Daniel's been thinking about the law of armed conflict — what most people call international humanitarian law — and he's landed on a tension that's uncomfortable. The idea is this: we have a legal code that governs the use of military force, something that by definition happens when normal law has broken down. He wants to know how clear these obligations actually are compared to domestic law, what enforcement mechanisms exist for people who say the rules were broken, and then the real question — if enforcement is often impractical or just fails, what's the point of having the code at all?
That last question is the one that makes people squirm. And it's the right question. I've spent more time than I'd like to admit reading ICRC commentaries and tribunal rulings, and I can tell you the enforcement gap is real and it's wide. But the premise that a law without a police force is pointless — that's where most coverage gets it wrong.
Go on.
Domestic criminal law works because there's a state with a monopoly on force standing behind it. You rob a bank, the state arrests you, prosecutes you, locks you up. The law of armed conflict has none of that. There's no world police. The International Criminal Court can prosecute individuals, but it can't send marshals into a war zone to grab a defendant. The International Court of Justice settles disputes between states, but it can't enforce its own rulings. So you've got a legal system where the primary enforcement mechanism is... the parties themselves agreeing to follow it.
Which sounds like having a speed limit with no traffic cops and hoping drivers feel bad about going fast.
That's not far off, but here's where the analogy breaks. Speed limits work because most drivers comply most of the time, and the minority who don't get caught often enough that the system holds. In armed conflict, compliance isn't about fear of punishment — it's about reciprocity, reputation, and something the ICRC calls "the dictates of public conscience." The Martens Clause, which has been in every major humanitarian treaty since the Hague Convention of eighteen ninety-nine, says that even in cases not covered by specific treaty provisions, combatants remain under the protection of principles of humanity and the demands of public conscience.
So the law has a built-in admission that it can't write a rule for everything, and it falls back on "don't be monstrous" as a legal principle.
And that sounds vague, and in some ways it is, but it's not nothing. The Martens Clause has been cited in actual rulings. It's been used to fill gaps where treaty law was silent. The clarity question Daniel raised — I'd say the obligations are surprisingly clear in some areas and maddeningly fuzzy in others.
Give me the clear part first.
The four Geneva Conventions of nineteen forty-nine and their Additional Protocols. These are the backbone. They protect the wounded and sick on land, the wounded and sick at sea, prisoners of war, and civilians. The rules about who you can target are explicit. You cannot target civilians. You cannot target medical personnel or facilities. You cannot attack someone who is hors de combat — out of the fight, wounded, surrendered. You cannot use weapons that cause superfluous injury or unnecessary suffering. These aren't suggestions. They're treaty obligations ratified by every single country on earth.
Every country.
Every country. The Geneva Conventions are the only treaties in history with universal ratification. One hundred ninety-six states. That happened in two thousand six when Montenegro and Nauru signed on, and it's remarkable. It means that at least on paper, the entire world has agreed that certain conduct is beyond the pale regardless of the conflict, regardless of the cause, regardless of who started it.
So the clarity exists at the level of principle. The fuzziness is in application.
Right. Distinction — telling combatants from civilians — sounds clean in a textbook. In practice, you've got fighters who don't wear uniforms, you've got civilians who take up arms intermittently, you've got dual-use infrastructure like roads and power grids that serve both military and civilian purposes. Proportionality says you can't launch an attack where the expected civilian harm is excessive relative to the concrete and direct military advantage anticipated. That word "excessive" is doing an enormous amount of work. What's excessive to one commander is acceptable to another. And the standard is judged based on what the commander knew at the time, not what we know afterward.
Which creates a built-in defense for almost any strike. "The intelligence I had said it was a command center."
And that defense is legitimate in many cases. The law doesn't demand perfect knowledge. It demands reasonable precautions and good-faith decision-making. The problem is that "good faith" is almost impossible to disprove after the fact without a paper trail showing deliberate targeting of civilians.
So we've got a body of law that's clear on the big prohibitions, fuzzy on the edge cases, and largely unenforceable against anyone who decides not to comply. Which brings us back to Daniel's real question. What's the point?
Let me give you three answers, and none of them are "because it's the right thing to do."
Thank you.
First, the law of armed conflict shapes military training and operations in ways that are real and measurable. Professional militaries — and Daniel specifically mentioned professional militaries — they don't just read these treaties and file them away. They build them into rules of engagement, into targeting doctrine, into the legal review process that every operation goes through. The US military has judge advocates — JAG officers — embedded at every level of command. They review target lists. They advise on proportionality. They can and do say no.
