A friend told me, gently, over coffee in Doha last spring, that my LLM was now of mostly historical interest. He meant it kindly. He was not the first person to say it and he will not I think be the last. The idea when it is made to me (and it has been made to me, in slightly different formulations, by approximately every politically-engaged friend I have acquired since October 2023) goes something like this. The Israelis have committed an innumerable list of egregious breaches of international human rights law, and the Americans have struck Iran’s nuclear facilities, and Trump has spoken openly about acquiring Greenland by means up to and including coercion, and the Head of State of Cuba was forcefully extracted from Caracas in a blatant disregard of the fundemental tenets of international law. The international legal order, my friend tells me, is dead, or was always a fiction, or has been revealed as the colonial instrument it always was, take your pick. He is sympathetic. He knows I spent a year reading the Vienna Convention and the ICJ Reports and Brownlie’s textbook in a small room in London. He just thinks the world has moved on and that I should perhaps consider doing the same.

I want to explain here why I think he is wrong and why I find the current iteration of the argument irritating. The irritation is partly an irritation with my own former self, who at twenty-two held something close to my friend’s view and would have made it with considerably more gusto. My dabbling in philosophy for the past half decade helped me come to the appreciation of a particular distinction which people I largely agree with politically fail to make it or find to be a trivial one. The distinction is between two kinds of questions; an empirical one and a conceptual one. The empirical one is on how the world is arranged at the moment, on who has the power to do what to whom, on what the patterns of compliance and breach look like, what the prevailing settlement about what counts as legitimate violence and where. This is a question about what Carl Schmitt called the nomos of the earth, i.e., the underlying spatial and political order that makes any given pattern of organised violence intelligible as such. Schmitt’s claim, which I think is roughly correct as a descriptive matter, is that legal norms presuppose a normal situation against which legality and illegality can be distinguished. When the nomos shifts, the legal categories created on top of it begin to misfire. They do not necessarily disappear but their grip loosens, their predictions fail, and the people who use them begin to sound somewhat like medieval scholastics arguing about angels in a building that is on fire.

I think the nomos is in fact shifting, and I think this is what people who deny international law as Real Law are responding to. The post-1945 settlement wherein a particular American-led configuration of power considered certain rules as binding (on others, mostly) and certain violations as scandalous (when committed by enemies, mostly) and certain definitions of war and territory as fixed (until it became inconvenient), is being contested in ways it has not been contested since their inception. The events of the past year or so that my friend listed are indeed part of a larger renegotiation about what counts as a legitimate use of force and a sovereign decision, and where the boundaries of permissible state behaviour now is. It would be quite silly to pretend otherwise and the international lawyers I respect have stopped doing so. The nomos is changing. This much is conceded.

The second question is conceptual, which is thus: granted all of this, is international law law? My friend, and many of his political fellow-travellers, equivocate the first question and the second. The rules are not being enforced; the rules are being applied selectively; the rules were always a tool of the powerful; therefore the rules are not really rules, and invoking them is participating in a fiction that merely flatters the powerful by dignifying their preferences with the language of legality. I want to insist that this is a non sequitur, and that the non sequitur is consequential, and that recognising it as such does not require running apologetics for any of the empirical claims that precede it.

The reason it is a non sequitur is that the question “is X a law?” and the question “is X a good law, evenly applied?” are different questions, and one cannot answer the first by merely gesticulating at the second. The Antebellum South had laws. British India had laws. Nazi Germany had laws. None of these were just laws when indexed to my standards; several were indeed monstrous; all were, in the relevant conceptual sense, legal systems, with primary rules and secondary rules and very easily recognisable practices of legal reasoning conducted by recognisable legal officials, and the fact that the rules they enforced were vile does not retroactively make them not rules. In other words, the moral evaluation of a legal system does not determine its ontological status. If the international legal order is selectively enforced, colonial in origin, and currently being violated with impunity by its principal creators, all of that may be true and all of it is orthognal to the question of whether what is being violated is law.

