Hi Marc
Keep in mind though that I was ref. CENTCOM's response to Peter's question - and it was oriented as you noted to the specific mission in Afghanistan. I do think it requires a regional perspective to really consider the capabilities and capacities required - and moreover to put the right capabilities in the right places at the right time. AfPak is a good example, however more broadly - what are the requirements to enable a strategy for interdicting enemy recruiting, traiing and basing grounds abroad, and how can you disrupt or interupt their movements (both physical and others) between locations - this quickly gets into all oj our USG resources, and is contingent to a large degree on muli-lateral support.
As such, I think there is a signifcant component of developing partner security forces to this end, however there are other areas that must be considered as well.
You are correct in bringing up the unique nature in each set of conditions and that sustainable security can take multiple forms - however they must be weighed in light of the outcomes you can tolerate.
Marc considered:
This can't be just about SFA (although as a developmental activity its a great place to discuss it) - this is much more broadly the issue of all the actions we take to achieve a policy end. I will say that design supports considering this more broadly and the risks associated with one COA over another (of which one may be doing nothing as to not make things worse).This tightness of focus will, IMHO, cause the adoption of some seriously flawed assumptions for SFA. The one that is running around in my head right now is the flawed assumption that "nationality" is the pre-eminent component of identity (vs., say, kinship, ethnicity, religion, etc.). If we assume that a lot of SFA is taking place in so-called "fragile states", i.e. ones that never really developed a strong, unitary "national character", then it strikes me that this is a fatal flaw, since those other elements of identity (think of them as the bases of motivational factors)
Lets asssume that design uncovered the issues you brought up - but your requirement to extend security in order to deny safehaven remained. The process of design may lead you alternative ways of doing this, and requirements that pop out of the SFA LOE and into the governance and/or economic (or whatever LOEs you are using in your campaign design). There might be a requirment for political accomodation with a tribe that is currently excercising a form of self government - but which might support some assistance in other areas. The possibilities are as numerous as the range of conditions, however our tolerance may not allow us to accept all of them.
SFA is really a force employment concept to support whatever ends are decided on, the process of campaign design though is what is supposed to frame how you can best achieve those ends. e.g. it may tell you that if you extend security in this area, you need to consider what are the implications to the adjacent areas. It may telll you that at the moment the conditions do not support a preferred COA, but you may be able to do other things that shape the outcome in the meantime - e.g. if country X says "no way" to your assistance - maybe he'll accept support from somebody else who is willing or desires your assistance. We often get myopic in the way we approach a problem and don't look at the alternative ways to solve it because it does not seem direct enough - design supports identifying the correct problem and then looking at that problem from multiple perspectives to consider the range of possibilities.
WRT to SFA - Design lends itself well to it because of the nature of development which may include a siginificant timeline where conditions can be greatly altered based on interaction. This I think is really beneficial when trying to establish a rational for generating one capability over another.
Hope that answered you questions - I'll forward you the UNCLASS design guide based on the work I did. Its not perfect - really more of a functional design, but it does get at the logic wrt identifying requirments.
Best, Rob
Bookmarks