A news story in today’s Nature notes that the US military are pumping more money into social science research which is considered to be an important ‘game changing’ component of 21st century warfare.
The unconventional wars now being fought by the US military have also bolstered interest in the social sciences. With the military trying to stave off a growing insurgency in Afghanistan, the Pentagon now believes that understanding cultural dynamics is at least as important as weapons. Consequently, Lemnios is ramping up funding in social-science projects, including a model developed by Los Alamos National Laboratory in New Mexico to simulate the opium trade in Afghanistan and analyse the effectiveness of efforts to combat it. The office is also supporting a project at the University of Chicago, Illinois, to model and predict potential conflicts.
Research to be used ‘on the ground’, like that described above, is likely to involve at least two important components. The first is the deployment of social scientists in conflict zones to use their skills to better solve problems that require the co-operation of local populations, along the lines of the Human Terrain System.
The other is the use of mathematical modelling to look at the structure of social networks to decide who are the key players or to infer (technically, to impute) the likely structure of the network that is not directly known about.
For example, within a large town there may be a small network of insurgents / terrorists / freedom fighters (take your pick) who your army wants to destroy. The traditional approach involves getting informers to reveal this network while you have to make a subjective judgement about how true this information is and how important each individual is, often before blundering in with troops, much to everyone’s alarm.
With social network analysis, you can create maps of the known network and mathematically analyse it for how accurate it seems and get an estimation for how important each person is for the overall structure. You can even add to the model with observational data (e.g. with traffic analysis – looking at communication patterns without knowing content) and make computational best guesses as to relationships and people you know must exist, but no nothing directly about.
The idea being that you can destroy the network by taking out the key people with the minimum of fuss – where ‘taking out’ could mean killing, arresting or bribing, and where ‘fuss’ could mean violence, risk or public knowledge.
Essentially, you are analysing the behaviour of networks and much the same principles (and indeed, some of the same laws) apply to other sorts of networks like transport and communication.
While this sort of research would help forces on the ground, other research, as funded by the Pentagon’s Minerva programme, seems aimed more at foreign policy and PsyOps operations, where the attitude and behaviour of very large populations need to be understood.
As has been the case with previous military enthusiasm for such ventures, this new level of funding is likely to cause additional concern that social science will become ‘weaponised’ and distort a field that has traditionally had a commitment to a ‘do no harm’ policy.
The Nature article also has a curious bit where it discusses the Pentagon’s interest in “how organisms sense and respond to stimuli ‚Äî such as chemicals, ions and metals, or electrical, magnetic, optical and mechanical impulses” to be able to develop “living sentinels”.
UPDATE: Thanks to Ian for sending a link to this Slate article series on ‘how the U.S. military used social networking to capture the Iraqi dictator’.
Link to Nature on Pentagon / social science tryst.