Daniel's been reading up on the history of international law, and he's sent us three questions that sound simple until you actually try to answer them. How old is what we call international law? When did customary practices between states first harden into something we'd recognize as functional law? And do any parts of it have actual formal channels for enforcing obligations?
That third question is the one that really bites. Because the whole thing rests on a paradox — international law calls itself law, but there's no global police force. So when does a promise between states stop being diplomacy and start being legally binding? I mean, if I promise my neighbor I'll trim my hedge and I don't, that's annoying but it's not a crime. If I sign a contract and breach it, there's a court that can seize my assets. Where on that spectrum does a treaty live?
That's exactly the tension. And the answer to the first question — how old is this thing — turns out to be older than most people think. We're not talking about the UN Charter or the Geneva Conventions. We're talking about a stone treaty carved before the pyramids were built.
Before the pyramids. That's not hyperbole. We're talking about a literal carved stone that says "this is the border, and we both agree to it." And what's wild is that it's not just a declaration — it has the structure of a binding agreement. There are witnesses named. There are consequences specified. The bones of contract law are all there, three thousand years before anyone wrote a legal textbook.
So we're going to trace this from a boundary agreement in Mesopotamia around twenty-one hundred BCE, through Roman law and early Islamic humanitarian rules, to Grotius and Westphalia, and then land on the enforcement question — courts, Security Council powers, and this strange thing called customary law that binds states without them ever signing anything.
And we'll end up where Daniel's question leads: a system where the most powerful states are effectively immune from the enforcement machinery they control. Which is either a design flaw or a design feature, depending on who you ask.
Let's start with the stone.
The earliest known treaty we have physical evidence for is between two Mesopotamian city-states — Lagash and Umma — around twenty-one hundred BCE. It's a boundary agreement, literally carved into a stone block. Two polities agreeing that this line here is the border, and we're both bound by that. And the reason we know about it isn't just the stone itself — it's that the stone was a monument. It was meant to be seen. It was displayed publicly so that everyone in both cities knew the terms. That's not just law, that's law with a publicity function. It's law that says "we are accountable to each other and to our populations."
Which is... startlingly recognizable. You could drop that into a modern territorial dispute and the basic legal impulse is identical. Two parties, a disputed line, a written agreement, witnesses, and some mechanism to make it stick.
And there's possibly an even older one — the treaty between Ebla and Abarsal, which some scholars date as far back as thirty-one hundred BCE, though the text is fragmentary. So we're pushing the origin of formal interstate agreements back into the fourth millennium BCE. That's before the Bronze Age collapse, before the invention of the alphabet, before almost everything we think of as civilization's infrastructure. But the one that really stops me is the Egyptian-Hittite Peace Treaty from around twelve fifty-nine BCE.
This is the one with both parties' versions still extant.
Both versions. After the Battle of Kadesh, which was essentially a stalemate that both sides claimed as a victory — some things really don't change — Ramesses the Second of Egypt and Hattusili the Third of the Hittite Empire sat down and produced what they called the Eternal Treaty. It's the oldest surviving peace treaty where we have copies from both sides. We have the Egyptian version in hieroglyphics and the Hittite version in cuneiform. And the contents are... look, if you showed this to a modern diplomat, they'd recognize the structure immediately.
What's actually in it? Walk me through the provisions.
A mutual defense clause — I'm quoting here: "If another enemy come against the land of Hatti, the great king of Egypt shall send his troops and his chariots and shall slay his enemy." And the reciprocal obligation is there too — if Egypt is attacked, Hatti sends its forces. It establishes eternal peace and brotherhood, respects each other's territory, and — this is the part that gets me — it includes extradition provisions for political refugees. If someone fled from one kingdom to the other, they'd be returned. But here's the nuance: the treaty specifies that the returned person shall not be punished for their flight. They're to be returned safely, without retribution. That's a level of humanitarian consideration that we don't usually associate with the Bronze Age.
Extradition with a non-punishment clause. In twelve fifty-nine BCE. That's more sophisticated than some extradition treaties from the twentieth century.
It is. And the enforcement mechanism was... A thousand gods. The treaty invokes a thousand gods of both empires — Egyptian and Hittite — to curse anyone who violates it. And these aren't generic gods. The text names them. It calls on the sun god, the storm god, the gods of the mountains and rivers, the gods of both royal families. It's an exhaustive list because they wanted to make sure no divine being was left unoffended by a violation. Which, in that cultural context, was probably a more effective enforcement mechanism than a lot of what we have today.
Because they actually believed the gods would punish them.