And that's not just window dressing.
It's not. There are documented cases where strikes were called off because the legal advisor said the collateral damage estimate was too high. That doesn't happen in every military, and it certainly doesn't happen in non-state armed groups, but for the professional forces Daniel's asking about, the law is an operational constraint, not just a rhetorical one.
Second answer?
Reciprocity. If you mistreat my prisoners, I mistreat yours. If you target my civilians, I target yours. That spiral is as old as war, and the law of armed conflict exists in part to short-circuit it. The Geneva Conventions are explicit that they apply regardless of whether the other side is complying. You don't get to torture prisoners because the enemy tortures yours. But the practical reality is that compliance is often mutual. When both sides treat prisoners humanely, both sides' soldiers come home. That's a tangible incentive that doesn't require a court.
It requires both sides to care about their people coming home, which — not always a given.
No, and that's a limit. But for state-on-state conflict between professional militaries, which is what the Conventions were primarily designed for, the reciprocity mechanism is real.
Third answer.
Post-conflict accountability, as uneven and unsatisfactory as it is. The international criminal tribunals for the former Yugoslavia and Rwanda actually happened. They convicted people. The International Criminal Court has secured convictions, though far fewer than its critics would like. The principle of universal jurisdiction means that some states — Belgium, Spain, Germany — have prosecuted war crimes committed in other countries by people who happened to be on their territory. A Syrian regime torturer traveling to Europe can be arrested and tried. That's happened.
But the ICC can't touch the major powers. The US isn't a party. Russia withdrew. China never joined. Israel isn't a party. So you've got a court that can prosecute warlords from small states and rebel commanders, but not the people commanding the world's largest militaries.
That's a fair summary, and it's the single biggest criticism of the system. The Rome Statute has one hundred twenty-three states parties, which sounds like a lot until you notice who's missing. India, Pakistan, Indonesia — not members. Most of the Middle East outside Jordan and Palestine — not members. The court's jurisdiction is real but it's patchy, and the Security Council referral mechanism means that permanent members can shield themselves and their allies.
So the enforcement is asymmetrical in exactly the way that makes it look like victor's justice to anyone on the receiving end.
And that perception problem is itself a threat to the system's legitimacy. If the law only ever prosecutes the losers, it starts to look like just another weapon the winners use. That's not entirely fair — the ICTY prosecuted people from multiple sides of the Balkan conflicts — but it's not entirely wrong either.
Let me push on something. You said professional militaries incorporate these rules into their operations. But Daniel's question was about clarity compared to domestic law. If I'm a soldier in the field, domestic law tells me exactly what I can and can't do — don't steal, don't assault, don't kill except in self-defense. The law of armed conflict tells me I can kill enemy combatants, but not if they're surrendering, and not if they're wounded, and not if the weapon I'm using causes unnecessary suffering, and not if the expected civilian harm is excessive. That's a much more complicated decision tree.
It is. And the comparison to domestic law is instructive in another way. Domestic law is precise because it governs normal life. The law of armed conflict is trying to regulate something that is inherently chaotic and violent, and it's doing it through rules that have to be applied in split seconds by people under extreme stress. A police officer's use-of-force decision is hard enough. A soldier's targeting decision in an active firefight is orders of magnitude harder.
So the clarity question — the answer is that the obligations are clear in the abstract and unavoidably fuzzy in application, and the fuzziness isn't a drafting failure, it's a reflection of what's being regulated.
That's right. And I think that's actually the strongest answer to Daniel's "what's the point" question. The point is that even fuzzy law is better than no law. A commander who has to think about proportionality, even if the standard is vague, is operating under a constraint that wouldn't exist otherwise. The alternative isn't perfect clarity — the alternative is no constraint at all.
There's a historical dimension here that Daniel alluded to. The law of armed conflict is relatively new. How new?
The first Geneva Convention was eighteen sixty-four. It was ten articles long and it basically said treat the wounded, protect medical personnel, and use a red cross on a white background as a protective emblem. That's it. Henry Dunant, a Swiss businessman, had witnessed the aftermath of the Battle of Solferino in eighteen fifty-nine — forty thousand dead and wounded in a single day — and he was so horrified that he organized local civilians to help the wounded regardless of which side they fought for. Then he wrote a book about it and lobbied governments until the Convention happened.
So the whole edifice starts with one guy who saw a battlefield and thought, this is unacceptable.
