Extremism is a topic of significant practical and ethical concern which, though it overlaps with research on terrorism (Gelber 2019), has only fleetingly been explored as a topic in its own right within analytic philosophy (Hardin 2002). Now, with Extremism: A Philosophical Analysis (2021), Quassim Cassam has provided the first major philosophical treatment of extremism. In a style similar to his recent work (2019), Cassam explores extremism through a range of rich examples including ISIS, Anders Breivik, Pol Pot, the IRA and Timothy McVeigh. It utilises resources from several disciplines and is lucid throughout.

The book can be seen as having two parts that address, respectively, what extremism is and how it relates to similar concepts, and whether people should be extremists and how extremism can be countered. This article will, first, outline Cassam’s main ideas on these issues. Then, it will explore a weakness in Cassam’s account, which is that, in two respects, it fails to do justice to the social and group features of extremism. Specifically, it fails to adequately distinguish between individual and collective forms of extremism, and does not adequately recognise the deep socio-epistemic nature of the radicalisation and deradicalisation process.

1 Extremism: What it is and how to Prevent it

The central idea that Cassam develops in the book is that there are three overlapping but distinct ways to conceive of extremism: ideological, methodological and psychological. Cassam’s account of ideological extremism is cashed-out in terms of one’s position in ‘ideological space’ (40). The most familiar instance of this idea is of left–right political ideologies. For instance, an extreme form of communism, such as the Khmer Rouge, would be placed on the extreme left-wing of this spectrum, whereas forms of ultra-nationalism could be placed on the extreme right-wing.

This conception faces two challenges. First, how do we determine what is meant by right- or left-wing? As some commentators have pointed out, ideologies are relative because their meanings are ‘regularly updated’ (Heywood 2017, 16; Cassam 2021, 48–9). We can see this given the tendency to conceive of ‘the right’ in terms of attachment to free enterprise, cultural conservatism and ownership of private property (Scruton 2007, 601), but also to ethno-national commitments ‘to preserve the superiority and dominance of some groups over others’ (Miller-Idriss 2020, 8). This makes the left–right spectrum somewhat relative. Second, there are a number of spectrums beyond left–right in terms of which one may be ideologically extreme. For instance, there could be a ‘Pro-Violence spectrum’, where, at one extreme are pacifists, and at the other, those that see violence as widely justified and useable (Cassam 2021, 54). Another option is an ‘authoritarianism spectrum’ (ibid., 57–8), which is determined by beliefs about the strict orderliness of society, and the extent to which infringements on authority can be severely punished.

The next conception is methods extremism – the use of extreme methods to achieve one’s objectives. Cassam’s approach here is not to give a necessary and sufficient conditions analysis, but a normative account of a paradigm methods extremist. Cassam says that

if there is an archetypal methods extremist, then it is someone who [1] uses unnecessary or disproportionate violence [2] in pursuit of an unjust but hopeless cause, and [3] directs their violence against illegitimate targets. (2021, 80. Italics and numbering added.)

Note that the definition is of an archetype, and so there will be methods extremists who are outliers to this account. Take condition [1]. Someone like Bobby Sands, who starved himself to death in prison to achieve his political aims, used an extreme method, but not (in this instance) one that was violent. But for the most part, groups using extreme methods, including the IRA (of which Sands was a member), tend to use violent methods, such as bomb attacks and mass shootings.

Conditions [1] and [3] in particular show how this account is normative. As Cassam points out (ibid., 74), Osama bin Laden often claimed that American citizens democratically elect their governments, and so their civilians are not innocent, making them legitimate targets. If he is right, then on Cassam’s definition, the 9/11 attacks may be acts of military retaliation rather than terrorism. But if he is wrong, then they are acts of terrorism. How we categorise acts, on the account, depends on how we normatively classify them on grounds of their proportionality and legitimacy.

