1. “How many soldiers will it take to over-run Taliban soldiers in this village?” is a highly technical military question where the officer corps provides expertise.
2. “What will it take to rally civilian support in that village and build a stable anti-Taliban community/government? ” is a very different question requiring different levels of expertise to answer.
3. And “Which of two competing answers to question #1 above is more credible” is something different too.
4. “Will 50,000 soldiers added to US forces in Afghanistan successfully produce a "victory”?“ is a question that mixes together components of questions of the first three types with additional issues of larger politics and economics. Lyndon Johnson’s administration proved conclusively that the Department of Defense of the time was unable to provide correct answers to questions of this type.
One would hope that answers to questions of type 4 provided to US officials by the bureaucracy (civil and military) would be informed by deep expertise in a number of topics – such as Afghan culture/languages/history, economics, government operations, and civil infrastructure engineering – in addition to expertise on purely military tactics. Such a comprehensive approach would not guarantee a correct answer, but the absence of that expertise almost guarantees that the answers are going to be wrong. The arrogant assumption that one could answer such a question without deep understanding of, for example, the ability of the Afghan government to provide basic services is good evidence that the answer won’t be worth anything at all.
None of this calls for people who are "expert” at demanding “resolute policy” or forecasting that sufficient yelling will intimidate all opponents.
See: Learned Nothing Forgot Nothing.
and Why US Foreign Policy is so stupid
Updated March 19 2014
krebscycle 
Leave a comment