Sherlock Holmes vs James Bond

...Is an analogy used in Fixing Intel: A Blueprint for Making Intelligence Relevant in Afghanistan. The comparison and contrast may have lost a little luster since Sherlock Holmes has been bad-assified in the new movie (thoroughly enjoyed the movie by the way. One point to ponder: in issuing the coup de grace hit on a hand-to-hand opponent, Holmes analyzes the efficacy of and then employs a liver punch, only it's wrong. So, film gaffe or safety measure so that kids don't run around exploding each others' livers?) but the paper is, in my humble opinion, a great overview of the shortfalls of our current information and intelligence architecture and a decent roadmap where it needs to go in order to better serve our national interests in Afghanistan.

In an effort to practice what they're preaching MG Flynn, Deputy Chief of Staff, Intelligence for ISAF and his co-authors have produced an eminently readable paper that "critically examines the relevance of the US intelligence community to the counterinsurgency strategy in Afghanistan." The authors avoid delving into egghead esoterica and succinctly flense the current modes and methods of procuring intel in Afghanistan and offer up sound recommendations on improving the system. The authors proclaim that

The most salient problems are attitudinal, cultural, and human. The intelligence community’s standard mode of operation is surprisingly passive about aggregating information that is not enemy-related and relaying it to decision-makers or fellow analysts further up the chain. It is a culture that is strangely oblivious of how little its analytical products, as they now exist, actually influence commanders.

The study demands that the intel community shed its current "Powerpoint" mentality and start banging away on MS Word. The authors point out that the intel community doesn't provide Commanders on the ground the breadth and depth of information needed for a Commander to matriculate enduring positive effects for Afghanistan. Instead, by focusing primarily on anti-insurgent (as opposed to counterinsurgent) intelligence, the intel community predisposes Commanders to adopt a "capture/kill" mentality against the Taliban, rather than an "isolate, diminish, and starve" approach. Thus

“A military force, culturally programmed to respond conventionally (and predictably) to insurgent attacks, is akin to the bull that repeatedly charges a matador’s cape – only to tire and even­tually be defeated by a much weaker opponent,” General McChrystal and U.S. Forces-Afghanistan Command Sergeant Major Michael T. Hall recently wrote. “This is predictable – the bull does what comes naturally. While a conventional approach is instinctive, that behavior is self-defeating.”

The changes that the intel community needs to make are to begin looking in depth across the breadth of Afghanistan and focusing on every dimension of Afghani society that might provide a lever or fulcrum to a Commander. The best analogy offered on how the system needs to change its perspective was a neo-Clausewitzian comparison of the COIN intel requirement to that of a political campaign in that

To understand the dynamics of this process, it is useful to think of the Afghanistan war as a politi­cal campaign, albeit a violent one. If an election campaign spent all of its effort attacking the oppo­sition and none figuring out which districts were undecided, which were most worthy of competing for, and what specific messages were necessary to sway them, the campaign would be destined to fail. No serious contender for the American presidency ever confined himself or herself solely to the “strategic” level of a campaign, telling the staff to worry only about the national and regional picture and to leave individual counties and elec­tion districts entirely in the hands of local party organizers, disconnected from the overall direction of the campaign. In order to succeed, a candidate’s pollsters and strategists (the equivalent of a J-2 staff) must constantly explore the local levels, including voters’ grievances, leanings, loyalties, and activities. Experienced campaign strategists under­stand that losing even one or two key districts can mean overall defeat. (Recall, for example, the defin­ing impact of two Florida counties – Miami-Dade and Palm Beach – on the national outcome of the 2000 presidential election.) To paraphrase former Speaker of the House Thomas P. “Tip” O’Neill’s famous quote, “all counterinsurgency is local.”

The proposal to build Stability Operations Information Centers, each with a healthy demographic of journeyman analysts that go out with every possible ground element in order to survey and collate information picked up by every available sensor, is sound. The folks who live and operate in a given area will know everything knowable about it--up to and including minutiae that they don't even realize will be important in gleaning a better picture of cause and effect within their respective AOs. As the paper says, "The soldier or development worker on the ground is usually the person best informed about the environment and the enemy. Moving up through levels of hierarchy is normally a journey into greater degrees of cluelessness."

The most important piece of the paper, though, is that it emphasizes, again and again, that intel is Commanders' business. Leaders need to constantly push their intel shops to collect, evaluate, and assess the relevant information and intelligence that they need to shape their campaigns. Too often, Commanders demand that their intelligence be diluted into bite-size bubble charts that may track what happened, but not how or--more importantly--why.

The Stability Ops Info Centers would in many ways replicate the intel procedures prevalent in current intel Fusion Centers. However, while Fusion Centers usually concentrate on "red" activity (enemy activity), The SOICs would concentrate on "white" activity (the population, economy, development and government). So, an example of the efficacy of the SOICs would be when

An NGO wanting to build a water well in a village may learn, as we recently did, about some of the surprising risks encountered by others who have attempted the same project. For instance, a foreign-funded well constructed in the center of a village in southern Afghanistan was destroyed – not by the Taliban – but by the village’s women. Before, the women had to walk a long distance to draw water from a river, but this was exactly what they wanted. The establishment of a village well deprived them of their only opportunity to gather socially with other women.

The authors also make a good argument for keeping the intel system bifurcated along the Fusion Center/SIOC dichotomy. Access, timeliness, and customer culture between the two intel venues differ so wildly that it is best to keep them separate.

I would push, though, for a trifurcated (yuh, I invent words as I go along) architecture. Rather than a red vs. white paradigm for collection, analysis, assessment, and dissemination, I'd like to see a "black, grey, white" paradigm. Black for lethal/bad guy targeting, white for all the DIME stuff, and grey for those actors about whom we're not real sure. Many of the persons, infrastructure, locations, and organizations on which we collect would probably be graded on a sliding scale along that spectrum, depending on which dimension was prioritized. For example, a local political figure might be pro-Coalition (white) and have a lot of influence with various tribal shuras (white), but he might have sticky fingers when it comes to aid money (grey), spend a little too much time with certain criminal elements (grey) and meet regularly with his brother's wife's cousin's husband, who just happens to be a Taliban bigwig (black). In order to get the holistic view that the authors want, the firewalls betwixt the different targeting systems would have to have access points so that analysts could braid the information coming from the three areas into a coherent picture. I don't think the introduction of grey intel necessitates a whole new third type of center, I just think that there's got to be an overarching means of getting the full picture of any particular person, place, or thing. Get this right, and Sherlock Holmes becomes the bad ass he needs to be without requiring James Bonds "double-0" status.