The Nuclear Necessity Principle: Making U.S. Targeting Policy Conform with Ethics & the Laws of War

The world will note that the first atomic bomb was dropped on Hiroshima, a military base. That was because we wished in the first attack to avoid, insofar as possible, the killing of civilians.

Truman said he had given orders to stop atomic bombing. He said the thought of wiping out another 100,000 people was too horrible. He didn't like the idea of killing, as he said “all those kids.”

Applying just war doctrine and the laws of war to planning for a nuclear war may seem like an impossible task. After all, the destructive power of nuclear weapons is so massive that most conceivable uses of such weapons, even against legitimate military targets, are likely to kill multitudes of innocent people. Strategists can imagine limited uses of nuclear weapons – a single detonation against a ship at sea or an isolated military target in the desert – that might meet stringent ethical and legal standards, but these are mostly imaginary scenarios, far removed from the real concerns of policy-makers and planners.³ This state of affairs naturally leads to grave doubts about whether the principles of distinction (or noncombatant immunity) and proportionality can be meaningful concepts in actual nuclear war planning. A valuable literature about ethics and nuclear weapons exists in philosophy and political science, but most of this work focuses on the question of whether it is ethical to make a threat that would be immoral to execute. And even reluctant supporters of nuclear deterrence have lingering doubts. “Nuclear weapons explode the theory of just war,” Michael Walzer has written elsewhere. “The reason for our hesitancy and self-doubt is the monstrous immorality of what our policy contemplates, an immorality we can never hope to square with our understanding of justice in war.”⁴

Yet it is this monstrous immorality that lies at the heart of how most policy-makers think and talk about deterrence. The dominant logic underpinning nuclear deterrence has been about punishment to noncombatants. The dominant language, however, has used more clinical or euphemistic formulations, like “unacceptable damage,” “assured destruction,” “countervalue targeting,” or “holding at risk that which an adversary values most.”

At the same time, the U.S. government has tasked the military with creating plans to deter, and to inflict damage if deterrence fails, in ways that comply with the law of armed conflict (loac). The Obama administration's 2013 guidance on the employment, or use, of nuclear weapons, for example, explicitly directs that “all plans must also be consistent with the fundamental principles of the law of armed conflict. Accordingly, plans will, for example, apply the principles of distinction and proportionality and seek to minimize collateral damage to civilian populations and civilian objects. The United States will not intentionally target civilian populations or civilian objects.”⁵

We believe these two requirements – the unstated threat to inflict high levels of civilian punishment to deter a nuclear attack and the official promise to respect the law of armed conflict and “minimize collateral damage” – are in deep tension with each other. When policy-makers implicitly view nuclear deterrence as including the ultimate threat of the pain and suffering inflicted on civilian populations, they are relying on something that the U.S. military has largely come to reject as a legitimate military objective. Those who are responsible for the practice of deterrence are left to resolve these incompatible positions as best they can – a task we like to think of as putting a square missile in a round silo.

We have no reason to doubt that military officers and U.S. government lawyers are largely seeking to follow the guidance given to them with skill and professionalism. But it is important to recognize that the law of armed conflict, to use a common legal theory metaphor, provides “standards,” rather than “rules,” to guide decision. For example, a standard would be a law telling a driver “do not drive recklessly,” while a rule would be a law saying “do not drive above sixty miles per hour.” Bright line rules are put in place when one wants strict adherence to a specific constraint on behavior; more flexible standards are put in place when one wants individuals to be able to use their judgment on how best to confirm with the purpose of the specific law. The laws of war are mostly standards rather than bright line rules. It is therefore critical to examine how the U.S. military and civilian lawyers alike have interpreted the standards they are required to follow under the relevant laws of war.

An understanding of the history of nuclear targeting and our reading of contemporary military guidance documents makes us deeply skeptical about the degree to which U.S. nuclear war plans actually conform to the principles of distinction, proportionality, and necessity. U.S. military officers surely want to follow the laws of war, seeking to be just warriors and not illegal killers, but they are in a nearly impossible position. The result has been an unfortunate expansion of the definition of “military objects” and the creation of many loopholes and exceptions to rules. Thus, the resulting war plans – which could produce tens of millions of noncombatant deaths – are still claimed to be consistent with the principles of distinction and proportionality found in just war theory and the law of armed conflict.