At the heart of the book is the third conception – mindset extremism. For Cassam, someone has an archetypal extremist mindset if they tend to have four aspects to their psychological profile. First, disproportionate (again, normative) preoccupations with purity, virtue, humiliation and victimhood. Second, disproportionate emotions, especially anger, humiliation, resentment and self-pity. Third, attitudes such as unwarranted lack of compromise, indifference, intolerance and anti-pluralism. Fourth, patterns of thinking that are conspiratorial, apocalyptic and catastrophic.

To elaborate, consider the extremist mindset of Nazism. Perhaps, for example, many Nazis will tend to have a strong preoccupation with racial purity (ibid., 92–3) and a sense of victimhood that, in the case of Nazi Germany, could go back to the perceived unjustness of the Treaty of Versailles. These preoccupations might give rise to salient, though disproportionate, emotions of anger, resentment and humiliation. Nazis are unwilling to compromise on their preoccupations, intolerant towards others, and unwilling to countenance a pluralistic society. And finally, they could be obsessed with anti-Semitic conspiracy theories and driven towards a corrupted Utopianism. This conception of extremism as a mindset goes beyond an ideology cashed-out in terms of one’s position in ideological space, towards an encompassing idea of a total interconnected psychological profile.

With accounts of the varieties of extremism in hand, Cassam then asks whether extremism is sometimes necessary or justified? As Cassam points out, Martin Luther King Jr. thought that some could be extremists ‘for love’ and for ‘the extension of justice’ (2018, 19–20), and some think of the nineteenth century American abolitionists as extremists and fanatics. Clearly, the actions and attitudes of such cases are justified, but that does not justify extremism. First, recall that the definitions Cassam offers for methods and mindset extremism involve a normative component where the extremism is in some way unnecessary, unjust, disproportionate, hopeless or illegitimate. Hence, it cannot be that extremism, as Cassam defines it, is necessary or justified. Second, drawing on the work of English (2016), Cassam claims that extremist methods do not achieve the political change desired by those who wield them. Rather, it is radical politics that has been successful. Instead of describing Luther King Jr. or the abolitionists as extremists, we can treat them as radicals. Labelling them ‘extremists’ is anachronistic.

So, if extremism is to be avoided, how can we prevent people from becoming extremists? This requires us to say how people become extremists and to work backwards from there. Cassam considers two pathways to extremism: psychological and social. The psychological pathway is to adopt an extremist mindset, which will act as an ‘enabling condition’ that will ‘ease the path’ to the use of extreme methods (2021, 175). In particular, ‘a preoccupation with victimhood and indifference to the consequences of violence’ (ibid.). Then, there will be some form of ‘trigger’, such as the ‘perception of grievance’, which will move the agent with the extremist mindset towards violent behaviour.

To elaborate, Cassam looks at the actions of Timothy McVeigh, and how he formed an extremist mindset, including emotions of anger and resentment, which acted as enabling conditions for when his grievance would be felt, to push him towards extreme methods. But there are still people with these mindsets who do not engage in extreme methods. So, why do so few people resort to acts of violence? Cassam suggests that McVeigh and others ‘see themselves as warriors for justice who have no option but to act as they do’ (ibid., 172).

For the social pathway to methods extremism, Cassam considers the claim that people are radicalised within insular groups that function like echo chambers. The downstream effects of this include the hardening of your beliefs and polarisation to extreme positions. Cassam objects to this ‘socio-epistemic’ account because people like McVeigh worked alone rather than operating within a larger group, as might be the case with members of ISIS.

This issue is explored further when Cassam directly considers how we might counter extremism. His way into this issue is to look at the UK government’s policy strategy of tackling extremism called ‘Prevent’. According to Cassam, the Prevent strategy is built on the myth that becoming radicalised is like catching a disease to which some people are more susceptible, which is spread by radicalisers (ibid., 193). The problem with this model is that, first, some people seek out radicalisation, and so it undermines the role of agency in the process (ibid., 197); and second, it ignores the role of arguments in radicalising people. These arguments take the form of narratives that make sense of someone’s experiences and grievances. Drawing on methods used in Saudi Arabia, Cassam argues that the way to deradicalise people is to offer compelling counternarratives that can provide an alternative way of making sense of these experiences that pushes people away from extremism. They must be supported by credible messengers, have relevance and be made accessible to the people they target.