They believed it. Deeply. This wasn't performative. The gods were real parties to the agreement. Breaking the treaty wasn't just a political betrayal, it was a cosmic offense. And here's what really lands for me: there's a copy of this treaty sitting at the United Nations headquarters in New York right now. A three-thousand-year-old peace treaty in the lobby of the modern institution that's supposed to do the same thing.
That's either inspiring or deeply ironic, and I'm not sure which.
I think it's both. It's inspiring because it shows how long we've been trying to solve this problem. It's ironic because the Bronze Age solution — divine enforcement — might have been more psychologically effective than what we've managed to build with courts and sanctions. But let me stay with the ancient world for a minute, because the Greek city-states developed something that looks even more like a legal system. They had rules governing interaction in peace and war — among themselves. They had arbitration procedures for territorial disputes. They had rules about the treatment of heralds and ambassadors. They had norms about when you could and couldn't attack. The catch is they explicitly excluded non-Greeks. If you weren't Greek, these rules didn't apply to you. You were outside the circle of legal protection.
So it's a precursor to the "civilized nations" framing that would dominate Western international law for centuries.
That's exactly what it is. The idea that law applies among us, but not necessarily to them. The barbarian is outside the law by definition. And that thread runs all the way through to the nineteenth century and beyond. The standard of "civilization" that European powers used to justify colonialism — that's just the Greek distinction between Hellene and barbarian, dressed up in suits and legal treatises.
What about Rome? You mentioned jus gentium earlier.
The Romans developed this concept called jus gentium — the law of peoples. It regulated interactions between Roman citizens and foreigners, which was a practical necessity when you're running an empire that spans half the known world and your citizens are trading with Egyptians and Syrians and Gauls. You need some common legal framework or commerce grinds to a halt. But what's interesting is how they justified it. They linked it to natural law — the idea that certain principles of fairness are universal, derived from reason itself, not from any particular legal system. So a Roman judge could say "this contract between a Roman merchant and a Greek trader is valid, not because Roman law says so, but because reason demands that promises be kept."
And that distinction — between positive law and natural law — that survives.
It survives completely. The concepts of equity and natural law that run through modern international law trace directly back to Roman jus gentium. When the International Court of Justice talks about fundamental fairness or principles of humanity, they're drawing on a tradition that's two thousand years old. They might not cite Cicero, but they're standing on his shoulders.
But the standard history of international law is heavily Western — Greece, Rome, Grotius, Westphalia. What's missing from that picture?
A lot. And this is where the traditional narrative really shows its blind spots. Early Islamic law under the Caliphate developed sophisticated humanitarian rules for warfare centuries before they appeared in Western law. We're talking about institutionalized rules for the treatment of prisoners of war — providing shelter, food, clothing, respecting their cultures, prohibiting execution, rape, and revenge. There were rules about the protection of civilians, about the treatment of the dead, about proportionality in combat. These weren't suggestions. They were codified obligations, taught by jurists and enforced by commanders who believed they were answerable to God for their conduct.
And this is before the Geneva Conventions by...
By over a millennium. The first Geneva Convention was eighteen sixty-four. These Islamic rules were being applied in the seventh and eighth centuries. And they weren't just theoretical — we have records of caliphs and generals being held to these standards. Abu Bakr, the first caliph, issued specific instructions to his military commanders about sparing civilians, protecting crops, and leaving monks in their monasteries. The Western narrative of international law tends to skip this entirely, but the historical record is clear. If you're tracing the genealogy of humanitarian law, the Islamic contribution is a direct ancestor.
So we've got treaties going back four thousand years, we've got customary rules hardening into something law-like in multiple civilizations. But when does this become what we'd recognize as modern international law? When does it stop being a collection of treaties and customs and start being a system?
That transition has a name, and it's Hugo Grotius. Sixteen twenty-five, he publishes De Jure Belli Ac Pacis — "On the Law of War and Peace." And this is the starting point for what scholars call the modern international legal system. What makes Grotius different is he argued that natural law derives from universal reason, not from a deity. He's breaking from the medieval framework where law comes from God. Instead he says: reason tells us two things. Promises must be kept, and harming another requires restitution. And he makes this move deliberately — he says these principles would be true even if God did not exist. That's a radical break.
That's the foundation? Two principles?
That's the foundation. Everything else builds from those two principles. Treaties are binding because promises must be kept — pacta sunt servanda. Reparations exist because harm requires restitution. And he's writing during the Thirty Years' War — one of the most brutal conflicts in European history, where religious violence was tearing the continent apart — so this isn't abstract philosophy. He's trying to establish rules that can constrain behavior in the middle of chaos. He's saying: even in war, there are limits. Even enemies have obligations to each other. That's a profoundly countercultural argument in sixteen twenty-five.
