This is a much, much better piece from Kim Kagan than usual. But there’s still this, when she addresses whether the U.S. and NATO can "win" in Afghanistan:

Some answer simply and sharply in the negative: They claim that Afghanistan has never been centrally ruled (which is wrong) and that it has been the "graveyard of empires" (which is true in only a specific handful of cases). Failure is not at all inevitable. The war in Afghanistan has suffered almost from the start from a lack of resources, especially the time and attention of senior policymakers. The United States prioritized the war in Iraq from 2007 until 2009, for strategically sound reasons. Some of this parsimony also comes from flawed theories of counterinsurgency: U.S. Secretary of Defense Robert Gates, for example, misreads the Soviet experience in Afghanistan, which has consistently led him to argue incorrectly against expanding the size of the force there, claiming that it increases the risks of failure.

1) Winning means what, ma’am? Establishing "central rule" by the Afghan government? Preventing al-Qaeda from returning to Afghanistan, as John Brennan recently put it? Fracturing the insurgent syndicate so that its al-Qaeda affiliates are marginalized and its non-affiliates do whatever? (For what it’s worth, this is the definition of success that most interests me.) Something else? Pick a definition, because the rest of the piece is rudderless without one. Why does Kagan find any utility in the term "win" at all?

2) Bob Gates is not perfect. But his career has in large part been defined by Afghanistan and by the Soviet Union. In the 1980s, while deputy CIA director, he focused intensely on repelling the Soviet occupation. And, on balance, the man contributed to precisely the outcome he sought. In what possible sense does he "misread the Soviet experience in Afghanistan?" Because he’s concerned about a tipping point past which Afghans will feel besieged by U.S. troops and not aided? Kagan doesn’t say. Nor, for that matter, does she say how the most counterinsurgency-promoting defense secretary in history is in the throes of "flawed theories of counterinsurgency."

But look. Gen. McChrystal says that losing the support of the Afghan people will be "strategically decisive." There’s, of course, more than one way to lose that support. Something McChrystal told the Journal yesterday resonates here: "I think it’s what you do, not how many you are. It’s how the force conducts itself." There’s a lot of merit to that. Having lots of troops doing nothing for Afghans builds antipathy, and so does having few troops doing nothing for Afghans — sentiments I heard first-hand from farmers in Paktia Province. But it would be irresponsible to suggest that policymakers shouldn’t consider that balance.

Kagan ends on a note that I will, in fact, co-sign for:

[A]s is often the case in these kinds of war, if you aren’t winning, you’re losing.

Absolutely correct. Counterinsurgency is the Talladega Nights of warfare. This is the ballad of Stan McChrystal.