Words Of No Power.

And the power of no words.

These essays will always be free, but you can continue to support my work by liking and commenting, and most of all by passing the essays on to others, and passing the links to other sites that you frequent. If you would like to take out a paid subscription I won’t stand in your way, (I’d be very honoured in fact) but I can’t promise you anything in return except a warm feeling of virtue. A special thank-you to those who have taken out paid subscriptions recently.

I have also set up a Buy Me A Coffee page, which you can find here.☕️ Thanks to all of those who have recently contributed.

And as always, thanks to others who tirelessly supply translation in their languages. Maria José Tormo is posting Spanish translations on her site here, and Marco Zeloni is also posting Italian translations on a site here, and Italia e il Mondo is posting them here. I am always grateful to those who post occasional translations and summaries into other languages just as long as you credit the original and let me know. And now:

**************************************

I have seen several worried articles recently, suggesting that because the current New START nuclear weapons treaty between Russia and the United States expires in a couple of weeks time, and is unlikely to be renewed, then the world is entering a dangerous new era.

The assumption seems to be that the lapsing of the Treaty will somehow release pent-up forces for nuclear escalation which have only been held in check so far by words on paper, and that, almost automatically, nuclear stockpiles will start to increase again. Put like that, perhaps, the argument does seem a little curious, but it exemplifies very directly what I see as a fundamental misunderstanding of the relationship between acts and texts in international politics, and one worth trying to correct here. Briefly, in international politics, words are generally a consequence of actions, and not the other way round. Texts exist to record underlying agreements (and sometimes differences) not to impose them. They are, if you like, Polaroid photographs of the current understandings, differences and power relationships between those who draft and sign them.

Yet this reality goes against a lot of inherited cultural assumptions about the power of words. For millennia, Jews and Christians have read in the first chapter of Genesis that God said “let there be light,” and light came into existence. In Islam God said “be” and the universe was. (There isn’t space for other creation myths today, sorry.) Subsequently, Adam gives names to all living things in the Garden of Eden. The Gospel of John famously proclaims that “in the beginning was the word (logos) and the word was with God and the word was God.” And later the word “was made flesh” in the form of Jesus. Now even allowing for the wide range of meanings of the word logos, there is a clear connection with the act of speaking here, and this is how Christians have always interpreted the sentence (the Latin Vulgate likewise has verbum). God spoke and something happened.

In fact, all societies at all times have assumed that speech can have a causative effect. Curses were intended to hurt others, after all. Most religions have the concept of “words of power:” indeed, prayers of all sorts are essentially based on the hope and expectation that supernatural forces will hear you and do what you ask. For centuries, Christians have repeated the XXIII Psalm in times of danger and difficulty. From Babylonian times until the present day, ships have been launched with naming and protection rituals designed to protect them from the dangers of the sea. And so on.

In some cases, words themselves were believed to be possessed of inherent power. Jewish tradition held that the real name of God (written YHVH) could not be pronounced, or the world would come to an end. (Yes, Arthur C. Clarke produced an ingenious variation on this idea, based on Tibetan mythology.) This was a reflection of the traditional belief in an underlying order of the universe, which could be expressed in words and, especially, in numbers. (Greek and Hebrew alphabets, remember, used letters as numbers, and so every word had a hidden numerical value.) The idea of individual words being powerful permeates western (and other) culture to the present day, in the form of spells and mystical chants, and in beliefs such as that placing a clay tablet with the name of God in the mouth of a golem to bring it to life (still a trope in modern culture, apparently.) The invention of writing, which was at the beginning of course itself an esoteric art, reinforced all these ideas

The modern equivalent of these beliefs and practices is political propaganda and advertising, both of which, curiously, are generally believed to be dangerously effective because of their quasi-magical use of words, although there’s often no real proof of this. In particular, the propaganda of Dr Goebbels is frequently credited with an almost magical effect on the German population, because of its use of myth and symbolism and its incantatory methods, yet Gestapo records of the time show how little effect it had on the thinking of ordinary people. And today, of course, western governments work themselves into a frenzy over foreign “disinformation,” which they believe will somehow exert magical powers over their own populations. Magical scrolls and grimoires with spells to bring about wealth, power or happiness are still current today, not least in popular culture, and their opposite (curse tablets) seem to have originated in the Graeco-Roman world, and are still found in some parts of the world even now.

