The punditocracy is still abuzz with the prospect of a summit between President Trump and North Korean leader Kim Jong-un. The buzzing slightly quieted as the new workweek began, in contrast to the mood this past Friday, when the White House appeared to have attached conditions to a meeting that made one less likely. So things are in limbo, or at best uncertain—but that is the way things usually are in this pale imitation of an Administration, whether it comes to foreign policy, or gun control legislation, or hastily prepared tax legislation. Don’t like how a policy sounds? Wait a few hours.

Given the attention-deficit disorder that seems to afflict many Americans these days (largely on account of the coherence-destroying power of addictive information technology gadgets), many may not realize that the idea of a U.S.-Nork summit is not new. It first erupted in early May of last year, remember? At that point it seemed to be the President’s idea alone, rather than something hatched in Korea and conveyed to the White House from Seoul.

I wrote about it at the time. I thought the idea had real potential if handled right, and I explained why. Since the subject is back in the news, I thought I would just repeat the basic analysis, about which I have not changed my mind. It appears below, bracketed by asterisks—after which I have appended a few notes to bring us back to present tense.

*          *          *

Monday’s presidential news tended to the personal and Asian: Philippine President Rodrigo Duterte brushed off President Trump’s invitation to visit him in Washington, but Trump surprisingly extended his hand for yet another face-to-face meeting, this one with North Korean leader Kim Jong-un “under appropriate circumstances.”

When word broke of Duterte’s brushoff, one may be sure that the President’s political handlers (and both Diplomatic Security and the Secret Service, as well) breathed deep sighs of relief. When word broke of the tentative tender to Kim, it took only minutes for several U.S. Asia-policy “hands” to engage their mouths in ridiculing the President. Thus spoke Ambassador Chris Hill, once in charge of the fruitless Six-Power Talks: “Kim Jong-un would be delighted to meet with President Trump on the basis of one nuclear leader to another. If I were Trump I would pass on that.”

But that need and ought not be the precondition of a direct meeting. The idea of a U.S.-North Korean summit, which is novel coming from the mouth of the President of the United States, bears more thoughtful scrutiny.¹ That said, it remains unclear if the President’s remark is just another random tweet-from-the-hip verbal ejaculation connected to no strategic thought at all, or if this time, since the Administration has formed an approach to North Korea, it may be more deliberate than it might seem. But however it got started, if handled well, the President’s idea (let’s give him the benefit of deserved doubt and call it that) could turn out to be brilliant.

How so? It could be the first step toward a potentially useful transaction in which U.S. diplomacy trades something Kim desperately wants for domestic purposes—recognition as a legitimate interlocutor by the most powerful state on earth—for a reduction in his outsized paranoia and, hence, some benefits down the road. The basic idea is that if we stop treating him as a pariah, it may reduce his incentives to act like one.

The beginning of wisdom in understanding the Korea tangle is the acknowledgement that Kim believes the United States seeks regime change in Pyongyang, for he knows that, if positions were reversed, that is what he would do. That is why we cannot, and never could, end or significantly curtail the nuclear program without using force: It is the regime’s last resort against dangers of the existential sort, and there is nothing we or any other power can give or promise that will balance off against that perception of an ultima ratio.

Regime change in Pyongyang would be nice, of course, and it will happen one day of its own accord. But no compelling reason exists for the United States to actively pursue it now, since none of our allies—notably South Korea—wants to deal anytime soon with the mess it would cause. So why not tell Kim so, directly but privately?

If that, along with the recognition that will flow from a summit, will calm Kim down, we too need to calm down long enough to recognize that, if North Korea ever resorts to nuclear first-use, anywhere, it would lose the ensuing war, its regime would be destroyed, and Kim personally would end up an irradiated cinder. If Kim is nuts, he’s murderously nuts, not suicidally nuts, so I lose no sleep worrying about an unprovoked North Korean attack on the U.S. homeland. Panicky public rhetoric on our part in that regard is unhelpful in every way. In other words, if both nations’ worst fears could be bounded and a bit assuaged, it would establish new parameters for deal-making: That is the promise of the President’s idea, if it is properly handled.