And the International Committee of the Red Cross, which he co-founded, is still the guardian of the Geneva Conventions today. They're the ones who visit prisoners of war, who negotiate access to conflict zones, who do the quiet, unglamorous work of reminding parties what they signed. The ICRC isn't a court and it isn't a police force, but it's the closest thing the system has to an institutional conscience.
Eighteen sixty-four. That's practically yesterday in the history of warfare. For most of human history, the rule was that the victor could do whatever they wanted to the vanquished. Rape, pillage, enslavement, massacre — all standard practice.
And it's worth remembering that the modern law of armed conflict emerged alongside the modern nation-state and the modern professional military. It's not a coincidence. Once you have standing armies rather than feudal levies or mercenary bands, you need rules to govern them. Discipline becomes a military necessity, and law is one form of discipline.
The paradox Daniel's pointing at — law governing the breakdown of law — maybe isn't a paradox at all. It's law governing the conduct of state-organized violence, which is a different thing from the breakdown of law. War isn't chaos. It's organized. It has chains of command, logistics, doctrine. Regulating it with law is weird, but it's not contradictory.
I'd go further. The fact that war is organized is exactly what makes it regulable. If it were pure chaos, law would be pointless. But because militaries are hierarchical institutions that train, plan, and issue orders, you can impose legal obligations on them and expect some level of compliance. The commander who orders a war crime is making a decision that could have been different. That's what criminal responsibility rests on.
Let's talk about restitution. Daniel asked what mechanisms exist for people who allege breaches. If I'm a civilian whose home was destroyed in an airstrike that I believe was unlawful, what do I actually do?
The short answer is that your options are limited and none of them are fast. You can try to bring a case in your own country's courts if the perpetrator is within reach. You can petition the ICC if your country is a state party or if the Security Council refers the situation. You can bring a civil suit in a third country under universal jurisdiction. You can seek reparations through a truth and reconciliation process or a claims commission after the conflict ends. The UN Compensation Commission after the first Gulf War processed over two point seven million claims and paid out more than fifty-two billion dollars. That's real money that went to real people.
But that was a unique situation — a clear aggressor, a clear loser, and a Security Council that was united enough to set up the mechanism.
Right. Most conflicts don't end with one side surrendering unconditionally and the international community setting up a compensation fund. Most conflicts end messily, if they end at all, and the people who suffered the worst abuses get nothing.
So the restitution mechanisms exist on paper but are practically unavailable to most victims most of the time.
Yes. And I think we have to be honest about that. The gap between the law's promises and its delivery is enormous. The law says you have a right to remedy. In practice, that right is contingent on geopolitics, on which side you were on, on whether the perpetrators are still in power, on whether anyone with jurisdiction cares enough to act.
Which loops us back to the question of what the law is for. If it can't deliver restitution, can't consistently punish violators, and can't stop atrocities from happening in the first place — what work is it actually doing?
I think it's doing several things at once, and they're not all about enforcement. The first is norm-setting. The Geneva Conventions establish a baseline for what civilized states do not do. That baseline matters even when it's violated, because it provides the standard against which violations are measured. Without the law, you can't even say "that's a war crime." With the law, you can name it, document it, and preserve the record for whatever accountability mechanism might eventually exist.
Even if that mechanism is just history.
History and politics. The law of armed conflict gives third parties a language for applying pressure. When the US tells Israel that a particular operation raises concerns under international humanitarian law, that's not just a moral complaint — it's a legal claim with specific content. The law provides the vocabulary for the conversation.
And for militaries themselves, it provides a framework for training. You mentioned JAG officers earlier. That's not a small thing. The fact that every US military operation involves lawyers at the planning table means the law is shaping decisions before they're made, not just judging them afterward.
There's a concept in legal theory called "internalization." The idea is that when people — or institutions — incorporate legal norms into their own decision-making processes, the law works even without external enforcement. Professional soldiers are trained that certain things are illegal, and most of them don't want to commit crimes. The law shapes their identity as professionals.
Most of them. The ones who do want to commit crimes aren't deterred by a treaty.
No, and that's the limit. The law of armed conflict is least effective where it's most needed — in conflicts involving non-state armed groups, in civil wars where one or both sides reject the legal framework entirely, in situations where the perpetrators have no expectation of ever being held accountable. The law works best when both sides are professional militaries that care about their international standing. It works worst when one side is fighting for survival or ideology and doesn't care what the ICRC thinks.
Which describes most conflicts since nineteen forty-five.
