The fallacy at the center of this concept is the classic strategic error of assuming the predictability and controllability of inherently unpredictable and uncontrollable events.  A nuclear show of force rests entirely on the hope that enemy leaders will clearly see a demonstration shot for what it is and not overreact. But misperception is a key part of international relations, and it is pure folly to assume that a nuclear explosion will have a clarifying, rather than a panicking, effect on the enemy.

4. We’re Dragged into it by Someone Else:

Sometimes, your worst enemies are your friends.

South Africa once had a nuclear arsenal. That’s not a widely known fact; the United States had its suspicions for years and tried in both Democratic and Republican administrations to put a stop to it, but to no avail.  It’s probably not a reality anyone wants to think about too closely today, especially after the white apartheid government gave them up before handing power to the black majority at the end of the 1980s. It’s never been clear what the South African whites thought they were going to do with them, but one theory is that the weapons were meant to be an insurance policy against being overrun by some putative alliance of Soviet-armed Africans in nearby states.

The goal, however, wasn’t to kill the invading armies. Instead, it was to reveal the arsenal during the conflict – maybe even engaging in one of those “demonstration shots” – and thus spark a superpower crisis that would drag the United States into the mess and (I guess) save Pretoria at the last minute.

Even if the United States has no plans to involve itself in a nuclear conflict, U.S. allies or other powers might have ideas of their own. South Korea, for example, a few years back floated the idea that the United States might want to consider returning tactical nuclear arms to the Korean peninsula, a notion from which the Americans quickly distanced themselves. (We removed them all from South Korea in 1991, and they’re not going back.)  If Iran gets a bomb Turkey or Saudi Arabia might follow suit. In each case, the presence of a nuclear weapon might be part of a smaller nation’s national defenses, but it is more likely to be bait for the U.S. to intervene before things go nuclear.

The step along this path to war involves the intervention of another power like Russia or China. In 1973, the Soviet Union threatened to intervene militarily in the Yom Kippur War a gamble that provoked an American move to a heightened state of nuclear readiness. What happens if fighting in the Middle East or Asia involves Russia or China, and smaller decides that a South African strategy of nuclearizing the conflict is the only hope of bringing the Americans into the fray?

Small states don’t have to develop their own arsenal for things to go awry. The path to nuclear war can always involve the traditional problem of alliances, and the constant danger, whenever nuclear armed powers are in proximity to each other, that one side or the other will see nuclear weapons as their trump card in a confrontation.

5. The “Sore Loser Scenario”:

Finally, there are paths to nuclear war that rely on the most durable source of war there is: human stupidity. If the major powers don’t bumble into a nuclear war, or get dragged into one by their friends, they can always just choose to launch one themselves.

During the Cold War, NATO’s strategy was actually quite simple. We can’t defeat you, we told the Soviets, and so if you invade Western Europe, you will be placing us in a position where we will have no choice but to repel you with battlefield nuclear weapons. You Soviets, having been nuked, will have no choice but to respond, at which point the U.S., Britain (and maybe even France) will turn the USSR into glass, even as you will do the same to us. So let’s not take that fateful step, because the first rifle fired in Wurzburg will inexorably lead to the last missile that falls on Vladivostok.

This chain of deterrent logic no longer applies to possible conflicts with Russia or China, in part because there is no longer a large battlefield between the U.S. and its nuclear opponents. If war breaks out over some smaller issue, there will be no way to pull back or even stabilize a military standoff, and a military loss by China or Russia is highly likely against a far superior (yes, even today) American force.

If China, for example, decides to press a claim in the Pacific and precipitates an open conflict with the United States at sea, it will almost certainly lose. At that point, China will have to make a choice: surrender whatever was at stake, or remove the U.S. fleet from the conflict by nuclear force. Likewise, if Russia and NATO come to blows in Europe – a scenario I thought ridiculous in the 1990s and now must reconsider – Russia will also lose, and like China will have no set of buffer states around it to prevent the fight from spilling back into Russian territory.

I call this the “sore loser scenario,” since the use of nuclear arms will serve only to make the victor pay a price equal to one the loser feels has already been suffered. Theoretically, the loser lashing out against the winner might create a kind of nuclear reset, but only the most optimistic Chinese strategist could hope that an act of the magnitude of a nuclear attack on a U.S. carrier could produce a military draw. (Put another way, it’s not a demonstration or a show of force if it involves instantly incinerating 5,000 U.S. military personnel.) The United States will be forced to respond, and then we’re off to the races.

Whether the Chinese really believe they could get away with this is unclear. But the Russians do, at least to judge by their own writings. In fact, the “sore loser” concept is embedded in Russian military doctrine. Russia is acutely aware of its conventional weakness; even as they torment Ukraine right under NATO’s nose, the Russians know that they have no chance against NATO without nuclear weapons, a role reversal between NATO and Russia whose irony has not gone unnoticed in the Kremlin. And so Moscow’s strategy, as analyst Nikolai Sokov and others have pointed out, is to use nuclear weapons in a “de-escalatory” capacity: that is, when they’re losing a war, they’ll engage in a limited nuclear attack to get the enemy to back down.

Sadly, this isn’t the product of some half-baked recent order from current Russian President Vladimir Putin. Back in 1999, when Yeltsin was still in charge and Putin had only just arrived on the Moscow political scene, the Russian military held a major war game, Zapad-99 (or “West-99,” in case anyone missed the point). In this scenario NATO, for some utterly inexplicable reason, seized the Russian city of Kaliningrad, an enclave on the Baltic coast. This idea is so crazy that I am loathe to believe that any Russian planner really believed in it, but in any case, NATO had its nefarious way with poor Kaliningrad, despite heroic Russian measures. And so the game’s final act was a set of four nuclear strikes on NATO, two on Western Europe and two in North America. This ended the war and forced NATO to give back the Russian enclave.

As in the Chinese scenario, it is difficult to imagine that any Russian planner worth his salt really thinks that even the most feckless U.S. president would back down after Russian nuclear strikes kill and injure millions in Europe and America. More likely, as even Russian commentators have suggested, this might be bravado for internal political consumption. It could also be what nuclear mavens call “declaratory policy,” or the things nuclear powers say they would do in hopes of having never to do them, and which may not be very close to what they really train and prepare actually to do in a nuclear war. That would be my guess, but Zapad-99 and subsequent Russia thinking on nuclear war is, by any standard, disturbing stuff.

Unfortunately, American exercises aren’t much better. In 2006, for example, the United States conducted a drill called “Vigilant Shield” in which we were at war with an assorted group of generic bad guys, including “Ruebek,” “Nemazee” and “Churya.” (If you can’t recognize Russia, North Korea, and China, you’re not using your imagination. I suppose it was easier than calling them “Ussia-Ray” and “Orth-Nay Orea-Kay.”) In the end, America settled everyone’s hash by launching a few strategic nuclear weapons at “Ruebek,” which brought the whole mess over whatever it was to an end.

One or two nuclear bombs, apparently, are aren’t all that destructive. In fact, “Vigilant Shield” included a hypothetical enemy nuclear strike against the Pentagon (which sits within sight of the Mall and just a short distance from the White House) that only killed 6,000 people. This remarkably low figure prompted journalist and Defense Department critic William Arkin to note that two of the core assumptions in the game were obviously that “nuclear warfare can break out for no particular reason at any particular time,” and that “small nuclear weapons, while bad, don’t really kill that many people.”