But, of course, it might be handled badly; as that famous 20^th-century philosopher, the Wicked Witch of the West, once put it: “These things must be done delicately.” So what’s the difference between a properly delicate handling and a bad one?

There are two main ways to think about a U.S.-North Korean summit: a way in which we explicitly attempt to link it to the North Korean nuclear program, and a way in which we avoid explicitly linking it, the better to implicitly link it. The former is the wrong way to proceed; the latter is the shrewdly delicate one.

The summit message from the President to Kim Jong-un should be as simple as possible: “I respect you as a national leader, and my Administration does not intend to undermine your regime.” Leave implicit, or for other, lesser officials to say, that North Korea’s conventional and unconventional military programs must never threaten its neighbors, or, as has been the case for decades, the United States and its allies will respond at a time and in a manner of their choosing to restore international peace and security. And then shut up, arm up, and let Kim figure out what it all means. Deterrence flows from deeds implied by power more than it does from declarations. So he will figure it out.

This dual-track approach is preferable to linkage because Kim will likely conclude that any attempt to suffocate his nuclear program is just a prelude to an attempt at regime change, whatever we may say. If we avoid putting the nuclear program front and center, perhaps later, with Kim’s ego suitably inflated and his arse feeling more secure, a differently shaped negotiating proposal designed to contain the nuclear program (not end it—it’s too late for that as long as the regime endures) might be possible.²

That is anyway as much as we can expect to accomplish anytime soon, for every other option we can think of is either unavailable or dangerous to excess. Let us count the ways.

Consider what we cannot do. We cannot negotiate our way to ending or even moderating significantly North Korea’s nuclear program. Leverage the Chinese to do it for us, since they have material leverage? We tried that in 2005, arguing that China would suffer most from a regime implosion in the North and suffer from the regime’s endurance, as well, because its behavior works as a goad to Japanese nuclearization. The prospect of a permanently denuclearized reunified Korea under the aegis of the democratic South, in the context of a revised U.S.-ROK relationship that would significantly reduce the U.S. military presence in and near Korea (and hence near China), struck some as a bowl full of tasty carrots from the Chinese perspective. Yet the effort got close to nowhere: The Chinese Communist Party’s worst fear concerning Korea is a unified democratic state lying astride its border, not the U.S.-ROK treaty relationship.

That most likely remains the case. When Xi Jinping “explained” to Trump at Mar-a-Lago last month that China could not press North Korea on account of several tons of historical baggage, he was not being entirely honest. The weight is there, true enough, but that’s not all that’s there. China does have decisive leverage over North Korean regime survival, but it fears to use it because China is a kind of hostage to North Korea at the same time. The tighter the Chinese pull the noose, the more the fallout from incipient regime collapse threatens China with multifarious mayhem: a refugee crisis, for certain, but, even more portentous, a scramble for who picks up the pieces in a melted-down Pyongyang. Beijing likely has other motives as well for not helping us.

So if we can’t negotiate our way out of the problem, wait for regime collapse, or directly impress Chinese power on our behalf, can we intimidate the North into backing its nuclear program down under threat of a preemptive war? That seemed to be the policy gist of Secretary of State Rex Tillerson’s message, speaking in Seoul on March 16, when he ruled out new negotiations and declared that, “The policy of strategic patience has ended.”

It was brilliant of Tillerson (or his speechwriter), and very generous too, to describe the futile policies of the past quarter century as a sagacious “policy of strategic patience.” But what he said still frightened people, including people in the very (politically roiled) city in which he uttered the remark. Worse, it seemed to relegate U.S. policy choices to feckless passivity, leaning on Beijing, or preemptive war. But we’d have no allies in such a war—neither South Korea nor Japan, for different reasons, would ever be publicly complicit in a military action that killed lots of Koreans—even as we unleashed unknowable but certainly very bloody consequences. It’s foolish to bluff such a course as a means of intimidation. And it won’t work as intimation anyway.

