It is a comforting notion many people have that good outcomes arise from good intentions, and that positive achievements flow from sound reasons. Sometimes this is true. Jonas Salk set out to create a life-saving vaccine against polio, and he knew his science well enough to reason himself to success. On the other hand, at the time—from the first onset of epidemic polio levels in the United States at the end of the 19^th century to the heights of the disease in the 1940s and 1950s—no one knew that our best intentions in cleaning up our water supply by introducing better sewage treatment and overall hygiene reduced natural exposure to the virus and hence crashed immunity to it, creating the conditions for epidemic disease. And thus the aphorism that the road to hell is, often enough at least, paved with good intentions.

So, too, the inverted version of the same logic. Bad people doing genuinely bad things sometimes lead to unanticipatable good outcomes, in due course. The Nazis built the paradigm of modern evil, although of course they fantasized that they were doing noble things in desperate defense of German corporate identity. But the Nazi social revolution destroyed the Junker class and so inadvertently paved the way for genuine liberal democracy to arise in its wake. So it is true, often enough at least, as Emil Cioran said in A Short History of Decay (1949): “History is irony in motion.”

What has the history of polio, or of Germany for that matter, to do with the Singapore Summit between the leaders of the United States (henceforth “us”) and the Democratic Peoples Republic of Korea (henceforth “the Norks”)? Potentially everything.

Donald Trump’s idea of diplomacy as a projection of his personality is exactly what it appears to be. He is selfish to and well beyond a fault, so that what mattered to him most in Tuesday’s Summit was the appearance of success, even at the expense of national and global security interests. That is characteristic of this reality TV presidency¹, and of his abiding narcissism. Trump’s craving for attention is so bottomless that his behavior often resembles that of an emotionally fragile four-year old, who will throw a controlled tantrum at the sight of good food just to seize modicum of control over her parents—a trait vividly on display at this past weekend’s G7 conclave in La Malbaie.

Trump clearly doesn’t understand complicated issues like East Asian geopolitics or Korean, Chinese, and Japanese history and psychology, and he doesn’t care to understand them. That is because he doesn’t really care about them, or for that matter about the U.S. equities bound up in the current U.S.-DPRK engagement. His insecurities are so enormous that he refuses to be briefed or to get experienced others to help him plan and prepare, so his style is the bravado of an outwardly confident know-it-all who improvises out of contempt for earned expertise of all kinds. He lives by the seat of his pants, utterly present-oriented, as a man who is proudly on record as being dismissive of history and of reading as well. This is a connect-no-dots, stream of semi-consciousness presidency that never looks back at the rubble left in its wake. The President himself could not have said it better: “I think, honestly, I think he’s going to do these things. I may be wrong. I mean, I may stand before you in six months and say ‘Hey, I was wrong.’ I don’t know that I’ll ever admit that, but I’ll find some kind of excuse.”

Is this a wildly irresponsible attitude for a leader of a state that designed and still supplies common security goods in a dangerous world? Damned sure it is. But could it lead to a positive outcome anyway, in ways our encyclopedically ignorant President can’t begin to imagine? Actually, yes.

Upon rising and reading the news from Singapore, I was stunned to see the under-headline in the New York Times: “Trump Agrees to End War Games with South Korea.” All the rest of it—vague and rather short formal statements, photo ops, incessant ego preening on the part of both principals, mediocre questions from reporters—could hardly raise a pulse for being all too common. But learning that Trump had promised to end the most critical practical connectivity that defines the U.S.-Republic of Korea (hereafter “ROK”) security alliance—and that he reportedly did so without the foreknowledge of either the ROKs or his own Secretary of Defense—was a bit of a surprise. That was the real slinging that burst forth from Singapore.

(Did National Security Advisor John Bolton know the President was about to do this? Given that one of John’s favorite witticisms about the Norks goes like this—“How do you know when the Norks are lying? Their lips move.”—my guess is that he did not. What he’s actually thinking about this whole business of trusting Kim Jong-un’s good intentions is anyone’s guess. Most likely he just enjoys the rarified air that comes with being National Security Advisor, and probably sees his role, at least with respect to the Korea portfolio, as being akin to that of a deep safety in the fourth quarter, playing back to guard against breakaways in order to protect a lead.)

The more germane question is, did Trump understand the importance of these military exercises before he gave them away? One is tempted to quickly answer “no, of course not.” But that’s not so clear, as becomes evident from looking at the actual statement in the transcript—the first line of which is as revealing as it is amazing coming out of the mouth of an American President—of the post-summit press conference:

At some point I have to be honest. . . . . I want to get our soldiers out. I want to bring our soldiers back home. We have 32,000 soldiers in South Korea. I would like to be able to bring them back home. That’s not part of the equation. At some point, I hope it would be. We will stop the war games, which will save us a tremendous amount of money. Unless and until we see the future negotiations is [sic] not going along like it should. We will be saving a tremendous amount of money. Plus, it is very provocative.

In other words, Trump’s ultimate intention is to get U.S. troops out of South Korea, and likely other places as well. He does not have conventional policies, but like it or not (I don’t) he does have a more or less coherent strategic vision.² Note, too, that the “war games” are more accurately suspended, not ended, because he clearly makes the promise contingent on the negotiation “going along like it should”—so this is not the irresponsible, irreversible giveaway it may first seem to be. As for who has been provoking whom all these years, only someone with Trump’s mindset could possibly get that as completely ass-backwards as he did.

