EDITOR’S NOTE: This article originally appeared at TomDispatch.com. To stay on top of important articles like these, sign up to receive the latest updates from TomDispatch.com.

Most people intuitively get it. An American preventive strike to wipe out North Korea’s nuclear bombs and ballistic missiles, or a commando raid launched with the same goal in mind, is likely to initiate a chain of events culminating in catastrophe. That would be true above all for the roughly 76 million Koreans living on either side of the Demilitarized Zone. Donald Trump, though, seems unperturbed. His recent contribution to defusing the crisis there: boasting that his nuclear button is “bigger and more powerful” than that of North Korean leader Kim Jong-un.

The president’s high-school locker-room braggadocio provided rich material for comedians and maybe for shrinks. Meanwhile, there remains the continuing danger of a war in the Koreas, whether premeditated or triggered accidentally by a ship seized, an aircraft downed, a signal misread… you get the picture. No serious person could dismiss this scenario, but even the experts who track the evidence closely for a living differ on just how probable it is. In part, that’s because, like everyone else, they must reckon with a colossal wild card—and I’m not talking about Kim Jong-un.

The Pessimists

On one side are those who warn that President Trump isn’t blowing smoke when he talks, or tweets, about destroying North Korea’s nuclear warheads and missiles, the infrastructure supporting them, and possibly even the whole country. By now, it’s common knowledge that his national security officials—civilian and military (the distinction having blurred in the Trump era)—have been crafting plans to strike before that country’s nuclear arsenal becomes fully operational.

No one who listened to PBS NewsHour’s Judy Woodruff interviewingNational Security Adviser Lt. Gen. H.R. McMaster just after the Trump administration released its National Security Strategy in December could simply dismiss the warnings as those of so many Cassandras. McMaster dutifully summarized that document, which included a pledge to “respond with overwhelming force to North Korean aggression and improve options to compel denuclearization.” When Woodruff then asked whether he believed war was becoming more likely by the day, he agreed, adding that “the president has asked us to continue to refine a military option, should we need to use it.”

Others who should be in the know have offered even scarier prognoses. During an interview with ABC News on the last day of 2017, former chairman of the Joint Chiefs of Staff Adm. Mike Mullen claimed that, while McMaster and Defense Secretary James Mattis had stayed Trump’s hand so far, their ability to continue to restrain such a “disruptive” and “unpredictable” president was diminishing. “We’re actually closer to nuclear war with North Korea and in that region,” he concluded, “than we’ve ever been.”

Then there’s Trump himself. He has long since moved from saying, as he did last May, that he would “be honored” to meet Kim Jong-un “under the right circumstances” to warning, in August, that if North Korea threatened the United States, it would “be met with fire and fury like the world has never seen.” In September, he upped the ante again in a speech to the UN, declaring that he would “have no choice but to totally destroy North Korea” if that were needed to defend the United States.

Left unspecified was Trump’s definition of “defend.” Would additional North Korean nuclear and missile tests pose a sufficient threat for him to order a preventive war? Was his red line a fully operational North Korean nuclear force? Or did he mean that he would retaliate in kind only if Pyongyang were to attack the United States, Japan, or South Korea with nuclear weapons? If either the first or second scenario represents his threshold, then Mullen’s dire assessment can’t be discounted as hyperbole. If it’s the third, the world can breathe a bit easier for now, since there’s no conceivable reason for Kim Jong-un to attack a country with nuclear weapons, least of all the United States, except in response to the potential destruction of his state.

In his latest gyration, having failed to scare Kim into denuclearization, Trump has welcomed talks between Seoul and Pyongyang that he had only recently discounted and, predictably, taken credit for a turn of events that has sidelined him. He even suggested that the United States could eventually join the negotiations, meant in part to prevent a conflict during the February Winter Olympics in Seoul, and reacted positively to the possibility that they might continue even after the games end.

Of course, this president can turn on a dime, so such words mean next to nothing and should offer no solace. After all, on two occasions he derided Secretary of State Rex Tillerson’s efforts to defuse the crisis through negotiations, declaring, “I told Rex Tillerson, our wonderful Secretary of State, that he’s wasting his time trying to negotiate with little Rocket Man. Save your energy, Rex, we’ll do what has to be done.”

The Optimists (Well, Sort of)

On the opposing side of the how-likely-is-war debate are the optimists, a different coterie of journalists, ex-officials, and policy wonks. Their basic point boils down to this: Yes, Trump has made fire-and-brimstone statements about North Korea, but chalk up the endless bombast to his problem with impulse control and his desire to feed red meat to his base, while scaring Kim.

Unfortunately, you can’t put much stock in this take either—not once you consider the accompanying caveats. Gideon Rachman, an Asia specialist and Financial Times columnist, is typical of this crew in concluding that war on the Korean peninsula is unlikely—only to liken the current atmosphere in Washington to the one that prevailed just before the 2003 Bush administration invasion of Iraq. For good measure, he adds that Lindsey Graham—super-hawk, Trump confidant (to the extent that anyone is), and member of the Senate Armed Services Committee—believes that war is “inevitable.” (This is optimism?) Rachman’s fallback suggestion is that Australia, Japan, and South Korea won’t support a preventive strike on North Korea. Now ask yourself this: How often does Donald Trump take others’ advice? When is the last time you heard him say “multilateralism”?

