Can science predict future terrorists?

Those who know me know I am not a fan of Tom Cruise for all kinds of reasons not relevant to this blog.  But one movie he appeared in is directly germane to today’s topic.  In Minority Report, a film based on a short story by US science fiction writer Philip K. Dick, a man of whom I am a big fan, a fascinating world is explored.  In short, police in 2054 are able to make preventative arrests before criminal acts take place thanks to the the foreknowledge provided by psychics called “pre-cogs” floating in pools.  In the movie, Tom Cruise’s character is accused by a pre-cog of intending to murder someone in the next 36 hours and, since Mr. Cruise is always the good guy, he flees this unfair “verdict” imposed on him.

I thought of this while reading an interesting article in this weekend’s Wall Street Journal.  Scientists in Sweden have found that “time discounting” can point to future criminal behaviour.  I know this sounds science-fictiony but essentially time discounting refers to how long a person is willing to wait for a reward.  Those who cannot put off having a desired object right away tend to value that object less and are apparently more susceptible to committing criminal acts in the future (the scientists did test 13,000 13-year olds on which to base their theory).  This work complements earlier studies tying a lack of self-control to future anti-social behaviour.  Intuitively this makes sense: if you are impulsive you are probably willing to go to extreme measures to get what you want regardless of the effect of your actions on others.

Is there anything here that can help us understand terrorism?  I am not so sure.  There are some significant obstacles to using “science” to predict violent extremist behaviour.  Among those obstacles are:

  • the fact that terrorism is a very low incidence phenomenon.  Despite headlines and general panic, acts of terrorism are very infrequent events and hence it is very hard to get large datasets upon which to test hypotheses.
  • unlike most social science experimentation, you cannot divide objects of study into terrorist and control groups – for obvious reasons
  • it is far from certain that there are any meaningful links between what we know about criminality in general and terrorism since many terrorists do not exhibit the classic crimogenic risk factors that the average criminal seems to do
  • every “causal factor” put forward to account for – or predict – terrorist behaviour has failed miserably upon closer scrutiny.  Terrorist scholars are no closer to explaining the decision to embrace violent extremism today than in the past and no where near being able to solving the vast numbers of false positives and false negatives that accompany any given paradigm or theory.

At the end of the WSJ article, Robert Sapolsky helpfully notes us that the findings of these scientists are statistical averages with many exceptions.  In other words, “it depends”.  This is a great reminder that when it comes to human behaviour we will never come up with models that can exhaustively predict what people will decide to do.

The purpose of this blog is not to discourage scientists and others from developing hypotheses and testing them when it comes to terrorism.  It is simply a plea to be humble in their work, see their tools and models as – one hopes – generally illustrative and contributing to a greater understanding, and not claim to be comprehensive and capable of prediction.  After all, short of a world of pre-cogs we humans will just have to muddle through.

By Phil Gurski

Phil Gurski is the President and CEO of Borealis Threat and Risk Consulting Ltd. Phil is a 32-year veteran of CSE and CSIS and the author of six books on terrorism.

Leave a Reply