Behavioral Problems and Disorders among Radicals in Police Files

by Anton W. Weenink

Whatever label we put on those who participate in this type of behavior, our core focus must be on the behaviors themselves, and the ways in which they develop.- John Horgan [1]


In this article we explore to what extent behavioral problems and disorders can be found in a sample of radical Islamists that are known to the police in the Netherlands as actual or potential ‘jihadists’. Our aims are, first, to assess whether the consensus in terrorism studies that terrorists are ‘surprisingly normal’ is justified when it comes to these subjects; second, we try to establish whether behavioral histories offer clues to the police on how to approach them. Personal details of 140 subjects, who are considered to have traveled from the Netherlands to Syria, or on whom police had information that they might be preparing to do so, were entered in police databases. Preliminary results indicate that individuals with histories of behavioral problems and disorders are overrepresented. The results are at odds with the consensus view on terrorists alleged ‘normality’. A focus on individual psychology could complement existing social-psychological approaches to radicalization. It may also assist in broadening awareness among policy makers and law enforcement officials that disengagement efforts need to be tailored to the individual, and that mental health specialists might have to play a role here.

Keywords: Jihadists, the Netherlands, terrorist psychology, mental health, criminology


The starting point behind this article was unease with the accepted wisdom that “the outstanding common characteristic of terrorists is their normality,”[2] and that terrorists do not “suffer from some sort of mental disorder.”[3] Data in terrorism and extremism files of the National Police of the Netherlands appear to provide evidence that several subjects did have a history of problem behavior, if not personality disorders or mental illness. Whether or not subjects with ‘troubled’ personalities were overrepresented was not immediately clear. Therefore, and because police – for whom prevalence rates are less important–need to know how to approach radicalized individuals, we decided to take a closer look at an issue many researchers consider settled.[4]

The Consensus

Since the early nineteen eighties, the view that terrorists tend to have ‘normal’ social backgrounds and are not mentally disordered has become widely accepted among terrorism researchers. Studies on Italian, German and Irish terrorists failed to provide evidence for any major psychopathology or terrorist personality; notably, many terrorists had unremarkable middle class backgrounds.[5] After 9/11, studies of jihadi terrorism seemed to confirm this image of terrorist normality. Dismissing Neo-Freudian theories that postulate terrorists to suffer from pathological narcissism, paranoia or authoritarian personality disorder, Sageman concluded, “terrorists are surprisingly normal in terms of mental health”.[6] Silke, in a review of studies into the psychology of jihadists–those by Sageman [7] and Bakker in particular [8]–arrived at the same conclusion.[9]

In an attempt to explain ‘radicalization’–defined as the process that puts people on the path to political violence,[10] or of becoming involved in terrorism[11] – some researchers use an alternative approach, social psychology. Sageman, for example, rejects micro and macro level theories that explain radicalization, and proposes a mid-range theory in which four elements are key: “a perceived war on one’s in-group; moral outrage at some salient major injustice; resonance with personal experiences; and mobilization by an already politically active network”.[12] The Internet fosters a bottom-up process of radicalization and self-recruitment.[13] The fact that terrorists are heterogeneous in several respects is utilized to confirm the impossibility of finding a terrorist personality, or unique properties of terrorists that would allow for making a terrorist ‘profile’.[14] Horgan and Bjørgo draw attention to the distinct motives and psychological processes that are at work at different stages, by which individuals engage in terrorism, remain terrorist, disengage from terrorism, or become recidivists; they also stress that different types of individuals get involved in different types of activities.[15] Different mechanisms explain how groups radicalize.[16] There is no need to assume terrorist irrationality, as there seems to be some logic in terrorist normality: The mentally ill would be a liability to clandestine organizations.[17] This explains also why Lone Actor Terrorists (‘Lone Wolves’), who seem to be the exception to the rule of terrorist normality, indeed operate alone.[18]

Undoubtedly, social psychological approaches have enriched our understanding of the process of radicalization. Acknowledging identity problems and grievances in immigrant youth, and differentiating between roles, phases and motivations can assist policymakers in responding more adequately to radicalization. Showing there is no ‘terrorist personality’, or that profiling of terrorists is impossible helps to prevent the introduction of policies that would be ineffective, or even counter-effective as they could lead to stigmatization.

This being said, we think the evidence for ‘normality’ in radical Islamist terrorists is less unambiguous than currently accepted. Bakker, for example, found that mental illness was overrepresented in a sample of European jihadists.[19] Venhaus, in a large sample of foreign fighters, did not find “signs of any clinical psychosis”, but did note that “antisocial behavior was clearly present in all”.[20] Merari likewise found that “although none of the would-be suicides was diagnosed as psychotic, most of them had personality traits which made them more amenable to recruiting for suicide missions”.[21] In the Netherlands, on the basis of files from criminal investigations into Dutch jihadist networks, De Poot and Sonnenschein found that many subjects had a criminal record, had been substance abusers, and had attained lower educational achievements.[22] In Germany, on account of interviews with 39 convicted extremists and terrorists of different ideological backgrounds (twenty-four right-wing, nine left wing, and six Islamist oriented extremists) Lützinger found that they “did not display any pathological features”; however, she did note that “all individuals in our group had experienced the same irregular developments and shown the same deviant behaviour during the first and second socialisation instances (family and school) as other offenders who did not necessarily become extremists or terrorists”.[23] Bouzar found that 40% of jihadists in a French sample had suffered from depression.[24] These studies diverge in design and results, but overall, suggest that the social and psychological background of terrorists tends to deviate from the average. Radical Islamists came from less privileged social strata,[25] and although most of them were not mentally ill, many had a history of problem behavior and troubled backgrounds. In this article we assess whether these findings also hold for the new wave of radical Islamists from the Netherlands: those who travel to the Middle East to join jihadist organizations.

