Reviewing the links between violent extremism and personality, personality disorders, and psychopathy

ABSTRACT

Many early published analyses of the terrorist placed psychopathy as the core explanatory variable for terrorist behaviour. This speculative opinion was derived mainly from popular culture, and the desire to attribute mental disorders to those committing such violent acts. Poor research designs and a lack of empiricism ultimately undermined these arguments in favour of terrorism being rooted in disorders of personality. Multiple studies supporting psychopathic and personality-level explanations were conducted in the absence of rigorous clinical diagnostic procedures. Despite the methodological issues, concluding remarks from this research continues to hold instinctive appeal across the research field. This incentivises a need for a rigorous synthesis of the evidence base. The objective of this systematic review is to assess the impact of personality upon attitudes, intentions, and behaviours in the context of radicalisation and terrorism. This paper follows the same systematic process as the Gill et al. paper in this special issue. However, we use the model to interrogate the existing empirical literature on personality and terrorism in terms of its coverage, common themes, methodological strengths and weaknesses and implications. The search strategy for the systematic review is based on the Campbell Collaboration method. Results and their implications are discussed.

KEYWORDS:

• Personality
• Psychopathy
• Radicalisation
• Violent Extremism
• Terrorism

View correction statement:

Correction

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. These are (1) ambivalence toward authority, (2) defective insight, (3) adherence to convention, (4) emotional detachment from the consequences of their actions, (5) sexual role uncertainties, (6) magical thinking, (7) destructiveness, (8) low education, and (9) adherence to violent subculture norms and weapons fetishes.

2. As given in both the PCL-R (R. D. Hare, 2020) and the description of antisocial personality disorder in the DSM-5 (American Psychiatric Association, 2013). Strentz (1988) defines that leaders cloak their paranoia through charismatic self-confidence and a commanding presence, whereas the activist-operators are defined by a history of criminal activity and desire for violence and hedonism.

3. Currently, there exists a distinction between the diagnosis of ASPD, dissocial personality disorder (DPD), and that of psychopathy (although the DSM-5 notes that these diagnoses are referred to interchangeably (American Psychiatric Association, 2013, p. 659)). Some authors, however, consider that, much like other personality disorders, ASPD and DPD diagnoses focus on observable behaviours, whereas the diagnosis of psychopathy also requires interrogation of personality traits (R. Hare, 1996; Ogloff, 2006; Venables et al., 2014). Whereas others argue that psychopathy and ASPD are at ends of the same diagnostic continuum (Coid & Ullrich, 2010).

4. For more information, see www.campbellcollaboration.org

5. For more information, see http://eppi.ioe.ac.uk/cms/Default.aspx?alias=eppi.ioe.ac.uk%2Fcms%2Fer4

6. Studies that did not measure significance, but reported other outcomes, such as effect size were assessed using the discretion of the coder. This predominately relied on assessing the core value of the statistics measured during the study and following appropriate guidelines regarding individual statistical tests.

7. Within the primary review team.

8. For example, the reviewer identified all use of the term ‘radical peers’, ‘radical friends’ and ‘social bonds’ and, after checking the source document to ensure accuracy in the meaning of the term, changed this to ‘social networks’ for greater consistency across the variables.

9. For example, Weisburd et al. (2001) identified differences between effect sizes of interventions between randomised experiments and quasi experiments. Weisburd et al. found that those studies with weaker methodological designs were more likely to find that an intervention was effective due to extraneous influences from confounding variables on offending.

10. The novelty of this review as compared to the work of Misiak and college is explained elsewhere in this issue (Gill et al., 2021)

11. These studies were not taken forward for analysis (and are covered elsewhere in this issue – see Gill et al., 2021).

12. As Jones did not employ a control group and was measuring attitudes using correlations (and the regression models identified that any significance disappeared), this indicates that there are high threats to internal validity and it is not possible to establish causal order between psychopathy and right-wing authoritarianism.

13. Bélanger et al. (2014) randomly assigned participants to a number of conditions, helping to remove threats to internal validity.

14. The activism-radicalism intention scale (Moskalenko & McCauley, 2009), the short Coolidge axis II inventory (Coolidge et al., 2010), cognitive complexity instrument (Bagdasarov, 2009), intolerance of uncertainty scale – short form (Carleton et al., 2007), rational decision-making style (Scott & Bruce, 1995), cognitive style index (Allinson & Hayes, 1996), the frustration-discomfort scale (Harrington, 2005), need to belong scale (Leary et al., 2013).

15. Although the authors employed SEM, which enables the testing of direct and indirect effects of relationships, no control group was employed, and the employed measure of radicalism (the activism-radicalism intention scale (Moskalenko & McCauley, 2009) has not yet been tested for validity across populations, and was only first tested for reliability in the Soliman et al. (2016) sample.

16. Although none of the below studies cite Bélanger et al. (2014), given their results, it is reasonable to assume that antisocial behaviours may be of utility in explaining radical behaviour.

17. As measured using the following proxy statements: ‘“I feel strongly British (English, Scottish, Welsh or Northern Irish) if that means standing up for yourself or your country”; “I feel more like people with my own religious, cultural or political beliefs than people who are British”; “I support the war in Afghanistan”; “I oppose the war in Afghanistan”; “I could fight in the British army in Afghanistan”; “I could fight against the British army in Afghanistan”’ It could be argued that these items do not accurately capture extremist attitudes (see Ozer & Bertelsen, 2018), and using these statements to measure extremist attitudes may introduce threats to internal validity.

18. The inclusion of comparable groups of offenders reduces threats to the internal validity of the study.

19. With standard deviations from the mean also not differing across groups (controls; 9.61, terrorists; 9.66, murderers; 10.46).

20. We note that a high proportion of the studies reviewed undertook self-report surveys and questionnaires. These designs are problematic for determining disordered presentations, so the conclusions from these studies are interpreted with regards to personality traits only and not clinical presentations.

21. In order to determine the most appropriate traits for inclusion in the review, the authors undertook substantial thematic work of wider personality literature to help guide the allocation of all personality traits identified as significant across all studies reviewed.

22. Pauwels et al. (2014) examined rates of self-reported political violence and extremist propensity in their sample of 4473 young people in Belgium and Antwerp. They identified that impulsivity was significantly associated with political violence to both persons and property. Schils and Pauwels (2016), L. J. Pauwels and Svensson (2017), and Pauwels et al. (2020) furthered this work, identifying that those with the lowest self-control showed the highest levels of extremist beliefs.

23. Superiority was assessed using four items, with the authors reporting that the validity for these items ‘was satisfactorily (.71)’ (Doosje et al., 2013., p. 593)

Additional information

Funding

This work was supported by the Department of Home Affairs (Australia); European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme [758834]; Public Safety Canada [8000-18884].