Louis René Beres[*]
[This essay is available in PDF at this link]
In early May 2019, U.S. President Donald Trump—responding to Kim Jung-un’s latest round of missile tests—sought to reassure the American public. His seat-of-the-pants comments, however, were based entirely upon the presumed importance of his personal relationship with Kim Jung-un and had nothing to do with any measurably refined strategic assessments or “preparation.” In these public comments, Trump has stressed that “attitude,” not “preparation” matters. This philosophy undermines U.S. deterrence and fails to account for strategic assessments that should influence the President’s understanding of the unprecedented threats posed by unpredictable nuclear regimes.
For U.S. strategic planners, there are few tasks more daunting than assessing an enemy’s nuclear capabilities. Among other things, this task involves using a holistic approach to accurately determine an enemy’s nuclear intentions. It should not be replaced with speculation based upon a “feel” for the country’s leadership. The Democratic People’s Republic of Korea (North Korea or DPRK) has regularly issued statements threatening South Korea, Japan, and the United States with its nuclear arsenal and continues to design rockets capable of striking the continental United States. President Trump’s handful of meetings with Kim Jung-un has seemingly led him to trust the North Korean dictator and somehow ignore the DPRK’s repeated rocket tests. On another front, Trump still alleges that Iran would likely use its nuclear weapons should it ever be allowed to gain this capability. Though this allegation could conceivably be plausible, there can be no properly logical or scientific basis for its correctness.
In essence, developing an effective approach to rogue actors like North Korea and Iran requires acting on diplomatic and strategic assessments that consist of diligently collected and suitably analyzed intelligence. But the President’s oft-expressed disregard for “preparation” in favor of “attitude” shows that such a “diligent” approach is conspicuously lacking in the current White House.
A state’s intentions and capabilities should be analyzed as codependent variables. For example, North Korea’s intentions to attack American or allied targets and relevant capabilities such as infrastructures, nuclear weapons, and inter-continental ballistic missiles, must be analyzed together. At some point, an enemy nuclear state such as North Korea might choose to adopt a more challenging atomic stance vis-à-vis the United States. Plausibly, this aggressive stance would be taken on the basis of Pyongyang’s own presumptively improving (or diminishing) nuclear capabilities.
Optimally, neither side would engage in a nuclear comparison that merely counts warheads and missiles or narrowly assesses the number of weapon ranges and yields. This too-limited type of approach simply looks at the size of each side’s “nuclear button” and not at the respective strategic cost of actually initiating conflict. Even though the DPRK is weaker, it also has less to lose. The United States has world standing, allies, and many theaters of operation that require substantial resource allocations. By contrast, North Korea has a very small area of operations, and its leadership could appear more willing to subject its population to the palpable horrors of nuclear war. Plainly, even a “small” nuclear attack on the United States or its allies by a “small” nuclear power could portend unimaginable casualties, both prompt and long-term.
Whether an adversary is a state, sub-state, or “hybridized” foe, U.S. threat mitigation should rest in large measure on whether that enemy is presumed to be rational. By definition, this means, among other things, that U.S. policy judgments should consider how much the adversary values its own collective survival. In this context, an irrational adversary is one that could accept a significant risk of annihilation in order to maximize some other more highly valued preference or combination of preferences, e.g., the fulfillment of “national destiny” or primary religious obligation.
The prospect of irrational decision-making tends to be more likely among certain sub-state or terrorist actors (e.g., Shiite Hezbollah or Sunni Al-Qaeda) than among any enemy states. But there is still no good reason to discount enemy-state irrationality out of hand. For one thing, history provides sobering examples of states that have behaved more or less irrationally (most obviously, the Third Reich), including some in which the foreseeable consequences were genuinely existential. Consequently, Washington must systematically extend its strategic attention beyond any militarily traditional or orthodox “order of battle.”
What might be more valuable to America’s prospectively irrational state enemies than their own physical survival? One possible answer is the avoidance of any presumed apostasy, shame, or humiliation. This scrupulous avoidance would include the unendurable charge that they had somehow defiled their most unalterably sacred religious or doctrinal obligations.
Furthermore, the United States’ nuclear (or prospectively nuclear) adversaries need not necessarily be purely rational or irrational. A third logical option exists. This is the destabilizing adversary that is actually “mad.” “Madness” is not the same as irrationality and must be dealt with analytically as a different problem. “Madness” is an “idiosyncratic” or manifestly altered ordering of preferences. While rationality would represent the usual ordering and weighting of preferences and irrationality their unusual ordering, “madness” would express the total lack of any consistent preference ordering.
