When it comes to nuclear weapons, advocates have often framed the discussion around how much is enough? But that is the wrong foundational question. It presumes a number can be calculated in the abstract. It cannot: The same arsenal is excessive for one purpose and inadequate for another.
The essential question is enough for what? What are these weapons expected to do? How one answers that question determines everything that follows, including an understanding of sufficiency, which is strategically defensible, as distinct from superiority, which is wasteful and counterproductive.
Nuclear weapons are not force applied but force communicated: material capability fused with political conviction into a case addressed to the judgment of another government’s leadership. That reframes sufficiency entirely. A threat persuades not because of the destructive capacity behind it, but because an adversary comes to believe that capacity reflects a genuine disposition to use it — and an adversary judges disposition by what a posture is built to do, not by what a head of state with their finger on the button claims to intend.
A force aimed at destroying an adversary’s ability to retaliate is postured for coercion — a call to act as the threatening state wishes, backed by capability alone. But capability without a credible will to follow through is not coercion; it is only danger. Such a force asks a jury to weigh not whether the threatening state could act, but whether its conduct leaves room to rule out that it intends to. That case must expand or adapt as the targeted state takes steps to resist it and secure its own survival, which is why a quest for superiority becomes a recipe for arms races without end — a compellent posture has no natural stopping point, only the next increment of pressure needed to keep the possibility of action alive in the target’s mind.
The problem this creates is visible in real time. In March 2026, US President Donald Trump demanded Iran’s “unconditional surrender,” rejecting any negotiated end to the war the United States had pursued alongside Israel. The demand was backed by evident capability — American and Israeli strikes had already killed Iran’s Supreme Leader and degraded its military and nuclear infrastructure.
A demand for unconditional surrender, though, is a call to act, not a call to refrain, and its persuasiveness rests on the same test any such case must pass: can Iran’s leadership find it plausible that the US intends to see regime change through — to include sustained ground presence, indefinite commitment, the costs of administering what comes after — or not? A campaign built on airpower and coercive diplomacy does not, on its own terms, lead to regime change. However real the capability behind it, a demand for total capitulation unaccompanied by conduct consistent with actually remaking the state is easily discounted — which is precisely why “superior” force offers no guarantee that a compellent case will hold.
By contrast, a force that can only survive an attack and answer it — one that cannot strike first without inviting its own destruction — makes a different case: not do this, or I may act, but do not do this, for I could have no reason to act unless you did.
That case persuades not because an adversary becomes convinced of the deterring state’s expressed conviction, but because a posture with no other plausible use leaves the possibility that the conviction exists and is impossible to safely dismiss. The distinction, in other words, is not between two verdicts but between two kinds of doubt. Declared conviction is cheap to set aside: Anyone can say they care deeply about a matter, and a jury weighing that claim risks little by discounting it. A posture an adversary can find no other use for is harder to set aside, not because it meets some external standard of rationality, but because the adversary’s leadership cannot construct a more plausible account of why the posture exists — and so the possibility that the disposition behind it is real stays open, unresolved, and weighty enough to dissuade from action.
This is what sufficiency actually requires: not superiority nor any absolute number, but a survivable retaliatory capability built for no purpose an adversary could interpret as anything except deterrence. That case holds only if the retaliation it threatens aims for appreciable harm. By contrast, damage limitation — a force built to destroy what remains of an adversary’s arsenal — reopens the very ambiguity survivability was meant to close, since a damage-limiting retaliatory force looks less like restraint and more like the search for advantage.
Applied to force structure, this points away from the nuclear triad and toward a monad: a nuclear arsenal built exclusively around the nuclear-powered ballistic missile submarine (SSBN) force. In addition to being the most difficult launch platform to locate, the dispersion of the SSBN force around the globe greatly complicates any adversary’s bid to neutralize it entirely.
With these facts in mind, warhead and delivery-vehicle counts should be set to a specific, statable standard: sufficient to inflict appreciable harm on the ten largest population centers of each nuclear-armed state, in a single round of retaliation, against all such states simultaneously if necessary. That standard is deliberately not calibrated to any adversary’s remaining military or nuclear forces — those are not the target set, though some will be unavoidably co-located near the population centers this posture is built to reach. Such a force has no plausible use beyond retaliation and no plausible growth path beyond matching new entrants to the community of nuclear-armed states; it does not expand as adversaries harden or diversify their own forces, because its purpose was never to chase them.
Declaratory policy should do nothing more than confirm what this posture already demonstrates: a formal no-first-use pledge, stated not as a concession extracted through negotiation but as an honest description of a force that could not credibly strike first if it wanted to.
If nuclear weapons exist to deter conflict that would otherwise threaten the body politic’s survival, and their use — the one contingency in which deterrence has failed — would inflict harm on that same body politic comparable to or greater than what they were meant to prevent, the question returns in a sharper form: What purpose do these weapons actually serve? The honest answer is both some purpose and no purpose at once — indispensable for as long as other states hold them, and self-defeating the moment they are used. A survivable retaliatory capability is accordingly a means to a temporary end, not a permanent settlement; arms control, difficult and slow as it is, holds out the only prospect of a lasting one.