Most conflicts by number, yes. Most casualties, probably. The big state-on-state wars — the kind the Geneva Conventions were designed for — have been relatively rare. What we've had instead are civil wars, insurgencies, counterinsurgencies, and asymmetric conflicts where one side doesn't wear uniforms and the other side struggles to apply distinction and proportionality in a way that looks lawful.
And the law has adapted, somewhat. The Additional Protocols of nineteen seventy-seven extended protections to victims of non-international armed conflicts. But Common Article Three of the Geneva Conventions, which applies to civil wars, is a single article. It prohibits murder, mutilation, torture, humiliating and degrading treatment, and the passing of sentences without a fair trial. That's the floor. Everything above that is in Additional Protocol Two, which applies only when the non-state group controls territory and has a responsible command — a threshold most insurgencies don't meet.
And even when they do meet it, the non-state group hasn't signed anything. They're bound by the law because it's customary international law, binding on all parties regardless of treaty status, but they haven't consented to it in any meaningful sense. Their fighters haven't been trained in it. Their commanders may not know it exists.
So we've got a legal framework that was built for one kind of war and is being applied to a very different kind of war, with predictably mixed results.
I'd say the framework has shown more adaptability than its critics give it credit for. The customary international humanitarian law study that the ICRC published in two thousand five — it took eight years and catalogued one hundred sixty-one rules of customary IHL. Most of them apply in both international and non-international armed conflicts. The law has evolved through state practice and opinio juris, not just through treaties. It's not static.
Eight years to catalogue the rules. That's a very Herman fact.
It's a monumental piece of work. And it matters because customary law binds everyone, not just treaty parties. So when people say the law doesn't apply to non-state groups, that's not quite right. It applies. The problem is enforcement, not applicability.
Let's circle back to Daniel's framing. He called the law of armed conflict paradoxical — a law to govern the situation in which law has failed. I think after talking through this, the paradox softens. Law hasn't failed when war breaks out. Law has been set aside in one domain — the prohibition on the use of force between states, which is the UN Charter's domain — but another body of law activates in its place. It's not a failure of law, it's a transition between legal regimes.
The UN Charter says states shall refrain from the threat or use of force against the territorial integrity or political independence of any state. That's the law against war. When that law is breached, the law of armed conflict kicks in. It's not saying war is legal or acceptable — it's saying since war exists, here are the minimum rules for conducting it without descending into barbarism.
The jus ad bellum and the jus in bello. The law governing whether you can go to war and the law governing how you fight once you're in one.
And they're deliberately kept separate. A soldier fighting an illegal war is still bound by the law of armed conflict. A soldier fighting a just war can still commit war crimes. The separation is one of the great achievements of the system, because it means the rules apply equally regardless of whose cause is righteous.
Which is also one of the things that drives people crazy about it. If the other side started the war, why do they get the same protections?
Because the alternative is that every war becomes a contest of who can be more brutal while claiming the moral high ground. The separation of jus ad bellum and jus in bello is what prevents "but they started it" from being a defense to war crimes.
So Daniel's question — what's the point of a legal code without enforcement — I think the answer we've landed on is that the law does several things that don't depend on enforcement. It sets norms. It shapes training and operations. It provides a vocabulary for accountability even when accountability is delayed or incomplete. It creates the possibility of reciprocity. And it preserves the distinction between combatants and civilians in a way that, imperfectly and inconsistently, actually does save lives.
The ICRC has documented that in conflicts where the parties have accepted IHL training and ICRC access, compliance with the basic rules — treatment of prisoners, protection of medical facilities, distinction in targeting — is measurably higher than in conflicts where they haven't. It's hard to quantify precisely, but the direction of the effect is clear. The law changes behavior at the margins, and in war, the margins are where thousands of people live or die.
At the margins. Not at the center. The center is still states doing what states do when their interests are at stake.
Yes. And I think we have to hold both thoughts at once. The law of armed conflict is real and it matters and it saves lives. And it is also routinely violated, selectively enforced, and powerless against the most powerful. Both things are true. The question isn't whether the law works perfectly — it doesn't. The question is whether we're better off with it than without it.
The answer to that is not close.
Hilbert: The word "professional" is doing too much work in this conversation.
Go on.
Hilbert: You both keep saying professional militaries as if that's a stable category. I've been in four of them and it isn't.
Four?
