Support The Bulwark and subscribe today.
  Join Now

We Don’t Know Anything About Nuclear War

Not how it would come about and not what it would look like.
March 11, 2022
We Don’t Know Anything About Nuclear War
(Jim Heimann Collection / Getty Images)

The potential for a nuclear war has been a prominent feature of every serious discussion about U.S. military intervention in Ukraine. That’s as it should be. But one thing that is too often missing from these discussions is an acknowledgement of how little we actually know about nuclear warfare.

On February 27, Vladimir Putin raised the alert level of his country’s nuclear forces—putting them on “special regime of combat duty.” The operational meaning of that heightened nuclear status is unclear, but the intended signal is plain: Be warned. The move came just days after Russia staged nuclear-weapons maneuvers intended to deter NATO intervention in Ukraine. That earlier round of muscle-flexing prompted the French minister of foreign affairs, Jean-Yves Le Drian, to remind Putin that NATO also has nukes.

Putin’s war on Ukraine has brought the world closer to nuclear war than any time since the Cuban missile crisis of 1962, itself a leftover from the Berlin crisis of 1958–61. (The 1983 Soviet false-alarm incident is a close runner-up.) When in 1961 Paul H. Nitze told former President Dwight Eisenhower that none of the Kennedy administration’s 128 contingency plans for the Berlin crisis had been realized, the soldier-statesman responded with one of his mottos: “Plans are useless, but planning is indispensable.” War is a human affair, and its prosecution and outcome are never perfectly foreseeable. Putin’s war on Ukraine is a reminder of Clausewitz’s famous observation:

War is a contest of wills; it is unpredictable; it is the domain of accident and contingency; nothing goes as planned; and events are smothered in a fog created by misinformation and fear.

That understanding of the unpredictability and uncertainty of war ought to inform the debates about the war in Ukraine—including the debates about the possibility of nuclear weapons being used. It’s fine to make predictions about what might happen. It’s fine to express fear or confidence about specific possibilities. But no one should feel great certitude in their positions.

Because, despite decades of theorizing, including by very smart people—like Nitze, Albert and Roberta Wohlstetter, Herman Kahn, Henry Kissinger, and Thomas Schelling—we really just don’t know what nuclear war would look like.

Only twice have conflicts erupted between two nuclear-armed states, and both instances (between China and Russia in 1969 and India and Pakistan in 1999) were minor border disputes rather than major wars, in great part because the states involved were wary of escalation to nuclear level. Nobody knows for sure how different actors would react if nuclear-armed states fought in earnest. Known unknowns are scary, and the known unknown of a war between the United States and Russia is frightening enough even before thinking about escalation to the nuclear level. The overuse by the press and Hollywood of such words as “armageddon” and “apocalyptic” has debased the available terminology, leaving the English language without a word sufficiently powerful and grim to convey the prospects of a massive thermonuclear war between the United States and Russia.

It would be wrong to assume that, if the U.S. military enters the conflict in Ukraine in some capacity, nuclear war would inevitably follow. First, it wouldn’t be the first time that Americans and Russians have killed each other. The Soviets reflagged their aircraft during the Korean and the Vietnam Wars and directly engaged with their American enemy. They also launched missiles at American aircraft during the war. The United States, on the other hand, has killed Russian mercenaries as recently as 2018 with no ramifications. In both cases, the Russians benefited from (im)plausible deniability.


Many theories and assumptions about nuclear war, including some that influenced major policy decisions, later came to be discarded or even discredited. Casualty estimates have moved up and down over the years, as knowledge from relevant fields has evolved and modeling techniques improved. In the mid-1970s, various bigwig analysts predicted nuclear war by the century’s end. In the 1980s, Carl Sagan popularized the concept of “nuclear winter”—the idea that a large-scale nuclear war would be followed by dangerous global cooling—but within years, other scientists started to reject the theory in favor of more modest estimates of nuclear war’s potential atmospheric effects. The doctrine of mutual assured destruction, which figured prominently in Cold War policy thought, is completely discredited now that people have realized that it takes many more nukes than previously assumed to annihilate the United States or Russia.

Advances in empirical data and modeling are not the only things that change the conventional wisdom about nuclear war. New theories about politics—about what motivates different political actors and how best to influence them—do too.

The things we can be most certain of are the things closest to the questions of physics and engineering. We know that the Russians come short of the United States in delivery capabilities of intercontinental ballistic and sea-launched missiles and bombs, and that—thanks to the forward deployment of American nukes to Europe—it is much easier for the United States to hit the Russian homeland than vice versa. We know that the U.S. arsenal is inferior in theater nuclear weapons—nukes with smaller blasts designed for battlefield warfare—and some commentators have suggested that Putin might use one such weapon, perhaps on a Ukrainian military installation.

If we are to think and debate more clearly about the possibility and the prosecution of a nuclear war between the United States and Russia, we should begin by asking four questions: First, would Putin order a nuclear strike? Second, would those in the Russian chain of command obey the order? Third, what are the chances that a small nuclear attack will escalate to thermonuclear weapons? Fourth, how can we manipulate the answers of the previous three questions in our favor?

Those are fine questions for study and debate. But nobody who claims certainty in answering them is a credible commentator. Perhaps no material question in human history has been accompanied by a wider gap between its potential effects and human knowledge of it. This begs for epistemological modesty.

Shay Khatiri

Shay Khatiri studied Strategic Studies at the Johns Hopkins University School of Advanced International Studies. He’s an immigrant from Iran and writes the Substack newsletter The Russia-Iran File.