The sceptic would point out that, sure, the Antebellum South, British India, Nazi Germany, and so on, had systems of law nevermind their normative status, nevertheless those are cases where the unjust rules were brutally and systematically enforced. The slave codes worked. The Nuremberg Laws worked. They were morally monstrous precisely because they worked. The point, the sceptic goes on to argue, is not that internationa law is wicked but effective; it’s that international law is selectively applied to the point of incoherence. The rejoinder to the sceptic is: most of international law works pretty well. ICAO standards keep planes from colliding over the Atlantic, the Universal Postal Union routes mail, diplomatic immunity holds, extradition treaties function, WTO dispute mechanisms still produce compliance even from states that grumble, innumerable shipping laws continue to organise delivering your stuff across oceans. The cases the sceptic cites, namely Gaza, Iran, Greenland-talk, the Maduro extraction etc., are the headline cases precisely because they are unusual enough to be scandalous. If they were the norm, they wouldn’t be scandalous. The sceptic’s evidence depends on the background regularity she is denying.

I am aware, while writing this, that I sound like I am making a small and pedantic point. I am not. The reason it is not pedantic is that the denial of ontological status, i.e., the equivication from “this law is being violated” to “this is not really law”, has a specific political consequence insofar as it disables the vocabulary in which the violations can be most precisely characaterised. If genocide is not a legal category but ‘only’ a moral one, then the indictment of those committing it loses the particular kind of force that legal indictment carries and it becomes one moral complaint among others, addressed to an audience that may or may not share the moralist’s premises. If the prohibition on the acquisition of territory by force is not a legal rule but ‘merely’ an ideological preference, then there is no asymmetry between Russia in Crimea and the United States in Greenland; both are exercises of power, and the language of law has nothing distinctive to say about either. The sceptic believes she is exposing the law as merely ideology, and that doing so is liberatory. But what she is actually doing is removing one of the few instruments her own side possesses for distinguishing between behaviours she opposes and behaviours she would like to oppose more effectively. The instrument is imperfect. It is also the only one of its kind currently in service and discarding it because it has been misused is the political equivalent of refusing to use a hammer because somebody once used a hammer to commit a murder.

Here I should say something about the philosophical literature, briefly, because the conceptual question has an actual history and I do not want to pretend I am inventing the answer. The view that international law is not really law because it lacks a sovereign with a monopoly on coercion is John Austin’s, from 1832, and it is a view that has been embarrassed by almost every serious legal theorist who has come after him, including ones who otherwise agreed with him about a great deal. Austin’s command theory cannot account for vast portions of municipal law (constitutional limits, customary law, broad standards, the ordinary operation of contract and property) and taking it as the criterion of legality leaves us with the result that most of what English lawyers do on most days is also not really law. This is not a serious position.

H. L. A. Hart’s view, which is the one most legal theorists now work with in some form, is more accommodating. Law on Hart’s account is a union of primary rules (which impose obligations) and secondary rules (which specify how primary rules are made, identified and changed). Hart himself thought international law had primary rules but lacked secondary rules and was therefore a kind of pre-legal system, on the model of a primitive society without a constitution. I think Hart was wrong about this, and I think the wrongness was a function of his writing in 1961, before the Vienna Convention on the Law of Treaties had codified the rules of treaty-making in a manner that any practising international lawyer now relies on as routinely as a domestic lawyer relies on her jurisdiction’s rules of evidence. The Vienna Convention is a paradigm secondary rule. So is the doctrine of jus cogens. So is the practice, settled in the ICJ and reflected across every major arbitral tribunal, of identifying customary international law through state practice and opinio juris. These are norms about how to identify the norms about how to behave, which is precisely what Hart meant by secondary rules. International law has them. It has had them, in a fully recognisable form, for at least sixty years, and quite probably longer.

What is true, and what Hart’s framework helps to make precise, is that the configuration of international law’s secondary rules is genuinely different from the configuration of municipal law’s. There is no global legislature and no international police force. The International Court of Justice has compulsory jurisdiction only over states that have accepted it, and even then only in a circumscribed way. Compliance is largely decentralised through a mixture of self-interest, reputational concern, treaty-based monitoring and the occasional spectacular sanction. None of this, however, is evidence that international law is not law. It is evidence that international law is a horizontal legal system between formally equal sovereigns rather than a vertical legal system between a sovereign and her subjects, and that the institutional shape appropriate to the former is not the institutional shape appropriate to the latter. The municipal model is not the paradigm of legality and the international system is not a deficient approximation. There is no such paradigm. There are different kinds of legal order suited to different kinds of community, and the international one is suited to the community of states in something like the way the domestic one is suited to the community of citizens. Each has its characteristic pathologies. Neither is a degraded version of the other.

