so we bomb a hospital, say "sorry", and just move on?

[Igel]

66. Internet arguing is difficult.

The premises keep changing as you argue. You start against one point, but mid-argument you find that the premises have changed and you need to refocus.

1. Bombing a civilian hospital is not necessarily on contravention of the Geneva Conventions. Protected civilian buildings are not legitimate targets, but if they're used by armed members of the opposing military they cease to be protected. That can include staging armed military in the building, using it as a safe building for launching attacks from, etc., etc. It must be demilitarized to avoid this risk.

That's part of the Conventions, pure and simple. We like to claim how important the Geneva conventions are, but really want to ignore those parts we find inconvenient.

One principle, though, is that the target has to be worth the damage. In other words, there must be some sort of proportionality. That doesn't mean "if they're shooting pistols, we can't use J-DAMs." It means that if there's one fighter we can't level the building and kill 500 civilians. The level of civilian damage and loss of civilian life has to be proportional in some undefined sense to the military value of the target. At that point the entire loss of civilian infrastructure and loss of civilian life falls clearly, squarely, and absolutely on those who caused the devaluation of the civilian target, not on those who targeted it.

2. What do you mean by "error"? Mistaking it for some other building? Nobody's claiming that.

Mistaking it for a legitimate non-civilian target because of bad intelligence that claimed it was a military target? That's what's claimed, and if the US has no forces on the ground to provide the intelligence then it has to decide to either trust the Afghan army or to decide that they're untrustworthy.

If they're untrustworthy, how can you bomb any target they identify?

Trying to decide on a target-by-target basis is to thread the needle from 10 000 miles away. Fiendishly difficult to do. And this is the kind of claim that can be ticklish to pull together when the info's in several countries across 12 time zones in several languages--it relies on numerous sources of information and cross-linked chains of command. Hard to collect, collate, synthesize in the 28 seconds' turnaround time that we can usually pull off in complicated situations. Expect a moving target of sorts in the explanation as a matter of course.

3. Or perhaps it was an honest error on the part of the Afghans that coordinate with the US. The higher ups got intelligence from those on the ground, but those on the ground misrepresented things. 4 fighters became 40 or 400 as they panicked or misjudged the situation. If so, it may wind up being the MSF folk versus Afghan fighters on the ground who are either dead or unavailable for the indefinite future. If this is so, then the risk is assuming that what we know is all there is to know--we hear from the MSF folk and can't hear a counterclaim, so the MSF (who benefit from the halo effect anyway) go unchallenged by any opposing narrative or our own critical faculties.

Note that the MSF have no great love for the US military, since the halo effect works here and in reverse; if they let their building, through weakness or indifference, be used by the Taliban then they have no grounds to make any claims; if they lost their building and the US is culpable, then they have claims for compensation.