Research Focus

Research question, definitions, and goals

This article is the result of explorative research that was conducted between February and November 2014. Where possible, we add recent findings. Our research question is: to what extent do jihadists have a history of problem behavior or mental disorder? We conducted a number of database searches in the Dutch National Police database by entering the personal details of known and suspected jihadists. We searched for information indicating that these jihadists have been diagnosed with a disorder or disability (conduct disorder, personality disorder, mental illness, cognitive disability), as well as for signs of problem behavior. The concept ‘problem behavior’ usually refers to ‘difficult’ behavior in children, as well as in adults; we use it as an umbrella term for conduct that deviates from the social norm and causes harm or distress to oneself or others; such as a history of quarrels, crime, and violence.[26] The focus is on maladaptive behavior that suggests a lack of self-control (see under Method as well).[27] When an individual persists in such behavior, it may flag an underlying ‘psychopathology’, such as antisocial personality disorder or a mental illness. ‘Jihadists’ is the term we use for radical Islamists who participate in what they perceive to be the ‘Jihad’: a holy war against perceived ‘enemies of Islam’. (We will not treat ‘radical behavior’, which could be considered a form of deviancy in its own right, as such as indicative of some mental health problem). In addition to the search for mental health problems, we also take a brief look at some background variables that may help us assess whether Dutch jihadists tend to have normal and ‘unproblematic’ social backgrounds or not: age, gender, nationality and, as far as our sources allow, socio-economic characteristics. We consider backgrounds to be more ‘problematic’ if a subject’s relations with family and friends are unstable, if he or she did not finish school, is unable to find employment, is homeless, etc.

Our research serves two purposes: to establish whether or not jihadists are relatively troubled, and to highlight individual cases, which might be of help in developing ways of approaching them. In a next phase the research could be extended to other forms of potentially violent radicalization, such as right- and left-wing and single-issue extremism.

The sample

The sample is a list containing personal details of radical Islamists from the Netherlands whom the Dutch police suspect of having joined the fight in Syria, or are considered potential travelers (for example, because they have expressed their intent to do so). The list is a national ‘List of Travelers’ (LOT), as compiled by the Counterterrorism and Extremism (CTE) team in the Central Unit of the Dutch National Police. The original data come from local police units. When local police believe that a person in their district has left for Syria (or is considered a potential traveler or returnee, etc.), they share the personal details of this person with the CTE Team. Although many of the subjects can be considered ‘foreign fighters’, the more neutral term ‘travelers’ will be used here. This allows for differentiating between those who leave to participate as a fighter and those who take on other roles (for example, as a housewife or a provider of humanitarian aid). Importantly, it also includes those who have not yet departed. A LOT from February 2014 was chosen as the research sample; this was the most recent List at the time the research started. This LOT will be referred to here as ‘List S’. List S contains the personal details of 140 individuals.

The number of subjects on the LOT, as well as their statuses, changes nearly on a daily basis. Joining the fight in Syria became popular amongst radical Islamists by the end of 2012; in 2014 Iraq became a country of destination as well.[28] By November 2014 the LOT contained the details of almost 300 ‘travelers’. Evidently, during the research, the sample size as a share of the total population of travelers declined from 100% to less than 50%. Additionally, travelers’ statuses changed as people moved in and out of the conflict zone, were killed or arrested, etc. In February 2014, eight travelers had already died in Syria; by January 2015, the number of casualties in the sample had risen to 16, and 21 in the LOT.


We use several police databases. To retrieve information we entered full name and date of birth of each subject from List S in an online facility that gives access to four underlying databases, with the following acronyms (which we do not translate here):

  • BVH: files from local police units; if there is a hit, it leads to police reports from community police officers, describing home visits, street quarrels, citizen’s reports to police, etc. BVH also contains many interviews with and interrogations of subjects and their acquaintances. BVH was our most important source.
  • HKS: all crimes adult subjects have been suspected of and for which police drew up a report for the public prosecutor; HKS does not show if the public prosecutor decided to drop a case. A single report may contain records of more than one crime.
  • GBA: citizen registration: address and migration.
  • VIP: residential history, including data that indicate periods of homelessness, institutionalization, detention.

We also consulted a database with the criminal record of subjects; ‘JD-Online’, which shows a subject’s life record of convictions for both crimes and misdemeanors.


It should be clear that we do not diagnose.[29] We do not have access to the subjects or their medical files, and police data are not detailed enough in order to consistently apply criteria from, for example, the DSM V classification of mental disorders. We merely assign subjects to the following categories on the basis of information found:

  • Category A No hit

For individuals without any traceable problem behavior or mental health problem, we created two subcategories:

A.1 No problem behavior found

A.2 No problem behavior found, but problematic social setting

Category A.2 allows for tracing back those cases where we found no information on problem behavior in an individual, but where there is evidence of serious social problems.

  • Category B: Problem behavior

B.1 Mild problem behavior: a limited number of petty crimes and misdemeanors, vandalism, high debts, etc.

B.2 Problem behavior: Persistent offending; child neglect (by the subject).

B.3 Serious problem behavior: e.g. persistent serious criminality; references that a subject is mentally impaired, or seems to suffer from a conduct or personality disorder, or mental illness, but where no indications of a diagnosis were found.

A subject with a criminal record containing less than 10 suspicions or convictions [30] for crimes that are not too serious, ends up in Category B.1 when there are no additional signs of problem behavior. An individual is categorized in B.1 as well when local police consider an individual to be a member of a criminal youth group. Category B.2 contains subjects that have been suspected of, or convicted of, 10 or more crimes, or of more serious crimes as well as cases of child neglect. Category B.3 covers individuals with apparently serious and persistent problematic behavior. Serious problem behavior can pertain to persistent offending, but with an extra factor – such as, repeated violent crimes, repeated escalating conflicts with family or neighbors, self-mutilation, child abuse or molestation etc. Category B.3 also comprises those with cognitive disabilities, as well as those who were institutionalized, but for whom no psychiatric diagnosis was found.

