Over the past two years, counter-radicalisation practitioners have regularly asserted a potential link between autism, mental illness and terrorism. But, underneath the bold headlines, those same agencies and researchers clarify that no causal relationship exists between the three, and that it is rare for people with learning disabilities or mental illnesses to become involved in terrorism. Why, then, is the association continually made? Andrew Silke has referred to Alice in Wonderland to describe this as a ‘Cheshire cat logic’ – one that cannot find evidence of clinical disorders in terrorists, but instead makes vague assertions about pathological personality characteristics. I argue here that by associating mental illness and autistic conditions with terrorism, mainstream society offloads its discomfort with violence. Society cannot understand how someone would commit terrible atrocities against civilians, so it is easier to assume the perpetrator is incapacitated or ‘not like us’ somehow.
The intersection of criminal justice and mental capacity has a long history, one which frequently touches upon terrorism. The ‘insanity defence’, where a defendant can claim they are not responsible for a crime due to mental illness, developed in response to a nineteenth century terror attack. In 1843, Daniel M’Naghton made an assassination attempt on British Prime Minister Robert Peel. He mistook Peel’s secretary for the Prime Minister and shot him in the back, killing him. The discussion of mental incapacity and criminal responsibility which later followed in the House of Lords became known as the ‘M’Naghton Rules’, establishing a precedent for the ‘insanity defence’. Here violence is made comprehensible as the result of pathology – it is explained as ‘one of those things that can’t be helped’ and ceases to disturb mainstream society after being linked to unavoidable illness.
In the contemporary era, we might recall the repeated psychiatric testing of Anders Breivik – who bombed Oslo Government Quarter in 2011 before executing 69 people (mainly teenagers) at a Labour Youth summer camp on Utøya Island. Breivik denied criminal guilt by claiming that he was the Commander of an anti-immigrant resistance movement and thus acting in the public interest against corrupt governments. The assumption was made that he could be psychotic – who else could murder children in cold blood? Psychiatric tests were ordered to ascertain Breivik’s fitness to stand trial, and he was initially declared insane. However, the public outcry in response to these findings led to the Norwegian Court ordering further tests. After the second set of tests, he was declared sane and fit to stand trial, eventually receiving the highest possible criminal sentence under Norwegian law. So, sometimes society recalls the determination of insanity, in order to hold an offender punishable for their crime.
It is difficult to determine whether ‘powerful delusions’ or extreme political commitments motivate a perpetrator, and such decisions are rightly made on the basis of individual cases which face the court. But in recent years, mental illness and learning disabilities have been made central to Countering Violent Extremism programs which attempt to divert people away from terrorism. These programs, like the UK’s Prevent Strategy, intervene on the basis of ‘risk factors’ for radicalisation. These indicators identify broad emotional states and social contexts (peer pressure, lack of self-esteem, anti-social behaviour) and connect them with radicalisation – replicating Silke’s ‘Cheshire cat logic’. They rectify social discomfort around our co-existence with perpetrators, by narrating pathways towards terrorism as interruptible and preventable.
Mental illness and autism are frequently invoked in public commentary on risk factors for radicalisation. However academic research on the prevalence of mental ill health amongst terrorists reaches uncertain conclusions. Studies repeated between 2012 and 2017 have repeatedly identified no common psychopathy or personality factors, and that the majority of terrorist attacks have nothing to do with mental illness. And while other studies identify that ‘lone actor’ terrorists are somewhat likely to have characteristics of mental illness – they do not identify a causal link to particular pathologies.
There is a similar absence of a causal pathway between autism and terrorism. In a recent study of lone actor terrorists, post-1991, only 3.3% of offenders were found to have autism spectrum disorder (ASD). Recently however, the learning disabilities of three attempted bombers in the UK have attracted significant media attention. Nicky Reilly, who had Asperger’s Syndrome, bombed the Giraffe Restaurant in Exeter in 2008 (injuring only himself); Damon Smith, who has an autism spectrum condition, left a bomb on a tube train in 2016 (it did not detonate); and Lloyd Gunton, who has autism, planned to attack a Justin Bieber concert in Cardiff in 2017 (but was arrested beforehand). The contrast with attacks by non-ASD perpetrators, and the carnage wrought by them upon public spaces, could not be more striking.