Before swerving back to more practical issues of politics, it’s worth just acknowledging that there are cases even today where pronouncing words actually has real life consequences. What are known as “speech acts,” are cases where, unsurprisingly, words themselves equate to, or bring about, actual changes in the real world. Thus, “I resign as Prime Minister” is a speech act. Similarly a jury that pronounces a prisoner guilty of murder changes that person’s objective status, and the judge who sentences them to life imprisonment does so by means of a speech act. Notice that, twenty years later, a judge saying “whoops we got that wrong, you are free to go” is another. Neither, of course, necessarily means the victim is either innocent or guilty in real life. (Speech acts are a complicated subject and here is a good short guide.) Finally, of course there are cases where certain utterances are at least treated as speech acts: traditionally, excommunication from the Catholic Church was believed to mean that the excommunicated person would finish in Hell.

But outside these very specialised examples, alleged speech acts often have little power. Many things that look like speech acts (“I promise …”) are in fact not, because they lack any element of constraint or automaticity. Likewise, many resounding statements are purely performative (“we will force Russia to the conference table!”) even if they look superficially like speech acts. At which point the connection between religious mysticism and power politics may become more evident. Many political statements are in fact purely performative (“We must stop Putin winning!”) even though they are intended as speech acts. Indeed, the consequences of mistaking one for the other are a major theme of this essay.

In order for texts of any kind to have a practical effect, they require one of two things, and ideally both. The first is an enforcement mechanism, since no text intended to regulate behaviour will ever be automatically accepted by everyone. Thus, the great English legal philosopher John Austin, argued in his Command Theory of Law that for a text to qualify as a Law it had to be enforceable. This theory has never been popular among legal theorists since it would take their bread and butter away, and is in especially bad odour today, but it’s hard to argue with. As we’ll see, this is an especial problem with international agreements, which are often treated as a kind of law. Alternatively, and ideally additionally, a Law should reflect a broad underlying normative consensus about rights and wrongs and justice, in which case people will obey the law because it corresponds to their own beliefs. This is why the current movement from laws reflecting norms to laws trying to impose norms is so dangerous.

Now, many texts and groups of texts do have enforcement mechanisms. Contract law can be enforced in courts, professional conduct can be evaluated according to agreed texts and penalties exacted for non-compliance, using the word “he” instead of “they” can get you dismissed from your job. But even then, as I’ve often emphasised, it’s actually precisely the unwritten codes of conduct and behaviour that are the most powerful and useful, and attempts to reduce them to texts with coercive powers almost always fail. That is just as true of international relations as of Humanities Departments of Universities.

How then do international political texts (of which treaties are one type) fit into this scheme? The first thing to stress is that such texts vary enormously in their scope and objectives, and must not be confused. At one extreme, the Prime Minister reports to Parliament on bilateral talks, or his their office releases a short summary. Or maybe after a bilateral meeting the two sides issue an agreed statement listing the points that they discussed. Or there is an agreed summary of oral decisions and promises (the often-cited Minsk documents are a good example of this.) Or there may be a quite elaborate joint statement by two or more participants, covering points of agreement, and possible further action, and usually prepared in advance. At the other end of a long spectrum is an actual formal treaty, usually taking months, sometimes years, to negotiate.

The distinction between these types of documents is generally supposed to be that a Treaty (for technical reasons some are called Conventions or Agreements) is “legally binding.” It is certainly true that treaty texts look different. There is generally a list of nations signing, followed by the magic words “have agreed as follows.” And the language of the treaty is generally coercive: not “will seek to,” for example, but “agree to.” All of this, together with the formality and the breathless coverage of important treaties by legal scholars, gives the impression of a special and important type of document, which superficially resembles a legal text. This isn’t entirely untrue, but it’s subject to a host of practical qualifications.