  • Category C: Diagnosed mental health problem

Category C contains individuals of whom we found information that they have been clinically diagnosed with a disorder. The main source here are police reports; indications of a diagnosed mental health problem in a police report we try to control in other police reports and in sources like VIP (residential history) or verdicts. Having a mental disorder does not necessarily imply more serious problem behavior, or that someone is to be considered more dangerous than individuals in Category B.

In a next phase, we hope to replace the use of categories with a formal coding system and procedure, in collaboration with forensic psychiatrists. Below, we compare our results with prevalence rates in DSM V, and with findings from a mental health service in the Netherlands. We do not yet have a control group.


A minimum of inter-subjectivity was realized by sharing the method and findings with experts. The plan and results were discussed with two social scientists, a forensic psychiatrist and researchers at the National Police. A behavioral expert advised on the categories we use to sort cases. Crime analysts, criminal investigators, and an expert in Arabic studies at the CTE Team, all with many years of experience in the field, provided information on subjects and their histories, and discussed our findings with us. Initial results were discussed with the CTE Team, other police researchers and forensic psychiatrists at the National Police. An expert from a psychiatric health service commented on our results and shared quantitative and qualitative findings.


Readers should be aware of several limitations. As mentioned, there was no access to psychiatric reports, and medical experts are not allowed to share mental health data on individuals with law enforcement, except under severe legal restrictions.[31] Neither are we permitted to discuss specific cases. Further limitations pertain to the completeness and quality of databases. Files from local police in BVH that are older than five years, cannot be consulted, and the police registration system of suspicions (HKS) in most cases does not contain data from the period before a subject reaches 18 years of age. HKS appeared to be incomplete as well, because several records appeared to be missing. Furthermore, police do not record socio-economic and educational statuses of subjects in a systematic way.


We present our results in two steps. First, we discuss demographic characteristics of travelers, followed by the discussion of behavioral and mental health issues.

Demographic variables

The subjects in our sample are not a cross-section of the ‘normal’ population when it comes to their nationality, age, gender, and other social and economic characteristics.

  • Nationality

In the Netherlands it is possible to have two nationalities. Having a second nationality can be considered an indication of immigrant status of travelers or their parents. In the sample, the first nationality is Dutch in 133 (95%) cases; 108 subjects (77%) have a second nationality; 25 (18%) are only Dutch and 7 (5%) are not Dutch. Evidently, most travelers come from immigrant families. First and second nationality was distributed as follows:

Figure 1 Nationality by number of travelers (n1=140)

The largest groups by second nationality are Moroccans (78 subjects, 56%) and Turks (13 subjects, 9%). These percentages are considerably higher than the share of their immigrant communities, that each account for roughly 2% in a population of 16.8 million, as of January 1st, 2014.[32] We can safely conclude that jihadists of Moroccan descent are overrepresented in the sample. Nevertheless, a total of 78 travelers also testifies to how rare Jihad travel is in the Moroccan community, as it is in others.

  • Age and gender

The age distribution of travelers is similar to findings in earlier terrorism research: The mean age in the sample is 24. On average men are older (25) than women (21). With 117 men (84%) and 23 women (16%), males dominate the sample. Travelers tend to be young, but Jihad travel is not at all ‘adolescence limited’, as criminality is.[33] On January 1st 2014 only ten subjects had not reached the age of 18. At 17%, women are more often underage than men (5%). Only one underage traveler, a sixteen-year-old boy, effectively joined a jihadist organization, and was killed.

  • Social and economic characteristics

Some findings cannot be quantified, such as that many subjects in our sample seem to come from broken families; six subjects had lost a parent, in two cases due to suicide. Additionally, educational achievements tend to be rather low. We could not find a subject with a completed higher education in the sample thus far. Subjects either did not finish high school or vocational training, or became unemployed afterwards. We did not find subjects who had had a steady career; when they were employed, it was mainly in irregular jobs. As mentioned, several individuals had been homeless – six were at the moment of registration on the LOP–for longer or shorter periods of time. Homelessness seems to be related to conflicts with parents and partners, or with finding no place to live after detention.

Demographic characteristics indicate that travelers do not form a cross-section of Dutch society. Males in their mid-twenties, with immigrant and socially vulnerable backgrounds dominate the sample.

Behavioral Problems and Disorders

  • Crime

Our crime data suffer from inadequacies that do not allow for discussing involvement in crime extensively. Of 66 travelers (47% of the sample) we found one or more reports in HKS; these 66 received 380 police reports. The overall registration level in HKS is 23% for young men younger than 22 [34] suggesting that male travelers are twice as often involved in crime as other young men. The percentage comes close to that of the group in the Netherlands that is criminally most active, that of young men of Moroccan descent, that stood at 54% in the study of Blokland et al. Subjects of Moroccan descent are overrepresented as we saw above, but this cannot completely explain the high crime level in the sample.[35]

Twenty six percent of the women had a record in HKS, compared to 5% of women under the age of 22 in the Netherlands as a whole, suggesting that female travelers are five times as often involved in crime as other women.[36] In fact, women travelers tend to be ‘more criminal’ than men in the wider population.

Curiously, the 47% with a HKS registration, is lower than that of travelers in JD-Online that have been convicted for a criminal offence (53%).[37] This indicates that HKS is incomplete because not all police reports lead to a verdict. Since HKS data is incomplete, we refrain for now from discussing the type of crime travelers have been involved in.

  • Behavioral problems and disorders

The figure below shows the score in the sample for each of our categories.

Figure 2 Behavioral problems and disorders, by category (n 1 = 140)

Our results indicate that a psychosocial problem can be identified in 60% of travelers. Just under half of the subjects exhibit problem behavior (46% in Category B). In one in five there appears to be either serious problem behavior (14% in Category B.3) or a diagnosed personality disorder or mental illness (6% in Category C). Below, we present details for subjects in categories B.3 and C.

