Since the advent of US and Russian nuclear-armed ballistic missiles and early warning systems, the danger has always existed that a false warning of attack—believed to be true—could cause either nation to inadvertently launch a responsive “retaliatory” strike with its own nuclear forces. Fear of a disarming nuclear strike, especially during a crisis, creates immense pressure to use-or-lose nuclear forces if an attack is detected. Because launch-ready ballistic missiles allow either side to launch a counter-strike before nuclear detonations confirm whether or not the perceived “nuclear attack” is real, the launch of a retaliatory strike would in reality be a preemptive nuclear first-strike, should the warning prove to be false—resulting in accidental nuclear war. This pressure applies to any nation that might develop the ability to launch before detonation; as a result, what the United States and Russia decide to do could conceivably act as a role model for others—depending, of course, on the unique circumstances of each country.
Consequently, there have been many calls to eliminate, or at least “de-alert,” these launch-ready forces—that is, to institute changes to the weapons systems that will prevent an overly hasty launch. This approach would make it physically impossible to start a nuclear war by accident, in response to a false warning of attack. Unfortunately, there has not been much enthusiasm in either the United States or Russia for de-alerting or eliminating high-alert nuclear forces.
Yet the recent, escalating tensions between the United States and Russia have increased the need for both nations to address the dangers posed by their launch-ready strategic nuclear weapons. Almost all US and most Russian silo-based intercontinental ballistic missiles (ICBMs)—as well as some of their submarine-launched ballistic missiles—remain at launch-ready status, capable of rapid launch within a maximum of 15 minutes after receiving a warning. These weapons are armed with strategic nuclear warheads, and the detonation of even one such warhead could kill hundreds of thousands of people.
There is another way to reduce the risk of accidental nuclear war: Russia and the United States could each independently adopt a policy of not launching their nuclear-armed missiles before confirmation of a nuclear detonation on their respective territories. Such a policy would make it impossible to launch a responsive or reflexive nuclear strike based upon a false warning of attack. To help them reach such commitments, the diplomatic world should address a factor that has spawned confusion and controversy: nuclear terminology.
More than a question of language. Unfortunately, diplomatic discussions about launch-ready nuclear weapons have been complicated by semantic disputes over the terminology used to describe how the United States and Russia might employ these weapons during the initial phases of a nuclear exchange. (Such disputes occurred routinely during the meetings of the First Committee at the United Nations during 2007 and 2008, when Switzerland, New Zealand, and Chile introduced resolutions calling for de-alerting.) In particular, there have been arguments about the meanings and implications of two key terms: “Launch Under Attack” (LUA) and “Launch on Warning” (LOW).
Both terms define when nuclear weapons are to be launched in response to a perceived nuclear attack, with the perception of the attack based either upon electronic signals from early warning systems, or upon unmistakable evidence provided by actual nuclear detonations. The actual meanings of LUA and LOW have, however, been confused both by usage and translation.
LUA was previously defined by the US Defense Department’s online military dictionary as the “[e]xecution by the President of the Single Integrated Operational Plan subsequent to tactical warning of strategic nuclear attack against the United States and prior to first impact” [emphasis added]. (In April, 2010, the Defense Department removed this long-held definition of LUA from its online dictionary; it now provides no definition for either Launch Under Attack or Launch on Warning.) Yet experts such as the Brookings Institution’s Bruce Blair noted in his book The Logic of Accidental Nuclear War that LUA has also been previously used to describe launch authority as “being withheld until after nuclear detonations had been detected.”
Note the difference: One source defined LUA as a retaliatory launch prior to the impact of an incoming enemy missile strike, while the other defines it as after the enemy strike has detonated and been detected.
Although no official definition for LOW was apparently published, the concept of launching nuclear forces upon warning of attack, “to provide the time necessary to ready our missiles so that they can be fired before they are destroyed” was discussed in a report to President Eisenhower in 1959. Many declassified documents reveal that LOW was subsequently a term commonly used in the US military during the 1970s and 1980s. Some in the military have, however, protested that the term LOW should not be used at all, because it implies that an essentially automatic or reflexive nuclear launch would occur upon tactical warning of strategic attack. In 2007, an email from the US Strategic Command to the Arms Control Association stated that US policy does not rely on a “launch on warning” strategy.