The first is that by definition only countries that have signed a treaty are bound by it. Moreover, most treaties contain withdrawal clauses, and indeed the consensus is that, in fact, you can withdraw from any treaty unless there is a specific clause forbidding it. This means in effect that treaties apply until they don’t. But the converse is also true: states may well decide to continue to observe the provisions of a treaty that has lapsed, for political reasons, because they think it is practically useful, or both. And here we notice, again, that what counts is not documents but political reality, which documents to some extent reflect. Moreover, treaties not only have to be signed by governments, but also ratified, usually by parliaments. Treaties can be stuck for years, or even forever, in the limbo between the two. And some states, especially small ones, consider that they have paid off debts and earned points from the international community simply by signing. Implementation can come much later, if at all.

The second obvious point is that treaties cannot be enforced. This does not mean that states will not suffer consequences if they violate treaties, but it does mean that there is no “international justice” mechanism that can enforce behaviour as domestic justice mechanisms do. (The ICJ only handles differences between states and only then by agreement.) To that extent, and following John Austin, it’s helpful to think of treaty documents not as legal texts, but as a kind of agreed compromise summary of what states are prepared to say, and do on a particular subject, albeit often using a vocabulary and structure that resemble genuine legal texts. Thus, when we say that a treaty is “legally binding,” we really mean that states have agreed to behave as though it is, until they no longer wish to. It’s worth adding that a number of treaties actually require signatories to pass domestic legislation to implement them in their own countries, and of course these are legally enforceable.

In practical terms, governments generally do abide by the provisions of treaties because it’s in their interests to do so. A classic is the Vienna Convention on Diplomatic Relations of 1961, which governs the mechanics of how international relations work and provides protection and safeguards. In the end, it’s in every state’s interest to observe it. Moreover, many governments do actually take treaty obligations very seriously in practice, and incorporate them into policy-making. This is becoming a real problem in the EU, where states are having to deal with unintended consequences of treaties signed decades ago, as interpreted today by European judges and bureaucrats, against which they have no appeal. Thus, treaties are observed most of the time, and help to bring a degree of order and coherence to international relations. Nonetheless, it’s important not to give treaties a status they don’t deserve. At best they are a writing down of normative aspirations and existing practice: they are not speech acts.

Things start to get difficult when treaty texts are subject to interpretation, especially by judges, and treated as though they were actually legal texts and could be analysed as such. Now of course this can be problem even with domestic law. In many countries, and especially with the profusion of new laws on every conceivable subject, laws can be badly drafted and incoherent, and indeed contradict other laws (including European and international law), of which the drafters were unaware. And since such laws are often the result of complex political bargains between groups, the clarity that judges hope to find about the meaning of legal texts was often never there in the first place.

At the international level it’s considerably worse, because of the number of actors and issues that have to be taken into account. Moreover, in many cases disagreements are essentially political rather than legal. In turn, this is because treaties are political documents, and reflect what the signatories are prepared to publicly commit themselves to do. A topical and much discussed example is Article V of the Washington Treaty, which was often qualified as being about “mutual defence,” and providing a “security guarantee;” It was only when people started to actually read the Treaty, after the beginning of the war in Ukraine, that they realised that the Article contradicts itself. Yes it does indeed say, in the phrase everyone remembered that “an attack on one is an attack on all,” but it also makes it clear that it is for individual states to decide how to react, if at all. So if an imaginary Court had to decide whether its government had a legal duty to support an ally that had been attacked, the only possible answer would be “not really, although the presumption is that it would, in some form or other, and taking account of the geographical limits in Article VI” (which of course no-one reads.) Judges, in other words, have nothing to contribute, but it’s not their fault.