  • Category B.3

In twenty individuals, five women and fifteen men, we found indications of serious problem behavior, or indications of a mental health problem of which we did not find a diagnosis:

  1. Mentally impaired (most likely).
  2. Psychiatrically examined. Institutionalized. Child neglect. ‘PTSD’ (Post Traumatic Stress Disorder).
  3. Special education. Radicalization after illness.
  4. Persistent offender. Stalker. Violent crimes. Fascination with extreme violence.
  5. Child abuse. Intimidation.
  6. Criminal. Mental capabilities under examination.
  7. ‘Borderline personality’.
  8. Psychiatric help to manage anger.
  9. Ignored advice to attend psychiatrist. Parent molested other parent and children.
  10. ‘Compulsive disorder’. ‘Autistic / schizophrenia’.
  11. Unmanageable as a child. Youth welfare work. Tantrums.
  12. ‘Autism’. ‘Childhood trauma’.
  13. Special education due to problematic behavior.
  14. Mild disorder, tantrums.
  15. ‘Strange’ behavior. Aggression against parent; other parent psychiatric patient.
  16. Mentally impaired. Criminal. Aggressive. Institutionalized repeatedly.
  17. Several times in a closed youth detention facility.
  18. Persistent offender. Homeless. ‘PTSD’.
  19. Psychiatric issues (most likely depression).
  20. Mentally impaired. Aggressive.
    • Category C

In eight individuals, two women and six men, we found a diagnosed disorder:

  1. Psychotic.
  2. Narcissistic. Substance abuse / addiction. Crime since age 13.
  3. AD/HD (Attention-Deficit/Hyperactivity Disorder). Medication. Petty crime, many misdemeanors, prolonged periods of homelessness.
  4. ADD (Attention Deficit Disorder), schizophrenia, refuses medication
  5. Psychotic; medication and psychiatric aid.
  6. Schizophrenia.[38]
  7. ASD/PDDNOS (Autism Spectrum Disorder / Pervasive developmental disorder - not otherwise specified). Aggressive.
  8. PTSD. Persistent offender. Homeless. Unmanageable as a child.

The descriptions reveal comorbidity of difficulties in individuals, and many individuals come from ‘multi-problem’ families.

Behavioral problems and disorders in a second group of travelers

In the period March 2014–January 2015, 158 additional travelers were listed. So far, we have analyzed the details of 43 travelers (27%) from this ‘second group’ (32 men, 11 women).


Subjects in group 2

Preliminary %

(n2 = 158)






































Table 1 Behavioral problems and disorders in a selection of 43 travelers from a second group (n2 = 158)

With 16.4% the total of Categories B.3 and C is somewhat lower, but we only used 27% of this second ‘group’. Therefore, the score of these categories can be expected to rise. Nearly all Category B.3 subjects in the second group received some form of psychiatric help. The cases in the ‘second group’ are relatively heavy because of selectivity bias: we typically studied them after members of the CTE Team brought them to our attention when they presumed these were potentially relevant cases. Hence, subjects without behavioral problems will be underrepresented. We did decide to show the preliminary results from the second group, as they reinforce the impression of widespread behavioral problems and disorder in jihadists.

Do the Troubled Persons make it to the Jihad Zone?

One reason why terrorists were supposed to be ‘normal’ was that mentally unstable persons are thought to be a liability to a terrorist organization. If this were true, we would not expect many troubled subjects to reach the conflict zone, let alone to join groups like the Islamic State (IS) or Jabhat al-Nusra. Some subjects on List S that we categorized as B.3 and C attempted to join, but indeed did not make it. However, in both categories 75% reached Syria or Iraq (15 in Category B.3 and 6 in Category C), which is slightly higher than the ‘rate of success’ in the sample as a whole (73%, 102 subjects).[39]

Involvement in Terrorist Activity

Police data and open sources suggest that jihadists from the Netherlands may have been involved in atrocities: beheading, stoning, and flogging, as well as the execution of prisoners of war. Two travelers from List S died in suicide attacks. From the ‘second group’ two subjects became suicide bombers as well. Thus, some jihadists from the Netherlands may have been involved in terrorist activities and in war crimes. Among the perpetrators we find Category C subjects (both from List S and from the group that joined later). They are terrorists who were diagnosed with a mental illness.



Our finding of eight persons or almost 6% of the sample with a diagnosed personality disorder or mental illness indicates that mental health problems are more common in jihadists than in the general public. In our sample, three people (2%), suffer from schizophrenia. The prevalence for schizophrenia in DSM V is 0.3 to 0.7%.[40] Two persons in the sample (1.4%) suffer from psychosis, which has a lifetime prevalence of 0.21 to 0.54%.[41] In the second group we so far found five people with psychosis. This indicates that with 7 subjects in the population of 298 travelers (and potential travelers), as of January 2015, prevalence of psychosis is 2.3%. Both afflictions seem to be overrepresented, but the numbers are too small to be definite. We have not yet checked our findings against a proper control group, which might diminish the gap. On the other hand, we do expect that closer scrutiny will reveal that some of the people in categories B.1 to B.3 and A.2 belong in Category C. This impression, that our results represent a lower limit, was confirmed by a mental health service.

Findings from a mental health service

In January 2015 we discussed our findings with an employee of a mental health service. The service employs psychiatrists who treat and guide psychiatric patients; it has a forensic psychiatric department that treats disturbed criminals. It participates in a multiparty policy platform on radicalization, through which it receives personal details of radicals provided by local authorities from one of the ‘larger’ municipalities in terms of the number of jihadists. The service is only allowed to reveal whether these radicals – travelers and other Islamist radicals – have been under psychiatric treatment on a ‘yes or no-basis’. In this municipality, the answer was ‘yes’ in 31 cases so far, which accounted for 60% of all cases submitted. Every week, two or three new cases can be added concerning radicals who have been under psychiatric treatment. In another municipality 76% (19 out of 25) of radicals had been under supervision of the Child Protection Service. The data from the service indicates that mental illness and disorders in radicals are more common than we found.[42]