More recent definitions of LOW and LUA were provided by a comprehensive analysis of de-alerting published in 2015, Global Zero Commission on Nuclear Risk Reduction: De-Alerting and Stabilizing the World’s Nuclear Force Postures. Global Zero made it clear that it equated LOW with LUA, as well as with “prompt response;” all three terms were used interchangeably in the text. But an earlier authoritative 2009 report sponsored by the Swiss and New Zealand governments, Reframing Nuclear De-Alert, stated that there could be a problem with this approach, because of existing confusion about the definition of LOW.
A further complication exists because previous Soviet and current Russian terminology use a term roughly analogous to LOW—but also seems to incorporate some of the elements of the meaning of LUA—to describe the launch of nuclear weapons in retaliation to a perceived attack. Reframing Nuclear De-Alert describes this difference:
In Russia, the Strategic Rocket Forces define the term “otvetno-vstrechnyi udar” (OVU) as “a form of responsive measures of Strategic Nuclear Forces ordered after analysis of all reconnaissance and early warning data so that the transmitting of launch orders to a major portion of delivery systems and the launch of those systems are carried out before the first impact.” This is similar to the US term LUA but the word “all” conveys a sense of comprehensiveness of information inflow. A Russian expert calls OVU a “retaliatory offensive strike”—as in “we are not going to be first but we are not going to be second, either."
LUA has also been described in Russia as a retaliatory strike that occurs “when the enemy missiles have already begun to hit targets.” This confuses LUA with another Russian term, “oventyi udar,” or OU, which is defined as a “responsive-confronting strike,” or second strike (“a missile-nuclear strike which is conducted during and after a missile attack”), as Austin and Muraviev noted in The Armed Forces of Russia in Asia. This multiplicity of terms (LOW, LUA, OVU, and OU) combined with differences in definition, translation, or interpretation, together act to create confusion—which is dangerous in a discussion of nuclear weapons and nuclear war.
A possible fix. One solution to this problem is to create a new, mutually understood set of terms, which would reduce ambiguity (and thus disagreement) when describing how the United States and Russia might respond to a perceived or confirmed nuclear attack. New terms can be crafted by using a common component of the two previously mentioned US and Russian military definitions (LUA and OVU), which describe the initiation of a retaliatory nuclear strike in response to a confirmed first nuclear attack.
Consider that the previously mentioned US Defense Department definition of LUA and the Russian Strategic Rocket Force’s definition of OVU both specify that the launch of strategic nuclear forces occurs before the attack is confirmed by “first impact.” The term “first impact” in these definitions means the impact and detonation of one or more nuclear warheads. Thus the US and Russian military both appear to identify an expected imminent nuclear detonation as the pivotal event which drives the launch process.
Most important, only nuclear detonation provides unequivocal proof that a nuclear attack—and not a false warning or a conventional attack—has actually occurred. Early warning systems cannot discriminate between conventional and nuclear warheads while they are still in flight; only the detonation of warheads will reveal if they are conventional or nuclear—or if they exist at all.
Without the final, unambiguous evidence provided by nuclear detonation, only the perception of an imminent nuclear attack exists. Before a nuclear detonation takes place, the decision to launch a nuclear strike must be based essentially upon electronic early warning system data and other forms of technical and strategic information. A nuclear strike launched before detonation is, therefore, essentially responsive and preemptive in nature.
In all these circumstances, the launch of a nuclear strike can be characterized on the basis of when it occurs—before or after the first nuclear detonation confirms that a perceived nuclear attack is indeed underway. This approach serves two functions at once: first, it allows an observer to simply describe the launch of a nuclear attack as a chain of observable physical events; and, second, it separates the launch of retaliatory nuclear forces into two general classes: Launch Before Detonation and Launch After Detonation. (In both cases, we are talking exclusively about nuclear detonations.)
Launch Before Detonation (LBD) categories. LBD includes two distinct categories of launch. Both are inherently preemptive and occur before the unequivocal proof of a hostile nuclear attack is obtained through nuclear detonation.
The first category of LBD is first strike, or the unambiguous first use of nuclear weaponry, ordered in the absence of tactical warning of nuclear attack, although perhaps after a strategic warning of attack. No apparent disagreement about this term exists, but it is mentioned in order to clarify the sequence of events in relation to the other terms being defined.
LBD’s second category is Responsive Launch Before Detonation (RLBD), or the preemptive launch of a nuclear strike in response to tactical warning of an incoming nuclear attack, but before one or more nuclear detonations provide unequivocal proof that the perceived attack is in fact a real nuclear attack.
