Known Unknowns

The likelihood of another catastrophic attack against the United States naturally is a big topic in defense circles these days. By extension, the issue has enormous defense budget implications. It is also provides a subtext for the question of what the Obama administration's strategy should be to combat terrorist safe havens worldwide, like in Afghanistan or Pakistan.

Earlier this year, Defense Science Board experts issued a report outlining steps the Defense Department should take to reduce the chance that adversaries could one day deal a surprising blow to America's security.

Today, the Federation of American Scientists' Secrecy News blog brings news of another DOD advisory panel, the JASONs, taking up the topic.

The October 2009 report, titled "Rare Events," examines how the chance of an attack with weapons of mass destruction against the United States could be determined -- if it can be predicted at all.

To find the answer, defense officials should stay away from strictly quantitative modeling schemes and instead turn to the scientific methods of the social sciences, the report states. After all, the issue has mostly to do with predicting the behaviour of humans, the authors argue.

The combined urgency of the rare event threat, the difficulty of evaluating rare event models, and the complexity of social sciences problems has led some to advocate the suspension of normal standards of scientific hypothesis testing, in order to press models quickly into operational service. While appreciating the urgency, JASON believes such advice to be misguided. The threat of “rare events” will be with us for a long time. Like finding a cure for cancer or predicting earthquakes, this is a difficult research area that will most likely make progress in many small steps.

And speaking of human behavior, JASON panelists urge defense officials to add "motivation" to the existing WMD threat evaluation scheme of "intent/capability/opportunity." Further studying the issue of terrorist attack motivation, and perhaps finding a way to eliminate it, could help anticipate "rare events," the authors argue.