Sovereignty and the Bomb

This is the chapter that makes people uncomfortable, and it should. It sets out a pattern in the historical record that is not seriously disputable, notes the conclusion that many governments have drawn from it, and then gives the case against that conclusion at full strength, because the chronicle owes the reader an argument rather than a slogan.

The record

Consider what happened to states that had, gave up, or never obtained the capacity to inflict unacceptable damage on a great power.

The pattern is not subtle, and it is not a fringe reading. It is what North Korea's foreign ministry says out loud, and it is what a large body of international relations scholarship (from the deterrence literature onward) would predict.¹

The bargain that was not kept

The Non-Proliferation Treaty of 1968 was a trade with three parts: the non-nuclear states would not acquire weapons; they would get access to civil nuclear technology; and (this is Article VI) the nuclear states would pursue negotiations in good faith on effective measures relating to cessation of the arms race and to nuclear disarmament.

Fifty-seven years later, the arsenals number in the thousands of warheads, all of the recognised nuclear states are running modernisation programmes measured in hundreds of billions of dollars, and the Treaty on the Prohibition of Nuclear Weapons (negotiated by the non-nuclear majority and in force since 2021) is boycotted by every state that holds them.

India's diplomats coined the phrase nuclear apartheid for this arrangement, and whatever one thinks of India's own conduct, the description is accurate: a legally entrenched division of states into those permitted the ultimate guarantee of sovereignty and those forbidden it, with the division corresponding closely to who held colonies in 1945.²

The African case

Africa's relationship with this question is distinctive and instructive.

The continent has a nuclear-weapon-free zone: the Treaty of Pelindaba, signed in 1996 and in force since 2009. And it has the only case in history of a state that built weapons and then destroyed them: South Africa, which assembled six devices and dismantled them between 1989 and 1991.

The timing of that dismantlement deserves to be stated precisely, because it is rarely mentioned. The apartheid government destroyed the arsenal before the transfer of power to a Black majority government. F. W. de Klerk's administration ensured that the incoming ANC government would not inherit them.³

Africa, in other words, is the region that most fully honoured the norm, and Africa is also the region whose states have been invaded, partitioned, occupied, sanctioned and left to be exterminated with the least consequence for anyone. Rwanda in 1994 is the sharpest instance in living memory: a government organised the extermination of its own population, and the Security Council responded by reducing the force on the ground to 270 men.

The inference, and the case against it

The inference many governments have drawn is straightforward: nobody is coming, guarantees are worthless, international law binds the weak, and the only reliable protection is the ability to impose unacceptable costs. That is the argument this chapter reports. Here is why the chronicle does not simply endorse it.

The Ukrainian case is weaker than it looks. Ukraine inherited warheads on its soil but not operational control of them: the launch authority and permissive action links remained with Moscow, and converting the inheritance into a usable independent arsenal would have taken years, enormous expense, and a confrontation with both Russia and the United States that a newly independent state could not have survived. "Ukraine gave up its nuclear weapons" is closer to "Ukraine gave up someone else's nuclear weapons that were parked in its garage."⁴

Acquisition invites the very attack it is meant to prevent. Israel destroyed Iraq's Osirak reactor in 1981 and a Syrian facility in 2007; Iran's scientists have been assassinated and its facilities struck. The dangerous period is not after you have a deliverable weapon; it is the decade while you are trying to get one, when you are conspicuous, sanctioned, and not yet protected.

The cost is prohibitive for the states most in need of protection. A credible weapons programme plus delivery systems runs into many billions of dollars and requires an industrial and scientific base that a country of thirteen million people with a GDP in the low tens of billions does not have. The same money spent on health systems, education and infrastructure buys more national security by almost any measure, as, in fact, chapter 28's Rwanda has demonstrated.

Weapons do not deliver the sovereignty advertised. The Soviet Union had 40,000 warheads and dissolved anyway. Pakistan has an arsenal and has had foreign drones operating over its territory. France and Britain have arsenals and have been unable to sustain independent foreign policies of consequence. Nuclear weapons deter invasion and conquest; they do not deter sanctions, coups, currency crises, debt traps, information operations or the extraction described in chapter 16.