And then Westphalia gives us the framework.
Sixteen forty-eight, the Peace of Westphalia ends the Thirty Years' War and establishes state sovereignty as the cornerstone of the international order. The idea that each state has exclusive authority within its own territory, and no external power can intervene. That's the framework within which all modern international law still operates. Every treaty, every court, every enforcement mechanism — it all rests on Westphalian sovereignty. And the reason that matters is that it creates a system of equals. At least in theory. No state is legally subordinate to another. The smallest principality and the largest empire are formally equal before the law.
Which creates the enforcement problem. If sovereignty means no external power can intervene, who enforces the law? If every state is equal and supreme within its own borders, who tells a state it's broken the rules and actually makes that stick?
Now we're at Daniel's third question. And the answer is: there are formal enforcement channels, but they're all gated by consent. Let me walk through them. The International Court of Justice — established nineteen forty-five, principal judicial organ of the UN, sits in The Hague. All one hundred ninety-three UN member states are parties to its Statute. It settles legal disputes between states and issues advisory opinions. It's got fifteen judges elected by the General Assembly and the Security Council. It looks like a court. It quacks like a court.
But a state has to agree to be bound.
Has to agree. The ICJ's jurisdiction is fundamentally consent-based. A state can accept the court's compulsory jurisdiction — but only about a third of UN member states have done so. And even those that have can attach reservations. A state can simply refuse to participate in a case, or refuse to comply with a ruling. The court has no bailiffs, no police force, no mechanism to enforce its judgments directly. It relies on the Security Council for enforcement.
And the Security Council has its own problem.
The veto. Five permanent members — China, France, Russia, the UK, the US — and any one of them can block enforcement action. Chapter Seven of the UN Charter gives the Security Council real power: it can impose economic sanctions, it can authorize military force, it can create international criminal tribunals. But the veto means the most powerful states are effectively immune from enforcement. If you're a permanent member, or a close ally of one, the enforcement machinery stops at your border.
So the states whose compliance matters most — the ones with the largest militaries, the largest economies, the greatest capacity to do harm — are the ones who can't be compelled.
That's the paradox in one sentence. The system has the form of law — courts, judges, written opinions, precedents, legal reasoning — but the reality of power politics. When the ICJ ruled against the United States in the Nicaragua case in nineteen eighty-six, finding that the US had violated international law by supporting the Contras and mining Nicaraguan harbors, the US simply blocked Security Council enforcement and ignored the ruling. It didn't argue that the court was wrong on the merits. It just... declined to comply. And that was that.
What does that do to the system's credibility? If the most powerful state can just shrug off a binding judgment?
It damages it, but it doesn't destroy it. Because the judgment still exists. It's still cited. It's still taught in law schools. The legal reasoning still shapes how other states behave. The US lost that case in the court of legal opinion, even if it won in the court of power politics. And over time, that matters. But it's a slow process, and it doesn't help the Nicaraguans who were harmed.
But there's another layer, isn't there? Customary international law. You mentioned this binds states without their consent. How does that work in practice?
This is the strangest part of the whole system, and in some ways the most powerful. Customary international law binds all states regardless of whether they signed a treaty. It doesn't require a signature. It doesn't require ratification. It emerges from how states actually behave and what they believe they're obligated to do. It requires two elements. First, state practice — widespread, consistent repetition of similar acts over time. Second, opinio juris — the belief that the practice is legally obligatory, not just a courtesy or a habit.
So it's not enough that everyone does it. Everyone has to do it because they think they have to. That's a psychological requirement. You have to get inside the heads of states.
And that's what makes it so tricky to prove. The ICJ established this test in the North Sea Continental Shelf cases in nineteen sixty-nine. They said the acts must amount to a settled practice, and they must be evidence of a belief that this practice is rendered obligatory by the existence of a rule of law requiring it. That's the quote. And once a customary norm is established, it binds everyone — even states that didn't participate in the practice, even states that came into existence after the norm crystallized.
Unless you persistently object.
That's the one escape hatch. A state can avoid being bound by a customary rule if it persistently objects during the formation of the rule. But that's a high bar — you have to object consistently and publicly while the norm is crystallizing. You can't just opt out later. You can't say "well, I never liked that rule" after it's already become law. You have to be on the record opposing it from the beginning, and you have to keep opposing it.
So if a new state emerges — say South Sudan in twenty eleven — it's bound by customary norms that crystallized decades before it existed, and it never had a chance to object.
That's the rule. New states inherit the existing customary framework. There's no clean slate. Which makes sense if you think about it — otherwise every new state would be a legal blank slate and the system would unravel — but it does raise questions about consent and legitimacy.