U.S. nuclear deterrence must also accurately assess enemy intentions, capabilities, and priorities. For analysts and policymakers, the expected adequacy of U.S. nuclear deterrence (1) is most apt to be tested by an already-nuclear North Korea or a prospectively nuclear Iran; (2) could be tested by adversaries with various conspicuous or inconspicuous ties to these states; and (3) may be tested by other near-peer nuclear adversaries that fall somewhere on the decipherable spectrum of rational, irrational, or mad.
Traditionally, strategies of deterrence have implicitly assumed adversarial rationality (with an estimated understanding of a prospective adversary’s skewed preferences). The pertinent reasoning behind this simplification is fully evident and unambiguous. Without such an assumption, predicting an adversary’s actions would be operationally difficult or impossible. Even irrational states may be subject to various alternative forms of deterrence based on how they structure their preferences. Thus, correctly recognizing an opposing state’s or leadership’s preferences in a timely manner is indispensable to a country’s national security. But the actions of a “crazy” or “mad” leadership are necessarily random and unpredictable. Consequently, facing a “mad” adversary in world politics could be substantially worse than confronting “just” an irrational one.
However, in those rare circumstances where an enemy state ranks certain outcomes higher than “staying alive” as a nation, trying to militarily deter that state from pursuing such outcomes could fail. In more serious circumstances involving nuclear deterrence, the plausible consequences of any such failure could become catastrophic for both sides even if one is verifiably dominant. For the Unites States, this problem is daunting but nonetheless clear. Washington could sometimes have to deal with adversaries that are neither habitually nor routinely rational. Generally, U.S. nuclear deterrence should continue to be fashioned and assessed vis-à-vis national or state adversaries rather than assorted sub-state enemies.
The ever-present strategic questions of rationality and irrationality are not abstract. They are profoundly real and (more or less) determinable, especially between the United States and North Korea. President Trump’s decision to somehow trust the North Korean leadership because it has the right “attitude” is exceptionally risky. Given his statements, it does not appear that Trump has dispassionately assessed whether Kim Jong-un’s preferences are capable of being deterred through force or placated via his current strategy of engaging in a starkly contrived U.S.–DPRK détente.
There is no reliable public information supporting President Trump’s odd assumption that Kim Jung-un would agree to bargain away Pyongyang’s extant nuclear weapons and delivery vehicles. Expressed charitably, any still-current U.S. presidential hopes for North Korean “denuclearization” are unrealistic. Thinking through the situation rationally, it is clear that a dictator who values personal power and legitimacy more highly than any other preference or combination of preferences would not willingly surrender the core source of his power: nuclear weapons.
Similarly, determining the intentions and capabilities of a not-yet-nuclear Iran rests on correctly assessing the Islamic Republic’s objectives and rationality. Though very different from the dynamics surrounding the U.S.–DPRK relationship, President Trump’s unilateral decision to exit the multilateral Joint Comprehensive Plan of Action (JCPOA) without carefully assessing how the Iranian leadership and the rest of the world would react was exceptionally imprudent. Even though the original multinational pact of July 2015 was correctly critiqued as weak and unpromising, President Trump’s fallacious reasoning on this issue showed that he had not considered the strategic ramifications of withdrawing from the deal. He mistakenly inferred from the deal’s evident imperfections that it needed to be scrapped altogether. This had three negative effects: first, it erased what little trust existed between the United States and the Islamic Republic, severely hindering future negotiations; second, it caused the Iranian regime to start rebuilding its nuclear capacity to hedge against any significant U.S. military operations, increasing the risk of a nuclear confrontation between the two powers; and third, the re-imposition of sanctions prompted the Iranian regime to engage in escalatory activity in the Persian Gulf, prompting certain U.S. allies in the region to consider building a nuclear capability of their own.
In these conflict-prone environments, successfully deterring an adversary from escalating is contingent upon first knowing that adversary’s preferences. But non-rational adversaries are not always deterred by U.S. threats of retaliatory action, and even rational national leaders could sometimes rank certain priorities over collective preservation or regime survival. When this occurs, a state can no longer be deterred by any of the usual forms of military action. Though there are no tangible examples of nuclear deterrence failure, the logical risk of such a failure increases as tensions rise and more states (especially those with irrational or even “mad” preferences) attempt to develop or enhance their respective nuclear capabilities.
For the moment, North Korea and Iran do not appear to be recognizably irrational or “mad.” Periodically harsh enemy rhetoric notwithstanding, Pyongyang and Tehran do not appear ready or willing to launch a first strike against the American homeland or the vital strategic interests of U.S. allies using either nuclear weapons or other weapons of mass destruction. For now, the most plausible explanation for this apparent restraint is that any such aggression would assuredly elicit a devastating reprisal.
