“Warfare is the greatest affair of state, the basis of life and death, the way to survival or extinction. It must be thoroughly pondered and analyzed.”-Sun-Tzu, The Art of War
In an impressively-prepared television documentary, The Fog of War, one-time US Secretary of Defense Robert McNamara warns about the limits of rationality in world politics. More precisely, according to McNamara, even perfectly-reasoning nuclear adversaries could suddenly or incrementally find themselves in a war that neither party had ever wanted. Secretary McNamara, we should recall, served President John F Kennedy as his principal military advisor during the 1962 Cuban Missile Crisis.
There is more. Concluding The Fog of War, McNamara warns openly and unambiguously about America’s diminishing prospects for nuclear war avoidance. In the final analysis, we may learn from the film, it was just “dumb luck” that prevented nuclear war with the Soviet Union back in October 1962. As for the future,[1] the 1960s era “SecDef” summarizes succinctly: “Rationality won’t save us.”[2]
The core problem for major states is still national survival in an international “state of nature.”[3] More than ever before, especially during tangible crises, the United States is apt to rely on variously complex dynamics of military threat. Among other things, this means constantly changing calculations about alliance obligations, retaliations and strategic deterrence. Significantly, because a nuclear war has never been fought, gainful defense policies will need to be based on abstract deductions rather than tangible historical events.[4]
Dealing with Expectations of “Escalation Dominance”
International crises and confrontations are essentially inevitable, and the only way for powerful states like the United States to remain powerful is by demonstrating capacity and willingness to dominate high-value escalations. To best ensure such a perceived capacity, this country will need to take exceptional risks, but – simultaneously – avoid nuclear warfare. Prima facie, such existential problems should be managed intellectually before they are displayed politically.
The odds of a nuclear war (deliberate and unintentional) are not scientifically calculable. This is because meaningful probability judgments in science must always be based on the discernible frequency of relevant past events. Ironically, though it is obviously good news that there has never been a nuclear war, this fact also means that science-based predictions of nuclear conflict odds are logically impossible.
How should the incumbent American president proceed? In protecting the United States from deliberate nuclear attack, American strategists will need to accept core assumptions of enemy rationality.[5] By definition, without such assumptions, there could be no coherent theory of nuclear strategy and nuclear war.[6] Nonetheless, these same assumptions have no identifiable basis in history, and would not hold up in all potentially-impending cases,
There are further details. Critical dangers could be created by enemy hacking operations, computer malfunctions (accidental nuclear war) or decision-making miscalculation (whether by the enemy, the United States[7] or both). In the plausibly indecipherable third-case scenario, damaging synergies could arise that would prove difficult or even impossible to reverse.
Historical Context and Present Threats
In these matters, history deserves some evident pride of place. Since 1945, the global balance of power has been transformed, in considerable measure, to a “balance of terror.”[8] In crisis circumstances where at least one adversarial state party to a conflict was already nuclear, a search for “escalation dominance” could enlarge the risks of inadvertent nuclear war. Although widely underestimated, such risks would include nuclear war by accident or decisional miscalculation and could be incurred in almost any part of the world. Where only one state party was already nuclear, these risks would reference an “asymmetrical nuclear war.”
The more-or-less transient “solution” to proliferating risks is not to wish-away the ubiquitous search for “escalation dominance” (any such wish would be contrary to the “logic” of anarchic world politics[9]), but to manage all prospectively nuclear crises at their lowest possible levels of destructiveness. Wherever feasible, of course, it would be best to avoid such crises altogether and maintain reliable “circuit breakers” against strategic hacking and technical malfunction. At the same time, especially in furtherance of nuclear war avoidance, hope can never be a correct strategy.
Among other things, existential conflict risks to the United States will be related to this country’s changing alliance arrangements. Accordingly, US defense planners should focus more explicit policy attention on the expected consequences of President Donald Trump’s breach with NATO over Ukraine[10] and on Israel’s changing ties with certain Sunni Arab states. These Israeli-Sunni Arab ties center on preventing a common enemy in Shiite Iran from “going nuclear.”
Israel’s own nuclear security decisions will have serious implications for the United States. Though Israel currently has no nuclear adversaries, the rapidly accelerating approach of a nuclear Iran[11] could encourage nuclearization by Saudi Arabia, Egypt and/or Turkey. Moreover, non-Arab Pakistan[12] will likely become a more direct adversary of the United States and Israel.[13] As a generally overlooked early warning, Pakistani jihadist group Lashkar-e-Taiba carried out the large-scale attack against India (Mumbai) back in 2008.
