I spent the first two days of this week conducting scenario based training for soldiers highlighting decision making challenges in a COIN environment. The soldiers we were scheduled to train were from an S2 section but we also took advantage of an MP platoon that had an unexpected gap in their training schedule and ran them through as well since we had extra time. Before I get into the actual ‘meat’ of the post, just a couple of observations:
- I’ve been in the military for 18 years now and I still get blown away by how motivated our young soldiers and junior NCOs are. I’ve taught or given presentations to a lot of different types of people and by far the most receptive (and respectful) audiences are soldiers. If you ever get a chance to instruct or talk in front of soldier jump at the opportunity, you’ll never get a better audience (unless you pay them).
- Riffing off the point above, soldiers seem more engaged in training than when I first came in. They’re asking tons of questions and you can see most of them really thinking through what’s being taught. It may be because we’re nine years into our wars and everyone is still facing deployments where they’ll have to do this stuff or the fact that so many people have deployed that even the newest soldier is going to have a bunch of people in her unit who can say ‘I was there and this is what happened’.
- Soldiers love to do soldier stuff. Sometimes, particularly with support soldiers, the system forgets that they chose to be in the army rather than work in an office for a reason. Give all but the most jaded chairborne soldier good instruction, equipment and a challenge and they’ll eat up whatever training you want to give them and beg for more.
With regard to this subject, however, the most important part of the exercise was when I got to talk to the soldiers about their training and deployments. Most of the soldiers were junior enlisted, lower ranking NCOs and junior officers. Most hadn’t deployed. Still, there were some common themes when talking to them.
- Everyone agreed that intel schools were teaching COIN. The problem is that while they’re taught that they need to do analysis like:
- In order to evaluate the people, the following six sociocultural factors should be analyzed:
- Social structure.
- Power and authority.
This analysis may also identify information related to areas, structures, organizations, and events.
- In order to evaluate the people, the following six sociocultural factors should be analyzed:
They don’t teach soldiers how to do analysis like that. I think asking an 18 year old kid (heck, even a 41 year old E-8) to do something like this who has just about zero orientation to the culture and country he’s supposed to analyze is going to be a pretty tough row to hoe. And it’s not like we’ve got a huge cadre of mentors out there who can demonstrate expertise in this sort of thing.
I suspect if this is going to be an expectation for analysts in the future we’re going to need to seriously rethink our selection and training process. We send people to language school for one year to gain an understanding of Dari or Pashtu. We give analysts what? A week or two (if we’re lucky) in instruction of Afghan culture, insurgency training, social network analysis, etc and expect them to be proficient? It just doesn’t make sense. But, when I started this series of posts I promised I’d recommend a solution that didn’t require massive changes to our system so we’ll have to leave this for later…
- Our intelligence systems and products are still sub-standard. The soldiers didn’t even realize this because, I suspect, the system we’ve got is just presented as the way it is. So, what’s wrong with it? Allow me to provide a hypothetical:
- Assume in village X, there’s a village elder named Abdul Rhaman. When a new unit comes into the area they probably will get some sort of briefing on him but it simply can’t cover everything. If you want to understand the guy you’ll probably want to search databases and get a more complete picture. You can search for him and you’ll get every report that mentions ‘Abdul Rhaman’. Let’s assume the search terms allow you to identify all the relevant reports (including misspellings, reports without the correct name, etc.) and weed out all the irrelevant reports. You’ll have a stack of reports that might go back years that you now have to read individually and develop an analysis.
- Flash forward 15 months…
- The next unit that enters the area wants an assessment of Abdul Rhaman. They now have to go through the very same process, conducting the same search and then read all those documents and re-analyze them. Now, some duplication of effort is good but at no point in the process you get even an attempt at something like a comprehensive narrative. And no unit is lucky to only have one personality or even one village to concern itself with. There’s simply not enough time to do this sort of work on every person in a unit’s AO every year.
- In short, our system is practically designed to eliminate institutional knowledge every year or so.
