To the editors:
In his essay, Jean-Pierre Dupuy writes that there is a remorseless logic to mutually assured destruction (MAD), a strategy employed by nuclear-armed states in an effort to keep themselves safe. He explores the topic using modal logic in a way that is rational, logical, and subtle, but also deeply troubling.
Dupuy’s approach to the topic mirrors the thinking and approach of the strategic community over the last seventy years. It has been an approach that ponders the costs and benefits of war, wrestles with the dangers of vast and devastating attacks, and that ends by concluding we are stuck with this appalling danger in perpetuity.
Both Dupuy and the strategic community have made assumptions that undermine the reality of their arguments.
Dupuy makes the same mistake as Thomas Schelling in assuming that the world of human affairs is rational. For many years, strategists have assumed that just as a mad dog causes reasonable people to steer clear, so feigning madness confers an advantage in a confrontation. But this also assumes that your adversary is reasonable. There can be no such assurances. The assumption that actors are rational is the cardinal mistake made by most Cold War-era nuclear weapons thinkers. Acting crazy can incite crazy actions in return. The Madman theory fails because acting like a madman could lead to your adversary losing his head.
Human beings are, for the most part, emotional. Sometimes we think and sometimes we think very clearly. But we always feel. And on many more occasions than we would like to admit, our feelings are what drive our actions. Many people think of their conscious, rational selves as a race car driver: deftly navigating a dangerous world whilst remaining alert and decisive at high speed. The mind, they say, is a miracle of control. But it might be more accurate to think of our rational mind as a friend trailing behind a drunk, apologizing and explaining away errant behavior. We are far better at retrospectively rationalizing our emotion-driven behavior than making rational decisions in the first place and sticking to them.1
If you don’t believe me, ask yourself why much of the developed world is overweight. How is it that despite all our knowledge about diet and nutrition, we are still unable to resist crème brûlée? It makes no rational sense to be heavier than your ideal weight, but the fact that so many people are overweight seems to indicate that rational sense isn’t driving the car.
Cold War nuclear weapons thinking embraces an Enlightenment view of human beings that was swept away long ago. It is no accident that this year’s Nobel Prize for Economics was awarded to Richard Thaler, whose work embraces psychology and examines the unpredictability of human choices. Rational actor models are increasingly viewed as unrealistic. Nation states rarely make purely rational decisions about whether to go to war. These decisions driven as much by emotion as any of our other choices.
Dupuy writes that nuclear deterrence involves a paradox. But there is nothing mysterious or paradoxical about nuclear deterrence. On balance, it works reasonably well. The problem is that a mechanism for holding back nuclear war that works reasonably well just won’t do. Nuclear deterrence has to be perfect. Even a single catastrophic nuclear war is unacceptable. But nuclear deterrence cannot be perfect. It is a process that involves human beings—we make the threats, we evaluate them, and we decide how to respond. If human beings are involved in nuclear deterrence, then nuclear deterrence is inherently fallible—it will inevitably fail. Given how overwhelming the desire for revenge can be, there is a fairly good chance that any deterrence failure will lead to a catastrophic nuclear war.
Consider the actions of Fidel Castro during the Cuban Missile Crisis. We now know that Castro wrote a long letter to the Soviet leader Nikita Khrushchev at the height of the crisis urging him, if war seemed imminent, to launch a nuclear strike against the United States.2 With his country in the crosshairs, Castro urged Khrushchev not to hesitate. Here then is a verified case of a world leader urging nuclear war in a situation despite the costs vastly outweighing any possible benefit. Any claim that decisions to go to war are always based on a cost-benefit analysis simply doesn’t hold up.
For many years, the nuclear-armed states divided the world into so-called reliable and unreliable states. Sergio Duarte, former High Representative for Disarmament at the United Nations, once listed twelve common criticisms of disarmament. Number eleven was that “[d]isarmament should only apply to states that are unreliable.”3 US President Donald Trump is proof that there are no reliable countries where nuclear weapons are concerned. If even stable democracies can elect a dangerously unsuitable leader, then nuclear deterrence is unsustainable in the long term.
When you think about nuclear weapons, it is not necessary to be appalled or horrified. It is merely necessary to be realistic and to act.
Ward Wilson
Jean-Pierre Dupuy replies:
There is nothing more painful for a writer, especially a philosopher, and, in particular, one who endorses the criteria of clarity and rigor of analytic philosophy, than to realize that he has been radically misunderstood. Ward Wilson lumps me together with the strategic community, accusing me of committing the “cardinal mistake made by most Cold War-era nuclear weapons thinkers” by assuming that “the world of human affairs is rational.” I cannot reproach Wilson for being unaware of my background. A text stands for itself, whoever the author may be. If it is misread, it is, most of the time, the writer’s fault. I have spent a good part of my intellectual life criticizing the rational choice paradigm—see my Economy and the Future: A Crisis of Faith (2014)—and an even larger part of my political life attacking my country’s nuclear policy. I am a board member of Initiatives pour le Désarmement Nucléaire (Initiative for Nuclear Disarmament), a French organization equivalent to the Global Zero movement in the US. It would also be unfair to blame Wilson for not having read the much longer version of my essay published in French as part of the same issue. Had he read the longer version, he would have seen that his critique was irremediably off target.