There is more. Pakistan is an already nuclear Islamic state with close ties to China. Like Israel, Pakistan is not a party to the Nuclear Nonproliferation Treaty (NPT). Nuclear China has never renounced its right or intention to “recover” Taiwan by military force.
Remembering Clausewitz
“Everything is very simple in war,” says Carl von Clausewitz in On War, “but the simplest thing is very difficult.” With America “in the loop,” Israel should consider those circumstances where issuing specifically nuclear threats against its still pre-nuclear adversary in Tehran would seem gainful. In part, at least, conclusions would depend on Jerusalem’s prior transformations of “deliberate nuclear ambiguity” (the “bomb in the basement”[14]) into recognizable variants of “selective nuclear disclosure.” Though such considerations would necessarily concern matters that are sui generis or without historical precedent, Israel has no analytic alternative to launching deductive policy investigations.
What is the probabilistic difference between a deliberate or intentional nuclear war and one that would be unintentional or inadvertent? Without carefully considering this core distinction, little of pragmatic use could be said about the calculable likelihood of any nuclear conflict. Any such lack of consideration would be imprudent ipso facto. Because there has never been an authentic nuclear war(Hiroshima and Nagasaki don’t “count”),[15] determining relevant probabilities must always represent a sorely problematic task.
Though with greater “informality,” capable analysts and decision-makers will still have to devise optimal strategies for predicting and averting a nuclear war. Inter alia, this calculation will vary according to (1) presumed enemy intention; (2) presumed plausibility of accident or hacking intrusion; and/or (3) presumed plausibility of decisional-miscalculation. When considered together as cumulative categories of nuclear threat, these three component risks of an unintentional nuclear war should be described as “high-urgency.”
There will be assorted linguistic clarifications. Any particular instance of an accidental nuclear war would be inadvertent. However, not every case of inadvertent nuclear war would be the result of an accident. Most worrisome would be preemptive strikes that leave the victim state unsure about follow-on strikes.[16] Here, in order to protect itself against such more-or-less plausible strikes, the victim state might launch a massive or all-out retaliation.[17]
A Double-Edged Sword
Designed to guard against a US preemption, adversarial protective measures could involve the attachment of “hair trigger” launch mechanisms to nuclear weapon systems and/or the adoption of “launch on warning” policies, possibly coupled with certain pre-delegations of launch authority. This means, incrementally, that the US could sometime find itself endangered by steps taken by an enemy state to prevent or minimize an American preemption. Plausibly, the United States would do everything possible to prevent such adversarial steps because of the expanded risks of accidental or miscalculated attacks against American populations.
Nonetheless, such steps could become a fait accompli, and Washington could calculate that a preemptive strike would be legal and cost-effective. This is because the expected enemy retaliation, however damaging, would still appear more tolerable than the expected consequences of enemy first-strikes. Ironically, in this case, the American preemption would have been generated by enemy failures of “anti-preemption” measures. In principle, at least, this same ominous scenario could be played in the other direction. Here, a security-seeking United States, by deploying similarly destabilizing anti-proliferation safeguards, would spur mistaken or premature preemptive attacks by aptly apprehensive enemy states.
Per Clausewitz (see epigraph, above), more fundamental issues will need to be “pondered and analyzed” in Washington. Above everything else, such existential matters should never be approached by American national security policy-makers as a narrowly political or tactical problem. Rather, informed by in-depth historical understandings and refined analytic capacities, US military planners should prepare to deal with a large variety of overlapping threat-system hazards. At times, the analyzed intersections could prove “synergistic”[18] or force-multiplying.
For the United States, there are specifically concerning “geographies.” The North Korean nuclear threat should come immediately to mind. In terms of survival enhancements, no concept could be more important than synergy.[19] Unless such interactions were reliably evaluated, the American president could sometime underestimate the total impact of any belligerent nuclear engagement. Looking ahead, the flesh and blood consequences of such underestimations would likely defy analytic imaginations and post-war justifications.
Staying the Collision Course or Advancing Beyond “Dumb Luck”
In any global “state of nature,” there is little likelihood that the corrosive dynamics of nuclear risk-taking and nuclear deterrence would fade away on their own. Operating rationally in our centuries-old world system of belligerent nationalisms, the US president and his counterparts in Russia, China, India, Pakistan, North Korea, Israel and elsewhere will seek to prevail in multiple and possibly-interrelated struggles for “escalation dominance.” Amid unrelenting global anarchy, these leaders would have no real choice but to stay tethered to a “scripted” geopolitical course.