This state of affairs reflects a gap, one that has existed since 1945, between the way we understand nuclear weapons, on the one hand, and plan for their use, on the other. The euphemistic and impoverished language of nuclear strategy – the ways in which we describe legitimate “military objects,” the “nuclear umbrella,” “collateral damage,” and “countervalue” and “counterforce” targets – that too easily obfuscates the real human consequences of nuclear use, helps identify that gap. This gap can lead to both public misinformation and private discomfort and denial, as seen in Harry Truman's public assertion that Hiroshima was a military target and his private concern about civilian casualties. Truman claimed that he never lost one night's sleep over the dropping of the bomb, but the fact that he said it so often, we suspect, is a sign that it was not true.

The consequence of this gap is a profound disconnect between what the Government Accountability Office politely described as an “indirect” relationship between the requirements and the practice of selecting targets and developing operational plans.⁶ It is inevitable that there will be some distortion as a large bureaucratic military organization attempts to turn the often vague and poorly considered aspirations and intentions of policy-makers into actual plans. While Truman emphasized the military nature of Hiroshima, he was nonetheless not informed that the targeting committee moved the ground zero aim point from the military factory area to a bridge at the center of the city.

It is time to close this gap by better matching the requirements and practice of deterrence. It is possible to revise the president's nuclear employment guidance to make nuclear targeting practice better conform to principles of just war doctrine and the law of armed conflict. To do so, the U.S. government should state clearly that the United States will not employ nuclear weapons against any military target that can be reliably destroyed with conventional weapons. This guidance would be a modern, nuclear version of what some just war theorists have called the “necessity principle” in warfare. In this case, the president would be committing to refrain from using nuclear weapons when they are not necessary to achieve legitimate military objectives and thereby minimize collateral damage fatalities to civilians.

There are healthy disputes in the legal community within the U.S. military about how to interpret the laws of armed conflict in the context of nuclear weapons. We make these proposals to encourage an informed public debate, not to settle such disputes. We understand that using conventional weapons as a substitute for nuclear weapons when feasible can create its own unintended consequences and challenges. But we want to have that debate about such ethical, legal, and security dilemmas in the open, rather than behind closed doors, where bureaucratic biases can too easily reign.

The U.S. Strategic Command (stratcom) has its origins in the Strategic Air Command (sac) whose legendary second commander, General Curtis LeMay, famously proclaimed during World War II that “there are no innocent civilians.”⁷ There have been other voices within the U.S. military, including the Air Force, that have held a different view of the principles of distinction and proportionality. The emphasis on avoiding attacks against noncombatants has increased over time, particularly now that the Geneva Convention, negotiated after World War II, has created a stronger set of agreed-upon laws of armed conflict.⁸ Still the attraction of strategic bombing against civilian populations remains strong, even if the advocates avoid explicitly admitting this. When, for example, one of us (Sagan) submitted the book Moving Targets in 1988 to classification review to ensure that it did not inadvertently reveal classified information, the Pentagon security reviewers initially concluded that there was one word in the following paragraph that needed to be removed:

The word that the Pentagon censors wanted removed was “innocent.”

This kind of open ideological and bureaucratic denial of the existence of innocent noncombatants in an adversary's country seems less likely in the U.S. military today. Yet when we are told that the United States does not “intentionally target civilians or civilian objects” and will “seek to minimize damage to civilian populations,” what does that really mean? What kinds of targets are left off the target list because of that guidance and which ones remain? How seriously does the “principle of proportionality” constrain war plans?

The bureaucratic process by which nuclear targeting decisions are made is an arcane and highly classified arena of military planning, but it is necessary to examine it seriously, to peer into the back rooms of nuclear planning and implementation, if you will, in order to understand the observations that lead to our skepticism. Despite the Obama administration's guidance that the principle of distinction will be followed, for example, official Joint Chiefs of Staff (jcs) planning documents continue to argue that there are legitimate exceptions to the prohibition on targeting civilians in warfare.¹⁰ According to the official U.S. 2013 Joint Targeting manual: “Civilian populations and civilian/protected objects may not be intentionally targeted, although there are exceptions to this rule. Civilian objects consist of all civilian property and activities other than those used to support or sustain the adversary's war-fighting capabilities. Acts of violence solely intended to spread terror among the civilian population is [sic] prohibited [emphasis added].”¹¹ The term solely in this passage implies that as long as there is a primary intent to destroy legitimate military targets, it is acceptable to have secondary intent to “spread terror” to affect the will of a government to continue the war. Similarly, the 2007 edition of the official document on Strategic Attack cites the 1999 Operation Allied Force strikes against Serbian electrical power plants, executed with the intent both to sever electric power to military facilities and to create popular opposition to the war to encourage Slobodan Milosevic to come to terms with nato, as an example of justified bombing of dual-use infrastructure targets.¹² A 2012 Air Force document, in a definition of “punishment,” also noted that the word is often used to refer to a strategy, “which attempts to inflict enough pain on enemy civilians so that they cause their leaders to change their behavior.” This document continued to argue that although “a punishment strategy may conflict with the LOAC, depending on the nature of a conflict, it may nonetheless be a feasible, if not always acceptable strategy.” It then listed Operation Allied Force as a successful example.¹³