The grand total of people killed and injured in these plots totals zero. Furthermore, research has shown that militant organisations avoid relying on those they perceive to suffer from ASD or mental illness as this is perceived to risk detection and capture. So why do Police Chiefs and Prevent officials highlight the tiny minority of people with impairments who go on to attempt terrorist crimes?
Firstly, it is easier for mainstream society to offload discomfort about terrorism by associating it with illness and impairment. Violence is ‘externalised’ as a result of pathology. Secondly, there has been a long term transition within public policy whereby acting upon predicted ‘risks’ has become normal. Interventions in crime and disorder are now staged pre-emptively, before they occur – on the basis of warning signs. This means they operate in the realm of possibility rather than certainty. This has transformed counter-terrorism. Governments now presume to identify potential militants before they attack. The idea of ‘radicalisation’ is crucial here, even though it has none of the evidence base of studies which pose alternative understandings of violent mobilisation. Counter-radicalisation identifies factors which could lead to terrorist activity, but which don’t necessarily.
By framing mental illness and ASD as ‘risk factors’ for terrorism, these two historical trends come together. Counter-radicalisation policies, like Prevent, treat terrorism as the result of an interruptible process called ‘radicalisation’. And in the UK, it is now the duty of teachers, social workers, nurses, dentists, doctors and prison wardens to report concerns about radicalisation to the Prevent program. The training provided to these public sector workers lasts an hour and identifies broad emotional and behavioural states (like withdrawing from social settings, scripted speech and/or experiences of trauma) as potential indicators of concern.
Public sector workers are accustomed to protecting the children and adults in their care from abuse. They look after their best interests and protect them. And because Prevent is framed as a safeguarding duty, professionals tend to accept that people with ASD and mental illness might be susceptible to radicalisation – despite evidence that they are rarely involved in successful terrorist attacks. Illness and impairment is framed as vulnerability, which is then linked to potential future criminality.
The blurring of ASD/mental illness with terrorism is the result of Prevent being framed as a safeguarding duty. Safeguarding emerged to protect children and vulnerable adults from abuse. But it has been hijacked to associate vulnerable people with radicalisation risk. Their vulnerabilities are deemed to make them ‘at risk of becoming risky’.
Blurring Prevent with safeguarding has created a geography of suspicion around those with mental illnesses and learning difficulties. As a result, those receiving treatment and assistance with their conditions are now more likely to be reported to Prevent. This year, it was discovered that four mental health trusts in the UK routinely screen all their service users for signs of radicalisation – rather than limiting such intervention to cases where radicalisation indicators were presented. They are profiling their service users as potential radicalisation risks. Similarly, concerns that people with learning disabilities are vulnerable to radicalisation have prompted Google and the Institute for Strategic Dialogue to fund projects which work to increase their resilience.
This push to protect people with learning disabilities and mental illnesses actually risks stigmatising them as potential terrorism risks. And this is not without consequences. Yousef Farooq was referred to Prevent by his learning support assistant, after he corrected his teacher on the historical meaning of jihad. His learning difficulties influenced the perception that he was vulnerable to nefarious influences. Here Prevent creates a geography of risk whereby religiosity and race combine with cognitive/psychological factors to stigmatise people who deserve better.
 Bhui, K. S., Hicks, M. H., Lashley, M., & Jones, E. (2012). A public health approach to understanding and preventing violent radicalization. BMC Medicine, 10, 186; Bhui, K. S., & Jones, E. (2017). The challenge of radicalisation: A public health approach to understanding and intervention. Psychoanalytic Psychotherapy, in press.
 Paul Gill, John Horgan and Paige Deckert (2013) ‘Bombing Alone: Tracing the Motivations and Antecedent Behaviors of Lone-Actor Terrorists’, Journal of Forensic Sciences 59(2), pp.425-35.
 Corner, E., Gill, P., & Mason, O. (2015) ‘Mental health disorders and the terrorist: A research note probing selection effects and disorder prevalence’, Studies in Conflict and Terrorism, 39, pp.560-568.
Charlotte Heath-Kelly is associate professor in Politics and International Studies at the University of Warwick.