It is in the context of our poverty of promising options that the President’s offer of a direct meeting figures to reshuffle a bum deck. It could thaw the frigid psychological environment we face to possibly useful ends.

Of course some will object that to recognize Kim as an interlocutor, even if done in a way that does not reward the existence of the nuclear program, would constitute a human rights atrocity—essentially a pass from the United States for the North Korean regime to do whatever it wants to its own people. Let me be twice blunt here.

First, within limits, the principle of the non-interference of states into the internal affairs of other states remains a wise predicate of the Westphalian system, because it dramatically reduces the sources of conflict and war among states: Cuius regio, eius religio makes as much sense in our ideological age as it did in a more religious one. If some people cannot discover the moral value of that consequence, I am sorry for them.

And second, the unfortunate truth is that there is very little the U.S. government can do to ameliorate human rights violations in North Korea, so to argue that symbolic verbal protests should have pride of place over a policy that might do some palpable good is to embarrass moral logic itself.

And if, after all, Kim rejects the invitation to a summit, he would look weaker and even more isolated than he already is. And Trump would look more magnanimous than he surely is. So come on, let’s do it.

*          *          *

A fair bit has happened in the past ten months with respect to U.S. policy toward Korea. Without going into detail, for a while it looked like the Administration was considering a so-called punch-in-the-nose military option. This seemed to me to be a crazy idea, a split-the-difference notion between a major preemptive attack and no attack. But it had, or would have had, most of the downsides of a much larger military action—specifically, the potential to touch off a disastrous war.

It is not often mentioned, but in all the years since the 1953 armistice the U.S. government has not so much as fired a single shot at or toward North Korea, notwithstanding all of the nasty and even deadly things the Norks have done to us and to others. So we have absolutely no idea how they would react to a punch, a kick, or even a sideways fart. Partly in response to what I considered to be a dangerously irresponsible drift toward war, I wrote a piece here, on January 10, that aimed to substitute a bold diplomatic initiative for a counterproductive violent spasm. It argued a controversial view, but one that I thought was supported by a fundamental logic.

The basic logic is that U.S. declaratory threats to use force against North Korea lack credibility because our South Korean ally will not permit it. Every time we raise the noise level, North Koreans just rings the sunshine bell and the South Korean government salivates like Pavlov’s dog. What that means is that the 30,000 U.S. troops in South Korea no longer act mainly as a tripwire to raise the price of a North Korean attack on the South, as they did during the Cold War; under changed circumstances they mainly amount to hostages to any outbreak of violence, which in turn reduces further the credibility of American declaratory policy. The only way to change that is to radically reconstruct the U.S.-ROK relationship so that the tripwire is removed, and the South Koreans take full responsibility for the ground dimension of their own defense.

I explain in the essay that merely freeing U.S. declaratory policy to give it more credibility will probably not suffice to drive the U.S.-North Korean relationship into a negotiation that has any chance of denuclearizing the North. But there are lesser benefits to be gained, so long as—and I stress—care is taken to make sure that the change is not perceived as a retreat of American power from the Indo-Pacific, but a refurbishing and strengthening of it, and that it is therefore not set in motion too rapidly.

I’m sure my essay did not change the Administration’s mind, but some testimony before Congress in late January may have helped do so. After Mike Green and John Hamre both poured icy water on the punch-in-the-nose option, it seemed pretty much to disappear—thank God. At least I hope it has disappeared.

And now? We’re pretty much where we were before. There is no viable military option that’s worth the cost in blood, and let us remind ourselves, too, that any U.S. use of force that the South Korean government does not wish would destroy the relationship rather than recast it, as I have proposed. That would be unwise, not least because it would transform in a negative way Japanese perceptions of U.S. reliability and wisdom.