2 The Social Features of Extremism

Cassam’s work on extremism is ground-breaking, and in particular, his account of mindset extremism appears to offer a plausible explanation for what draws people into using extreme methods, and has not been so concisely expressed in other literature. However, the account is not without its weaknesses. In this section, I focus on the general problem that Cassam’s account fails to do justice to the social and group features of extremism. This plays out in different ways. The first way that I will explore is that it fails to adequately distinguish between individual and collective forms of extremism.

One way to see this is to, first, briefly consider Cassam’s distinction between extremism and fundamentalism. Fundamentalism is here understood as a way of being religious that aims to preserve a distinctive identity given the view that one’s religious group is beleaguered by modernising forces of secularisation and globalisation (see Almond et al. (2004); Ruthven 2007). This response tends to manifest in a strict scriptural literalism, an unwillingness to compromise on core religious dogmas, an insular community aiming to preserve the beliefs of its members (Malcolm forthcoming), and Messianic beliefs that good will eventually overcome evil. Now, Cassam proposes a number of overlaps between fundamentalism and mindset extremism that he says ‘are so striking as to make the two virtually indistinguishable’ (2021, 142). These include preoccupations with purity, virtue, persecution and victimhood; attendant emotions of resentment and anger; tendencies to be uncompromising; and Manicheaistic and apocalyptic thinking. He concludes that ‘fundamentalists are mindset extremists’ and the reverse is not true only because not all mindset extremists are religious (ibid.).

But the claim that fundamentalists have extremist mindsets is a significant overstatement. Note that ‘religious fundamentalism’ is an exceptionally broad label, taken to include groups as large and disparate as the Christian Right in the USA; Salafi Islam in the Middle East, Africa and Europe; Haredi Judaism in Israel and North America; some Christian groups in Africa and South America; and some Hindu factions in India. Now, the religion in each of these cases do exhibit the similarities identified by Almond et al. (2004), notably, being a reactionary movement opposed to modernisation. However, the people in each case may lack any similar mindset, and certainly not one that is reactionary to modernisation. These people may have no real understanding of the modernising forces that have led to the creation of the fundamentalist community or movement to which they belong. Those are historical forces that are part of the genealogy of the group, but may not be understood by the people within it. As such, they may have no preoccupation with persecution or victimhood, and no attendant emotions of resentment or anger, connected to modernisation or any other historical forces.

We can extend this point to mindset extremists generally. A group like ISIS could have the kind of mindset that Cassam describes. But that mindset might only be held by a few group members – the majority merely belong to the extremist group. Whilst the identity of the group has the features described by Cassam’s conception of an extremist mindset, in reality, it only makes for the psychologies of a few members. I suggest that this makes space for a distinction in the conception of an extremist mindset: there can be an individual or a group extremist mindset. Just as, on some accounts of belief, a group can believe something even if only some of the members of the group believe it (Lackey 2021, 48–9), so too can there be a group extremist or fundamentalist mindset that is only embedded in the psychologies of a few members of those groups.

Cassam appears to be describing an individual extremist mindset. So, what would it take for there to be a group extremist mindset? One way to develop this concept is by looking to the work on group belief. One account of group belief by Gilbert (1987) is in terms of a joint commitment to believe p as a body. Gilbert maintains that ‘The members of a population, P, collectively believe that p if and only if they are jointly committed to believe that p as a body.’ (Gilbert and Pilchman 2014, 197; c.f. Gilbert 1987, 195). There are two main clauses to draw out from this analysis. The first is the idea of joint commitments. On this account, a joint commitment is one that cannot unilaterally be rescinded. If I agree to meet a friend for lunch, I cannot unilaterally withdraw this commitment: if I decide not to meet her, then I need to give my reasons as to why I am breaking the commitment. If I don’t, then my friend is justified in directing moral blame towards me. As such, joint commitments exert normative pressure on individuals to follow through on them, which is enhanced when the failure to come through on the commitment can have significant negative consequences.