Over time, no matter how carefully, responsibly and rationally each state’s security preparations are carried out, an international order based on incessant power struggles will fail. For the moment, the principal risk of such catastrophic failures stems from unintentional nuclear war. It follows, recalling Sun-Tzu’s timeless wisdom, that such existential risk “must be thoroughly pondered and analyzed.” Properly, this analytic task is a matter solely for disciplined thinkers and strategic theorists.[20] Under no circumstances should any primary intellectual responsibilities be handed off to politicians or government officials. Next time around, prima facie, America could run out of McNamara’s “dumb luck.”
[1] Recalling McNamara’s experience-based warning, it is a grave mistake for analysts and politicians to assume that a nuclear war between states is preventable or manageable by deliberate and rational decision-making processes. Plausibly, the greatest current risks of a nuclear war (NATO/Russia, US/North Korea; India/Pakistan, China/Taiwan) are ones of inadvertence. Such an unintentional war is more or less probable even if only one conflict party is operationally nuclear. Technically, this oft-overlooked scenario references an “asymmetrical nuclear war.”
[2] Years later, during the Fall of 1967, McNamara, then President of the World Bank, spoke at a small private Princeton gathering of former Kennedy-Johnson era national security officials. At that event, the present writer, then a first-year graduate student of international law at Princeton, found himself seated next to the speaker and asking McNamara about what had happened back in October 1962. Even then, nodding laconically to me, his answer came down to one portentous word: luck.
[3] Seventeenth-century English philosopher Thomas Hobbes notes that although the “state of nations” is in the anarchic “state of nature,” it is still more tolerable than the condition of individuals coexisting in nature. With individual human beings, Hobbes instructs, “…the weakest has strength enough to kill the strongest.” But with the continuing advent of nuclear weapons, there is no persuasive reason to believe that the state of nations remains more tolerable. Now, nuclear weapons are bringing the state of nations closer to the true Hobbesian state of nature. See, in this connection, David P. Gauthier, The Logic of Leviathan: The Moral and Political Theory of Thomas Hobbes (Oxford: Oxford University Press, 1969), p. 207. As with Hobbes, philosopher Samuel Pufendorf argues that the state of nations is not quite as intolerable as the state of nature between individuals. The state of nations, reasons the German jurist, “lacks those inconveniences which are attendant upon a pure state of nature….” In a similar vein, Baruch Spinoza suggests “that a commonwealth can guard itself against being subjugated by another, as a man in the state of nature cannot do.” See, A.G. Wernham, ed., The Political Works, Tractatus Politicus, iii, II (Clarendon Press, 1958), p. 295.
[4] In the words of mathematical strategist Amatol Rapoport, “Formal decision-theory does not depend on data…. The task of theory is confined to the construction of a deductive apparatus, to be used in deriving logically necessary conclusions from given assumptions.” (Strategy and Conscience; 1964).
[5] Expressions of enemy irrationality could take different and/or overlapping forms. These include a disorderly or inconsistent value system; computational errors in calculation; incapacity to communicate efficiently; random or haphazard influences in the making or transmittal of particular decisions; and 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 otherwise willing capacity to act as a single (unitary) national decision maker.
[6] “Theory is a net,” says philosopher of science Karl Popper in The Logic of Scientific Discovery (1934): “Only those who cast, can catch.” Interestingly, Popper drew this metaphor from the German poet Novalis.
[7] See by this writer, Louis René Beres, at Bulletin of the Atomic Scientists: https://thebulletin.org/biography/louis-rene-beres/
[8] This term was originally popularized by distinguished political scientist Albert Wohlstetter in his classic article “The Delicate Balance of Terror” (1959): Vol. 37, No. 2 (Jan., 1959), pp. 211-234 (Council on Foreign Relations).
[9] See by this author, Louis René Beres, at JURIST, 2022: https://www.jurist.org/commentary/2022/01/louis-beres-international-law-state-of-nature/
[10] Regarding this breach, nothing of strategic consequence was changed by the May12025 rare earth minerals deal between the United States and Ukraine.
[11] On deterring a potentially nuclear Iran, see: Louis René Beres and General John T. Chain, “Could Israel Safely Deter a Nuclear Iran?” The Atlantic, August 2012; and Professor Louis René Beres and General John T. Chain, “Israel and Iran at the Eleventh Hour,” Oxford University Press (OUP Blog), February 23, 2012. General Chain (USAF/deceased) had served as Commander-in-Chief, U.S. Strategic Air Command (CINCSAC).
