By Noah Shachtman
Danger Room, November 9, 2007
The Pentagon is paying Lockheed Martin to try to predict insurgencies and civil unrest like the weather. It’s part of a larger military effort to blend forecasting software with social science that has some counterinsurgency experts cringing.
Lockheed recently won a $1.3 million, 15-month contract from the Defense Department to help develop the “Integrated Crises Early Warning System, or ICEWS. The program will “let military commanders anticipate and respond to worldwide political crises and predict events of interest and stability of countries of interest with greater than 80 percent accuracy,” the company claims. “Rebellions, insurgencies, ethnic/religious violence, civil war, and major economic crises” will all be predictable. So will “combinations of strategies, tactics, and resources to mitigate [against those] instabilities.”
DARPA, the Pentagon’s bleeding-edge research arm, laid out the case for ICEWS this summer at itsconference, held outside of Disneyworld. “Commanders will always need to have an accurate picture of enemy positions, as well as friendly units and allies,” David Honey, who heads the agency’s Strategic Technology Office, told confab-goers in Anaheim, California.
“But increasingly it’s social, cultural, political and economic information, foreign language capabilities and other clues – that are proving essential.”
Figuring out how to find those clues won’t be easy, his colleague, Sean O’Brien, warned.
He has a three-part plan for how ICEWS might get it done, however. It tracks, roughly, to how meteorologists piece together long-range weather forecasts.
Step one: dump everything we know about a country like Iraq, and “create [software] agents that mirror the actual communities.”
Not only does that mean identifying “government leaders[‘]
propensity to defuse or exacerbate potentially volatile situations,”
O’Brien explained in a call for proposals.
It also requires a determination of “how a country’s macro-structural conditions (social, demographic, economic) affect the way in which the country’s citizens interact with its government.”
What’s more, according to an
article in the Military & Aerospace Electronics trade journal, the ICEWS system should be designed to “capture and process vast quantities of data from digitized news media, Websites, blogs, and other sources of information that reflect the dynamic and rapidly changing character and intensity of interactions between people and governments.”
Step two in the ICEWS plan: make these agents even more realistic, by “leverag[ing] the hundreds of social, cultural, and behavioral theories” about why people act the way they do. Step three: let commanders run mock battle plans against these modeled Iraqis, to see how they might react.
Experts on counterinsurgency are, to say the least, skeptical.
“Wait a minute, you can’t tell me who’s going to a win a football game. And now you’re going to replicate free will?” Lieutenant Colonel
John Nagl, who helped write the Army’s manual on defusing insurgencies, tells DANGER ROOM.
“They are smoking something they shouldn’t be,” retired Lieutenant General Paul Van Riper recently quipped to Science Magazine.
The military has been trying, for decades, to use social science data to forecast what might happen next on the battlefield — or around the world. In 1976, for example, Pentagon-funded researchers put together a “integrated crisis warning system… comprised of (1) quantitative military, political, and economic crisis indicators; (2) quantitative indicators of U.S. military, political, and economic interests abroad;
(3) a unified multi-method forecasting capability; and (4) a computer base.”
More recently, DARPA has funded efforts to “anticipate the societal/regional indicators that precipitate instability.”
For these projects, the agency turned to Carnegie Mellon University professor Kathleen Carley and Massachusetts research firm Aptima, Inc. VISualization of Threats and Attacks in Urban Environments, or “VISTA,” was their company’s attempt to “utiliz[e] cultural models and other social network analysis techniques to assess and forecast nation state instability and conflict.” The “Anticipatory Culture-Based Modeling Environment,” or ACUMEN, toolkit created a “simulation engine” based on “theories from psychology, social psychology, sociology, organization science, political science, and economics.”
ACUMEN modeled political, military, social, religious, and insurgent groups as agents, along with their relationships regarding hostility, support, membership, and more. ACUMEN modeled the profiles of agents and geographic regions (at the state and province levels)
within specific test states using a set of social, political, economic, health, and demographic indicators. In all there were 150 indicators for the state, 60 indicators for each province, and 30 indicators for each agent.
But that was all in the lab. The goal of ICEWS is to eventually bring the tool to war. As Military & Aerospace Electronics notes, “The third phase will involve a live, in-theater test of the system.”