Then there is the eternal if forlorn hope that somehow we can get the Chinese to fix the problem for us. As I explained in the January 10 essay, that is just not going to happen. I put forth some reasons, but, as I hinted then, there is more to say. It’s not just that the Chinese are afraid of what the collapse of North Korea might mean in terms of refugee flows and other annoyances. And it’s not just that the Chinese do not want to see a unified democratic Korea along their frontier, even if it is a denuclearized Korea by international guarantee. Something more fundamentally strategic is at play.

It is safe to say, I think, that the long-range Chinese geostrategic objective is to drive the United States away from the Asian mainland. There is nothing mysterious or even necessarily sinister about this objective: If a major power were breathing down our neck just off the coast of California, we would not feel much differently. But what this means in practical terms as far as Chinese aims are concerned is what experts call decoupling: The Chinese would like to drive a very large wedge between the United States and all its Asian littoral allies, and that includes South Korea and especially Japan. But how to do this?

It is possible that, for all the trouble North Korea causes China, the North Korean nuclear program has the potential to create a decoupling of South Korea from the United States that clearly serves Chinese interests but does not bear Chinese fingerprints. Indeed, it has a potential that even China’s own nuclear weapons does not have, because the latter implicate China in a war should deterrence fail, whereas North Korea’s weapons do not. The logic is simple: If North Korea can destroy targets in the continental United States, then the U.S. nuclear shield over South Korea becomes less credible. It’s the old extended deterrence dilemma that we saw in Europe throughout the Cold War: Would Washington sacrifice a major American city in order to protect Paris? Now it may soon be: Would Washington sacrifice Los Angeles to save Seoul?

For a variety of reasons that transcend geopolitics narrowly construed, the answer to the Cold War question was “yes,” or at least enough people believed that the answer was “yes” to serve the purpose. Asia in the post-Cold War context poses a different and more difficult problem for extended deterrence. Many believe that decoupling is inevitable now that North Korean capabilities are nearly to the point where they can credibly threaten North America. I do not agree that decoupling is inevitable; I do agree that it is threatened and that we need to respond by making some changes to our strategic posture.

So in a way, the Chinese are sitting pretty in a win-win situation. If the United States does not destroy North Korea nuclear arsenal in a preemptive strike, then the capabilities of that arsenal could portend decoupling, but if the United States does destroy North Korea’s nuclear arsenal, or tries to, then decoupling takes place anyway because that act would almost certainly destroy the U.S.-South Korean alliance, and that in turn would represent a clear degradation of Japan’s strategic circumstances. If that is an accurate assessment, then the Chinese obviously have no reason on earth to help us out of our lose-lose dilemma.

Two things can change the strategic vector: a North Korean regime collapse, and a U.S. policy that responds effectively to changes in the strategic environment (as outlined in my January 10 essay). If the North Korean regime does collapse, predictions get difficult; a lot depends on how and over what kind of timeframe a collapse would take place. But a policy predicated on U.S. active efforts to undermine the North Korean regime would have predictable consequences: It would terrify the South Koreans, and it could create such a level of panic in Pyongyang as to touch off the very behaviors we want to deter. So that approach, too, would be unwise. In this light, the Obama-era policy of so-called strategic patience doesn’t look all that bad, despite its obvious imperfections.

All of which is to remind us that the North Korea nuke problem is truly a problem from hell, a problem so hard that successive administrations of both major parties over four decades have not been able to figure a way around, over, or through it. And as already indicated, it is a problem that has not stood still, but that has gotten progressively worse.

This is another way of saying that it is a serious problem, serious because the lives of hundreds of thousands if not millions of people are at stake, and serious because how it gets resolved may presage the future of American power, not just in Asia but in the world. With this in mind, let us now return to the events of this past week with which we began.

