Sunday, December 06, 2009

Afghan Highs

For some reason these two bits of intel on Afghanistan seemed worth putting side by side.

First: Afghanistan is not a country; but a collection of tribal areas. It may be defined as that part of central Asia that has historically been resistant to conquest by empires. One unifying factor is that most of the men are constantly stoned on the locally produced high-potency opium and hashish.

(via Digby) Is this accurate? Dunno, sounds plausible, if I had to live in the Afghan tribal areas I'd probably want to be high most of the time. On the other hand, it's hard to square this picture with the Afghani's reputation as fierce fighters.

So, that's a view of Afgahnistan from the ground level. Meanwhile, perched on the commanding heights of a modern-day massive bureaucratic war machine, the American would-be benevolent rulers get high on insane business graphics (via Yglesisias). Something this fucked up could only be generated by people who were huffing too much white-board marker fumes.

Full PDF report here. I have an ongoing debate with a colleague about the value of graphics and visualizations, including some tangled network diagrams like this which are common in systems biology. I'm generally pro-pretty-pictures, but this is a great piece of evidence for his side. Measurement is pretty tricky in biological systems, but at least you know there is something real going on that could in theory be measured, like the amount of mRNA being made off a particular gene. In this chart, you have somthing like "coalition force density" -- which presumably could be an objective number, albeit not a very meaningful one -- is combined with variables like "indivdual competence, judgement, and ability to execute" and "ANSF Corruption & Tribal Favoritism", which can't possibly be defined tightly enough to be measured. So what is the point of this, other than to highlight what a hopless morass fighting a counterinsurgency warfare in mountainous country half a world away where we don't speak the language, understand the culture, or have an idea for an endpoint or exit strategy? None of those arrows point to an exit.

No comments: