My position on this question is that trying to have consistent principles on this issue is bad actually, and I will unapologetically evaluate self-determination issues on a case-by-case basis.
I disagree that letting whatever nontrivially-sized place secede is good, because states getting smaller and more numerous makes coordination problems on large scales worse (I'm aware there are plenty of arguments for the reverse, but I'm not going to expand on this issue for now). And a norm that every group gets a right to self-determination in the future if and only if they already have it now offers a lot of advantages in terms of stability; not redrawing borders at all can cut down on warfare. But neither of these heuristics seem like good reasons not to have taken away Serbia's ability to genocide Kosovar Albanians. Russia taking Crimea from Ukraine doesn't change how many different countries need to be involved in large-scale coordination challenges, and that particular operation didn't even involve any bloodshed, so you could make a case that that undercuts my argument that border changes are bad because war is bad. But this won't stop me from opposing Russia's annexation of Crimea, because Russia agreed not to do that without Ukraine's consent as part of an agreement for Ukraine to give up nuclear weapons, and undermining incentives for states to give up nuclear weapons is bad.
If you come up with a consistent principle, you'll inevitably encounter situations where it turned out your principle was missing something important. The actual principle I'm using here is "a group of people gets self-determination if and only if it is best for the world for them to get self-determination", but I'm not counting that as a real principle because it's too underspecified. Why should I adopt a different principle instead? It seems to me that if "what's best for the world" ends up conflicting with some more well-specified principle, I should go with what's best for the world.
One possible answer to this is that different people have different opinions about what's best for the world, and can end up in conflict over it, but if they can all agree to follow certain consistent principles instead, this can avoid conflict. I agree this is an issue (and isn't even the only source of conflict here; some people will simply have more provincial concerns than what's best for the world), but using consistent principles doesn't actually solve this, because there will be conflict over which principles to use. If you think self-determination is generally good, and I think too much self-determination creates too much coordination-problem headaches to be worth it, then "any at-least-city-sized group of people who want independence gets it" does not work as a compromise. I'd actually prefer just letting you choose every time, since then if there's some situation where letting some specific city declare independence ends up being obviously terrible, you might notice this and put a stop to it. Groups with conflicting interests can negotiate compromises without having clear abstract ethical principles behind their compromises, and that's okay. If the principles you come up with so that there doesn't need to be any conflict doesn't include anything like "... unless it's on this side of this arbitrary line, because we need to placate France", you're doing it wrong, and should probably stop trying to follow principles.
My position on this question is that trying to have consistent principles on this issue is bad actually, and I will unapologetically evaluate self-determination issues on a case-by-case basis.
I disagree that letting whatever nontrivially-sized place secede is good, because states getting smaller and more numerous makes coordination problems on large scales worse (I'm aware there are plenty of arguments for the reverse, but I'm not going to expand on this issue for now). And a norm that every group gets a right to self-determination in the future if and only if they already have it now offers a lot of advantages in terms of stability; not redrawing borders at all can cut down on warfare. But neither of these heuristics seem like good reasons not to have taken away Serbia's ability to genocide Kosovar Albanians. Russia taking Crimea from Ukraine doesn't change how many different countries need to be involved in large-scale coordination challenges, and that particular operation didn't even involve any bloodshed, so you could make a case that that undercuts my argument that border changes are bad because war is bad. But this won't stop me from opposing Russia's annexation of Crimea, because Russia agreed not to do that without Ukraine's consent as part of an agreement for Ukraine to give up nuclear weapons, and undermining incentives for states to give up nuclear weapons is bad.
If you come up with a consistent principle, you'll inevitably encounter situations where it turned out your principle was missing something important. The actual principle I'm using here is "a group of people gets self-determination if and only if it is best for the world for them to get self-determination", but I'm not counting that as a real principle because it's too underspecified. Why should I adopt a different principle instead? It seems to me that if "what's best for the world" ends up conflicting with some more well-specified principle, I should go with what's best for the world.
One possible answer to this is that different people have different opinions about what's best for the world, and can end up in conflict over it, but if they can all agree to follow certain consistent principles instead, this can avoid conflict. I agree this is an issue (and isn't even the only source of conflict here; some people will simply have more provincial concerns than what's best for the world), but using consistent principles doesn't actually solve this, because there will be conflict over which principles to use. If you think self-determination is generally good, and I think too much self-determination creates too much coordination-problem headaches to be worth it, then "any at-least-city-sized group of people who want independence gets it" does not work as a compromise. I'd actually prefer just letting you choose every time, since then if there's some situation where letting some specific city declare independence ends up being obviously terrible, you might notice this and put a stop to it. Groups with conflicting interests can negotiate compromises without having clear abstract ethical principles behind their compromises, and that's okay. If the principles you come up with so that there doesn't need to be any conflict doesn't include anything like "... unless it's on this side of this arbitrary line, because we need to placate France", you're doing it wrong, and should probably stop trying to follow principles.