The second clause to draw out is the idea of believing p as a body. According to Gilbert, believing p as a body is quite different from believing p as an individual, in which one has a mental state that represents the world as true. Rather, when believing p as a body, ‘the parties are jointly committed to emulate, in relevant contexts, a single believer—a single party who believes that p—by virtue of the actions, including the verbal utterances, of each’ (ibid. 198). So, it is not the adopting of a mental state that a group agrees to form, but rather the kinds of behaviours and dispositions characteristic of holding a belief that the group is committed to emulating. For, someone who believes p will be disposed to act as though p in relevant contexts, for instance, to affirm p when asked whether p, to behave as though p in contexts where p’s truth would guide action, or to stand up for p if p’s truth is challenged. On Gilbert’s account, the same will be true for collective belief.

This conception gives us the groundwork for an account of group extremism in terms of a joint commitment to various actions salient to the extremist group. And in particular, to emulate a single person with an extremist mindset, who may or may not be an actual agent. These actions could involve giving verbal support to the group’s cause, adhering to a set of doctrines or rituals, or even engaging in violent behaviour. People with joint commitments to these actions or norms may not have Cassam’s individual extremist mindset, but they nevertheless have a group extremist mindset in virtue of their joint commitment to imitate the behaviours of someone who has an individual extremist mindset.

This point may be even easier to see in the case of what we could call group fundamentalism. Fundamentalists generally have commitments to scriptural literalism and an unwillingness to compromise on the truth of core dogmas.Footnote 1 This describes a kind of commitment to a way of interpreting a religious text within a particular hermeneutic framework. Such a commitment, if joint, may be sufficient for one to belong to a fundamentalist group – to be a group fundamentalist. In the case of group extremism, the shared psychological profile amongst people could simply be a joint commitment to the norms of a group. This does not exclude the reality of the extremist mindset in the way that Cassam has described – it’s just that such a mindset may be quite rare for individuals.

The second issue with Cassam’s account is it does to not seem to fully acknowledge the deep socio-epistemic nature of the radicalisation and deradicalisation process. Recall that even in cases of individuals like McVeigh becoming radicalised, his pathway was still inherently social. He read gun magazines and The Turner Diaries which, as Cassam notes, were pieces ‘of writing that expressed views he already held’ (2001, 181). He also spent several years in the military with people of likeminded views. Both what he read and who he met would have been part of the social atmosphere that led to the formation of his mindset. So, even if the radicalisation process cannot always be captured by socio-epistemic accounts, like online echo chambers, there are other socio-epistemic concepts or accounts that can be deployed to explain that process – radicalisation does not happen in a vacuum.

There are similar oversights in the account of the deradicalisation process. Again, the features that Cassam draws attention to are inherently social in nature, but the account can go further to develop on the social epistemology of de/radicalisation. Recall that Cassam draws on methods used in Saudi Arabia as a way to deradicalise people by offering compelling counternarratives that can provide an alternative way of making sense of their experiences which push people away from extremism. But these processes depend on the creation of narratives by others; modes of access to people through books, peaching or online resources; communication and testimony from credible agents; and group acceptance of a new narrative.

Acknowledging these social features of de/radicalisation creates, in turn, space for a social epistemology of extremism that, in some cases, might show that the psychological profile of some extremists is quite rational since they adhere to a narrative that makes good sense of their world. But it can also show where there are problems of rationality, particularly where the social processes that have radicalised someone are unreliable (e.g. misleading authority figures), cultivate bad intellectual character or problematic ways of thinking (e.g. epistemic insouciance), or are generated within echo chamber-like environments (Malcolm forthcoming; Malcolm and Ranalli forthcoming).

Both the group and the social features of extremism can be brought further to the surface to allow greater understanding, and epistemic critique, of extremism. But despite these critical points, Cassam’s work genuinely advances the debate on extremism and brings it into the philosophical discussion.Footnote 2