independent reasons for belief, such as deference to de facto intellectual authorities (the Pope for Roman Catholics for example). Conjecturing about beliefs based on the content of those beliefs seems, in such cases, misses the point and is unlikely to succeed.

A non-epistemic-based account of radicalisation

Direct and indirect engagement are, I have argued, unlikely to succeed in preventing radicalisation in many of the kinds of cases they are intended to address. In the next section, I will argue that there is one important exception to this. In this section, I begin to argue for a different emphasis to preventing radicalisation. That emphasis is rooted in a different account of individual motivation than the ignorance/knowledge thesis.

To sketch this alternative account, it is useful to begin with the question of why some beliefs are so ‘sticky’ – or resilient to being overturned by rational persuasion. One answer to this question is that cherished beliefs are (a) held on the basis of non-epistemic reasons and (b) are central in framing and giving content to a person’s identity. That is, an individual will be particularly likely to ‘dig in’ in defence of their beliefs to the extent that those beliefs are seen as a central aspect of who they are and where those beliefs are held not on the basis of the force of the better argument, but because of some other basis which is beneficial to the person. When beliefs are held on non-epistemic grounds, they are particularly impervious to being dislodged through reasoned persuasion. When those beliefs are central to one’s identity, any threat to them is perceived as a threat to one’s core self. Giving up such cherished beliefs would require giving up part of one’s identity (Hannon, 2022: 34–35). This is a painful process, one we prefer to minimise by ‘preserving the integrity of the self-image’ (Tajfel, 1969: 92; see also Abrams and Hogg, 1988). Identity-defining beliefs, that is, drive motivated reasoning as a mechanism of self-defence against threats to our self-understandings (Williams, 2022: 7).

Beliefs can become identity-forming in this way because of the goods that affiliation with a cause, group or movement can bring to the individual (Williams, 2022). These benefits include feelings of belonging, self-esteem, solidarity or being valued by others (see Bjørgo and Horgan, 2009 [2008]: 3; Pettinger, 2017: 21; Williams, 2022: 2). Where someone receives these goods through affiliation, and their identity is increasingly shaped by that relationship, they will also come to increasingly adopt and espouse the defining beliefs of that group. ‘Social attachments are the horse that pulls the cart of ideological change’ (Iannaccone, 2006: 7). Because social movements, religious groups, cults, activist groups, political parties and sects are a considerable source of social and material benefit for people, they ‘routinely join and stay within belief-based coalitions for reasons independent of the truth of their beliefs’ (Williams, 2022: 9).

While this explains the relationship between identification with a movement, cause or group and the adoption of a set of related beliefs, it does not provide an explanation of why some people come to identify so strongly with particular causes, movements or groups – extremist or otherwise. One possible answer is: socialisation. That is, we are born into a social or ideological identity in the same way as we inherit religious, ethical and political views, or football team allegiances, from our parents: ‘In sports and politics, people become fans through socialisation’ (Hannon, 2022: 34).

On reflection, this is too hasty. Undeniably, patterns of affiliation and belonging are deeply influenced by social and familial ties. But to view this as the only cause – eliminating space for individual choice – would be an error. Setting choice and socialisation in opposition and concluding that it is socialisation all the way down would be to deprive socialisation of any explanatory worth. It would deprive us of an explanation of why some people are radicalised, and others in the same circumstances are not. Instead, we can see circumstances as shaping the range of options from which individuals then choose. These choices are then made based on the individual’s wants and preferences (see Williams, 2023: 100). These include desires for belonging, friendship, self-esteem and so forth. Most people receive these benefits via multiple sources – family, workplaces and secondary associations. Some receive them from a narrower, more concentrated set of sources – sometimes a single, convenient and intense source.

In turn, this provides an explanation for some people becoming highly committed – fanatical – even where such commitment is extremely costly or demanding. Affiliation with a small group of like-minded, highly dedicated, people – what Williams calls ‘belief-based coalitions’ – create a very intensive, rewarding experience for participants (Williams, 2022: 3). Being part of a coalition triggers positive feelings (Iyengar et al., 2012: 407; Tajfel et al., 1971). Members receive affirmation, praise and encouragement (Williams, 2022: 2). They gain a sense of superiority over outsiders who fail to ‘get it’, or who are not awake to the truth. High costs serve to signal commitment. Free riders – those who would accrue the benefits of membership without shouldering their fair share of the costs and who threaten the fragile stability of cooperation essential to coalitions – are excluded by high costs (Williams, 2022: 4).

The absence of social ties and networks, feelings of exclusion or marginalisation, or experiences of discrimination are well-known push factors for individuals to seek the goods of belonging, friendship and self-worth in more unitary and intense relationships. It is well established, for instance, that large degrees of economic inequality and feelings of economic disenfranchisement breed social issues such as violence, addiction, crime and gang membership (Pickett and Wilkinson, 2010, 2020 [2019]). Inequality operates as a ‘deep social fact’ that causes ‘relational inequalities’ such as status harms, relationships of social domination, the breakdown of social solidarity and the erosion of individual self-respect (O’Neill, 2008: 150–151, 2010: 403).

What this explanation gives us, therefore, is an upstream account of why some people are drawn to extremist ideologies and groups that is prior to downstream accounts of reasoned engagement. Social and material factors, including discrimination, inequality, the unavailability of educational and work opportunities, create grievances and fuel identification with small, radical groups and views, where belonging and self-esteem can be achieved (Pavlović, 2023: 18; see also Gurr, 1970).