And then there's jus cogens.
Jus cogens — peremptory norms. These are fundamental principles from which no derogation is permitted. No treaty can override them, no state can opt out, no persistent objection works. They're non-negotiable. The list includes the prohibition on genocide, slavery, torture, piracy, crimes against humanity, and wars of aggression.
That sounds like real law. No exceptions, binds everyone, can't be contracted around.
It sounds like it. And in theory, it is the closest thing international law has to truly enforceable obligations. The Vienna Convention on the Law of Treaties says any treaty that conflicts with a jus cogens norm is void. Not voidable — void. From the start. As if it never existed. If two states signed a treaty agreeing to commit genocide, that treaty would be a legal nullity. It wouldn't just be illegal — it wouldn't be a treaty at all.
But enforcement still depends on political will. Someone still has to do something about it.
It always does. The International Criminal Court can prosecute individuals for genocide, crimes against humanity, war crimes, and the crime of aggression. That's a formal channel for individual criminal enforcement — not just state responsibility, but actual people in the dock. We saw that with the Yugoslavia tribunal, with Rwanda, with the ICC's cases against African warlords. But the ICC's jurisdiction is also consent-based, and several major powers — including the US, China, and Russia — aren't parties. So the court can reach a mid-level perpetrator from a weak state, but it can't touch the leader of a nuclear power.
So we have courts that can't compel the powerful, a Security Council where the powerful hold vetoes, and customary norms that bind everyone but still depend on someone being willing to enforce them. It sounds like the whole system is built on a foundation of sand.
And yet. Here's the counterintuitive part. The Vienna Convention on the Law of Treaties from nineteen sixty-nine is so widely accepted that even non-parties follow it. Its key rule — treaties should be interpreted according to the plain meaning of their language, in context, in good faith — that's become functionally universal. States follow it not because they're forced to, but because the alternative is chaos and nobody wants to be the state that says "we don't interpret treaties in good faith." If you say that, nobody will sign a treaty with you again. You've torched your own credibility.
The reputational cost.
Reputation is the real enforcement mechanism. Always has been. The Egyptian-Hittite treaty had a thousand gods. We have diplomatic pressure, economic sanctions, and the fear of being labeled a pariah. Different tools, same basic logic. The enforcement isn't physical — it's social. You comply because the cost of non-compliance is that other states stop trusting you, stop trading with you, stop treating you as a reliable partner. In a world of interdependent economies, that's a real cost.
Which is exactly the right moment to bring in someone who spent years inside the enforcement machinery. Hilbert, you were at the Committee against Torture in the nineties. What did that look like from the inside?
Hilbert: The thing that got me was the gap between the language and the reality. You'd sit in these conference rooms in Geneva, very formal, very procedural, and you'd read these state party reports — "here's what we're doing to comply with the Convention Against Torture" — and then you'd compare them to what the NGOs and the victims were saying, and the gap was enormous. Sometimes the state report was basically fiction.
Hilbert: I spent two years translating for the Committee against Torture in the nineties. Geneva. State party reports would come in, we'd read them, we'd draft these careful paragraphs — "concluding observations," they called them — saying a state was in violation of the Convention Against Torture. Very precise language. Very diplomatic. You'd spend three days negotiating a single sentence because you didn't want to say "this state tortures people" — you wanted to say "the Committee expresses concern regarding certain practices that may be inconsistent with the state party's obligations under Article One." And then we'd send them off.
Hilbert: The state would just... not respond. There was no mechanism to make them. No follow-up with teeth. The file went into a drawer. And you'd sit there thinking, I just spent two weeks on this report and it's going into a drawer in a filing cabinet in Geneva and nobody will ever read it again.
Hilbert: But the states that cared about their reputation — the ones that wanted to be seen as good international citizens, the ones with domestic constituencies that paid attention to these things — they'd actually change their laws. Sometimes. Slowly. So the enforcement wasn't legal. It was reputational. It was the fear of being named and shamed in a report that would get quoted by human rights organizations and opposition politicians and newspaper editorials.
Hilbert: That's been true since the Hittites. The thousand gods were probably more effective than our concluding observations. At least the Hittites and Egyptians believed the gods would do something. Our reports went to a committee that had no power to do anything except write another report.
That's a grim benchmark. A three-thousand-year-old curse outperforms a modern UN treaty body.
Hilbert: The curse had believers. Our reports had... an audience. A small one. Mostly other lawyers. I used to joke that the Committee against Torture's reports were read by approximately twelve people, and eight of them were the authors.