Air Force lawyers have developed arcane arguments to provide legal justification for such secondary, but intended, targeting of civilian populations, even with nuclear weapons. In 1997, then-Colonel Charles Dunlap, who was the Staff Judge Advocate at STRATCOM, published an important article in which he noted that the “special political and psychological dimensions of nuclear weapons” posed a dilemma for the lawful use of nuclear weapons. “Although using nuclear – or any other – weapons merely to terrorize noncombatant civilians is contrary to international law,” Dunlap argued, “affecting the mental state of an adversary, degrading his morale, and eroding his will to continue the conflict, can all constitute legitimate military objectives.” Dunlap frankly admitted the difficulties associated with such amorphous standards. We can find no meaningful distinction between “terrorizing” noncombatants and the allegedly lawful objects of affecting their “mental state” and “eroding their will to continue the conflict” other than a rhetorical one. Dunlap's solution was to shift responsibility. “To avoid such dilemmas,” Dunlap continued, the relevant guidance documents should consider “affecting an adversary's ‘perception of U.S. will and resolve’ as an employment (as opposed to targeting) consideration. In other words, under U.S. doctrine a particular target must first be justified in orthodox military terms independent of the psychological or political ‘message’ the use of nuclear weapons might produce.”¹⁴

What Dunlap means by the distinction between employment and targeting considerations is important. The reference to an employment decision, in context, means that such considerations are left to the president, or National Command Authority. This frees targeteers to simply ask whether the destruction of the target represents a valid military objective. If the president knows that the real “benefit” or “intent” of attacking a radar facility next to a major city with multiple nuclear weapons, for example, is the terror produced among the surviving civilian population, then that specific policy decision is his problem. Targeteers may simply satisfy themselves by stating that the target is, itself, a legitimate military objective and that the very large number of nuclear weapons required to destroy it is little more than the application of a formula to achieve damage criteria created to implement the president's employment guidance. While there are other views that have been expressed by Department of Defense lawyers, we note that it is Dunlap's view that appears in educational materials prepared for Air Force officers.¹⁵

Furthermore, the U.S. military has expanded the definition of a legitimate military object to include civilian objects that have the potential to be used by the military in the future. As STRATCOM Deputy Staff Judge Advocate Theodore Richard has noted, the 2013 Joint Targeting guidance introduces “definitional flexibility” when describing the purpose or use of a military object:

In other words, any target that could conceivably be used by an enemy's military in the future, such as a civilian airport, could be deemed a legitimate military object. Any “civilian object” that could potentially contribute to an enemy's war effort, even if “indirect” or “discrete,” could be deemed fair game as a “a military object.”

The U.S. military's inclusion of “war-sustaining” rather the just “war-supporting” industry targets arguably produces a definition of legitimate military targets that stretches beyond the reach of what is permissible under the Additional Protocol to the Geneva Conventions. (Although the United States is not a party to Additional Protocol I, the targeting standard in article 52(2) is accepted by the United States as customary international law.) The dispute concerns whether it is permissible only to target industrial facilities or other economic objects that contribute directly to military production (such as a tank factory or munitions facility) or broader “war-sustaining” industrial targets (such as oil export facilities or other industries that merely provide tax revenue supporting the state's military efforts). The 2013 Joint Targeting document cited earlier legitimized attacks on “war-sustaining” targets, not just “war-fighting” or “war-supporting” targets. We agree, in contrast, with international law experts Janina Dill and Yoram Dinstein who point out that using such expansive “war-sustaining” criteria, rather than direct military support criteria, opens up the possibility that all civilian economic targets could become legitimate targets.¹⁷

Unfortunately, the principle of proportionality does not provide a check on such an expansive interpretation of the principles of just war and the law of armed conflict to nuclear weapons use. The military planner is required to weigh the costs of unwanted collateral damage against the benefits of destroying the military target. If, however, the contribution of a nuclear attack is to limit nuclear damage to the U.S. population or to maintain the United States as a viable society after a nuclear war, then almost any degree of collateral damage could be deemed acceptable under the proportionality principle. In short, following the principles of proportionality and distinction alone still make it too easy, too acceptable, even too legal, to kill many, perhaps millions, of innocent civilians.