The explanation for the broken-backed nature of Article V is, of course, political. The Article was supposed to do three things at once: reassure nervous Western European populations that the US was still interested in European security at a time of fear and uncertainty, make the Soviet Union more careful in its dealings with Western Europe, and reassure isolationists in Washington that the United States would not be dragged into another European war so soon after 1945. Different messages for different audiences, but a single text. This is common in politics, but for that very reason there can be no definitive analysis of what Article V “means.” In any event, if there actually had been a crisis, its handling within NATO would have had nothing to do with the contents of Article V.

Just as domestically there’s the bloke down the pub complaining that “there ought to be a law against it,” so internationally activists of all sorts are continually demanding not just treaties to cover this or that, but the application of treaties to areas they were never intended for, and their use in ways that are best described as “creative.” Sometimes this seems be a product of sheer fantasy, as in the suggestion that there is a doctrine in international law of “pre-emptive attack.” There is no such doctrine and no such right, but I have seen Very Serious People debating whether, for example, another Israeli attack on Iran could somehow be justified by this non-existent doctrine. What they are actually reflecting I suspect, is the fact that, in spite of what the United Nations Charter says, aggressive acts against other countries have been frequent since 1945 and are continuing, thus opening a massive gap—here as elsewhere—between theory and practice.

In that connection, there’s an amorphous concept called “customary international law,” which has sparked a huge literature, largely trying to decide what it is. In theory it is often described as “customs having the force of law,” or to be more honest “customs that states agree to pretend have legal force” (inasmuch, of course, as any international law has force.) More practically, it would be defined as “stuff we all think it’s right to do, mixed with stuff we can get away with,” even if there’s no explicit legal basis. Whilst it would be controversial to say that customary international law permits aggressive wars (as of course it did until 1945), I think, it’s nevertheless true, insofar as that concept has any meaning at all. But then “international law develops,” as one weary government lawyer said to me many years ago.

As I have pointed out frequently, a fundamental problem for the liberal-international doctrine that largely governs the international system is that the world is not organised as it ideally should be. Bad things happen, good things fail to happen, bad people prosper, bad regimes endure. As with our man down the pub, the Professional and Managerial Caste (the PMC) wants Something to be Done. There should be a law against it. If there isn’t one, we should nonetheless act as though there were. (As one US government official said in my hearing in 1999, “if international law stops us from bombing Kosovo, there must something wrong with international law.”)

This is silly, but perhaps not quite so irrational as it sounds. The PMC, and its shock-troops in the NGOs and the Human Rights industry have a normative, emotional view of the world, mostly unencumbered by facts and experience. So when the world does not function according to their normative aspirations, something must be done to correct it. The law is, of course, the basic PMC answer to every problem, and it regards laws, including treaties, as speech acts. Once promulgated, will by themselves resolve the problem since they will necessarily command obedience. Conventions on the Climate, for example, are hailed as “successes” or “failures” depending on what documents are produced, whereas of course all that matters is what governments actually do.

The assumption that documents lead to actions, rather than the reverse, is thus one of the biggest problems with the PMC mentality. Its un-finest hour was probably the conclusion of the Rome Conference on the Statute of the International Criminal Court, hailed by NGOs as “the end of impunity” and a new beginning for the human race. No, I’m not exaggerating: the atmosphere among NGOs in Rome (who to their chagrin were not allowed into the negotiations) resembled that of a religious revival. That said, we should remember that for the PMC the law is a moral discourse rather than just texts: it’s a way of talking about the world, and a vocabulary in which to express its desires and hatreds, and so it’s the sprit, not the letter, that matters. It wasn’t really surprising, therefore, that many who loudly cheered the bombing of Kosovo in 1999 nonetheless opposed the invasion of Iraq in 2003, and tied themselves in conceptual knots trying unsuccessfully to argue that the two were legally different.