The risks are not hypothetical. Every additional arsenal adds accident risk, command-and-control risk, theft risk, and the risk of a regional exchange whose atmospheric effects would fall hardest on countries that had nothing to do with it. Stanislav Petrov's judgement call in 1983 and the Norwegian rocket incident of 1995 are the near-misses we know about.

And the moral objection is the one this chronicle cannot dodge. Twenty-six chapters have condemned the mass killing of civilians: Rome's, Spain's, Belgium's, Germany's, France's, Rwanda's. A nuclear weapon is an instrument whose entire function is the mass killing of civilians. A chronicle that spent its length arguing that empires are illegitimate because of what they do to people cannot end by recommending the acquisition of the most efficient device ever built for doing it.

What the argument actually shows

So the chronicle's conclusion is not a recommendation. It is a diagnosis, and it is more uncomfortable than the recommendation would be.

The fact that this inference is rational is the indictment. An international order in which the most reliable route to being left alone is possession of a weapon of mass destruction is an order that has failed at its stated purpose. It means the Security Council does not protect; that treaties are enforced downward and not upward; that the Genocide Convention was optional in 1994; that written assurances given in exchange for disarmament were not honoured; and that the states which lecture most loudly about the rules are the ones that hold the exemption.

Everything in this chronicle has been an account of one thing: a boundary between the people to whom obligations are owed and the people to whom they are not. Chapter 1 found it in the deep past as the price of group survival. Chapter 2 found empires making it mobile. Chapter 8 found it manufactured into race to run an ocean-wide labour system. Chapters 13 and 14 found it taught in schools and dressed as biology. Chapter 24 found it printed on a Rwandan identity card. Chapter 25 found it in a Security Council vote on 21 April 1994.

Nuclear weapons are that boundary in its final form: the club of states whose destruction is unthinkable, and everyone else.

The alternatives are unglamorous and slow, and they are what small states actually pursue: regional blocs with real defence and trade content; diversified partnerships so that no single patron can dictate; domestic capacity so that food, medicine and energy are not levers; a professional and disciplined military that makes occupation expensive without threatening anyone; participation in and improvement of the multilateral institutions, precisely because the weak need rules more than the strong do; and a truthful public history, so that a population cannot be told that the last catastrophe was somebody's imagination.

That last one is why this chronicle exists.

The end of the chronicle

The argument closes where it began. In chapter 1, human groups survived by telling themselves stories that were fierce at the boundary and restrained inside it. Every atrocity in the twenty-six chapters since has come from moving that boundary outward (inventing new categories of people to whom nothing is owed) while the technology for acting on the distinction improved beyond anything the deep past could have imagined.

Rwanda is in this chronicle not as an exception but as the place where the whole apparatus can be seen working end to end, in a single small country, within a single century: a race invented in Europe, taught in a mission school, printed on a card by a colonial administration, weaponised by a departing empire, armed by a former colonial power, ignored by a Security Council, and then (this is the part that the model cannot absorb) survived, and answered by people who had been written off entirely.

That is the fact the hierarchy has to keep explaining away. It is also the reason to keep the record straight.

Footnotes

1. On the deterrence literature and the proliferation debate, Scott D. Sagan and Kenneth N. Waltz, The Spread of Nuclear Weapons: An Enduring Debate, 3rd ed. (New York: Norton, 2013): Waltz arguing that proliferation stabilises, Sagan that organisational failure makes it dangerous. ↩
2. Treaty on the Non-Proliferation of Nuclear Weapons (1968), Article VI; Treaty on the Prohibition of Nuclear Weapons (2017, in force 2021). Jaswant Singh, "Against Nuclear Apartheid," Foreign Affairs 77:5 (1998), pp. 41–52. ↩
3. On South Africa, David Albright, "South Africa and the Affordable Bomb," Bulletin of the Atomic Scientists 50:4 (1994), pp. 37–47; Nic von Wielligh and Lydia von Wielligh-Steyn, The Bomb: South Africa's Nuclear Weapons Programme (Pretoria: Litera, 2015). ↩
4. On Ukraine's inheritance and the limits of its control, Mariana Budjeryn, Inheriting the Bomb: The Collapse of the USSR and the Nuclear Disarmament of Ukraine (Baltimore: Johns Hopkins University Press, 2022). ↩