Lastly, we note that the bureaucratic slide into legitimizing civilian targeting is too often hidden by a kind of Strangelovian doublespeak in the nuclear planning process. The Nixon administration, for example, stated that the United States did not target civilian populations “per se.” Nevertheless, National Security Directive Memorandum 242 (nsdm-242) guided planners to hold at risk the “enemy's postwar power, influence and ability to recover” from a nuclear war. In implementing this “counter-recovery targeting” planners increased the number of weapons targeted on Soviet industry, reportedly including fertilizer factories. While the United States did not target “population per se,” it did seek to destroy the Soviet population's food supply, which would starve the population.¹⁸ As lawyer Theodore Richard has argued, “at this stage … law-of-war rules appeared to have minimal impact on nuclear targeting considerations. If anything, the law appeared to maintain its Second World War incarnation with vague notions of military objectives and toleration for civilian casualties.”¹⁹

Although the United States no longer targets food supplies, the episode illustrates how easily advocates of “strategic bombing” can create thin legal and moral justifications out of public view for targeting practices that would result in millions of deaths as “collateral damage” in a nuclear war. Policy-makers can too easily accept this state of affairs in the name of deterrence. Indeed, much of the discussion in the Washington policy community, and in scholarly circles alike, about how nuclear deterrence operates rests on the ultimate prospect of killing millions of innocent persons in retaliation to a nuclear strike. This is as true of dovish “minimum deterrent” advocates calling for “assured destruction” and “countervalue” targeting as it is for hawkish advocates of massive nuclear superiority. When the threat of massive collateral civilian fatalities is the basic building block of deterrence, however, those expected fatalities should no longer be considered collateral. It is worrisome, therefore, that some policy-makers appear to count on the terror of nuclear weapons against civilians for deterrence to operate, while targeteers attempt to sidestep the moral and legal complexities by narrowing their inquiries into whether the target is military or not.

We do not want to focus attention only on the U.S. military here. The root of the problem is that the fundamental conception of nuclear deterrence through threatened punishment to civilians is incompatible with just war principles, the law of armed conflict, and international humanitarian law. Some of the most trenchant critiques of these problems are by military lawyers. And we understand that junior officers questioning the lawfulness of orders is a concern for their commanding officers: it could result in a loss of discipline and an erosion of deterrence. The question is further complicated because the moral and legal harm rests in the creation of plans on paper – plans that, even if unethical or unlawful, most expect will never be executed. In our view, however, some advocates embrace the claim that U.S. nuclear war plans are compatible with ethical principles with excessive exuberance, helping to perpetuate the current state of affairs. Dunlap, for example, has argued that “should deterrence fail, our forces are – and must continue to be – ready to immediately execute orders of the national command authorities to employ nuclear weapons. Those who carry this gravest of responsibilities are entitled to be secure in the knowledge that plans they must execute honor the highest ideals of the country they have sworn to defend.”²⁰ We also want soldiers to be willing and able to execute all legal military orders. But we believe that very different guidance and practices are necessary to produce new nuclear and conventional war plans that are consistent with the highest ideals of our country.

We have argued that following the principles of distinction and proportionality do not appear sufficient to deal with the legal and ethical challenges posed by nuclear weapons. It is important to note that the “principle of necessity” was entirely omitted from the U.S. nuclear weapons employment guidance. The principle of necessity, however, features prominently in two legal findings relating to the threat or use of nuclear weapons: Ryuichi Shimoda et al. v. The State (1963) and the International Court of Justice advisory opinion on the Legality of the Threat or Use of Nuclear Weapons (1996). While these opinions represent a relatively diverse set of views about the legal questions involving nuclear weapons, they articulate important alternatives to how we think about the use of nuclear weapons.

