The American Physical Society has a newspaper call the APS News. Each issue, there is a one page opinion piece at the end written by different authors titled “The Back Page”. It covers various items, but this last one, written by Neil Johnson titled “New Terrorism Reveals New Physics” contains a lot to think about.The article expands on a previous Back Page article he’d written where he hypothesized that because people are people, it’s possible to find common behaviors, or signatures, of human groups. More specifically, he wrote about how insurgency or terrorist groups tend to act like specific types of complex systems modelled in physics (most specifically polymers).
While all this sounds fine and good, there is both important and sobering information in these two pieces. I’ll touch on some of the issues here.
Now, the news is currently awash with articles talking about global terrorism, suicide bombers, lone gunmen and the like. Right now, there is a sea of articles talking about ISIS (or ISIL or Daesh) and the wreckage they sow both in the Middle East and abroad. I’ll start here.
Social Media seems to be the one of the ways (maybe the primary way, currently) that subgroups that share a common belief aggregate and organize. Extra-national, there is no “there” there. All one needs is a way to get on line to be part of a group. Professor Johnson and his group looked at tons of data from just one foreign social media platform, VKontacte, and in less than a 6 month span of time, identified and analyzed almost 200 pro-ISIS groups with near instantaneous resolution. Turns out that these groups all show some common signatures one of which is membership is described by something called a power law.
While the details of this aren’t really important to this discussion, some aspects of it are. It has been found that small groups don’t really do much other than talk, and big groups do big things (often with very lethal results). The first take away hear is that if one can disrupt the growth of the group, one can reduce their potential lethality. So many of these social media platforms try to do just that. If they can identify the growth of a potentially dangerous group, one can just turn off everyone’s account and make the individuals scatter, much like turning the light on in the proverbial kitchen to make all the roaches run for cover. Some good news is that Facebook is very fast at disrupting these groups, but VKontacte isn’t as fast (that’s why Prof Johnsons group used their data.)
One thought that occurred to me is that most of the media talk about how to attak these groups is just wrong. The implications of this work is that the likelihood of any one of these groups actually committing violence is probabilistic in nature and predicting which group or which individual will act is impossible. As the groups increase in size and mature, it’s the inner couplings of the group that increase the intensity of the fervor, and eventually, one of the members decides to act. Think about the San Bernardino couple who went on a shooting rampage. They had practically no links to groups, no history of violent action, and yet they went on a rampage that killed 14 and injured 22. Yet we keep hearing about “cutting the head off the snake” and the like. Killing Ben Ladin didn’t end Al Qaeda, nor has targeted drone killings of leaders stopped terrorist organization. If we get the number one, the number two just steps in.
To a physicist, this is somewhat obvious. These are many-body problems, yet targeted killings is a single-particle approach to deal with a collective behavior. This has very important implications for anti-terror resource allocation. In Prof Johnsons study of VKontacte, there were only two hundred groups, yet there were over 100,000 individuals. Monitoring and disrupting 200 groups takes a lot less effort than monitoring and disrupting 100,000 individuals.
While Neil’s article in the APS News is a bit geek-speak and inside baseball with regard to physics vocabulary, it’s really, really worth a read. The implications are huge.
Facebook and social media is currently grappling with what to do about “Fake News”. Similar dynamics are at play here. How can we moderate social media content to prevent stuff that’s just wrong from influencing society? Is it possible to monitor social media to prevent domestic terrorists from aggregating to a size where the probability of one converting from a talker to an agent of carnage becomes significant? Is it even constitutional to try to do this? If not, how to we keep the rot of idealism at bay and not become a cancer that shreds our social fabric? These are big questions with no easy answers.