Jeffrey Lewis, a well-regarded expert on nuclear weapons, discounts the likelihood of war for a different reason. He thinks Trump’s bombast is so much bluster, designed to jangle Kim’s nerves and drive the North Korean leader to relinquish his nuclear cache lest an out-of-control American president vaporize his regime. Given what we now know about the present occupant of the Oval Office, that might be a modestly convincing thought if Lewis didn’t introduce his own qualifiers. He believes Trump’s faith that China, in hopes of getting economic rewards from the United States, will eventually persuade (or coerce) Kim to denuclearize is misplaced because Beijing lacks the necessary clout in Pyongyang. Indeed, Kim doesn’t trust China and has killed or sidelined those whom he suspects of being pro-Chinese.

Lewis also lays out a range of possibilities, each of which could trigger a spiral toward war. These include North Korea shooting down an American reconnaissance aircraft or sinking a South Korean naval vessel, both of which, he reminds us, Pyongyang has done in the past (the first in 1969, the second in 2010)—when it still lacked nuclear weapons. So Lewis’s American-style optimism doesn’t offer any more grounds for cheer than Rachman’s British variant.

Where does this lack of consensus on the likelihood of war leave us? The answer: No one can really assess the gravity of the danger, particularly because the man who occupies the White House is arguably the most volatile president we’ve ever had.

It’s no pleasure to quote once–Secretary of Defense Donald Rumsfeld, but when it comes to the probability of war in the Koreas, it’s hard not to be overwhelmed by the “known unknowns.”

What We Do Know

The inability to fathom just how close we may be to war there doesn’t mean we know nothing about the Korean crisis that’s worth knowing.

We know that North Korea has long been committed to building nuclear weapons and produced small quantities (six to 13 kilograms) of weapons-grade plutonium as early as 1992.

We know that North Korea withdrew from the Nuclear Non-Proliferation Treaty (which it joined in 1985) in 2003; that it detonated its first nuclear weapon in 2006 during the rule of Kim Jong-il, the father of North Korea’s current leader; and that it has conducted five other tests since then in 2009, 2013, 2016 (twice), and 2017—four of them after Kim Jong-un took power in December 2011.

We know that North Korea has been no less dogged in building and testing ballistic missiles, beginning in 1984, and that the Hwasong-15, test-fired last November (with an apogee of 2,800 miles and an estimated range of 8,100 miles), has the capacity to strike the continental United States. And Pyongyang has gone beyond liquid-fueled missiles (that require prolonged, telltale preparations to launch), testing solid-fueled variants, which can be fired at short notice.

We know that Pyongyang is close to producing, or has already produced, a warhead that can be placed atop an intercontinental ballistic missile and survive the heat and stress encountered on reentering the earth’s atmosphere. In other words, North Korea is without question effectively a nuclear-weapons state, which means Kim Jong-un’s claim, in his 2018 New Year’s Day speech, that he has a nuclear button on his desk may not be an idle boast (even if no literal button exists).

Finally, we know that American threats and military maneuvers on and around the Korean peninsula, a series of UN Security Council sanctions since 2006, and behind-the-scenes diplomacy by China and Russia have not induced Pyongyang to change course, even though China, in particular, recently imposed draconian limits on energy exports to that country, which could potentially cripple its struggling economy.

The Denuclearization Fantasy

No one (outside of Pyongyang) could celebrate a nuclear-armed North Korea, but no one could reasonably be surprised by it either. Nuclear weapons have long served as a symbol of exclusivity for great powers and their regional cohorts. It’s no accident that all the Security Council’s permanent members are nuclear states. Having accorded such weaponry supreme prestige, who could be shocked that other countries, even relatively small and poor ones, would try to acquire them as well and refuse to be cowed by political or economic pressure.

Despite various campaigns for nuclear disarmament, the current nuclear states have not shown the slightest inclination to give them up; so the promise of a nuclear-free world rings hollow and is unlikely to persuade states that really want nukes not to build them. Beyond conferring status, these weapons make attacking a country that has them dangerous indeed, providing a de facto guarantee against regime change.

The North Koreans have made this point more than once, citing the fates of Iraq’s Saddam Hussein and Libya’s Moammar El-Gadhafi, each of whom gave up his country’s nuclear program and then was taken down by the United States. The idea that the leaders in Pyongyang are simply paranoid maniacs or can’t possibly believe that they face such a threat from the United States (which already fought one war on the Korean peninsula) is preposterous. If you were Kim Jong-un, you’d probably build nuclear weapons.