In 1955, several Japanese nationals who were residents of Hiroshima and Nagasaki at the time of the bombing brought an action against the Japanese state for damages. Ryuichi Shimoda lost four daughters and a son in the bombing. He survived, but experienced severe health problems that left him unable to work. While the Tokyo District Court found that individuals could not claim damages, it also reached a number of conclusions about the legality of the bombings. The Tokyo court concerned itself only with the narrow question of whether the bombings of Hiroshima and Nagasaki were legal, avoiding the broader question of whether the use of nuclear weapons might ever be permissible. Yet the court found the 1945 bombings to be illegal. The court noted that indiscriminate bombing of undefended cities was prohibited because there was no justification in terms of military necessity for such an act. While recognizing that there were legitimate military objectives in the two cities, the court concluded that the atomic bombing was indistinguishable from the indiscriminate bombing of the two cities. The court also specifically rejected the notion that total war invalidated any distinction between combatants and noncombatants.²¹

The second case appeared in the International Court of Justice (icj). In 1994, the United Nations General Assembly passed a resolution seeking an advisory opinion from the icj on the Legality of the Threat or Use of Nuclear Weapons. (The World Health Organization had sought a similar opinion starting in 1993, but the icj eventually held that the question fell outside the scope of the who's mandate.) Like the Tokyo Court in Shimoda et al., the icj opinion avoided a sweeping ruling, arguing that “the Court cannot conclude definitively whether the threat or use of nuclear weapons would be lawful or unlawful in an extreme circumstance of self-defence, in which the very survival of a State would be at stake.” But the icj opinion held, among other things, that “the threat or use of nuclear weapons would generally be contrary to the rules of international law applicable in armed conflict, and in particular the principles and rules of humanitarian law.” Central to the court's reasoning were the notions of proportionality and necessity, with the court citing an earlier ruling on the “specific rule whereby self-defence would warrant only measures which are proportional to the armed attack and necessary to respond to it, a rule well established in customary international law [emphasis added].”²²

In both cases – Shimoda and the icj advisory opinion – the criteria of “military necessity” was not settled by simply asking whether the object of the nuclear use was a legitimate target, but also whether the means themselves were necessary to destroy the target and were thus permissible. The view expressed by both opinions is similar to the pithy summation offered by William Taft: “The condition of ‘necessity,’ rather, requires that no reasonable alternative means of redress are available.”²³ Emphasizing the necessity of the means, not merely the ends, is an essential aspect of improving our nuclear doctrine and policy. Absent this emphasis, “the principle of necessity” loses its meaning as a constraint.

To see how the principle of necessity might constrain nuclear war planning, consider the many “soft” targets, such as military-related industrial targets, that appear to remain in the nuclear war plans today. Such targets could be destroyed, if necessary, by conventional weapons. One might ask, therefore, what would be the purpose of using a nuclear weapon against a target when a conventional weapon would suffice? Would this not seem to be a reasonable test of whether the actual intention was to terrorize civilians instead of achieving a legitimate military objective? What are the practical benefits of using a nuclear weapon in place of a conventional one? It is this question we turn to next.

An awareness of the weak constraints created by the proportionality principle in practice is what led Michael Walzer to propose, in Just and Unjust Wars, that soldiers and leaders alike must take active measures to reduce collateral damage, including taking risks upon themselves. This concept of taking active measures to reduce collateral damage has been called the “due care” principle by political philosopher Steven Lee, and Sagan and Benjamin Valentino.²⁴ Others, like Seth Lazar, argue that using only the minimal amount of force necessary to destroy the target should be called the “necessity principle” in jus in bello.²⁵ It does not really matter which term is used, but the principle should be that the minimal amount of force needed to destroy the target with a reasonable prospect of success should be used at all times, even if this means accepting some risk. In terms of nuclear war planning, this would require the president to state as a matter of law and national policy that the United States will not employ nuclear weapons against any target that could be reliably destroyed with conventional weapons.

If we agree that nuclear weapons may not be used for other purposes such as their “unique psychological impact” – simply a euphemism to invoke the terror nuclear weapons cause – perhaps there will remain practical arguments that the use of nuclear weapons might be justified on some other grounds, such as economy of force or their ability to destroy hard and deeply buried targets.

In contrast to the economy argument, however, we believe there are many sound strategic reasons to prefer an increased role of conventional forces in maintaining deterrence today. The traditional benefit of nuclear weapons has been their explosive power; but the value of explosive power is a diminishing one. As weapons have become more accurate, leaders have preferred weapons that can reliably destroy a target through precision, while avoiding the mass casualties that nuclear weapons or indiscriminate bombing might inflict. The large yields associated with nuclear weapons are no longer a benefit, but rather an inhibition against their use. Political leaders are reluctant to use weapons that do not discriminate between combatants and noncombatants and produce undesirable physical effects like radiation. Nor are political leaders eager to transgress the “tradition of non-use” that has grown since 1945 for fear that doing so might encourage further nuclear proliferation and the potential use of these weapons by other nations.