Sunday’s New York Times, in a front-page story entitled “With Snap ‘Yes,’ Trump Rolls Dice on North Korea,” describes how cavalierly the President acted upon being presented with an invitation to a summit delivered by Chung Eui-yong, the South Korean national security adviser carrying a message from the North. According to the article, the President’s closest national security advisers were stunned and dismayed by the President’s instantaneous, seat-of-the-pants reaction.

As well they should have been. Maybe the President remembered his own statement from May of last year, or maybe not—one never knows with this man. But what the President’s closest national security advisers, and others in and around the White House, have no doubt figured out by now is that it is in the nature of a narcissist to care so much for himself that there isn’t much room for caring about anyone else. That is the basis upon which a man like Donald Trump can make deeply portentous decisions without even considering the consequences for others.

Yes, we have known for a long time that Donald Trump lacks any experience in foreign and national security policy. Yes, we know he doesn’t read, and so has a very limited potential to grasp such issues. Yes, we have evidence that on issue after issue he changes his mind because he believes whatever the last person he spoke with said to him—gun control is the most recent example, but there are dozens anyone can cite. But it’s not clear that most of us have figured out what is really at the bottom of all this. What is at the bottom of all this is that the man has no convictions or considered views about most policy areas because they don’t really directly affect him.

It is therefore easy for the President to dismiss the tortured history of an issue like Korea. As someone who distrusts experts—and for that matter who distrusts science when it tells him something he doesn’t want to hear—he can say that his policy instincts, however untutored, are just as good as anyone else’s. That the experts have not fixed the problem does not mean that the problem is hard, in his view; it just means that they’re not good experts, not bold, not “winners.” So any decision might work, just as any decision might not work. As far as he’s probably concerned, it’s close to random; so why bother taking it seriously enough to wrack one’s brain over it?

What does matter to him is that he appears to be in charge, to be a winner, to be able to yank everyone’s attention toward him—not that he’s made a decision that will work to solve a problem. If a decision turns out badly he’ll just blame someone else. He doesn’t really understand or care about the problems themselves, which is another way of saying that he doesn’t take them seriously. He doesn’t know the meaning of the word.

So what is the prospect that a U.S.-Nork summit, if it happens, can do any good? I still think there is a useful deal to be had—if serious people can get and keep their hands on the major decisions to be made, and keep the Flip-Flopper-in-Chief away from the table for as long as is necessary to make and implement them.

It looks like this: We agree to recognize provisionally the North Korean regime, establish diplomatic relations with it first at the DCM level, and come to some formal legal agreement to end the Korean War; and as a part of the deal we pledge no active regime change efforts—which is giving away nothing because we are not pursuing that course now anyway, for reasons already noted. We can also pull back on the sanctions throttle as we deem necessary.

In return we will not achieve a verifiable denuclearization of the country, and we should not insist on that as the make-or-break point of a deal. That is because their nukes are their last-resort weapons: They will never give them up in a negotiation, and any initial negotiation that insists on it is doomed to fail from the get-go. But if we pledge no active effort toward regime change, they will then have less incentive to pursue their military options at breakneck speed, and we might get in return a freeze and partial rollback of the program, to include both weapons, missiles, and the testing of both. That would break the present momentum of their programs, calm our nerves and war-planning velocity, and cool the decoupling fever on the loose in the region. That may be the best we can get for the time being.

Perhaps most important, if we devise a more flexible objective than impossible full denuclearization, we will cause all sorts of problems for the North Korean internal propaganda machine, which has been ceaselessly demonizing the United States for decades. Indeed, in the past, as with the Six-Party Talks, the North Korean agitprop authorities had huge difficulties reconciling that episode of face-to-face diplomacy with the image they had created of the West, in particular of the United States. If diplomatic relations are set up and there is an embassy in Pyongyang, with real live Americans in it, these guys will have a problem from hell, too.

Might that lead in due course to regime change without our really trying actively to bring it about? Stranger things have happened. I’m serious.