The 1997 Exceptional Case Study Project, which investigated every assassination or attempt since 1949 failed the make a connection between those who threaten and those who attack.
This was further corroborated by David James and others in 2007 when the Fixated Research Group investigated serious attacks on European politicians. Whilst they evidenced that in nearly 50% of the cases warning behaviours were evident, they could find no evidence of directly communicated threats.
If further corroboration was required, it was provided by German study in 2011 which looked at all non-terrorist German attackers of public figures between 1968 and 2004. Their research also found that a directly communicated threat before an attack was extremely rare.
In 1991 Dietz and others published a significant report in which they found that those subjects who wrote threatening letters to members of the US Congress where less likely to make a physical approach. In fact, they went further by stating that their findings suggested that the presence of a threat in any communication actually lowered the threat. A further study by Scalora found that if threats were apparent in communications to US Congress, an approach was less likely. Despite these findings, neither study rules out an attack.
In their study of Public Figure attacks in the US between 1995-2015 Meloy and Amman state that when looking at warning behaviours, direct threats was one area that they could code. This produced a frequency of just 5%. This, they suggest, supports the view that directed threats are infrequent in respect of violent approaches.
Calhoun and Weston argue that whilst ‘howlers’ intention is to frighten, to disturb and cause psychological harm, they choose to do so in a manner that avoids any physical risk to themselves. In a study of nearly 4000 inappropriate communications towards Judicial figures they found that less than 9% followed up their threats with actions. They further suggested that their research demonstrated that those who communicated from a distance, intended to maintain that distance. They further identified that if the person of concern became known due to suspicious activity, they were increasingly likely to conduct a violent approach, as opposed to those who came to notice via communications.
Furthermore, persons of concern who withheld their identity, were acting with others, or affiliated to a group were more likely to make a violent approach.
Another indicator was those who were driven by an ideological belief, not related to any specific legal case, but make reference to personal issues were more likely to conduct a violent approach. However, what their research also says it that 92% of those who send threatening communications won’t follow this up with action.
Those who work within the US Capital Police Threat Assessment Section (TAS) identified that whilst in general those who make a threat do not pose one, the growing field of content analysis indicates that there are a number of concerning behaviours which may signal an increased chance of an approach and potential attack. They found that those who made multiple communications using different methods, such as email, letter, phone, and where evidence of target dispersion, i.e making contact with a number of different targets, had an increased in chance of a problematic approach.
If the threats were part of a significant change in behaviour, and escalation pattern then again there was an increased chance of approach. They found that those who emailed were more likely than letter writers to engage in obscene, shorter and threatening communications. They also found that those who followed up their threatening communications with an approach, were more likely than not to be suffering from a mental disorder.
Sharon S. Smith, formerly of the FBI Behavioural Science Unit, whilst recognising that even though the chances of threatening communications turning into violent approaches is low, these communications have a significant impact on those who receive them, especially if they are in repeated receipt.
Evidence from Lord Bew’s recent report on ‘Intimidation in Public Life’ recognised that intimidatory behaviour impacts Parliament and our political system and ‘threatens the integrity of our political process’ To further evidence, we saw at the last general Election several MPs standing down stating their reason was the level of abuse they received. It is for this reason that those who do make threats should be investigated and where appropriate, prosecuted.
Smith’s research shows that an important minority of threateners do in fact become violent approachers. Developing on the work of the TAS, Smith identified that the way people use language can be a predictable factor. In support of the TAS theory, she demonstrated that those who make repeated contacts increase the possibility of a problematic approach.
Dr. Smith’s research identified 8 linguistic elements which increased the threat and formed a process now known as Threat Triage.
a) Conceptually concept language
b) Expressions indicating paranoia
c) Revealing religious prejudice
d) Using a polite tone
e) Mentions of love, marriage or romance (tied to the intimacy factor)
f) Indicating who the target is
g) Detailing the threateners return address
h) Making specific reference to weapons
In supporting the above indicators, Dietz et al suggested that taking the role of a special constituent, casting the member of Congress in a benefactor role, such as a rescuer or benefactor, expressing a desire to be rescued and requesting a face to face meeting increased the threat. From these indicators we are left with something of a paradox. Those who threaten may not pose a threat, but those who appear friendlier and less threatening, may in fact be the ones to worry about.
To conclude, the research overwhelmingly demonstrates that those who make threats are less likely to pose a threat.
What is clear is that those who do not make threats cannot nor should they be ignored. Other than the fact that they are likely to commit offences, they may pose a threat. The evidence whilst overwhelming, cannot be said to be conclusive and the it could be used against itself by way of an individual making a threat in order to persuade the security manager that they do not pose one. In fact, Dietz and Martell went further and concluded that “Those who rely on the presence or absence of threats in making judgements about what to do are making a serious mistake”
To combat this ambiguity the growing field of context analysis may prove effective but needs more research and whilst it is usually the domain of psychologists, should be made available to and taught to law enforcement and those with a protective responsibility. This provides a challenge for the security manager, who will have no problem encouraging Members of Parliament and other public figures to report threatening correspondence, but perhaps be less successful in persuading them to report other less obviously concerning communications. All material which evidences a change of behaviour or any of Dr Smith’s 8 linguistic elements, should be reported, and the threat assessors should then triage the communications, implementing lessons learnt from content analysis research.