However, faulty calculations or errors in information could sometime lead a perfectly rational enemy state to believe that a previously credible deterrent threat will no longer be deployed against it. From a purely military standpoint, the risks posed by a weaker enemy making this type of strategic blunder could be reduced or perhaps avoided altogether if the United States engaged in certain preemptive action. On occasion, President Trump has explicitly discussed it with his advisors.
Yet any reasonable expressions of anticipatory self-defense are difficult to imagine as strategically cost-effective against North Korea, which is already a nuclear state, or not-yet-nuclear Iran, against which such action would likely produce a debilitating protracted conflict. Though plainly “less powerful” than the United States, North Korea, even with a “small” nuclear reprisal against American or allied targets, could produce devastating outcomes. Similarly, Iran could produce a consequential wave of state-sponsored terrorist activity by activating Hezbollah or Revolutionary Guard Corps operatives in the United States or U.S.-allied countries.
Thus, all things considered, protracted nuclear deterrence, not preemption, is Washington’s best available “remedy.” The reasoning here is based on the well-founded presumption that North Korea is determined to remain nuclear and that stable nuclear deterrence between the two countries could produce a long-term equilibrium. Nonetheless, such equilibrium requires that each party knows the other maintains a credible retaliatory capability. For the moment, President Trump seems unable to understand this, especially when he earlier threatened North Korea with “fire and fury” and (more recently) Iran with “obliteration.”
This does not mean that the United States is incapable of destroying a “required” number of Iranian or North Korean nuclear assets and infrastructures, but rather that the net costs of any such calculated venture would likely exceed the gains. Already, such a conclusion applies in the case of striking Pyongyang, where any nuclear or non-nuclear reprisal could bring grievously catastrophic harms to any one of several U.S. allies. In the case of Iran, which is not yet nuclear, virtually all critical assets and infrastructures have been deeply hardened, widely dispersed, and substantially multiplied. For America to act preemptively against Iran, there would be considerable political and diplomatic costs attached to unilaterally exercising any preemption option.
Certain alternative forms of preemption, including targeted assassinations of nuclear scientists or cyberwarfare could still be useful or necessary, but it is unlikely that exercising any such options could permanently obviate either Iran’s or North Korea’s military nuclear capabilities.
Former Israeli Minister of Defense Moshe Dayan once declared: “Israel must be seen as a mad dog; too dangerous to bother.” If America’s enemies could still be presumed rational in the ordinary sense of valuing their physical survival more highly than any other preference, or combination of preferences, Washington could exploit certain latent strategic benefits of “pretended irrationality.” Were the United States to recognize that in certain strategic situations it could indeed be rational to feign irrationality, its faux irrationality could then be used to render particular adversaries less willing to challenge U.S. interests.
But pretended irrationality could quickly become a double-edged sword. Feigned irrationality is a conceivably feasible strategy against only a rational adversary that prioritizes regime survival above all else. In those circumstances where enemy states are irrational or expressly concerned about being the victim of preemptive U.S. military action, pretending to be irrational could sometime cause that adversary to escalate rather than back down.
The true purpose of the United States’ nuclear doctrine must always be deterrence ex ante. To even suggest otherwise, as this President has done, would be to fundamentally misunderstand the most basic nature of American national security. Still, there remain some readily identifiable circumstances in which nuclear exchanges could become unavoidable. Here, logically, some forms of an actual nuclear war could ensue, so long as (1) the enemy state’s first strikes launched against American assets would not destroy America’s second-strike nuclear capability; (2) the enemy state’s retaliations for an American conventional preemption would not destroy U.S. nuclear counter-retaliatory capability; (3) conventional U.S. preemptive strikes would not destroy the enemy state’s second-strike nuclear capability; and (4) American retaliations for the enemy state’s conventional first strikes would not destroy the latter’s nuclear counter-retaliatory capability.
Nuclear deterrence is a “game” that America’s national security leaders must play systematically, but to compete effectively, any would-be “winner” must first assess (1) the expected rationality of each opponent; (2) each opponent’s preferences; and (3) the likelihood that those preferences would change. These are undoubtedly complex, interactive, and glaringly imprecise forms of assessment. But, just the same, they constitute a much-needed foundation for America’s long-term nuclear security. Doctrinally, it is time for them to become a much more integral part of the current Administration’s nuclear strategy.