- Intel analysts need to (at least periodically) get outside the wire and see the area they’re analyzing. I was fortunate enough to see this play out during the training where two fire teams went through a ‘village’ where one half was unfriendly and the other half was friendly. One group reported it just like that. The second group reported one neighborhood unfriendly and the other neighborhood ‘highly suspicious and pretending to be friendly’ (and, in fact, the soldiers were more hostile to the civilians in the ‘friendly’ neighborhood – think of the repercussions for a COIN strategy of that!). In addition to everything else, analysts will need to have some understanding of their collectors in order to determine their reliability and credibility. I can’t imagine there’s anybody talking about that sort of analysis.
So, what does all this mean? I’ve tried through the posts in the series to point out how we don’t set up our intelligence analysts for success in COIN. We have two strikes against us to begin with given our unfamiliarity with the language/culture generally and the specific area units operate in specifically which, for example, most law enforcement agencies don’t have to contend with. We don’t have a system of institutional knowledge, making the intelligence game really one in which we fight a series of one year wars rather than one multi-year one. Our information is not collated properly forcing our analysts (potentially) to search and collate intelligence on the same subject time and time again.
Analysts aren’t going to develop the expertise in one year to write valuable, in-depth analysis of a particular AO and commanders aren’t going to take the time to read long form reports.
What we need is a system which can allow analysts to build upon the work of those who came before them. We do have a system in place that can handle that work…it’s called Intellipedia. Units should begin creating wiki pages for their AOs, perhaps pre-formatted with the elements of IPB to assist in the creation of the collection plan. As information is gathered, the S2 would fill in the gaps. It would be up to them (their primary mission) to maintain that page and associated pages (since you could link to pages of personnel, events, etc. just like wikipedia).
So, for example, take this vignette from the Flynn report:
Each night, the deputy intelligence officer hosted what he called “fireside chats,” during which each analyst radioed in from his remote position at a designated time and read aloud everything learned over the last 24 hours. Using this approach, daily reports incorporated a wide variety of sources: unclassified patrol debriefs; the notes of officers who had met with local leaders; the observations of civil affairs officers; and classified HUMINT reports. The deputy intelligence officer typed up a master report of everything called in by analysts and closed each “chat session” by providing them with an updated list of questions – called “intelligence requirements” – for the companies to attempt to answer.
Imagine instead of typing this up into a report which will soon be lost in a file or email folder somewhere, these information was entered into a wikipage. The commander (and everyone else on the network) could, at his leisure, see the newly updated information. Units scheduled to take over the next rotation could view the document weeks if not months before coming into country giving them a better orientation of what would be facing them.
Best of all, this thing would be the consistent product which would provide the institutional knowledge that units need. We’ve been in Afghanistan for going on 9 years now. I’d be willing to lay down some serious cash that there’s no product that talks about the evolution of relations among elders/commanders in the Bagram area between themselves and with the coalition over the past nine years in anything approaching the detail that you can find about the evolution of Deep Purple. Why is that? Might it be important? Especially as attacks in the area have increased lately? We’ve had in excess on 10,000 soldiers there for almost a decade so it can’t be because of a lack of potential collectors. It’s simply a lack of a way to capture information about the operating environment over time and without that ability you aren’t going to be able to develop an understanding of the environment that you need in order to produce good analysis.
Now ideally, this sort of thing should be widely accessible so that as many people as possible could view and contribute to it (although the S2 of the area in question should be the ultimate administrator/editor) which means there might be multiple versions (unclassified, secret, etc) so different groups could view as much information as possible. At the lower levels, a pointer system could be used to point readers to the existence of information at a higher level of classification (as is commonly used in many databases now).
This would pose a couple of hurdles. We still keep coalition partners away from our information systems and I think that goes double for non-military partners (NGOs, et. al). We’d clearly need to overcome the former hurdle. If the point is to capture institutional knowledge it’s just dopey to not have coalition partners (who may very well be followed by American forces in a particular area) contribute. Non-military contributors/users would be much more tricky for a whole host of reasons but there should be some way to get/provide information without compromising security or their status as non-belligerents.
So, can it really be this simple? Winning the war through a wiki? Well…no. But I am confident it addresses the issue of maintaining institutional knowledge in a (much) better way than we currently do and in a format that will seem easy to adopt by many of our intel soldiers and easy to use by many potential consumers. And maintaining institutional knowledge as well as collating/cataloging information in a central location is a huge step in understanding the operating environment. Understanding the operating environment is a big step in being able to implement COIN.