My friend Mike O'Hanlon has provided in these spaces an accurate summary of recent security and political trends in Iraq. O'Hanlon has long been one of the best sources of good data on Iraq, presented in a form making it easy for the public to chew on. The first point O'Hanlon makes is that Iraq remains much less violent than a few years ago, notwithstanding some recent bloody attacks that have gotten what he regards as undue emphasis partly because of “sloppy journalists.” He seems to be anxious to counter any argument to the effect that if security in Iraq is going downhill than the military intervention there is a failure and ought to end. The implications of any trend in violence, up or down, for the advisability of continuing the U.S. military mission there, however, have never been self-evident. If the security situation is good, does that mean what we are doing must be working and we should continue, or does it mean the mission has been accomplished and we should leave? If the security situation is bad, does that mean we have more work to do in Iraq and should stay, or does it mean the mission is a failure and the sooner it ends the better? One can find versions of each of these arguments, which says more about the predispositions of people advancing the arguments than about implications of a security trend considered by itself. So one needs to look at other trends as well.

That gets to O'Hanlon's second point, which is that Iraq's politics, unlike the security situation, are going “backwards,” with plenty of discord and an inability of the sectarian and confessional communities to reach agreements. Accurate again. Sunnis greatly distrust Shiites, Kurdish ambitions are incompatible with Arab ones and so on. O'Hanlon rightly observes that Iraq's civil war “cannot yet be considered definitively over.”

So the externally assisted imposition of increased security doesn't correlate positively at all with movement toward resolution of Iraq's internal conflicts and toward a more stable political order. Insofar as the political situation is moving backward, the correlation is negative. It is most glaringly negative with regard to aspects of the current conflict that are not just a delayed playing out of contests that the Baathist dictatorship had forcibly suspended but are direct consequences of the U.S.-initiated war. This is especially true of one of the indicators of instability that O'Hanlon mentions: a significant presence in Iraq of al-Qaeda, which wasn't there at all before 2003.

Put these observations about current trends in Iraq together and the conclusion should be clear: because the U.S.-assisted security effort is not bringing Iraq any closer to political reconciliation and stability, the end-of-year deadline for extracting U.S. troops from Iraq should be observed, save perhaps for the approximately 3,000 personnel that the Obama administration envisions keeping there in a mostly training role. And yet, O'Hanlon's concluding point is much different: "Three thousand U.S. troops in Iraq is too few.”

How can a conclusion so contrary to the implications of the premises be offered? Partly it is the light-at-the-end-of-the-tunnel syndrome. O'Hanlon talks about “buying time” for Iraq—time for “wounds to heal,” “trust to be built” and so forth. But the required time would be, in terms of what is politically meaningful for Americans, essentially infinite. The wounds to be healed and the distrust to be overcome go back centuries in the case of Sunni-Shia relations.

We are also witnessing another example of the tendency to view most peoples as harmonious by nature if immediate points of friction could be overcome, fears could be assuaged and appropriate legal formulas could be devised. Contain immediate threats to safety and security, the thinking goes, and people will find ways to resolve their differences. This was the premise underlying the surge in Iraq, and subsequent experience clearly showed the premise to be invalid. Similarly invalid thinking has been involved in peacekeeping missions in some other countries. The flaw in the thinking is that sometimes a conflict of interests is so substantial that no amount of reassurance or friction-reducing measures will resolve it. Sometimes the only thing that would resolve it is a test of strength in a civil war—a war that is allowed to be fought to a finish and not interrupted by outsiders, which only delays the inevitable. Edward Luttwak made similar observations more than a decade ago, arguing that sometimes the best thing for outsiders to do is to “give war a chance.”

Iraqis gotta do what they gotta do, even if that means waging more civil war. It ultimately does them no good, and it certainly does the United States no good, to keep intervening in their affairs with the vain hope of avoiding the unavoidable.