For these reasons, the threat of conventional weapons use is not just a more ethical choice, it is also, in most scenarios, a more credible choice. This policy might produce some risk that an adversary discounts the destructiveness of a conventional attack or retaliation. As conventional military options are more likely to be ordered in a conflict, however, we believe that they are generally a more credible threat. Although the common perception is that nuclear weapons are needed for stable deterrence because they are so terrible, in fact the opposite may be the case: nuclear weapons struggle to deter because the threat of their use is so incredible, since the consequences would be so terrible.

A second category of targets may be hard and deeply buried targets that require a very large yield nuclear explosion that sends a shockwave through the ground to crush a deeply buried bunker. There is, at the moment, a considerable debate about the effectiveness of conventional weapons against such targets. Many such targets are defended by uncertainties in their location as much as their depth. Once a site is discovered, it may have other vulnerabilities, such as ventilation, that are a more reliable route to destruction than simply hoping to crush them with a nuclear shockwave. And other targets may be too hard or buried too deeply for any form of munition, nuclear or conventional.

On balance, we believe setting a requirement that nuclear weapons never be used against any legitimate military target that could be reliably destroyed by other means would result in a substantial reduction in nuclear weapons, threatening the much smaller set of targets, if any, that remain immune to conventional attack. Most of all, we think the proposal conforms with simple common sense. It is hard to imagine a circumstance in which it would be either ethically permissible or wise from the perspective of our security interests to use a nuclear weapon when a conventional one would suffice. U.S. officials have made similar statements in the past. In 1996, then-Secretary of Defense William Perry was asked about whether the United States would allow Libya to complete construction of what appeared to be a large chemical weapons production facility near Tarhuna. Perry answered with one word: “No,” and then when asked about the use of force, he replied, “I wouldn't rule anything out or anything in.”²⁶ This was quickly interpreted as a nuclear threat, something Perry corrected by stating, “I would never recommend nuclear weapons for that particular application.”²⁷ The United States could easily generalize Perry's statement, reserving nuclear weapons only for those targets that cannot be reliably destroyed otherwise and letting the effectiveness of U.S. conventional military capabilities shoulder the burden of deterrence.

We understand that any U.S. policy change emphasizing the deterrent and war-winning potential of conventional capabilities could result in unintended consequences. Would this doctrine require a massive increase in conventional arms spending? Would it require the development of new nuclear weapons with smaller yields? How would the resulting shifts in doctrine influence the likelihood of escalation in a crisis? Would potential adversaries find this shift reassuring or threatening, and how would allies perceive the change? In particular, two U.S. nato allies, France and the United Kingdom, also have nuclear weapons and have, at least in the recent past, explicitly targeted cities in the Soviet Union and Russia, as their main deterrent threat.²⁸ Wouldn't the U.S. adoption of the nuclear necessity principle also suggest that these two nato nuclear allies must change their doctrines to conform to just war doctrine and the law of armed conflict? We believe such a policy would ultimately enhance extended deterrence to our friends and allies, by making U.S. commitment rest on threats that the U.S. president would execute if deterrence failed. But we are mindful of the need for dialogue with our allies and partners.

Finally, would this step be consistent with efforts to encourage the United States and other nuclear powers to honor their commitments under article VI of the Non-Proliferation Treaty to work in “good faith” toward nuclear disarmament? On its face, emphasizing an additional legal constraint on nuclear targeting should be welcomed by those seeking progress toward disarmament. And yet, in prohibiting some behavior, a rule permits others – in this case the targeting of nuclear weapons for specific purposes. Our view is that focusing on the limited and declining utility of nuclear weapons is a necessary step toward creating a world without nuclear weapons that is not predicated on unachievable conditions of world peace and general disarmament.

These are all legitimate issues for debate. It would behoove us, however, to debate such subjects with transparency and frankness in academic and government circles, rather than allow a gap to persist between how we talk about nuclear deterrence and how we practice it. Placing conventional weapons at the center of debates about the future of deterrence would help focus the policy discussion on realistic scenarios in which our military power might actually be used. And it would more faithfully honor the just war principles of distinction, necessity, and proportionality, by placing them at the heart of our deterrence and security policies, where our highest ideals belong.