The upshot: Short of a war, there’s no chance of denuclearization. That, in turn, means: Were Trump and his generals to launch an attack on North Korea’s nuclear arsenal and even a single warhead capable of striking the United States survived, Pyongyang might well use it to retaliate. According to the experts who engage in such grisly estimates, a 15-kiloton nuclear weapon (equivalent to “Little Boy,” the atomic bomb the United States dropped on Hiroshima, Japan, on August 6, 1945) that lands somewhere in, say, Los Angeles would kill more than 100,000 people immediately and yet more thereafter. To put this in perspective, bear in mind that the estimates of the yield of the warhead North Korea tested last September run as high as 250 kilotons. And don’t forget that, even if it couldn’t effectively reach the United States, the North could still target either South Korea or Japan, causing a devastating loss of lives and sending shock waves through the global economy.

And even if Kim couldn’t retaliate with nuclear weapons, he could still order the thousands of artillery pieces his military has trained on the South Korean capital, Seoul, to fire. The metropolis and its satellite towns are home to nearly 25.5 million people, half of the country’s total population, so the death toll would be enormous, even taking into account the limitations of the North’s artillery. And given that some 28,500 American troops and nearly 137,000 American civilians are based in South Korea, many close to the border, Trump’s reported remark to Lindsey Graham that, in the event of such a war, people will “die over there” is not just callous in its disregard for Korean lives, it’s ignorant. Even an American commando raid into North Korea could trigger a wider war because the North Korean leadership might reasonably regard it as a prelude to a larger attack.

The bottom line? Trump could fulfill his vow never to allow North Korea to become a nuclear-armed power only by resorting to a preventive war, as Pyongyang hasn’t been and is unlikely to be moved to disarm by sanctions or other forms of pain. And a preventive war would be calamitous.

Stopping the War Machine

Here’s a prerequisite for avoiding war in Korea: Stop believing in the North’s denuclearization, attractive and desirable as it might be (if achieved through diplomacy).

It doesn’t follow, however, that war can’t be avoided. Kim Jong-un and his inner circle are not, in fact, irrational beings immune to deterrence. Their paramount aim is to ensure the survival of the North Korean state. Starting a nuclear war would destroy it. Yes, many people have perished in North Korea (whether due to repression or famine), but deterrence worked in the cases of Soviet dictator Joseph Stalin and China’s Mao Zedong, both of whom enacted policies that killed millions. Mao supposedly even boasted that China could survive a nuclear war because of its huge population.

Coming to terms with the reality of a nuclear-armed North Korea and trusting in deterrence may not sound like a perfect ending, but under the circumstances it’s undoubtedly the best way to avert catastrophe. And that, unquestionably, is the urgent task. There are other ways, down the line, to make the Korean peninsula a better place through dialogue between the two Koreas, by drawing the North into the regional economy, and reducing troops and weaponry on both sides of the Demilitarized Zone. These shouldn’t be ruled out as infeasible.

For them to happen, though, South Korea would have to separate itself from Trump’s war plans by refusing to allow its sovereign space (land, sea, and air) to be used for such a preventive war. The symbolism would be important even if Trump could strike in other ways.

Seoul would also have to build on two recent positive developments that emerged from a surprise January 9 meeting between the Koreas. The first is the agreement on Kim Jong-un’s proposal (initially advanced by the South last June) to send a North Korean contingent to the February Winter Olympics in Pyeongchang, South Korea. The second flowed from South Korean President Moon Jae-in’s follow-up idea of restoring the hotline between the countries and beginning discussions of how to tamp down tensions on the peninsula. (Pyongyang shut down the hotline in February 2016 after South Korea’s conservative government closed the Kaesong joint industrial zone located in the North, which then employed more than 50,000 North Koreans.) Moon’s suggestion doubtless eased the way for the subsequent agreement to hold future military talks aimed at reducing the risks of war.

There are further steps Seoul could take, including declaring a moratorium on military exercises with the United States—not just, as now (with Washington’s consent), during the February Olympics and the Paralympics that follow and end in March, but without a preset time limit. While such joint maneuvers don’t scare Pyongyang, moves like flying American B1-B bombers and F-15C fighter jets in international airspace off North Korea’s coast do ratchet up the tension. They increase the chances of one side’s concluding that the other is about to attack.

Trump may continue his threats via Twitter and again denigrate the value of negotiations with Pyongyang, but South Korea is a powerful country in its own right. It has a $1.4 trillion economy, the 11th largest in the world (versus North Korea’s paltry $32.4 billion one), and ranks sixth in global exports. It also has a formidable military and will spend $34 billion on defense in 2017—more than North Korea’s entire gross domestic product. It is, in short, anything but the Asian equivalent of a banana republic for which Donald Trump should be able to write the script.

Trump’s generals and the rest of the American foreign-policy establishment won’t welcome independent initiatives by Seoul, as witness the condescending remark of a former official about the hazards of South Korea “running off the leash.” Predictably, mainstream warnings have already begun. Cunning Kim Jong-un wants to drive a “wedge” between the United States and South Korea. He’s trying to undo the sanctions. Agreeing to talks with Pyongyang will only communicate weakness. The United States must demonstrate its resolve and protect its credibility. And so it goes.

Policies based on these shibboleths, which portray South Korea as an American dependency, have brought us to the brink of war. Continuing them could push us over the edge.