Nowhere is this disconnect more obvious than in treaties dealing with subjects which arouse strong emotions and political controversy, and on which I’ll concentrate here. There are treaties that try to deal with bad things in the world, and I’ll discuss some, partly because they are topical, partly because they are a good example of my general thesis, and partly because I have some small experience of the problems they may cause. Several of them have the interesting nuance that, whilst it’s states that are parties to the treaty, the text itself is actually about what ordinary people might have done.

I’ll take the currently fashionable Genocide Convention as my main example, since it is widely cited, if little read. The text could be described as a political success (it was agreed) but a conceptual shambles (because of the process of agreement itself.) This outcome is actually quite common. The text is actually relatively short, as often the case with difficult and contentious subjects, and very largely concerned with technical legal issues. Its operational content can reasonably be summarised as “we agree that genocide is a bad thing and we will do what we can to prevent and punish it.” The operational and definitional elements between them take up only half a page of the three-and-a-half pages of the official text. The rest is bureaucracy.

Those who have never read the Convention are often surprised to learn that genocide does not mean “killing large numbers of people,” and so they sometimes argue that it should, and if it doesn’t then the text needs to be changed. In fact, the definition is complex and hard to understand, as is often the case with compromise texts fought over in different languages, and where there are no depths, only varieties of shallows. To begin with, genocide is not an act in itself, but an interpretation of one. Just as domestic criminal codes distinguish between voluntary and involuntary killing, or even criminal acts and genuine mistakes or carelessness, so the Convention requires “intent” to “destroy in whole or in part, a national, ethnical, racial or religious group, as such” before genocide may be said to exist. And “destruction” is limited to five methods: killing (obviously), causing harm, creating impossible conditions of life, preventing births, and transferring children to other groups.

Without labouring the point too much, nobody really knows what this all means. But that doesn’t actually matter—or didn’t in 1948—because the primary purpose of negotiating the treaty had been met, and governments could say that genocide would now be prevented and punished. There was no question of operationalising it, and no obvious way in which that could be done, beyond the normal application of criminal law in the countries of the signatories. But of course episodes of mass violence tend to take place precisely in areas where the law cannot be applied. This isn’t cynicism, by the way, or at least not only that: it’s how the system works, and this is one of the earliest modern examples of magical thinking producing assumed speech acts. However, I’ll just point out some of the more obvious difficulties that prevented practical application, observing in passing that books have been written and conferences held to try to decide what some of the text means, at least in theory.

Take “intent,” which is the keystone of the whole intellectual edifice. How would you know? At the time, the world was still under the spell of the Second World War and in fear of yet another devastating conflict. It was therefore convenient to both East and West to stress the unique evil of the Nazi regime, to argue that a page in history had definitively been turned and, in the case of the Soviet Union, to efface memories of its collaboration with the Nazis until 1941. In the immediate post-war era, the idea of some kind of long-term Master Plan implemented by Germany from 1933, an Intent which other countries had failed to understand and prevent, was widely credited. For political reasons it dominated at the time, and still has its political supporters today. In reality, Nazi Germany was institutional and political chaos, full of competing wild schemes for everything, few of which were ever implemented and none in full. The Nazis very much made it up as they went along. But of course if there had been a Master Plan, those who had not spotted it could be held politically responsible for the immense suffering of the War, and a satisfying morality tale written for the edification of future generations. (This is the ultimate origin of all those demands for action to “stop” the latest western hate-figure, most recently Mr Putin.)

The terrible exactions of the Nazis, and in particular those directed against the Jews were an important part of the Master Plan thesis: it seemed incredible that so many could have been slaughtered without some great Plan, and anyway, everybody knew that the Germans were racially programmed to be meticulous planners. There was thus much excitement about the discovery after the war of the portentously-named Wannsee Protocol, believed to The Plan. Yet on examination the document turns out to be a briefing by Reinhard Heydrich on a project to move Jews from Poland and elsewhere in Europe and work them to death as slaves in the newly-conquered territories. The plan was never implemented because the military situation turned against Germany. It can’t seriously be doubted that the Nazis intended to wage a war of extermination in the East, because there are lot of documents that said so, notably the infamous General Plan East, which foresaw the killing of sixty million Jews, Slavs and others as part of a colonisation programme, but was never really implemented. And they certainly did a lot of exterminating. But in spite of the horror, much of the Nazi state was at cross-purposes and daggers drawn with itself most of the time: modern historians, for example, see even the plan to exterminate the Jews developing in a haphazard fashion over a period of years.