At the start of the Cold War, the United States had only one nuclear adversary and, over time, was effectively able to reasonably discount decisional irrationality in Moscow. Hence, a posture of “mutual assured destruction” was successfully sustained. Now, however, to apply a suitable metaphor first introduced by political scientist Albert Wohlstetter, there are more than two “scorpions in the bottle.” Nuclear proliferation is all but certain, and the increased risk that a future adversary will be irrational or genuinely “mad” is deservedly worrisome. With this in mind, U.S. strategic planners must soon take far-reaching steps to ensure that any coherent and purposeful U.S. deterrence strategy will capably deal with all possible varieties of enemy, including rational, irrational and even prospectively “mad” ones.
Washington should never forget that U.S. actions could always have an impact on adversarial nuclear intentions. A more explicit U.S. policy accepting limited nuclear war might have a determinative effect upon North Korea’s or Iran’s nuclear intentions toward the United States. A sudden shift in U.S. policy or posture, especially with this President, might lead North Korean strategists and decision-makers to believe that the U.S. President is planning a preemptive strike to remove North Korean nuclear capability. Inter alia, such a belief could encourage a North Korean nuclear first-strike, regardless of the expected retaliatory consequences. In essence, a North Korean presumption concerning U.S. presidential action could “feed back” to Pyongyang and provoke it to take actions that would appear to be irrational or “mad.”
The principal conclusion to be drawn from this very concerning scenario is that the U.S. President has an overriding obligation to take nuclear strategy and diplomacy much more seriously. Whatever the precise shape of any future Washington-Pyongyang or Washington-Tehran diplomatic negotiations, this American strategy should never be regarded as analogous or comparable to ordinary forms of commercial bargaining. Looking ahead, our nuclear future is never merely about “attitude.” It must always be about “preparation.”
Louis René Beres, Complex Determinations: Deciphering Enemy Nuclear Intentions, Harv. Nat’l Sec. J. Online (Mar. 24, 2020), https://harvardnsj.org/wp-content/uploads/sites/13/2020/04/Beres_Complex-Determinations_v2.pdf.
[*] Emeritus Professor of International Law, Purdue University. Ph.D., Princeton University, 1971.
 This brings to mind Clausewitzian “friction,” which refers to the inherently unpredictable effects of errors in knowledge and information concerning strategic uncertainties on U.S. and Iranian under-estimations or over-estimations of relative power position; and on the unalterably vast and largely irremediable differences between theories of deterrence and enemy intent “as it actually is.” See generally Barry D. Watts, McNair Paper 68: Clausewitzian Friction and Future War 1–8 (rev. ed. 2004), https://apps.dtic.mil/dtic/tr/fulltext/u2/a427577.pdf [perma.cc/XL5K-9MX7].
 See, e.g., Ben Wolfgang, North Korea’s 2017 Nuke Test Was 16 Times Bigger Than Hiroshima Bomb: Report, Wash. Times (June 3, 2019), https://www.washingtontimes.com/news/2019/jun/3/n-korea-2017-nuke-test-16-times-bigger-hiroshima/ [perma.cc/A6NM-WA9U].
 Recall, in this connection, the widely cited observation of Sigmund Freud (in his book about America’s Woodrow Wilson): “Fools, visionaries, sufferers from delusions, neurotics and lunatics have played great roles at all times in the history of mankind and not merely when the accident of birth had bequeathed them sovereignty. Usually they have wreaked havoc; but not always.” Sigmund Freud & William C. Bullitt, Thomas Woodrow Wilson: A Psychological Study, at xvi (1967).
 Observes playwright Luigi Pirandello’s Henry IV: “Do you know what it means to find yourselves face to face with a madman—with one who shakes the foundations of all you have built up in yourselves, your logic, the logic of all your constructions? Madmen, lucky folk! construct without logic, or rather with a logic that flies like a feather.” Luigi Pirandello, Henry IV, in Three Plays 73, 130 (Edward Storer & Arthur Livingston trans., E.P. Dutton 1922).
 Even more specifically, expressions of decisional irrationality in world affairs could take various different and overlapping forms. These include a disorderly or inconsistent value system; computational errors in calculation; an incapacity to communicate efficiently; random or haphazard influences in the making or transmittal of particular decisions; and the internal dissonance generated by any structure of collective decision-making (i.e., assemblies of pertinent individuals who lack identical value systems or whose organizational arrangements impact their willingness to act as a single national decision maker).
 Such questions derive especially from Donald Trump’s occasional excursions into threats of gratuitous belligerence, threats that have no plausible connections to maximizing preferred U.S. outcomes. In principle, however, these excursions could more accurately reflect certain pertinent intellectual deficiencies than any genuine lapses into leadership irrationality.