With regard to travelers, our respondent observed that nearly all have experienced domestic violence – as a victim, a perpetrator, or both. Among radicals that have not or not yet joined the Jihad, many suffer from mental illness or personality disorder, but one should differentiate between types. There are ‘ideologues’ and ‘thrill seekers’. Ideologues, among whom many converts can be found, tend to suffer from a weak ego. Thrill seekers are merely attracted to radical Islam for the kick; they tend to have behavioral problems or are mentally impaired, and typically have a criminal record. Returnees from Syria have been damaged by their warzone experience and need to be ‘de-escalated’ (rather than de-radicalized) in some way.[43] Our respondent made two final observations. Firstly, there are cases of people suffering from psychosis who use the rhetoric of radical Islam because it is a sure way of attracting attention. They should be distinguished from genuine radicals. Secondly, many of the problems found in radical Islamists can be found in right-wing extremists as well, which confirms some of our first impressions of this group.[44]


The preliminary answer to the research question is that at 6% in our sample people with a diagnosed mental health problem are overrepresented. Another 46% displayed problem behavior. In 48%, we found no signs of problem behavior or mental health problems, but 8-percentage point of this group seems to have a problematic social background. We tend to think that our findings represent a lower limit of the prevalence of behavioral and mental health problems because 1) we had no access to medical files; 2) police sources are incomplete, and 3) a mental health institution found that 60 percent of radical Islamists in their files received psychiatric treatment.


Our results and the categories we use are purely descriptive and do not reveal causes. Behavioral and mental health problems may spring from any of the ‘biopsychosocial’ factors that affect individual development. Nevertheless, because the normality thesis does not appear to hold for the jihadists studied, we discuss some tentative answers as to why this could be the case.

  1. The thesis of terrorist normality may have been premature, considering the lack of data many terrorism researchers complain about.[45] Victoroff observed: “the total number of published theories exceeds the number of empirical studies - an imbalance that may be of more than academic import … the much-cited claim that no individual factors identify those at risk for becoming terrorists is based on completely inadequate research.”[46]
  2. Theories of terrorist psychology have been refuted on the ground that there is no single psychological characteristic that allows for defining a ‘Terrorist Personality’ or a ‘Terrorist Profile’. Our results confirm that mental issues diverge, but we see no reason why a psychology of terrorism should be about finding similar peculiarities in every single case. An example of this is Silke, who was not impressed by a finding that 33% of West-German terrorists had lost a parent: “It would be one thing if 92% had lost a parent before 14, but 33% is of uncertain significance. It still leaves 67% with both parents, but who nevertheless went on to become terrorists”. But what if the other 67% had different troubles that contributed to their radicalization?[47]
  3. Individual and social psychology are often regarded as rival explanations, whereas we consider them to be complementary levels of analysis.[48] Regardless, social psychological approaches are ill equipped to explain why the overwhelming majority of youngsters from perceived social risk groups abstain from radicalization, and do not go to Jihad.[49] Apparently, only few individuals are vulnerable to radicalization. Here we would like to hypothesize that a causal nexus between psychosocial problems and radicalization might be that ‘problem behavior’ tends to isolate individuals, not only from society, but from relatives and peers as well. As social psychological approaches suggest, social isolation is often a starting-point for radicalization. Isolated individuals are in need of companionship, and in a radical group they find identity, structure or even an alternative family.[50]
  4. The focus in the literature on mental illness may have lead to an undervaluation of the role of minor mental issues. Although mental illness is overrepresented in our sample, minor problems seem to be more common, as was suggested in Merari’s work and several studies on European radicals we referred to. Silke, however, considers the whole idea of minor mental problems ‘insidious’ and ‘dangerously misleading’. In 1998, he saw a ‘second wave’, a “second major attempt to set terrorists psychologically apart from the rest of the population”; the effect of this ‘attempt’, according to Silke, could only be to immunize the theory that terrorists are abnormal.[51] We disagree. Immunization is a risk only when a focus on minor issues would preclude generating new, testable hypotheses, but the opposite seems to be the case here. Perhaps the discussion should be broadened even further, to include the question whether variation in individual characteristics or temperament affects vulnerability to different radical ideologies and behaviors – which brings us to our next point.[52]
  5. In 2005, Victoroff proposed to incorporate insights from neuroscience in the psychology of terrorism, but to our knowledge his proposal found no resonance.[53] Here, terrorism research lags behind developments in, for example, criminology, where biosocial criminologists try to identify neurobiological risk factors for antisocial behavior and violent crime in particular.[54] These risk factors could play some role in terrorist violence as well. Points of mutual interest between terrorism studies and neuroscience are not difficult to find, especially when it comes to explaining specific cases of murder and a propensity for cruelty in some individuals.[55] Another subject of mutual interest could be the role of child abuse and neglect, of which many Islamist radicals seem to have been a victim, and which are known to affect neurological development.[56] This is not to say that we expect neuroscience to identify biological risk factors that contribute directly to radicalization or terrorist behavior,[57] but we do think such factors should be considered important pieces of the puzzle.[58]
  6. The world has changed to the extent that we would hesitate to retroactively generalize our finding of relative ‘abnormality’ in radicals to terrorists in older studies, like the members of the RAF, the Red Brigades, or even the Mujahedin of the 1990s. The difference the Internet makes can hardly be overestimated. The Internet, as Sageman stressed, facilitates a bottom-up ‘recruitment’ process.[59] Nowadays isolated youngsters have a means of finding companions that troubled loners could only dream of twenty years ago. After meeting offline with their online radical friends, group dynamics take over. Furthermore, for radicalized individuals, finding a ‘bridge to the Jihad’ has become much easier as social media allows for real time communication with those already on the battlefield. The vastly extended opportunities for communication might also explain why we find many cases where radicalization was not a gradual process at all, as it was supposed to be in earlier studies. In a matter of weeks, people radicalize and attempt to travel to Syria or Iraq. On the receiving end of the recruitment process things have changed as well. An organization like IS does not resemble so much a ‘traditional’ clandestine terrorist organization that has to survive in a hostile society, as an insurgent movement fighting for control over territory. To an insurgent organization, troubled foreigners need not be a liability. They can be used as cannon fodder, for sex, and as ‘willing executioners’ – either in the conflict zone, or as Lone Actors in their own country. For people suffering from behavioral and mental problems, barriers-to-entry to organized terrorist activity have become much lower.