[12] Pakistan recently reaffirmed its right to “fist-use” of nuclear weapons, and has been deploying nuclear warfighting weapons for the past several years. https://www.jurist.org/commentary/2025/04/pakistans-first-use-nuclear-policy-in-conflicts-with-india/ In this connection, conspicuous preparations for nuclear war fighting could be conceived not as provocative alternatives to nuclear deterrence, but rather as essential and integral components of nuclear deterrence. Some years ago, Colin Gray, reasoning about U.S.-Soviet nuclear relations, argued that a vital connection exists between “likely net prowess in war and the quality of pre-war deterrent effect.” (See: Colin Gray, National Style in Strategy: The American Example,” INTERNATIONAL SECURITY, 6, No. 2, fall 1981, p. 35.) Elsewhere, in a published debate with this writer, Gray said essentially the same thing: “Fortunately, there is every reason to believe that probable high proficiency in war-waging yields optimum deterrent effect.” (See Gray, “Presidential Directive 59: Flawed but Useful,” PARAMETERS, 11, No. 1, March 1981, p. 34. Gray was responding directly to Louis René Beres, “Presidential Directive 59: A Critical Assessment,” PARAMETERS, March 1981, pp. 19 – 28.).
[13]See by this writer, Louis René Beres, https://moderndiplomacy.eu/2021/09/12/can-israeli-nuclear-threats-protect-against-non-nuclear-attacks/#_ftn
[14] See, by this writer, Louis René Beres, at Strategic Assessment (Israel): 2014: https://www.inss.org.il/wp-content/uploads/systemfiles/adkan17_3ENG%20(3)_Beres.pdf
[15] The atomic attacks on Japan in August 1945 represent nuclear weapons use in an otherwise conventional war.
[16] In law, any such preemption (a defensive first-strike) if permissible, could be considered as “anticipatory self-defense.” The normative origins of such defense liein customary international law, more specifically in The Caroline, a case that concerned the unsuccessful rebellion of 1837 in Upper Canada against British rule. Following this landmark case, the serious threat of armed attack has generally justified certain militarily defensive actions that conform to the laws of armed conflict (jus in bello). In an exchange of diplomatic notes between the governments of the United States and Great Britain, then U.S. Secretary of State Daniel Webster outlined a framework for self-defense that did not require an antecedent attack. Here, the jurisprudential framework permitted a military response to a threat so long as the danger posed was “instant, overwhelming, leaving no choice of means, and no moment for deliberation.” See: Beth M. Polebaum, “National Self-defense in International Law: An Emerging Standard for a Nuclear Age,” 59 N.Y.U.L. Rev. 187, 190-91 (1984) (noting that the Caroline case had transformed the right of self-defense from an excuse for armed intervention into a legal doctrine). Still earlier, see: Hugo Grotius, Of the Causes of War, and First of Self-Defense, and Defense of Our Property, reprinted in 2 Classics of International Law, 168-75 (Carnegie Endowment Trust, 1925) (1625); and Emmerich de Vattel, The Right of Self-Protection and the Effects of the Sovereignty and Independence of Nations, reprinted in 3 Classics of International Law, 130 (Carnegie Endowment Trust, 1916) (1758). Also, Samuel Pufendorf, The Two Books on the Duty of Man and Citizen According to Natural Law, 32 (Frank Gardner Moore., tr., 1927 (1682).
[17] Conceptually, it is necessary to distinguish preemptive attacks from preventive attacks. Preemption is a military strategy of striking first in the expectation that the only foreseeable alternative is to be struck first oneself. A preemptive attack, on the other hand, is launched by a state that believes enemy forces are about to attack. A preventive attack is launched not out of genuine concern about “imminent” hostilities, but for fear of a longer-term deterioration in military balance. In a preemptive attack, the length of time by which the enemy’s action is anticipated is presumptively very short; in a preventive strike, the anticipated interval is considerably (though not measurably) longer. A related problem here for the United States is not only the practical difficulty of determining “imminence,” but also that delaying a defensive strike could prove “fatal.”
[18] By definition, the “whole” of any synergistic effect would be greater than the sum of its “parts.” Accordingly, focused attention on pertinent synergies should become a primary national security objective for the United States.
[19] See by this author, at Harvard National Security Journal (Harvard Law School): Louis René Beres, https://harvardnsj.org/2015/06/core-synergies-in-israels-strategic-planning-when-the-adversarial-whole-is-greater-than-the-sum-of-its-parts/
[20] Rabbi Eleazar quoted Rabbi Hanina, who said: “Scholars build the structure of peace in the world.” See: The Babylonian Talmud, Order Zera’im, Tractate Berakoth, IX.