RLBD does not imply that political or military leaders would or should respond with a launch of their own. However, the capability to initiate a RLBD is a primary attribute of all operational launch-ready nuclear weapons.
RLBD would certainly not be intended as a nuclear first-strike. Rather, it would be made upon the presumption that an incoming nuclear attack was under way, but without confirmation of nuclear detonation(s) from the perceived attack. RLBD thus creates the danger that a false warning of nuclear attack might be accepted as true and would trigger a responsive nuclear strike, which would in fact be a nuclear first-strike. Launch-ready nuclear forces make RLBD possible; de-alerting or eliminating launch-ready nuclear forces would preclude Responsive Launch Before Detonation. (The approximately 30-minute flight time of intercontinental ballistic missiles, as well as the correspondingly shorter flight time of submarine launched ballistic missiles, creates the compressed decision-making time that drives the responsive launch process; de-alerting nuclear forces so that they are unable to be launched in less than 30 minutes—or longer—precludes a responsive launch that could trigger an accidental nuclear war.)
RLBD could easily replace the term LOW, as well as LUA. Since no official military definition exists for LOW, no effort would be required to edit or eliminate it from military texts. The previously mentioned complaints about LOW make it reasonable to assume that there would be little official objection to the removal of LOW from the military lexicon.
Launch After Detonation (LAD) categories. LAD, or a nuclear strike ordered in retaliation for an attack confirmed by one or more nuclear detonations, could be used to replace the term “second strike.” LAD also follows logically from LBD. In the early 1960s, the United States actually adopted a policy of “launch after impact,” which of course would be synonymous with LAD.
The universal adoption of a policy of launching only after an attack is confirmed by one or more nuclear detonations would make it impossible to launch a responsive or reflexive nuclear strike based upon a false warning of attack. (In Russian.) Because technical failure, human error, and deliberate sabotage—or some combination of these circumstances—can all cause early warning system and nuclear command and control systems to issue false warnings of attack, a launch based upon a false warning could cause an accidental nuclear war. Adopting a policy of launching after detonation would also prevent the launch of nuclear weapons in reaction to an incoming strike of conventional warheads.
The accidental or unintended launch of a nuclear weapon (or weapons) has previously been considered a low probability event. For more than a decade, however, some 20 nations, including North Korea, have developed dedicated computer attack programs that deploy viruses designed to disable, confuse, and delay command and control systems and networks. Deliberate acts of terrorism, as well as the effects of cyberwarfare, have become variables that make it impossible to calculate odds in this equation of ultimate risk.
Maintaining launch-ready strategic nuclear weapons makes an accidental, unauthorized, or inadvertent launch possible. Furthermore, the risk of such an occurrence is is growing with international tensions such as the increasingly dangerous US/NATO-Russian confrontation in Ukraine and Eastern Europe. Given that an accidental launch or a single failure of deterrence could lead to apocalyptic consequences, it is imperative that the United States and Russia address the danger of an accidental nuclear war created by their launch-ready nuclear weapons.
We suggest, therefore, that it is in the interest of both the United States and Russia to adopt a policy of not launching their nuclear-armed missiles before confirmation of a nuclear detonation on their respective territories, as this would eliminate the danger of an accidental nuclear war caused by a false warning of attack. Because this policy could be adopted even with weapons in high launch readiness, it is not an alternative to de-alerting. However, it is our hope that the new terminology will offer a constructive risk-reduction policy in the absence of de-alerting—and ultimately one that could be applied to the control of not just nuclear forces in the United States and Russia, but to the nuclear arms of other countries.
Our proposal does not imply that we accept the logic or wisdom of launching a counter-attack once an initial attack has been confirmed. Retaliation does nothing to stop missiles that are already on the way. And once an attack has started, deterrence is history … and so may be human existence, if US or Russian launch-ready strategic nuclear forces are detonated.
Nonetheless, the adoption of a launch-after-detonation policy would be a small step to reducing the growing risk of accidental nuclear war, and could buy time to achieve the paradigm shift needed to ensure non-use of nuclear weapons through their abolition.
The authors would like to thank Alan Phillips, who pioneered the concept of launch after detonation in the 2004 Bulletin article “Let’s go No-Low.” They would also like to thank Col. Valery E. Yarynich of the Soviet Strategic Rocket Forces, who worked tirelessly to eliminate launch-ready nuclear weapons (see “Smaller and Safer: A New Plan for Nuclear Postures”).