Some possible contributions by schools to preventing violent extremism

The arguments above provide two possible areas where schools may pursue an anti-radicalisation agenda with some reasonable hope of success. This section outlines both in turn.

First, consider the social and material circumstances that I have argued play a significant role in driving radicalisation. While addressing significant material inequalities and inequalities in opportunity for education and employment are issues that require wider societal changes, schools can (and do) nevertheless play a fundamental role in creating environments that provide the grounds for improved self-esteem, belonging, solidarity and so forth, and in a more equal manner. Schools can do this in a practical sense, from the school ethos up. Forms of governance, student input, teacher–student relationships and school-based curricular and extracurricular activities can, if properly arranged, serve to foster feelings of worth, belonging and respect.

Some brief examples serve to illustrate. First is investment in extracurricular activities that create the circumstances for individual students to feel a valued part of a group, club or team. Such opportunities foster identity-creating attachments and diversify the base from which individuals will seek the goods of belonging and inclusion by providing other sources. Second is not unnecessarily pitting students against each other (Hussain, 2020). Competition is valuable and important and there is a place for it in schools. But designing competitive frameworks that unnecessarily set children against each other or make the realisation of one set of interest dependent on the setting back of the interests of others should be avoided. Where possible, pupils should be encouraged to share in the successes and failures of their fellow students, so that they come to identify with their shared interests (Fishkin, 2014; Hussain, 2020: 102–105). Third is a degree of reinforcement of those values through curriculum and classroom-based activity and learning. What is notable by their absence from the list of fundamental British values is any mention of socioeconomic values, such as seeing society as a cooperative venture for mutual benefit, where everyone is entitled to a fair share of economic goods and welfare – the values that underpinned the creation of the British welfare state and the NHS.

Second, the argument against direct reasoning and indirect conjectural reasoning turned on the claim that both were likely to trigger motivated reasoning in defence of threatened beliefs. Conjectural reasoning, even though accepting for the sake of argument the basic premises of the target’s view, risks being perceived as biased or lacking credibility. There is, however, one exception that is based on the argument for this claim. This exception stems from who is doing the persuading. Recall, for Rawls (1997), conjectural reasoning is a process where we ‘reason from what we believe, or conjecture, may be other people’s basic doctrines, religious or philosophical’ (p. 591). Rawls does not say in this passage who the ‘we’ is that is attempting to do the persuading. However, it is possible to distinguish two broad categories of persuaders: persuaders internal to the basic doctrine held by the target of that persuasion, and persuaders external to that particular doctrine.

What marks the difference between internal and external persuasion is whether the would-be persuader does in fact share some – or all – of the basic premises of the conception held by those they seek to persuade. To extend Wong’s example, in attempting to persuade a Christian who rejects one or more of the fundamental liberal principles that their interpretation of their faith is mistaken in reaching their conclusion, internal persuaders would include, predominantly, other Christian citizens (lay and clergy) who share similar presuppositions or articles of faith. External persuaders, by contrast, would lack this shared set of presuppositions, commitments, or beliefs, and engagement would take place in the conditional or conjectural manner outlined earlier.

So far, the discussion of reasoned engagement in schools has made no such distinction between internal and external persuaders. ¹⁰ Yet, this distinction gives rise to a possible division of labour that would be applicable within schools. Staff who share moderate variants of religious, philosophical and ideological views and memberships with students may be better placed to engage in the kinds of reasoned persuasion that Prevent seems to have envisioned. Internal persuaders are far less likely to be perceived as disingenuous, biased or lacking in credibility because of their shared beliefs, doctrinal commitments and affiliations. Moreover, these commonalities of belief, commitment and experience can generate trust between teacher and students (see Laden, 2024). Consequently, their interventions are likely to have greater traction than attempts to persuade made by those outside of those communities of beliefs or traditions. Of course, as we have seen, not all doctrines or ideological views can be approached in this way, because some forms of extremism do not admit of a liberal interpretation or variant. Nevertheless, for those where liberal versions do exist, refocusing support and resources away from more general values and knowledge teaching to more specific interventions by those professionals who are internal to particular views would be a more worthwhile use of valuable resources. ¹¹

Conclusion

I have argued that school-based attempts to prevent the radicalisation by reasoned engagement with the ideological beliefs that underpin violent extremism are likely to be unsuccessful. This is because the adoption of such beliefs is usually on non-epistemic, rather than epistemic, grounds. That is, we come to hold cherished beliefs that are constitutive of our identities because of the social interactions and affiliations that we inhabit. Consequently, attempts to dislodge such beliefs using epistemic or rational persuasion do not aim at the correct target, namely, the reasons people are drawn to involvement in specific groups, movements or causes. Providing more information, teaching towards more liberal values or contesting the arguments of apologists serves to trigger motivated reasoning in defence of those beliefs, entrenching them further. Instead, educational resources would be more effectively targeted at those conversations where the interlocutors share the same fundamental comprehensive commitments, as well as fostering the background conditions that provide all school-aged children with the grounds for achieving feelings of belonging, friendship, self-esteem and a sense of being valued. What is needed is a little less talking, and a little more action.