I have laboured this point a little, gruesome though it is, because the entire edifice of the genocide discourse rests, not on the general intention to do harm, however great, but on the existence of a precise plan to do between one and five specific things to “destroy” one of four specified groups. The trouble is that “national, ethnical, racial and religious” groups have no objective existence, so it’s impossible to know whether victims belong to them. The idea of a “national” group is the strangest. Are the Belgians a national group? Are Czechs and Slovaks one group of two? The answer may be that in some languages (especially Slavic ones) the same word means both “nation” and “people,”and this may be a clue to why the term was adopted. An “ethnical” (we would say “ethnic”) group is probably what was meant by “nation,” but these days ethnicity is what scholars call a “contested concept,” and is surrounded by so many caveats as to be largely useless. “Racial” groups are generally agreed not to exist in the sense that the early twentieth century racial “science” behind the Convention thought they did. And “religious” groups, as opposed to cultural affiliations, tradition and degrees of observance, are very hard to define.

At the time, none of this mattered. After all, it was unlikely that anyone would ever be put on trial for intending to destroy such groups. That said, may of the early backers of the Convention were right-wing nationalist exiles from Eastern Europe, and it was clear that Soviet actions in that region during the horrific years of generalised revenge after 1945, including widespread killing and forced movements of populations, as well as persecution of the Catholic Church, could be used for political accusations of “genocide.” Thus, some of the categories become more explicable (although of course other groups promoted the Convention as well). And in that chaotic era, when the West still hoped that the Soviets could be driven out of the region once more, some could always fantasise about criminal trials of Soviet officials for genocide: a kind of Nuremberg 2,0.

These groups may be destroyed “in whole or in part.” How big is a “part?” Nobody knows and it is impossible to decide. And these groups were targeted “as such.” (En tant que tel in French is a little clearer.) Nobody is sure what that really means either, if it means anything. But again, whilst different answers have been proposed and entire books written on these questions, there’s no “real” meaning to unearth, because the text, like all such texts, has no real meaning: it is simply where the negotiators wound up when they were able to agree.

None of this would really have mattered—the problem of genocide had been “solved”after all—had not attempts been made from the 1990s to put individuals on trial for genocide, first after propaganda and media reports from Bosnia, then as a result of the terrible events in Rwanda in 1994. The problems were obvious. A historian can infer “intent”perhaps from documents and government policies, but proving to a criminal standard of proof what was in somebody’s mind, and showing that the same person was at least indirectly responsible for specified horrors was, in the end, too difficult, so it was sidestepped. As a prosecutor in the Rwanda Tribunal put it to me, the argument was that the scale of the killings in 1994 was such that they must have been planned. So what needed to be proved could simply be assumed. And of course without intent, you have no guilt.

In the end, Courts effectively rewrote the Convention to make convictions possible. The Rwanda Judges, full of western assumptions about African tribalism and savagery, handed down convictions for genocide against the “Tutsi ethnic group,” until embarrassed specialists pointed out that the Tutsi/Hutu distinction was social and economic, akin to the Indian caste system, and not ethnic at all. Never mind, the Tribunal decided that these ethnic groups didn’t actually have to exist, as long as people thought they did. (So much for “as such.” ) And the conviction of General Radislav Krstic over the Srebrenia killings in 1995 surprised everyone—not least the prosecutors—because the victims were adult males who had fled the town, and around half were soldiers of the Muslim 28th Division. Reading the tortured prose of the judgements, especially of the Appeals Chamber, with their newly-minted legal theories, it was clear that in the end the judges were primarily trying to convince themselves, in order to make a political point about the undisputed awfulness of the incident.