Did you ever see a case where the reputational pressure actually worked? Where the report led to real change?
Hilbert: Israel changed its interrogation practices in nineteen ninety-nine after a committee report. The Israeli Supreme Court ruled that certain interrogation techniques — sleep deprivation, stress positions, violent shaking — constituted torture and were prohibited. Took a few years. But the Supreme Court ruling cited the Convention Against Torture. It cited the committee's general comments. So the words we wrote in those conference rooms ended up in a real judgment, in a real court, with real consequences for how interrogations were conducted. Took six years from the first report to the ruling.
Hilbert: Six years is fast, in that world. Most of the time you're pushing on a string. You write the report, nothing happens, you write the follow-up, nothing happens, you write the follow-up to the follow-up, and eventually everyone retires and a new committee starts over. But every once in a while, the string catches. Someone picks it up on the other end. And that one case makes you think the whole enterprise might not be pointless.
Hilbert: I still have a copy of one of those concluding observations somewhere. In a box in the attic. The language is so careful you can barely tell it's condemning torture. "Expresses concern." "Recommends the state party consider." "Notes with interest." That's as hard as the language gets. And even that was too much for some governments. I remember one delegation walking out of the room because we used the word "systematic." They could accept "isolated incidents." They could not accept "systematic." That one word was a diplomatic crisis.
The system knows what it's doing. It just can't say so plainly.
Hilbert: The system is designed to not say so plainly. That's the whole point. Diplomatic language is the lubricant that lets states admit they're violating the law without admitting they're violating the law. It gives them room to save face while making changes. Once you strip that away — once you say "you are torturing people and you need to stop" — the whole thing seizes up. The state stops participating. They stop sending reports. They stop showing up to the hearings. And then you have nothing.
Which connects back to something I was saying earlier. The Vienna Convention on the Law of Treaties works because everyone follows it voluntarily — the cost of compliance is low, and the benefit of having a shared framework for treaty interpretation is enormous. The torture convention... works less well, because the cost of compliance is higher — you actually have to stop torturing people — and the reputational penalty for non-compliance is lower, because the states that torture usually aren't that concerned about what a committee in Geneva thinks.
Hilbert: Torture's harder to hide than treaty interpretation. And the states that do it usually don't care what a committee in Geneva thinks. They care what their own security services think, what their military thinks, what their domestic political base thinks. A report from Geneva doesn't move the needle. Unless there's a domestic constituency that picks it up and runs with it. That's the key. The international mechanism only works if there's a domestic audience that amplifies it.
Hilbert: Anyway. That's what two years in a UN conference room looks like. A lot of very careful language, a lot of filing cabinets, and occasionally, a real change in the world that you can point to and say "we helped make that happen." But you have to be patient. Very patient.
I want to add one detail I couldn't fit in earlier, because it puts the whole timeline in perspective. The term "international law" itself is surprisingly young. Jeremy Bentham coined it in seventeen eighty-nine, replacing the older phrase "law of nations." Before Bentham, nobody said "international law." The concept — binding rules between sovereign entities — is four thousand years old. The name is barely two hundred. We've been doing this since before the pyramids, but we only named it after the American Revolution.
Which is worth sitting with. The practice predates the label by millennia. And the label itself — "international law" — carries assumptions that Bentham baked in. The "inter-national" part assumes nations are the primary units. But the Mesopotamian city-states weren't nations in the modern sense. The Hittite Empire wasn't a nation-state. The practice is older and broader than the label suggests.
The open question Daniel leaves us with: if enforcement depends on consent and political will, is international law actually law in the same sense as domestic law? Or is it a different category entirely — a system of binding promises with no sheriff?
I think it's the latter. But I also think that's not nothing. The fact that states feel compelled to justify their actions in legal terms — even when they're violating the law — that's a constraint. It's not a police force, but it's not meaningless either. When Russia invades Ukraine, it doesn't say "we don't care about international law." It produces legal arguments. It claims self-defense. It claims humanitarian intervention. Those arguments are mostly nonsense, but the fact that they feel the need to make them — that tells you the law has some pull. Even the violators want to be seen as complying.
As climate change and AI governance create new global problems that require binding rules, the ancient tension between sovereignty and enforcement is only going to intensify. How do you regulate carbon emissions when the biggest emitters can veto enforcement? How do you govern autonomous weapons when the major military powers won't sign the treaty? Daniel's question is more urgent than it's ever been.
We're going to need better answers than a thousand gods and a file drawer in Geneva.
Thanks to our producer Hilbert Flumingtop. This has been My Weird Prompts. Find us at my weird prompts dot com.
We'll be back soon.
