Revising FM 3-24: What needs to change?

[William F. Owen]

On a note of caution, I would not combine the obvious short comings of Effects Based Operations (EBO) as a short coming of using a systematic, well defined, repeatable, and well understood assessment process to help drive counter insurgency operations. FM 3-24, in order to be a good document to truly help commanders and staffs in the counter insurgency fight, needs a well understood, documented, and step-by-step assessment process to help military organizations track their progress, determine their successes, target their shortcomings, in order to provide counter insurgency operations a compass towards successful conclusion.
.

My personal experience of a systematic, well defined, repeatable, and well understood assessment process, is that it is of limited use, when carried out by human beings dealing with incomplete, changing and ambiguous data.

Some old guy called Carl Von something called it "friction" and I think specifically warned against trying to quantify it.

However, as someone passionate about military science and thought, I'd love to see you post of one these cards or some deeper explanation so we may better assess it.

[Entropy]

My personal experience of a systematic, well defined, repeatable, and well understood assessment process, is that it is of limited use, when carried out by human beings dealing with incomplete, changing and ambiguous data.

Some old guy called Carl Von something called it "friction" and I think specifically warned against trying to quantify it.

However, as someone passionate about military science and thought, I'd love to see you post of one these cards or some deeper explanation so we may better assess it.

I mentioned this briefly before in a similar topic some weeks ago, but I think the methodology used by the strategic warning community, often called "indications analysis" might be useful for some of the difficulties in judging effectiveness and or progress toward various goals in coin. Indications analysis is designed to be carried out by "human beings dealing with incomplete, changing and ambiguous data" - it comes with the territory of the strategic warning problem.

Probably the best primer on the subject is Cynthia Grabo's now-declassified and updated text from the 1970's. In a simplistic nutshell, indications analysis works backward from a particular end-state. A series of indicators, or signposts, on the way to that end-state are developed and then monitored. Although developed for warning, any end-state or scenario can be broken down and analyzed using this methodology. The advantages are two-fold: It can be used to make predictions (which is what it was originally designed to do) but it also can provide a means to analyze and compare various courses of action one might take.

To borrow from Mark's post below, indications analysis is both an "art" and a "science" and tries to combine the strengths of both philosophies. Although the framework might be science-based, the nature and ambiguity of information requires human judgment, particularly since indicators and indications are not limited to hard, quantifiable data. From Grabo:

An indication can be a development of almost any kind. Specifically, it may be a confirmed fact, a possible fact, an absence of something, a fragment of information, an observation, a photograph, a propaganda broadcast, a diplomatic note, a call-up of reservists, a deployment of forces, a military alert, an agent report, or anything else. The sole provision is that it provide some insight, or seem to provide some insight, into the enemy’s likely course of action. An indication can be positive, negative or ambiguous (uncertain).

and

An indicator is a known or theoretical step which the adversary should or may take in preparation for hostilities. It is something which we anticipate may occur, and which we therefore usually incorporate into a list of things to be watched which is known as an “indicator list.’’ Information that any step is actually being implemented constitutes an indication. The distinction between expectation and actuality, or between theory and a current development, is a useful one, and those in the warning trade have tried to insure that this distinction between indicators and indications is maintained. Many non-specialists fail to make this careful distinction.

Instead of "preparation for hostilities" you can substitute any theoretical end-state you desire. In fact, Indications Analysis within the intelligence community has expanded beyond the traditional role of warning of hostile actions by adversaries to monitoring a variety of issues of interest to the US. "Warning problems" have been established on a variety of topics that have little to do with the potential of an adversary's attack. I see no reason why this tested framework cannot be used in analysis of COIN and LIC, particularly since in my experience so much effort is put towards current intelligence, which has little value for this kind of estimation and analysis.

I do see two potential problems however. First, indications analysis requires a lot of resources, time and effort. Often, the strategic warning community is under-resourced in lieu of other requirements - it seems likely that a COIN-focused effort would suffer to an even greater extent.

