The real question is how we keep a pullout from looking like a surrender.

How do we win in Iraq—or at least get out of there without triggering catastrophe, handing a victory to the most radical insurgents, and possibly destabilizing the Middle East?

This is the question of the moment, among the war's supporters and critics alike. Blogs, op-ed pages, and brow-furrowing journals are filled with plans, some laid out in exacting detail, for eking some measure of success from this tragic venture. But all of them evade a basic reality about Iraq these days—that the United States is no longer in control. President George W. Bush could follow the best of these plans, or devise a better one, and it wouldn't make much difference, because there's nothing to suggest that the Iraqis would go along.

Three of the most widely discussed proposals are a Washington Post op-ed piece by Gen. Wesley Clark, an article in the current issue of Foreign Affairs by military historian Andrew Krepinevich Jr., and a 10-point plan by blogger-professor Juan Cole.

All three criticize the Bush administration's handling of the war. Yet they also all oppose simply pulling the troops out now, a move that could trigger a major blood bath, a civil war, perhaps a regional war, as well as a major victory for terrorist groups and a severe setback—on strategic, political, and moral grounds—for the United States.

So, what do they propose instead?

Gen. Clark should have the most to say on this subject. A retired Army four-star general, he's led troops on the battlefield in Vietnam, led whole wars as NATO's supreme allied commander during Kosovo, and helped negotiate the settlement of the war in Bosnia. And yet his Washington Post piece of Aug. 26, titled "Before It's Too Late in Iraq," is stunningly nebulous. The essence of Clark's plan is this:

The United States should form a standing conference of Iraq's neighbors, complete with committees dealing with all the regional economic and political issues, including trade, travel, cross-border infrastructure projects and, of course, cutting off the infiltration of jihadists.

The problem is that none of Iraq's neighbors has displayed any inclination to get involved in such matters. It is unlikely that Iran or Syria would agree to sit on any committee organized by the United States—or, if they did, that their interests would coincide or that they'd be interested in compromising with each other as well as with Turkey, Kuwait, Jordan, Saudi Arabia, and the others. It is at least as likely that discussing these issues at formal meetings could harden their disputes and conflicts.

Clark's military proposals are, surprisingly, no less vague. "The vast effort underway to train an army must be matched by efforts to train police and local justices." (This is under way.) "Canada, France, and Germany should be engaged to assist." (Oh? How?) "Military and security operations must return primarily to the tried-and-true methods of counterinsurgency: winning the hearts and minds of the populace through civic action, small-scale economic development and positive daily interactions." (How, if insurgents continue to sabotage all such efforts?) "Ten thousand Arab-Americans … should be recruited to assist as interpreters." (Good idea for the CIA, FBI, and military intelligence, but would so many want to join the Army? Why haven't they already?) "A better effort must be made to control jihadist infiltration into the country by a combination of outposts, patrols and reaction forces reinforced by high technology." (Isn't that what we're doing already?) "Over time U.S. forces should be pulled back into reserve roles and phased out." (What does "over time" mean? At what bench marks?)

If Gen. Clark is thinking about running for president again, he needs to do better than this.