Our study offers little room for the tabula rasa approach of social psychologists who maintain that radicals are people just like you and me, and that environment is all that explains radicalization. For policy makers and law enforcement officers, one implication seems to be that there is no one-size-fits-all approach of preventing radicalization, or of stimulating de-radicalization and disengagement. For police, our findings underline the importance of collaboration with providers of mental health services in dealing with radicalization, and in approaching radicals. Therefore, we think it would be valuable if police forces in other countries would allow researchers to study the behavioral and mental histories of radicals as well. Perhaps a greater sensitivity in terrorism researchers to psychopathology in radicalization could speed things up somewhat. There is some urgency here.

About the Author: Anton Weenink is a Senior Researcher in the Central Unit of the National Police of the Netherlands.


The author would like to express his gratitude to the experts and reviewers, who made many useful comments on the research and an earlier draft of this article. Naturally, the final version remains the full responsibility of the author.


The views expressed are those of the author and do not necessarily express those of the National Police of the Netherlands.


[1] John Horgan (2014), The Psychology of Terrorism. Revised and updated second edition, London & New York: Routledge, p.159.

[2] Martha Crenshaw (1981), “The Causes of Terrorism”, Comparative Politics, Vol. 13, No. 4 (Jul., 1981), 1981, p. 390. The point has been repeated by several authors:

- Randy Borum, (2004), Psychology of Terrorism, Tampa: University of South Florida.

- John Horgan, (2003), “The Search for the Terrorist Personality”, in: Andrew Silke (ed.), 2003, Terrorists, Victims and Society: Psychological Perspectives on Terrorism and its Consequences, London: John Wiley & Sons, Ltd.; see also Horgan (2014).

- Roel W. Meertens, Yvonne R.A. Prins & Bertjan Doosje, (2006_, In iedereen schuilt een terrorist. Een sociaal-psychologische analyse van terroristische sekten en aanslagen [“A terrorist is hidden in all of us. A social-psychological analysis of terrorist sects and attacks”], Schiedam: Scriptum; cf. “Schokkend normaal. Interview Bertjan Doosje” [“Shockingly normal. Interview Bertjan Doosje], De Psycholoog, May 2013.

- Jerrold M. Post (2005), “When Hatred is Bred in the Bone: Psycho-cultural Foundations of Contemporary Terrorism”, Political Psychology, Volume 26, Issue 4, pp.615–636.

- Marc Sageman, (2004), Understanding Terror Networks, Philadelphia, Pennsylvania: University of Pennsylvania Press; 2008, Leaderless Jihad. Terror Networks in the Twenty-First Century, Philadelphia, Pennsylvania: University of Pennsylvania Press; 2014, “The Stagnation in Terrorism Research”, Terrorism and Political Violence 26, pp.565–580, http://dx.doi.org/10.1080/09546553.2014.895649).

- Andrew Silke, (1998), “Cheshire-cat logic: The recurring theme of terrorist abnormality in psychological research”, Psychology, Crime & Law, Volume 4, Issue 1; 2008, “Holy Warriors. Exploring the Psychological Processes of Jihadi Radicalization”, European Journal of Criminology, Vol. 5 (1), pp.99-123.

[3] Sageman (2014, p.566).

[4] Our research proposal was accepted without much ado. In the Netherlands, several researchers have had access to closed terrorism investigation files:

- C.J. De Poot, & A. Sonnenschein et.al., 2011, Jihadi terrorism in the Netherlands. A description based on closed criminal investigations, The Hague: WODC & Meppel: Boom Juridische Uitgevers; www.wodc.nl;

- Jasper L. De Bie, Christianne J. de Poot, & Joanne van der Leun, 2014, “Jihadi networks and the involvement of vulnerable immigrants: reconsidering the ideological and pragmatic value”, Global Crime, http://dx.doi.org/10.1080/17440572.2014.930349;

- Bart Schuurman, Quirine Eijkman & Edwin Bakker, 2014, “The Hofstadgroup Revisited: Questioning its Status as a “Quintessential” Homegrown Jihadist Network”, Terrorism and Political Violence, http://www.tandfonline.com/loi/ftpv20.

[5] J. Horgan (2003); M. Sageman (2004); Alex P. Schmid (Ed.), 2011, The Routledge Handbook of Terrorism Research, pp.214-217; J. Post (2005), pp.616-617).

[6] Sageman (2004, p.83).

[7] Sageman (2004).

[8]Edwin Bakker, (2006), Jihadi terrorists in Europe, their characteristics and the circumstances in which they joined the jihad: An exploratory study. The Hague: Clingendael Institute.

[9] Silke (2008).

[10] Cf. Sageman (2004).

[11] Horgan (2014).

[12] Sageman (2008); cf. John M. Venhaus, 2010, Why Youth Join al-Qaeda, Washington D.C.: United States Institute of Peace.

[13] Sageman (2008).

[14] Andrew Silke, 2001, “Chasing ghosts: Offender profiling and terrorism,” Sex and violence: The psychology of crime and risk assessment. London: Routledge.

[15] Horgan (2014, and earlier studies); Tore Bjørgo and John Horgan (Eds.), 2009, Leaving Terrorism Behind. Individual and collective disengagement, Oxon and New York: Routledge; Horgan (2014).

[16] Clark McCauley and Sophia Moskalenko, (2008), “Mechanisms of Political Radicalization: Pathways Toward Terrorism”, Terrorism and Political Violence 20, pp.415–433, http://dx.doi.org/10.1080/09546550802073367.

[17] Sageman (2004), p.81); Silke (2008), p.104).