Hilbert: British Army, nineteen eighty-two to eighty-seven. Rhodesian African Rifles for about eighteen months — that was a mistake and I left. French Foreign Legion, two years. Then a private military contractor in the nineties that I won't name because the NDA apparently still holds. That's four. The point is, every one of them had legal advisors, rules of engagement cards, the whole apparatus. And every one of them had moments where the apparatus got shoved aside because the operation demanded it and someone senior enough said make the problem go away.
The law is real until it's inconvenient.
Hilbert: The law is real in the training rooms and the planning sessions. It's real when there's time. It's real when the operation is routine and the chain of command is watching. It gets less real at three in the morning when the intelligence is ambiguous and the window is closing and the colonel says he'll take responsibility. And "taking responsibility" means nothing. Nobody ever takes responsibility.
That's... a darker picture than what I was painting.
Hilbert: It's not darker. It's just what happens. The law exists. The training exists. The lawyers exist. And then there's the moment, and the moment doesn't care about any of it. That doesn't mean the law is pointless. It means the law is a constraint that operates most of the time, and the times it doesn't operate are the times that produce the worst outcomes. Which you both already said. I'm just saying the gap between the planning room and the ground is wider than you made it sound.
Does the training matter in that moment? When the colonel says make it go away, does the fact that you've been trained in IHL change what you do?
Hilbert: Sometimes. Depends on the soldier. Depends on the unit culture. Depends on whether anyone in the room is willing to say no and mean it. Most people aren't. The ones who are — they're not popular, but they're the reason the system isn't entirely hollow.
That matches what I've read about the My Lai massacre in Vietnam. The only reason it stopped was that a helicopter pilot, Hugh Thompson, landed his aircraft between the soldiers and the civilians and threatened to open fire on his own countrymen if they didn't stop. One person. The entire legal apparatus of the US military was theoretically present in that operation, and it took one person.
Hilbert: Thompson was a warrant officer. Warrant officers are the ones who actually know how things work. The rest is paperwork.
You said four militaries. Was the gap between law and practice consistent across all of them?
Hilbert: The Legion was the most honest about it. They didn't pretend the law applied when it didn't. The British were the most committed to the process — if the process said no, the answer was usually no, at least in my experience. The Rhodesian thing was... that was a different kind of war entirely. The law wasn't even in the room most of the time. And the contractor work had no law at all. You did what the client paid for and you didn't ask questions. I left after fourteen months.
I want to connect that to something we said earlier. The law works best when both sides are professional militaries that care about their standing. You're describing a spectrum even within "professional." The British Army cared about the law. The Legion cared about it selectively. The contractor world didn't care at all. So the category is doing less work than we implied.
Hilbert: That's what I said. The word is doing too much work.
Yet you stayed in the British Army for five years. Something kept you there.
Hilbert: The pay was terrible and the food was worse. I stayed because the people I served with believed in something. Most of them couldn't have named a single article of the Geneva Conventions, but they knew you don't shoot prisoners and you don't target civilians. They knew it as a matter of who they were, not what they'd read. That's the internalization you mentioned. It's real. It's also fragile, and it doesn't survive every pressure, but it's real.
I think that's actually a hopeful note to end the discussion on. The law works not just through courts and treaties but through identity — through soldiers knowing that certain things are beneath them.
Hilbert: It's not hope. It's just what I saw.
The single biggest misconception people hold about the law of armed conflict is that it's a set of rules imposed from outside that constrain what militaries would otherwise do. The reality is that professional militaries helped write these rules, train their people in them, and enforce them internally — however imperfectly — because it's in their interest to do so.
Because without those rules, the institution you're serving in stops being a military and starts being something else. The law isn't just a constraint on military power — it's partly constitutive of what a professional military is.
We've been talking around the enforcement gap for most of this episode, and I think where we've landed is that the law of armed conflict matters even when its formal enforcement mechanisms fail. It matters because it shapes training, because it enables reciprocity, because it provides the standard against which violations are named and remembered. It's not a substitute for a police force and it was never designed to be one.
Daniel asked what the point is of a legal code without enforcement. The point is that enforcement isn't the only way law works. Norms, identity, reciprocity, institutional design — these are all mechanisms of legal effect that don't depend on a court issuing a binding judgment. The law of armed conflict is a thinner kind of law than domestic criminal codes, but thin isn't the same as nonexistent.
Thanks to Hilbert Flumingtop for producing, and for the reminder that the gap between the planning room and the ground is where the law actually lives or dies.
This has been My Weird Prompts. You can find every episode at my weird prompts dot com, and if you want to send us your own weird prompt, email the show at show at my weird prompts dot com.
We'll be back soon.