See Min Joo Kim, North Korea Rejects Denuclearization Unless U.S. ‘Nuclear Threat’ Is Eliminated, Wash. Post, (Dec. 20, 2018, 6:42 AM), https://www.washingtonpost.com/world/asia_pacific/north-korea-rejects-denuclearization-unless-us-nuclear-threat-is-eliminated/2018/12/20/fcf642a2-0438-11e9-b5df-5d3874f1ac36_story.html [https://perma.cc/AF25-GBUG].
 See, with special reference to Israel, Louis René Beres & John T. Chain, Could Israel Safely Deter a Nuclear Iran?, Atlantic (Aug. 9, 2012), https://www.theatlantic.com/international/archive/2012/08/could-israel-safely-deter-a-nuclear-iran/260947/ [https://perma.cc/EC6U-YKXJ]; Louis René Beres & John T. Chain, Israel and Iran at the Eleventh Hour, OUPblog (Feb. 23, 2012), https://blog.oup.com/2012/02/israel-iran-nuclear/ [https://perma.cc/398J-F48K].
 See, e.g., Dan Smith, The US Withdrawal from the Iran Deal: One Year On, Stockholm Int’l
Peace Res. Inst. (May 7, 2019), https://www.sipri.org/commentary/expert-comment/2019/us-withdrawal-iran-deal-one-year [https://perma.cc/7FG2-YANQ].
 According to Thomas Schelling:
[T]heory that is based on the assumption that the participants coolly and “rationally” calculate their advantages according to a consistent value system forces us to think more thoroughly about the meaning of “irrationality.” Decision-makers are not simply distributed along a one-dimensional scale that stretches from complete rationality at one end to complete irrationality at the other. Rationality is a collection of attributes, and departures from complete rationality may be in many different directions. Irrationality can imply a disorderly and inconsistent value system, faulty calculation, an inability to receive messages or to communicate efficiently; it can imply random or haphazard influences in the reaching of decisions or the transmission of them, or in the receipt or conveyance of information; and it sometimes merely reflects the collective nature of a decision among individuals who do not have identical value systems and whose organizational arrangements and communication systems do not cause them to act like a single entity.
Thomas C. Schelling, The Strategy of Conflict 16 (1980).
 For the codified crime of aggression under international law, see G.A. Res. 3314 (XXIX), annex, Definition of Aggression (Dec. 14, 1974), https://undocs.org/en/A/RES/3314(XXIX) [https://perma.cc/SZ6Y-8FCJ].
 Under international law, a permissible preemption would be known as “anticipatory self-defense.” See Anthony Clark Arend, International Law and the Preemptive Use of Military Force, Wash. Q., Spring 2003, at 89, 90. This principle of customary jurisprudence has its modern origins in the so-called Caroline Case, which concerned the unsuccessful Canadian rebellion against British rule in 1837, during which British troops sank Caroline, an American ship that was allegedly assisting the Canadian rebels. See id. Following this landmark case, even the serious threat of an armed attack can sometimes be taken as sufficient justification for defensive military action. In more narrowly technical jurisprudence, the criterion of permissibility revolves around a danger presumed to be “instant, overwhelming, leaving no choice of means, and no moment or deliberation.” Id. at 90. Of course, during the first third of the nineteenth century, there could have been no conceivable thought of forestalling a nuclear aggression.
 See Chris Bolan, The Best Bad Option: Nuclear Deterrence & North Korea, U.S. Army War C.: War Room (Sept. 19, 2017), https://warroom.armywarcollege.edu/articles/best-bad-option-nuclear-deterrence-north-korea/ [https://perma.cc/39W7-LDZK].
 See Louis René Beres, Israel’s Nuclear Strategy: The Rationality of Pretended Irrationality, Arutz Sheva 7 (Nov. 4, 2016, 6:45 AM), http://www.israelnationalnews.com/Articles/Article.aspx/18702 [https://perma.cc/SY77-E6XE]. These conclusions necessarily follow from the operational definition of rationality in world politics—that is, a preference ordering that always values national self-preservation more highly than any other preference or combination of preferences.
 To suggest anything else here would be to accept the proposition that a nuclear war could conceivably be cost-effective. Although (fortunately) we don’t have any corroborative empirical information concerning the effects of a nuclear war, the “small bomb” experiences of Hiroshima and Nagasaki ought still to be adequately supportive. See Louis René Beres, Apocalypse: Nuclear Catastrophe in World Politics 123–27 (1980).
 In this connection, it must continuously be borne in mind that international law remains a part of U.S. jurisprudence. Recall the words of Justice Gray, speaking for the U.S. Supreme Court in the Paquete Habana: “International law is part of our law, and must be ascertained and administered by the courts of justice of appropriate jurisdiction . . . .” 175 U.S. 677, 700 (1900).