Edited by author 12-02-2003 07:31 PM
TM


Your strategy page link was a little disappointing, though in it’s own unintended way it points out a big problem whether the lessons learned are being reviewed by insiders or outsiders.



You work with computers, did you ever do any programming? From a systems design viepoint the amateur tends to write “spaghetti code.” Start with all the bottom level problems and build up the system up around them. When you trace the logic it looks like a bowl of spaghetti. For small programs this works very well, but for large programs it produces programs that often don’t work, are difficult or impossible to maintain, and are usually over budget. The better way to do it is to use a “structured top down” approach; engineers call this a “black box approach. Start at the top, define the problem inputs and outputs and the major steps in the process. You then have a structure at the top consisting of a “black box” for each step but you haven’t written the code for the step. Then you take each step ooor “black box” and do the same to it. You end up with a program does what it supposed to do, if there is a problem it is relatively easy to find it and fix and is much more predicable on the resources required to accomplish the task.



The article you linked is a prime example of spaghetti code thinking. Lots of good data (mostly correct) from the bottom level of the structure but he does not really look at the top of the system to see what the structure of the problem is.

For example he has an excellent description of the operational viewpoints of ground and air units with regard to air support. There are two things he does not consider that should come prior to his discussion.

First what is the overall mission of the Armed Forces? He realizes the Army and Air Force see their missions differently. Depending on how the mission is understood will change the relationship between the ground and air forces. If the mission is to “nuke them ‘till they glow” the emphases should be on the air force solutions, and if it is to throw out the bad guys and liberate their oppressed populations a ground force solution is more appropriate. Or should we maintain both capabilities? Put a little differently is the lesson learned to avoid ground operations in favor of strategic bombing or provide ground forces with the means to occupy and hold. Depending on how you see the goal both an air and ground lesson can be drawn from the article's evidence.

Second in the early 1940’s Congress decided and still thinks that the Army and Air Force should be split. Partly this was a result of their strategic vision the time, and partly the old policy of having agencies with overlapping functions so disputes must go back to congress for resolution. If Congress had decided instead to leave them in one service and force a more combined approach, there would be more of a common viewpoint. The Marine Corp witch combines tactical air and ground forces has much less of the problem he describes. Both approaches can work. There are problems and lessons learned with each approach, but if one do not see that the different viewpoints result are the indirect result of a Congressional decision, one is likely to make irrelevant or wrong recommendations.





The problem of using a wrong approach to look at lessons learned is common to military and civilians, insiders and outsiders. After all most of us live near the bottom of what ever system we are in, the “spaghetti code process" works for most of our day to day decisions and it is hard to transfer thought processes. This problem is solved by being sure persons making force structure decisions by personality or education tend to look at the problem top down.


The author correctly realizes that there is a consistent problem in applying lessons learned, and has lots of interesting data. However he does not correctly see where the problem lies. If one uses the wrong approach to analysis even “a totally dispassionate look at the lessons learned.” will not produce reliable results. However as he points out in the end, truth and logic will have their way. and if his approach is used the truth of this statement will be all to apparent.

cynical joe - I need the capability to destroy a city.

nuclear capability = 1 missile with one nuclear warhead
conventional capability = x troops, y ammunition

The combat nuclear capability and the combat conventional capability are equivalent by definition. I'm using x and y for the amounts for two reasons, first that I don't know the x and the y and second, if I did, I'd likely still get an argument which is irrelevant to the main point.

Your argument that construction + maintenance cost of the nuclear capability is greater than the construction and maintenance cost of the troops + ammunition. I think you're dead wrong.

To make an MX missile, for example, is a $250M expenditure with a much lower expenditure for maintenance, perhaps $10M. The army's budget is about $90B to maintain 10 divisions or about $9B per division including overhead (I believe nukes are in the air force budget so we're not double counting). For conventional forces to equal an MX missile, the destructive capacity equivalent of that missile would be <1% of one division including training, basing, arming, and overhead. I submit that these values for x and y are simply too small and they are too small by a large margin.

Hank - What I was trying to get at was the importance of renewing the military/civilian conversation. The aims of the armed forces are subordinated to civilian intentions by constitutional design. The growing disconnect between the military and the larger society is dangerous all around.

I'm not a programmer by trade, but come in from the administration perspective. We end up dealing with programmer's messes in the real world all the time. I do understand about design and black box development. I wouldn't call the underlying article spaghetti code but rather a nice prototyping effort for the lessons learned project. Prototypes don't have to be pretty and they're certainly not final but they focus attention on what are the essential elements to be solved.

TML:

I share Hank's evaluation of the Strategy Page piece. In particular, the CAS/DAS argument assumes a false dichotomy between those places where a ground controller is available to provide final control, and those where one is not. It also ignores the relative strategic mobility of lighter forces, and the greater likelihood those will be able to be employed-as well as their greater ability to provide the ground control discussed above.

The piece also has a flawed view of the lessons learned system. Its main design is to allow guys at the unit level to pass their experiences on to others who may conduct similar operations. It's not a tool for guiding acquisition policy. I'm all for conversations, but in order for it to be meaningful, both sides have to understand the basics. And the "civilian control" argument would be more convincing if most of the proposals weren't bids to avoid the reforms of the current civilian leadership (i.e., Sec Rumsfeld).

On the nuclear cost, the usual rule-of-thumb is that a nuclear force costs about 10% of a similar strength conventional force (though obviously it's less versatile, and not useful in most scenarios). However, the cost-effective argument probably shouldn't rest on how often a weapon system is used. If that were the case, Hitler's Wermacht or Napoleon's Grande Armee would be among the most cost-effective forces in history . . .

Cecil Turner - Hey, in baseball, .500 is a fantastic average. Thanks for filling some details in.