[18] Ramón Spaaij, (2012), Understanding Lone Wolf Terrorism. Global Patterns, Motivations and Prevention, Dordrecht, Heidelberg, London, New York: Springer, p.50. Cf. Paul Gill, John Horgan, and Paige Deckert, (2014), “Bombing Alone: Tracing the Motivations and Antecedent Behaviors of Lone-Actor Terrorists”, Journal of Forensic Sciences, Vol. 59, No. 2, 425-435 (p.432). McCauley & Moskalenko (2008, p.419); Stefan Bogaerts, Pinar Okur, Michelle Willems & Leontien van der Knaap, m.m.v. Marinus Spreen & Ivo Aertsen, (2012), Solistische dreigers: Ontwikkeling van een instrument voor risicotaxatie van solistische dreigers, [Solistic threateners. Developing an instrument for risk evaluation in solistic threateners], Tilburg: Tilburg University & Den Haag: Ministerie van Veiligheid & Justitie / WODC, p.43.; An extensive review on Lone Actor Terrorism can be found in Raffaello Pantucci, Clare Ellis, and Lorien Chaplais, (2015), “Countering Lone Actor Terrorism – Literature Review”, paper presented at the Definitional Workshop on Lone Actor Terrorism, Leiden University Campus The Hague / International Centre for Counter-Terrorism, The Hague, 14-15 January 2015.

[19] Bakker (2006), p.40).

[20] Venhaus (2010), p.4).

[21]”Two main personality patterns were discerned: dependent-avoidant personality (60% of the would-be suicides) and impulsive-unstable personality (26,7%). Merari, Ariel, 2010, Driven to Death. Psychological and Social Aspects of Suicide Terrorism, Oxford: Oxford University Press, p.268.

[22] De Poot and Sonnenschein et.al. (2011).

[23] Saskia Lützinger (2012), The Other Side of the Story. A qualitative study of the biographies of extremists and terrorists, Wiesbaden: Bundeskriminalamt, p.60.


[24] Dounia Bouzar, Christophe Caupenne & Suleyman Valsan (2014), La Métamorphose opérée chez le jeune par les nouveaux discours terroristes. France: le Centre de Prévention contre les Dérives Sectaires liées à l’Islam (CPDSI). http://www.bouzar-expertises.fr/metamorphose

[25] Bouzar’s cases are an exception, they mainly came from middle class families. However, perhaps this is due to a selection issue because her sample is based on cases that were brought to the attention of Bouzar’s institute by subject’s relatives themselves.

[26] Obviously, not all unpleasant conduct can be considered ‘problematic behavior’: “Occasionally, a child will have a temper tantrum, or an outburst of aggressive or destructive behaviour, but this is often nothing to worry about. Behavioural problems can happen in children of all ages. Some children have serious behavioural problems. The signs to look out for are if the child continues to behave badly for several months or longer, is repeatedly being disobedient, cheeky and aggressive; if their behaviour is out of the ordinary, and seriously breaks the rules accepted in their home and school. This is much more than ordinary childish mischief or adolescent rebelliousness. This sort of behaviour can affect a child’s development, and can interfere with their ability to lead a normal life. When behaviour is this much of a problem, it is called a ‘conduct disorder’” (“Behavioural problems and conduct disorder: information for parents, carers and anyone who works with young people”,


[27] Cf. Matthew B. Robinson and Kevin M. Beaver, 2009, Why Crime? An Interdisciplinary Approach to Explaining Criminal Behavior, Durham NC: Carolina Academic Press, pp.14 -15.

[28] [AIVD, 2014, Transformatie van het jihadisme in Nederland. Zwermdynamiek en nieuwe slagkracht, [Transformation of jihadism in the Netherlands. Swarm dynamics and new clout] Den Haag: Algemene Inlichtingen- en Veiligheidsdienst.

[29] APA, 2013, Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Arlington VA: American Psychiatric Association.

[30] When HKS turned out to be incomplete (no suspicions in HKS, but convictions in JD-Online), we used the number of convictions.

[31] When a psychiatrist considers a patient to be a serious danger to the self or others, the psychiatrist can decide, after consulting colleagues, to share details with law enforcement.

[32 ]There are 375,000 citizens from Moroccan and 396,000 from Turkish descent (Central Bureau of Statistics (CBS statline), accessed 10 December 2014).

[33] Cf. Moffitt’s distinction between life-course persistent and adolescent limited offenders (Terrie E. Moffitt, 1993. “Adolescence-limited and life-course persistent antisocial behavior: a developmental taxonomy.” Psychol. Rev. 100, pp.674–701).

[34] “23 percent of men and 5 percent of women born in 1984 and the twenty-second year were at least once recorded in HKS. Young people of non-Dutch origin are overrepresented in police records. Of all non-Dutch origin groups men of Moroccan origin are over-represented the most: 54 percent of them between twelve and 22 years come at least once in contact with the police, a third five times or more.” (Arjan Blokland, Kim Grimbergen, Wim Bernasco & Paul Nieuwbeerta (2010), “Criminaliteit en etniciteit. Criminele carrières van autochtone en allochtone jongeren uit het geboortecohort 1984” [Crime and ethnicity. Criminal careers of native and immigrant youngsters from birth cohort ,1984], Tijdschrift voor Criminologie 52/2, pp.122-152.

[35] 66 Subjects or 47% are men of Moroccan descent; 34 of this group–or 52%–has at least one registration in HKS.

[36] Blokland et.al. (2010).

[37]Another 8% was convicted of minor offenses or misdemeanors only.

[38]We use the labels as we found them, and abstain from discussing the controversies that seem to haunt psychiatric classification; one of the issues is whether schizophrenia should be considered a ‘normal’ psychotic disorder or not.

[39] This includes travelers who have ‘left’ and ‘returned’, or are ‘deceased’.

[40] APA (2013, p.102).

[41] APA (2013, pp.116-117). Psychotic disorder may be ‘medication/substance induced’. The prevalence rate is for “Psychotic Disorder Due to Another Medical Condition”. The prevalence rates here “are difficult to estimate given the wide variety of underlying medical etiologies.”