The thing that ultimately convinced me is the phenomenology of how international legal practice actually proceeds. States that intend to violate a rule almost never say so. They produce, instead, legal arguments, sometimes terrible ones, sometimes ingenious ones, sometimes ones that posterity will judge harshly, explaining why what they are doing is permitted, or required, or falls within an exception, or does not properly count as the act it superficially resembles. Israel, in defending its conduct in Gaza, has not said we acknowledge that the laws of armed conflict prohibit what we are doing and we propose to do it anyway. It has produced an enormous and continuous output of legal argument (about distinction, about proportionality, about the law of occupation, about the meaning of genocidal intent, about the status of Hamas as a non-state actor, etc.) most of which I find unpersuasive, some of which I find offensive, and all of which is recognisably legal argument of the kind that lawyers in any system would recognise as such. The same is true of the United States in defending its strikes on Iran, and of every other major recent violation. This is a remarkable fact about the world. It would be inexplicable if international law were not law. It is exactly what one would expect to find in a legal system that was being violated under stress, by powerful actors, in a period when the underlying nomos was contested.

The sceptic can grant this and reply that the legal-argument-output is mere pretext, a fig-leaf draped over decisions taken for entirely non-legal reasons, and that the ubiquity of the fig-leaf shows how thoroughly the law has become a tool of legitimation rather than a constraint. I think this reply concedes more than the sceptic realises. To say that powerful actors find it necessary to dress their conduct in legal language is to say that legal language has, in their judgment, a kind of authority that no other vocabulary in the international system possesses. They do not dress their conduct in moral language, or in religious language, or in the language of historical destiny. They dress it specifically in legal language, because legal language is the one their counterparts and audiences take to be binding in a way the alternatives are not. This is not the behaviour of actors who think the law does not exist. It is the behaviour of actors who think the law does exist and that they are required, even when violating it, to participate in its public forms. The fig-leaf is not nothing. The fig-leaf is, in fact, considerable evidence for exactly the proposition the sceptic wants to deny.

So: Hart’s secondary rules work, violators still play the game and the unglamorous bulk of international law works in a perfectly routine way. International law is law.

I want to end by saying something about the political stakes of getting this right because I do not think this is a debate that can be conducted entirely within the confines of the ivory towers. The international legal order is, in fact, in a very bad trajectory. The nomos as I said is shifting, and the new settlement, if there is one, may be considerably worse than the old one. The most serious threat to international law is not the conceptual sceptic who thinks it was always a fiction; it is the actual, on-the-ground erosion of the institutions and practices that have given it whatever traction it has had. A serious response to this requires defending the legal vocabulary against its violators, not joining the violators in dismissing the vocabulary as illusory. Iran was pushed into an illegal war, but it has also violated international law several times; and excusing those violations by proclaiming that ‘international law is now dead’ is a grave mistake. The Palestinians are not better served by being told that the laws their occupiers are breaking were never really laws. They are better served, I think, by the patient, unfashionable and often thankless work of insisting that the laws are laws, that the breaches are breaches, and that the cost of those breaches must be paid in the only currency international law has historically possessed, which is the slow accumulation of cases, judgments, opinions, sanctions and shifts in the structure of recognition that, over time, change what powerful states can do without consequence. This work is going on now. The ICJ proceedings on genocide are part of it. The arrest warrants are part of it. The various national-level prosecutions under universal jurisdiction are part of it. None of this is a substitute for politics. All of it requires, as a precondition, that the people doing it not be told by their political allies that they are wasting their time on a fiction.

I know that this argument is not going to convince my friend. He will read it and say I have made a clever distinction that does not survive contact with the situation in Rafah, and I will not have a good answer, because the situation in Rafah does not survive contact with most of the distinctions that civilisation has produced. But the question of what we say afterwards, whether we say the law was a lie or the law was broken, is a question about what tools we will still have available when this is over, and on that narrower question I have a clear view. The degree was not useless. The degree was the slow acquisition of a vocabulary that, on the worst days, is the only one that gives me anything precise to say about what is happening. I would rather have it than not have it. I would rather my friends had it too.