However, in this war, "training" is done on the job—not so much by drilling and exercising the Afghan soldiers on bases (though there is some of that) but rather by leading, observing, and fighting alongside them out in the field. In other words, the line between "support troops" and "combat troops," ambiguous to begin with, is fuzzier still here. And at least in the short run (for the next few years), it's unlikely that enough Afghans can be trained quickly enough or thoroughly enough to secure the country on their own.
So we come to the option that President Obama is reportedly going to take, to some degree, in some fashion, in his speech Tuesday night (though press leaks of this sort haven't always been accurate): to send tens of thousands more troops—maybe not the 40,000 extra that Gen. Stanley McChrystal, the U.S. commander in Afghanistan, wants, but some number not much smaller.
The key question here is not so much how many more troops Obama sends but, rather, what he decides they should do (and we don't yet know his decision on that point, either). Still, some questions can be raised in advance.
If he decides on a counterinsurgency strategy (which emphasizes protecting the population more than chasing terrorists), the Army field manual's calculations suggest that something like 400,000 troops would be needed—and, even under the most optimistic assumptions, there's no way that U.S., NATO, and Afghan armies combined will amass anywhere near that many forces anytime soon, if ever.
This is why much of the strategy will likely involve cultivating Pashtun tribal leaders to fight the Taliban and prodding relatively moderate Taliban groups to turn against the more militant ones—in short, buying key people off, whether through persuasion, money, weapons, ammunition, logistical support, or the supply of basic services.
Gen. David Petraeus, the head of U.S. Central Command, has long been saying that success in Afghanistan has to involve, to some extent, striking a deal with enemies. "This is how you end these kinds of conflicts," he said in a speech at the Heritage Foundation in October 2008. There is, he added, "no alternative to reconciliation."
Petraeus is very agile at this sort of enterprise, as he demonstrated in 2003 in Mosul as commander of the 101st Airborne Division, and in 2007, with the "Sunni Awakening," as commander of all U.S. forces in Iraq.
But two concerns arise when mulling the transfer of these notions to Afghanistan. First, Petraeus had something to offer the Iraqi Sunnis. In Mosul, he handed out jobs (for as long as the money lasted, which, alas, wasn't long). In the Awakening, he provided military alliance after the tribal leaders (who initiated the contact) recognized that al-Qaida terrorists posed a greater threat than did the U.S. occupiers. He and McChrystal are now trying to reprise these sorts of deals in Afghanistan, but it's unclear whether they can offer much that's compelling to insurgent or fence-sitting Pashtuns.
Second, as smart as those two generals (and many of their advisers) are, how much do they really know about Afghan tribal politics, which (as they do know) are far more complex than Iraq's ethnic fissures and whose leaders are known to switch sides, and switch back again, at whim or the slightest provocation? (On this latter point, see the opening chapters of Dexter Filkins' 2008 book The Forever War.)
The United States has never fought this kind of war before (unless you count the Philippines, which lasted 40 years and involved a level of brutality that would never be countenanced today). We haven't been fighting this kind of war even in Afghanistan. (As the saying goes, we haven't been fighting for eight years but, rather, for one year, eight years in a row.) Starting to do so now, as even some of the advocates of escalation admit, is a large gamble with short odds.
So here's what it comes down to: This option might be a good idea if it worked, but the chances of its working are slim (though not zero); all the other options seem to be bad ideas, but they might cost less money and get fewer American soldiers killed (though not necessarily).
Which road is less unappetizing? I don't know. That's why I'm ambivalent.
My guess is that President Obama held so many meetings with his national-security advisers on this topic—nine, plus a 10th on Sunday night to get their orders and talking points straight—because he wanted to break through his own ambivalences; because he needed to come up with a reason (not just a rationalization) for doing whatever it is that he's decided to do, some assurance that it really does make sense, that it has a chance of working, so he can defend it to Congress, the nation, and the world with conviction. Let's hope he found something. A columnist can be ambivalent; a president can't be.