[42] Percentages close to 50% are not what one would expect in ‘normal’ populations, but in prison populations: “There is no doubt that the prevalence of personality disorder, which is around 4 per cent of the general UK population (Coid et al., 2006) is markedly raised in criminal populations. In a systematic review of 62 surveys of prisoners in 12 countries, Fazel and Danesh (2002) reported that of 18,530 men, 60 per cent had a personality disorder and 47 per cent had an antisocial personality disorder (ASPD). The equivalent figures for women were 42 and 21 per cent.” (Richard Howard and Conor Duggan, “Mentally Disordered Offenders. Personality Disorders”, in Graham J. Towl and David A. Crighton (2010), Forensic Psychology, Chichester: BPS Blackwell / John Wiley and Sons, pp.320-321.)

[43] Cf. J. Horgan (2014), p.133.

[44] So far, there are six right wing extremists on the ‘open’ list – one in Category B.1, four in Category B.2, zero in Category B.3, and one in Category C. Domestic violence seems to be one commonality.

[45] Sageman (2014), pp.569-570) is a recent example.

[46] Jeff Victoroff, (2005), “The Mind of the Terrorist. A Review and Critique of Psychological Approaches.” Journal of Conflict Resolution, 2005/49, p.35.

[47] Silke (1998), p.65-66); cf. Sageman’s (2004) discussion of several factors in Freudian explanations, neither of which is a sufficient explanation of radicalization, seems prone to the same logic. We would be interested to learn how many did not score on any of these variables.

[48] Cf. Victoroff, who recommends seeking “a middle ground between the reductionist position that proposes a single psychology of terrorism and the nihilist position that denies any explicit psychology of terrorism” (2005, p.31).

[49] Cf. Boutwell et al. who distinguish a third category besides the adolescence limited and life course persistent offenders: abstainers, people who never commit crime (Brian Boutwell, J.C. Barnes, Raelynn Deaton, Kevin M. Beaver, 2013, “On the evolutionary origins of life-course persistent offending: A theoretical scaffold for Moffitt’s developmental taxonomy”, Journal of Theoretical Biology, 322, pp.72–80.

[50 ]Lützinger (2012), p.56); cf. McCauley & Moskalenko (2008).

[51] Silke (1998), p.66).

[52 ]For example, we suspect that the prevalence and nature of psychosocial problems may differ between radicals with different ideological agendas, or between single-issue and other terrorists. The difference in personality traits of ideologues and thrill seekers deserves attention too. Specific ‘minor’ issues like weak self-control in thrill seekers may be a risk factor for radicalization. We agree with Freilich et.al who think the characteristics and motivations of terrorists have become ‘under-explored’ in terrorism research (Joshua D. Freilich, Steven M. Chermak & Jeff Gruenewald, 2014, “The future of terrorism research: a review essay”, International Journal of Comparative and Applied Criminal Justice. http://www.tandfonline.com/loi/rcac20 23 May 2014.)

[53] Victoroff (2005, pp.34-35).

[54] Cf. Moffitt (1993, p.680): “the link between neuropsychological impairment and antisocial outcomes is one of the most robust effects in the study of antisocial behavior”. Cf.:

- Kevin M. Beaver, and Anthony Walsh (eds.), 2011, The Ashgate Research Companion to Biosocial Theories of Crime, Franham (UK) and Burlington (USA): Ashgate;

- Noémie Bouhana, 2013, “The reasoning criminal vs. Homer Simpson: conceptual challenges for crime science”, Frontiers in Human Neuroscience, 23 October 2013).

- C.H. De Kogel, 2008, De hersenen in beeld. Neurobiologisch onderzoek en vraagstukken op het gebied van verklaring, reductie en preventie van criminaliteit [Brain mechanisms and antisocial behavior. Neurobiological research and issues concerning the explanation, reduction and prevention of criminality], Onderzoek en beleid 270, Den Haag: Ministerie van Justitie / WODC en Boom Juridische Uitgevers.

- Adrian Raine (2013), The Anatomy of Violence. The Biological Roots of Crime, New York: Pantheon Books.

- Dick Swaab, (2010), Wij zijn ons brein. Van baarmoeder tot Alzheimer [We are our brains. From the womb to Alzheimer’s], Amsterdam & Antwerpen: Uitgeverij Contact.

[55] Cf. Victor Nell (2006), “Cruelty’s rewards: The gratifications of perpetrators and spectators”, Behavioral and Brain Sciences 29, pp. 211–257.

[56] “Studies show that abuse and neglect have consequences for the development and regulation of the hormonal stress system (an overactive stress system can lead to depression, while an underactive stress system can lead to antisocial behaviour), for social information processing (for example, being too quick to interpret situations or other people’s behaviour as threatening or hostile)” (De Kogel, 2008, p.147).

[57] Van der Gronde et.al show in an extensive research study that, although biological risk factors for human aggression can be identified, understanding individual cases always demands accounting for environmental factors and life events as well. “In the future, new information from neuroscience, when integrated into the information already available from sociological and psychological assessments, could contribute to the development of better risk assessment tools, treatment and cures for offenders, reducing recidivism as well” (Toon Van der Gronde, Maaike Kempes, Carla van El, Thomas Rinne & Toine Pieters, 2014, “Neurobiological Correlates in Forensic Assessment: A Systematic Review”, PLOS One 9(10): e110672. Doi: 10.1371/journal.pone.0110672).

[58] “Different biological, psychological, and social risk factors can interact in shaping either violence or self-sacrificing heroism. Violence and terrorism are not just low physiological arousal, yet this is certainly one of the active ingredients that, when combined with other influences, can move us toward a more complete understanding of killers like Kaczynski” (Raine, 2013, p.133).

[59] Sageman (2008).

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.


Perspectives on Terrorism is  a journal of the Terrorism Research Initiative and the Center for Terrorism and Security Studies

ISSN  2334-3745 (Online)

Disclaimer, Terms and Conditions