Wilson might be surprised to discover that he is infinitely more rationalistic than I am. Whenever he wants to make the point that we humans are not rational, Wilson shows us driven by all kinds of disruptive passions—envy, jealousy, resentment, hatred, desire for revenge, weakness of the will, etc.—that render us flawed and inherently fallible. The fact that we are flawed implies that we are not rational. It follows logically that if we were rational, we would not be flawed. For Wilson, to be rational is always a positive predicate. To be irrational is to lack some ideal although inaccessible faculty. This is profoundly wrong.
The kind of rationality Wilson asserts we lack is the the cost-benefit analysis of homo economicus, otherwise known as the strategic rationality of game theory, or the instrumental rationality of moral philosophy. If we were rational in this sense, does he really believe that nuclear deterrence would work perfectly and peace would be guaranteed?
To be rational in this sense can lead to abominations. Consider Caiaphas addressing himself to the pure reason of the high priests and Pharisees. He admonishes them: “Ye understand nothing. Nor consider that it is expedient for us, that one man should die for the people, and that the whole nation live.”4 To be rational, one must follow the example of Caiaphas and make the sacrificial choice. The appeal of this rationality is momentous. Sacrifice is not only the choice dictated by a calculus of consequences. Let us not forget that Jesus is a Jew. Whether his sacrifice takes place or not, he is eliminated. This kind of sacrifice, which maximizes efficiency, is called Pareto-optimum in economic theory. It is a choice that appears the height of infamy to us today. This is precisely the conclusion I reach about nuclear deterrence. As I write in the French version of my essay, “It is indeed possible to provide rational foundations to the efficiency of nuclear deterrence. And that conclusion is horrendous.”
The sharp opposition that Wilson draws between rationality and craziness or madness, or between reason and sentiments or passions, has long been shown to be untenable. Cognitive science has been convinced of this since its beginnings in cybernetics.5 More recently, Antonio Damasio has shown that far from opposing reason, emotions and feelings are the terrain from which reason emerges.6 I have argued elsewhere that human reason retains traces of its origins in the sacred.7
The MAD logic of nuclear deterrence embodies this paradox. During the summer of 2017, after observing the escalating threats exchanged by Trump and North Korean Leader Kim Jong-un, many commentators referred to the Madman theory, which they ascribed to US President Richard Nixon. The Madman theory adds nothing to the MAD doctrine, which asserts that in order to maintain the so-called equilibrium of terror between two nuclear powers, each must threaten to reduce the other to radioactive ashes whilst also committing suicide.8 To commit such an act would be madness, but forming the intention to commit it would be the height of rational prudence. Wilson is wrong in maintaining that the standard theory of nuclear deterrence requires the assumption that actors are rational. As Martin Hellman notes in his response to my essay, irrationality is required. Each actor must be at the same time rational and crazy. The capacity to put oneself at a distance from oneself is needed—a capacity which, according to Diderot, defines the art of the actor.9 Wilson isn’t fond of the word paradox. Indeed, if his thinking is anything but paradoxical it is because it shirks the complexities of the problem. One could argue that Presidents John F. Kennedy and Nixon were capable of this duplicity. Trump, on the other hand, seems entirely incapable. This is our misfortune.
When one reaches the level of the metaphysical foundations of nuclear doctrines, concepts, and theories, things become much more complex. This response is not the place to repeat or summarize what I have tried to demonstrate in my essay. In the same way that there exist several axiomatizations of geometry that are all consistent and complete, there are several mutually exclusive ways to axiomatize the logic of human action plunged in time. A different concept of rationality corresponds to each. The only rationality Wilson is taking into consideration will be familiar to economists, game theorists, planners and decision makers of all stripes. But it is unable to provide foundations for the standard theory of nuclear deterrence. I’ve shown that another doctrine, existential deterrence, could be grounded in a very different metaphysics of time. It belongs to a philosophical tradition that includes names such as Henri Bergson, Martin Heidegger, and Jean-Paul Sartre, amongst others. Its major categories have nothing to do with rational choice or cost-benefit analysis. They include notions like fate and accident, and imply the absolute renunciation of strategic thinking. In this metaphysics too, being guided by reason is not a guarantee of goodness. In its light, nuclear deterrence appears to be grounded in reasons; it is nonetheless abominable.