Secondly, indications analysis requires a focused collection effort that may not be available in COIN scenarios. IOW, indicator development and monitoring are not possible if there isn't a significant body of baseline knowledge and the ability to get information in the necessary areas.

Of course, both these limitations apply equally to any other methodology or analytical framework that one might use.

[Rex Brynen]

Probably the best primer on the subject is Cynthia Grabo's now-declassified and updated text from the 1970's. In a simplistic nutshell, indications analysis works backward from a particular end-state. A series of indicators, or signposts, on the way to that end-state are developed and then monitored. Although developed for warning, any end-state or scenario can be broken down and analyzed using this methodology. The advantages are two-fold: It can be used to make predictions (which is what it was originally designed to do) but it also can provide a means to analyze and compare various courses of action one might take.

I've always thought that this quote from Grabo should be framed and hung on a number of walls (especially in the NGO/aid "early warning" community):

The researcher should take care not to be trapped in a rigid system which cannot be readily expanded or modified as new developments occur. The system should be designed to serve the analyst, not to have the analyst serve the system.

[marct]

Entropy, thanks for posting the link to the Grabo text - I hadn't seen it before. It's an interesting blend of inductive and abductive logics; I think I will probably use it in my course this fall.

[Jedburgh]

.....indications analysis requires a focused collection effort that may not be available in COIN scenarios. IOW, indicator development and monitoring are not possible if there isn't a significant body of baseline knowledge and the ability to get information in the necessary areas.

Entropy, thanks for posting the link to the Grabo text - I hadn't seen it before. It's an interesting blend of inductive and abductive logics; I think I will probably use it in my course this fall.

Grabo's Anticipating Surprise is a classic in the intelligence field and a great read. However, because of its fairly tight focus on warning in conventional conflict, the principles she relates may be more difficult for inexperienced readers to absorb and mentally shift into the COIN/UW context.

Entropy, I don't feel that indicator development is impossible or overly difficult in COIN/UW. But keep in mind that Grabo discusses strategic warning in her book, with the product intended to alert policymakers to emergining threats - and she also cautions about the difficulties of convincing them (and of others in the IC) of the real dangers embodied in emerging threats that are outside of their current perceptions. Warning intel in COIN/UW is most effective at the unit level.

Putting aside the formality of lists for a moment, just consider that all soldiers operating on the ground in such an environment develop their own personal indicator lists in their head. To use a cliche, but real, example - the sudden absence of locals from a normally lively street in town is usually taken as an indicator that something bad is about to happen. In the COE, such indicators range from being simple and broadly applicable as in that example, to the much more complex and focused on narrow, local context.

(Recall the bit from Go Tell the Spartans, where the analyst demonstrates to the cynical commander the ability to predict which village the VC are going to hit next.)

At an even more personal level, I used to train my HUMINT'ers in the principles of indicator analysis for interrogation. The baseline of information regarding kinesics, cognition and emotion is gathered during the first phase of the interrogation (or, if the situation allows, during the first screening interview). In this case, the indicators developed are used, not for "warning" in the standard sense, but to alert the interrogator to deception, potential leads and openings for manipulation of any one or all of the three mentioned aspects of the source.

To get back to analysis, the unfortunate truth is that many analysts at the tactical level have neither the training nor the experience to effectively implement a warning system for their units. Another obstacle is that, even if they develop the best list of indicators available in-country, unless they have an effective system for monitoring incoming information specifically for indicators and disseminating immediate warning, it ends up being a waste of valuable time of a critical asset. Just as Grabo relates for the strategic level, to be effective at the unit level it would also have to be a full time gig - and I know of few units that can spare an analyst to do nothing more than the warning job. But most competent analysts are still able to integrate elements of the warning discipline into everything else they've got on their plate. I'm sure many of those on the board can think of examples.


FYI: Several other good pubs from NDIC, along with the Grabo text, are available for download through the link posted on an earlier thread