What a fascinating document is General Stanley McChrystal's Initial Assessment on Afghanistan, delivered to President Obama on 30 August and made public today for the first time, with redactions.
McChrystal, who is America's top military man in Afghanistan, has written a document that must be one of the most unusual ever to come from a military commander.
He often sounds more like an anthropology professor than a general as he grapples with the tricky problem of what is going on in Afghanistan. "The conflict in Afghanistan can be seen as a set of related insurgencies, each of which is a complex system with multiple actors and a vast set of interconnecting relationships among those actors", he states. This is true, of course, but it is a pity it has taken eight years for America's senior brass to work out this most fundamental of facts.
Until recently there was no recognition by the military of the importance of the Afghan tribes to understanding events in the country. It was not understood, for example, that for insurgents to move from one part of the country to another required agreement from any tribes inhabiting the land in between.
Or that some of the Pashtun tribes were only fighting in Afghanistan, while others were also fighting against the Pakistan government. Or that there are both Sunnis and Shias amongst the Pashtuns. Or, until recently, that there are at least three (according to McChrystal) different Taliban factions (others would say there are five, or even seven).
The precise relationship between opium production and the insurgency has not been investigated. Opium eradication has been seen as a public health issue, not a matter of military strategy. And no effort has been put into producing propaganda that highlights the hypocrisy of the Taliban for producing drugs that are killing thousands of Moslems.
And how is it that the Taliban faction in Quetta in neighbouring Baluchistan is left undisturbed in this war? Everyone knows where they are and yet they appear to be untouchable.
In a remarkable admission, McChrystal acknowledges that the Taliban is better at propaganda than the Coalition: "Major insurgent groups outperform GIRoA and ISAF at information operations...They have carefully analysed their audience and target products accordingly. They use their Pashtun identity, physical proximity to the population and violent intimidation to deliver immediate and enduring messages with which ISAF and GIRoA have been unable to compete."
This is deeply depressing stuff. There are companies and consultants being paid millions of dollars to come up with solutions for problems like these, and yet the results are risible.
Having thoroughly (and correctly) trashed the Karzai government and the international community, McChrystal advocates what he calls population-centred counter insurgency. ISAF, together with the Afghan security forces, must shelter Afghans from violence, corruption and coercion. Military officers must also gain a much greater understanding of the country and its people.
This means local language training, a remarkable turnaround, considering that the British Foreign Office, for example, does not presently have any Pashto speakers.
"All ISAF personnel must show respect for local cultures and customs and demonstrate intellectual curiosity about the people of Afghanistan," he says.
It will be necessary to build personal relationships with the local people: "To gain accurate information and intelligence about the local environment ISAF must spend as much time as possible with the people and as little time as possible in armoured vehicles or behind the walls of forward operating bases."
This is all a very tall order and the question that will be on everyone's mind is whether or not it has all come too late. It may simply make more sense for the military to think of ways they can force the Taliban - in all its complexity - to the negotiating table.