There is an obvious military aspect to the reconstruction of Iraqi security forces, but this is essentially a training mission. Then there is the matter of providing for security and order as the basis of all progress towards political (and economic) goals. Already, the administration has held up the Sunni Arab insurgency as an excuse for why elections cannot be held quickly.

Even if we don't accept this line, suppressing the insurgents is clearly necessary. First, it serves to protect the Shi'ite community rather than waiting for acts of communal self-defense or retaliation that would lead directly to civil war. Second, unless the insurgency is suppressed, there is little hope for shifting actions in the name of Sunni communal interests from violent to non-violent forms.

But this cannot be the last thought anyone thinks about the use of force in Iraq. The U.S. military must also provide to Iraqis a model of the proper relationship between security forces and citizenry in a free society -- admittedly, a very difficult task while a low-intensity conflict continues, but not optional while building a democracy is considered a near-term goal.

The U.S. Army expects to hand off to Iraqis eventually. Should the Iraqis behave in the same fashion, like an occupying force? If so, it becomes increasingly difficult to understand how they will ever pass the key political milestones of limits on government power and conciliatory leadership.

Simply put, armed men who suppress one community for the benefit of another are already a familiar feature of the Iraqi experience. Even if the American version is mild by comparison, it's not enough to start repairing Iraqis' broken-down doors at the conclusion of raids based on bad tips, hoping this will mend wounded feelings. It's past time to reconsider the strategy of breaking down every door available in hopes of netting the maximum possible number of adversaries.

MORE FUNDAMENTALLY, it's past time to revisit how we got to a point where fixing the doors we break down became our last, best hope for winning the hearts and minds of average Iraqis. At this distance, there is no good way to identify the operations and tactics that will strike the right balance between firmness and conciliation. Possibly there is no way to achieve that balance. But we can at least revisit how these decisions are made in the first place.

Back in November, I received a letter from an Army friend in Baghdad, who almost despaired of the Army's ability to pursue a pacification mission. He wrote:

You know, the Army is a land-fighting unit, and that's it. [...]overall we just suck at handling what we're trying to make a large-scale friendly occupation, and rebuilding of a government. Honestly, some of the officers who are in charge of figuring out what's going on out there know surprisingly little about the culture and Islam.[...] No Army is ever going to invade a completely foreign country, and know exactly what's going on culturally. But I do think we can put someone in charge who's better than these infantry officers who are so simple-minded and ignorant ("Wahabi bad. Attack Wahabi."). Or can we? Who would we use? Do we have an army of trained diplomats?

Our area is run by an infantry battalion, and the only way they know how to handle problems is attack things. Here we are, [...]three people, trying to build relationships with key leaders and key communicators, and all our supported unit [the infantry battalion] wants to do is figure out what to attack. Did you know some areas are controlled by field artillery units? They're even worse. Fortunately, at least they try to raid places on foot, instead of firing shells.

Instead, the Army takes direction from U.S. Central Command, headed by General John P. Abizaid. Abizaid, in the famous formulation of Washington Post defense correspondent Dana Priest, is a "proconsul," a regional authority figure more powerful than any ambassador, let alone the head of the permanently overtaxed CPA. We should reverse this wrongheaded order of precedence. First, for the purposes of Iraq, Abizaid's CENTCOM, which in Pentagon lingo is a "supported command," should become a "supporting command" -- one that provides forces and logistics to another body, in this case the U.S. military headquarters in Iraq. (U.S. Forces Korea (USFK) already enjoy a comparable arrangement with U.S. Pacific Command.) Second, and crucially, the "USFI," currently led by Lt. Gen. Ricardo Sanchez, should answer to Bremer, the local civilian chief, not to Abizaid, the regional military honcho.

This approach defies two American traditions. First, in theory, civilian authority over the military is vested more or less exclusively in the President and the Secretary of Defense (prior to the present administration, they were known jointly as the National Command Authority). Nor (second) do Americans conceive of their country as the sort of old-fashioned colonial state that once might have created such arrangements.

But in practice, leaders delegate. (There's no mention of any Secretary of Defense or "National Command Authority" in the Constitution.) And old-fashioned or not, it's what the situation calls for. Differences about day-to-day operations in Iraq should not have to be mediated through an incessant stream of appeals to officials in Washington.

A stronger objection is that it's too late, that CPA was born defective, too weak and embattled to carry out its current functions, let alone to run an actual country. But giving either CPA or a suitable successor body the authority and resources to carry out such a mission seems necessary if the U.S. is even to attempt to carry out a strategy in Iraq consistent with the administration's stated policy. The alternative, seemingly, is to let the current "autopilot" approach linger on until it collapses.

Of course, a civilian-first approach is not a panacea. It would only be a start. Worse, all of this discussion is academic. It's not this administration's style to consider creative solutions. And the democratization of Iraq would be painfully unlikely to succeed even with the best-run military operations of all time. Getting Iraqis to accept foreign control, followed by power-sharing among themselves, is fundamentally not a military issue.

Analyst

Addendum. A response to the comments below is here.