What do networks tell us about white extremist terrorism?
In analysis published in The New York Times last week they presented evidence of a growing network of white extremist terrorists. The article said:
In a manifesto posted online before his attack, the gunman who killed 50 last month in a rampage at two mosques in Christchurch, New Zealand, said he drew inspiration from white extremist terrorism attacks in Norway, the United States, Italy, Sweden and the United Kingdom.
His references to those attacks placed him in an informal global network of white extremists whose violent attacks are occurring with greater frequency in the West.
An analysis by The New York Times of recent terrorism attacks found that at least a third of white extremist killers since 2011 were inspired by others who perpetrated similar attacks, professed a reverence for them or showed an interest in their tactics.
In a great infographic the New York Times then examined the links between white terrorists, in particular showing who more recent terrorists have stated they were influenced by.
One thing stuck out to me however: these very rarely actually spoke to each other. The Times only presented evidence of two terrorists who actually spoke, where “a school shooter in New Mexico corresponded with a gunman who attacked a mall in Munich.” As the Times notes, together these two shooters killed eleven people. Apart from this however all the links presented are ones of influence. The connections are created based of readings of texts and manifestos, with terrorists citing influence from those who went before them, but not citing actual meetings, discussions or organising (one potential exception not noted in the article is the Christchurch shooter, who says in his manifesto that he got a blessing for his attack from Anders Breivik. However, as far as I'm aware, we still do not have physical evidence this occurred).
So reading this piece I was left wondering: does this actually represent a network? If it does, what does this network tell us about white terrorist extremism today?
To answer this I’ve turned to social media network analysis. This is relevant, not just because my thesis is based in it, but also as the Times noted, much of the influence cited by these terrorists comes from social media. The methodological tools are therefore useful.
In a useful introduction to these topics, Web Social Science, Rob Ackland (who, disclaimer is also on my supervisory panel) defines a network to be “a set of entities called nodes (or vertices) and a set of ties (or edges) indicating connections or relations between the nodes.” In other words, a network represents particular actors (nodes), in this case white terrorists, who are connected via edges, in this case the edges being influence one terrorist has had over the next (or in one case the edge being actual conversations between the two).
Simples. We can therefore certainly argue that this is a network. But, what type of network is it?
This is where I think things get more interesting, and where this analysis from the Times can really tell us something useful. I would argue this network is what Ackland (in the same text) describes as an information network. An information network “involves a one-way transmission of information between actors, where the information is being sent by one actor is being broadcast to many actors.” Another way to consider this would be to look at a network we may be more familiar with. We could consider this network to be somewhat like Twitter, one in which a user sets up an account and is followed by others. This user tweets information but never actually engages with the users who follows them through the reply function. In an information network the users who follow this initial user would also not engage in this way. This is what we call a ‘directed network’, which defines networks where an edge “has a clear origin and destination and is not necessarily reciprocated”. One user follows another, but that edge is not reciprocated, either by follows or by interactions.
This is exactly what is happening in the network presented by the Times, although with much more severe consequences. In this network actors such as Anders Breivik and Dylan Roof are sending out 'information' -- through their awful attacks, online posts and manifestos. This 'information' then influences other people, who conduct their own attacks. This network is a very informal one. There are no reciprocal ties (except for the exception of the two terrorists who did speak), with the network edges only formed through terrorists following (in a very loose form) other terrorists. It is even difficult to figure out how to measure the influence described, and whether later attacks would have occurred or not without the initial influence.
Why is this important?
I think what is really valuable about the New York Times analysis is not the labelling of a white extremist network, but how it highlights the particular type of network that exists.
The fact that previous terrorists attacks influence the ones that come after is not new information. In fact, research shows that even the reporting of mass murders can influence, at least in part, others to take similar action. Moreover we've heard about this influence directly from the horse's mouth. The Christchurch shooter for example directly references Breivik and Roof amongst others, as influencers, in his manifesto.
What I think is interesting about this network therefore is what's missing -- that is evidence of direct links between terrorists (apart from the two already noted). This is, by all measures, a very weak network. It is not a network of actors working actively together to plan and conduct terrorist attacks. It is not a formal organisation of people working together to plan terrorist attacks. It is, as I’ve argued previously, more a collection of ‘lone wolves’ – people plotting these attacks on their own, although clearly influenced by the political, social and economic situation, as well as the actions of those who have gone before them.
Understanding the ‘network’ in this way is important as it helps us understand the particular threat and the challenges in dealing with it. Quite simply this form of network is much harder to infiltrate or break apart in a way that that is feasible with a more formalised organisation. It is in fact almost impossible to even determine the ties that exist in this network until after the fact, as many of them go un-noted until manifestos are placed online.
Potentially more importantly, this analysis once again highlights the different threat posed by these individual terrorists compared to far right, neo-nazi and fascist organisations. What is notably missing from this analysis is the role of any of these organisations. I do not know enough about all the 350 cases identified by the analysis to know whether individuals were involved in far right groups, and for the smaller instances it is almost certain there will be cases where people did have connections to far right groups. But based on the analysis presented these same far right groups and organisations are not actively plotting these sorts of terrorists attacks, particularly the larger scale ones (One exception is the murder of Heather Heyer, which occurred during the Charlottesville protests. While those protests were organised by a range of organisations it is important to note that these same organisations did not organise the car attack that occurred on the same day). These groups are certainly engaged in protests, riots, and other forms of violence, absolutely. They are certainly a threat. But this analysis does not show organisations that are plotting terrorist attacks. This is important again as it shows the differences of the threats we’re facing, and in particular the challenge of dealing with such an informal network of terrorists, one in which people are literally not even talking to each other.
This is the value that a network analysis can bring, and it is why I think what the New York Times has done here is really important. This analysis shows to us just how informal this terrorist network is, and in turn how much these actors are working on their own. This does mean they are not a threat. Of course they are. But the threat is unique and difficult to address. Understanding it in this way can help us try and tackle it.