The first principle is the obligation to provide these civilian decision-makers with unvarnished military advice.
The second is the requirement, enshrined in the Uniform Code of Military Justice, that all military personnel obey "lawful orders." In fact, it is a crime, punishable by court-martial, not to obey. The qualifier—"lawful orders"—is important: It pre-empts the Nazi defense of war crimes, and it is a legitimate way out for conscientious soldiers who don't want to take part in atrocities like My Lai or torture sessions like those at Abu Ghraib.
What should an officer do if the civilian authorities reject his advice and then issue orders that are lawful but, in the officer's professional opinion, unwise or damaging to the national security?
There are three types of disagreement, raising three separate issues.
First, by almost everyone's assessment, whether he's right or wrong, no military officer has any business speaking in public about matters of broad policy while a presidential review of that policy is going on. When Gen. McChrystal was asked the question in London, he should have demurred.
Second, if, after the review, President Obama decides to pursue a strategy less ambitious than counterinsurgency, and if his advisers calculate that this alternative strategy doesn't require as many troops, then the general's role is to salute and carry out the strategy as best as he can. There are reasons why Obama might make such a decision (the impossibility of propping up a regime that the Afghan people regard as illegitimate, a calculation that counterinsurgency's costs are too high and the chance of success too low, a conclusion that a Taliban defeat in Afghanistan won't preclude al-Qaida from working elsewhere). Maybe they're good reasons, maybe they're bad; the point is, this is a decision for presidents, not military commanders.
But the third type of issue is trickier. If Obama decides to pursue a full-blown counterinsurgency strategy but to send far fewer troops than McChrystal thinks are necessary to succeed, then the general and other senior officers who agree with him are in a hard spot. This would be analogous to the position of the generals in 2002-03 who advised Rumsfeld that he was sending too few troops to Iraq; their dispute wasn't about the mission (a political decision) but rather the means to accomplish it (a matter of professional military judgment).
Most Army officers have read Dereliction of Duty, a 1997 book by then-Maj. (now-Brig. Gen.) H.R. McMaster, which argued that the Joint Chiefs of Staff abrogated their responsibilities during the Vietnam War by not pushing back against the misguided policies of President Lyndon Johnson and Secretary of Defense Robert McNamara. When the chiefs folded under pressure from Rumsfeld in the lead-up to the Iraq war, many junior officers wondered whether their superiors ignored McMaster's lessons.
A crisis of conscience took hold in some Army circles, with officers asking what they would have done in the chiefs' place and what they should do if they someday found themselves in the same place.
In July 2007, two retired colonels, Leonard Wong and Douglas Lovelace, wrote an article dealing with this issue in the newsletter of the Army War College's Strategic Studies Institute. Titled "Knowing When To Salute," it laid out nine options short of disobedience that a senior officer might follow when political leaders resist military advice—including requesting reassignment, taking early retirement, testifying before Congress, and, if all else fails, resigning.
By the way, there is a huge distinction between retiring and resigning. When officers retire, they do so with full benefits, health care, and continued membership in the fraternity of the officer corps—and full allegiance to the code of military justice. When they resign, they're freed from that code's restraints—but they also give up all the benefits. In part for this reason, no U.S. general has resigned in more than 40 years.
Not every officer agrees with all nine of Wong and Lovelace's options. Some officers draw a line between disputes that involve policy and those that involve military tactics and strategy. Some note that the line between policy and strategy is a fine one, and they wonder where to draw it. In part for this reason, some officers believe that they shouldn't publicly criticize a president's policies under any circumstances.
One difference between the Bush and Obama administrations is that, when Rumsfeld ran the Pentagon, dissent of almost any sort was discouraged and in some cases punished. At least Obama's strategic review is giving all views a hearing. That fact alone may ease the frustration of those who disagree with the decision he winds up making.
Meanwhile, McChrystal's public pronouncement and Gates' indirect rebuke fall nowhere near the realm of a constitutional crisis. But they do hint at the sorts of tensions that the war might sire and that Obama and Gates will have to damp down.