Customary international law

How can a rule bind every nation when no legislature ever enacted it?

Ambassadors are immune from arrest. A coastal state controls the fisheries two hundred miles out. Neither of these came from a world parliament, because there is none. Yet a court will apply them against a state that never signed anything, and lawyers will say without embarrassment that the state is bound.

The tidy answer — that states consented — collapses on inspection. Consented when? A state born in 1965 was bound by rules already old. It never voted. So either the claim is a fiction, or law here is arriving by some route other than enactment. Which would mean a rule can come into existence without anyone making it. How would that even work?

Try building one from nothing. Suppose no rule exists about a strip of sea. One state starts behaving in a particular way — policing fishing out to a given distance. What is that, legally? Nothing yet. It is a fact about a state's conduct.

Now a second state does the same. A third. Others do not protest, or begin doing it themselves. At some point people start saying that this is what states must do rather than what they happen to do. Ask the awkward question: on which day did that change?

There is no such day, and that is the first real finding. The rule does not switch on. It condenses. What the law does instead of naming a moment is name two ingredients, and check whether enough of both are present. The formulation in Article 38 of the International Court of Justice's Statute is "international custom, as evidence of a general practice accepted as law", and lawyers read that as two elements.

The first is state practice — what states actually do and say: legislation, diplomatic protests, naval conduct, votes, published legal opinions. The Court held in the North Sea Continental Shelf cases (1969) that this practice must be extensive and virtually uniform, must particularly include the states most affected, though it need not run for very long if it is dense enough.

The second is opinio juris — the sense that the practice is done out of legal obligation, not habit, courtesy or convenience. Notice why this element cannot be dropped. States exchange gifts on national days without anyone thinking a state that stops has breached international law. Practice alone cannot distinguish the rule from the routine.

And here is the difficulty everyone runs into next, so let us not dodge it. Opinio juris asks states to act because they believe a rule already binds them. But the belief is what generates the rule. So the earliest adopters must have believed something that was not yet true. That circularity is real, it is not a trick of exposition, and it has no agreed solution — theorists variously call it a useful legal fiction, treat early practice as a claim rather than a belief, or argue that the two elements are less separable in practice than the doctrine pretends.

What of the state that refuses? There is a narrow exit: a state that objects clearly and consistently from the time the rule is forming — a persistent objector — is generally held not to be bound by it. Notice the timing condition. Object early and you are outside the rule; object once it has crystallised and you are simply in breach. And even that exit closes for peremptory norms, the small set of rules from which no derogation is permitted at all.

Three clarifications.

First, codification is not creation. When a treaty writes a customary rule down, the customary rule does not disappear into it; it continues to run in parallel and binds non-parties. This matters practically: a state outside a convention can still be held to the customary version of the same rule.

Second, do not confuse "general" with "unanimous". The test has never required every state to participate. It requires enough, in the right places, without significant contrary practice — which is why the conduct of maritime powers weighs more heavily on a law-of-the-sea question than the conduct of landlocked states.

Third, and worth saying plainly, this process is far less determinate than the vocabulary suggests. Identifying custom means an advocate or a court assembling evidence of practice and belief and judging that it suffices. The International Law Commission produced a set of conclusions on how to do that identification in 2018, which is a good sign of how much practical uncertainty needed managing. Where a rule sits near the boundary, reasonable lawyers disagree about whether it exists at all — and the honest answer is that the question is decided, not discovered.

• Jus cogens — how a subset of custom becomes non-derogable, and who gets to say which rules qualify.
• Whether the persistent objector doctrine has ever actually saved a state in practice, or functions mainly as a theoretical safety valve.
• Instant custom, and the argument over whether a single dense burst of practice can produce a rule in a few years.

Every term the collection defines is gathered in the glossary.