There is nothing especially enlightening in watching the eschatological fervour of liberal good-thinkers turn to anger and vindictiveness when the anticipated new world fails to arrive. But it is nonetheless a substantial factor in politics these days and has to be allowed for. The belief in Words of Power, the faith that a document alone can change history, and if it doesn’t it’s the fault of somebody who must be identified and punished, is deeply ingrained in PMC thinking. (There isn’t space to discuss the sometimes-terrible consequences of flawed or premature peace treaties here, although I have done so elsewhere. )

Sometimes the deception is almost comic. The 1997 Ottawa Treaty outlawed (for signatories), the “Use, Stockpiling, Production and Transfer of Anti-Personnel Mines” and demanded their “Destruction.” Now anyone who remembers that far back will recall that these words have nothing to do with the actual problem, which was that for several decades the Soviet Union and China supplied large stocks of mines to liberation movements and certain governments, mostly in Africa, and that these were laid promiscuously with few or no records of where they were. The toll on the civilian population in certain areas was awful. So how did Barbados and Ireland signing the Treaty help the problems of post-conflict Africa? Come to that, how could a country like Sierra Leone (where the war was still in progress) hope to carry out the duties of the Convention, especially destruction, when it had neither the resources nor the money to do so? But again, that is to misunderstand the purpose of the Treaty which was to pretend that a problem that has no real solution had nonetheless been Dealt With. And of course, a decade later the Internet was full of angry articles demanding why, ten years after Ottawa, people were still being killed and maimed in Africa by mines. Well, if you can’t be bothered to read the text ….

And that’s the problem. People see what they want to see, and often assume that if a word appears in the title of a treaty, it must mean what they want it to mean. So people ask why Ukraine has so many foreign fighters, or “mercenaries,” when it is a signatory to a 1989 Convention outlawing them. To which of course the reply is: read the text, and there you will find a very narrow and limited definition of “mercenary” which the terms of service of foreign nationals in Ukraine seem specifically to have been written to avoid. “But that’s a technicality!” some have protested. And I have no patience with people who ask “why didn’t the ICC indict Tony Blair for the Iraq War?” since they have clearly not read the Statute, nor spent five minutes researching the background to the 1998 negotiations.

Concluding on this point, we can say that the lesson for liberal good-thinkers and the PMC is, Be careful what you ask for, you might get it. You want to put people you dislike on trial? Fine, but then boring things like evidence, witnesses and admissibility of evidence have to be taken into account, and courts might get the wrong answer. When the first war crimes trials were held in Arusha and The Hague in the 1990s, human rights activists were outraged to discover that the defendants were being accorded human rights, including the presumption of innocence! Their lawyers were allowed to cross-examine witnesses! Proof was required of guilt! Some defendants were acquitted because of insufficient evidence! What was the world coming to? And as long as people have exaggerated normative expectations, and attribute to texts magical powers they don’t have, this kind of reaction will continue.

But I’m going to end, for once, on a cautiously optimistic note. On the one hand, a lot of international agreements, the kind that don’t attract unrealistic eschatological fantasies, are negotiated by experts who know what they are doing, and understand that what is in the treaty has to be made operational. I’ve deliberately limited this essay to other cases, where treaties are presumed to have supernatural powers they don’t possess. In any event though, as I have stressed, words on paper mean nothing unless the will and capability exists to implement them. But the reverse is also true: many treaties simply set out what is done anyway, and what would continue to happen in the absence of a treaty. And the international system runs, not by rule books, thank goodness, but through intelligent collective self-interest. In the end, the New START Treaty is not really the issue. Either side could already have denounced it, but they haven’t. There’s no reason to believe that the end of the Treaty will change anything substantial, so long as the two sides see advantage in limiting their arsenals. If they don’t, then treaties are irrelevant anyway.