But here, exactly, is where historical irony smirks from the mist that eternally hovers over the political present. Trump is right to want to remove U.S. forces from South Korea, even if his reasons are not right. This ambition has outraged experienced U.S. Asia hands, but as Trump has pointed out, these experts and their former political masters in previous administrations have not exactly draped themselves in honor in dealing with North Korea. Every once in a while, we must admit, the man says something that is true.

Now, what are the right reasons for wanting to withdraw U.S. ground forces from South Korea, and what are Donald Trump’s reasons?

The right reasons, as I have explained before³, concern two critical changes that have come about during the tenure of the U.S.-ROK security treaty: the end of the Cold War, and the advancement of North Korea’s nuclear weapons program to the point that, in a year or two, it will likely possess the capability of attacking the continental United States with nuclear weapons. These changes are critical to the logic of both the security treaty and the manner in which it can be made credible. And it is just shy of astonishing that nearly all of our so-called experts on this subject have failed to understand the meaning of these changes—which only shows, yet again, that it is possible to become so inured to a situation that habit can overpower actual thought, just as the band on the Titanic reportedly continued to play even as the ship was sinking.

Since North Korea invaded South Korea once, in June 1950, and nearly succeeded in conquering it before the Inchon Landing saved the day, it stood to reason that a defense alliance was mainly created to prevent a second occurrence. It was designed not only to deter North Korea from another attack, but also to influence the Soviet Union and the Peoples Republic of China away from encouraging another attack. So much for the military rationale. On the political side, the arrangement was designed to protect the hoped-for evolution of South Korean democracy, and with that evolution to strengthen democratic political cultures throughout Asia—a goal for which Japan was of course the linchpin.

All of that worked, and with the end of the Cold War it beggars imagination to contemplate Russian or Chinese support for a Nork invasion of the ROK. Would an unhinged North Korean leader perhaps do that anyway? Only if he were suicidal. And besides, even if such a very improbable thing were to occur, U.S. ground forces would not be instrumental in defeating an attack. U.S. air power would matter a lot more, and no one is talking about sterilizing the Korean Peninsula from the reach of U.S. airpower, as if an agreement of that sort could possibly be created given the inherent mobility of air power.

Not only is the original purpose of deploying U.S. ground troops in South Korea obsolete, those troops actually function as an impediment to U.S. security interest because of the other major change noted above. If U.S. mainland comes under threat of North Korean nuclear attack, it is the beginning of wisdom, if not also of standard prudence, to want to manage that challenge without shackles on one’s options. So take careful note: Every time, earlier in his presidency, that Donald Trump tried to threaten North Korea so as to deter it, it tripped off what it is fair to call “the sunshine effect.” North Korea simply turns on the charm temporarily, aimed at South Korea, in order to in effect undermine U.S. declaratory policy. The ROKs fall for it every time, because politically they must. The way this relationship is set up, any South Korean government essentially has a veto over what the United States may do militarily in the peninsula. The “in together, out together” arrangement essentially mutes the signaling aspect of any independent U.S. effort to manage the North Korean threat.

And it is even worse than that. Short of a very improbable future nuclear exchange between the United States and the Norks, the presence of U.S. troops in South Korea gives the Norks a convenient target to hit or threaten in the event of severe pressure being brought to bear against them. Unless the Norks are crazy enough to attack U.S. forces in Okinawa, or elsewhere in Japan, the troops on the ground in South Korea are by far the softest target set. They are therefore, in effect, hostages more than they are useful tripwires—they are, in other words, a tie that binds. They deter us as much or more than they deter the Norks.

Removing them, and presumably their dependent families with them, would not entirely eliminate the hostage problem because there are lots of other Americans living in South Korea, largely for business-related purposes. But it would help a lot in terms of numbers, and in the obvious fact that non-military American targets have low symbolic value and the Norks cannot hit them without also hitting large numbers of South Korean civilians, which they have no reason to do.

Removing U.S. ground forces from South Korea, and turning over the ground component entirely to the South Koreans, is well within the capability of South Korea. And doing so need not imply the abrogation of the treaty relationship itself. U.S. air power and a nuclear deterrence umbrella both could and should remain in effect at least for the foreseeable future. This is not peculiar; other examples exist of security relationships in which the United States extends its power without need for expeditionary ground forces (Australia, for example), and in at least one case (Saudi Arabia) it exists tacitly even after a ground contingent was removed.

The treaty itself does not need amendment, because it says nothing about troops; rather it is the relationship itself that needs some adjustment. Additionally, because of the optics involved, new energies must be invested in the U.S.-Japanese relationship, and in building up a stronger strategic posture elsewhere in Asia, in order to manage impressions of U.S. intentions and credibility. Those who argue that any change in the status quo in the U.S.-ROK relationship will amount to decoupling are therefore not only wrong, but they are implicitly making a breathtaking assertion: namely, that contingent arrangements for security may never be adjusted despite changes in the environment.

This is easy to demonstrate. The Soviet Cold War deployment of SS-20s in Europe did not lead to decoupling within NATO, because we had ways to prevent it, including an offsetting deployment of our own missiles; and the cancellation of the deployment with the success of the INF Treaty did not lead to decoupling either. Those who make mechanistic arguments—rising powers always pose a Thucydides Trap; North Korean ICBMs that can reach the U.S. mainland must lead to decoupling—lack respect for history and logic alike. We have choices, plenty of good ones among them.

That’s the good news. Then there are Donald Trump’s reasons for wanting to remove U.S. ground forces from South Korea.

Trump by now may or may not understand something of the foregoing rationale, but even if he does it plays a weak second fiddle to more fundamental beliefs that he apparently holds. Trump is a dyed-in-the-wool unilateralist, which is not the same thing as an isolationist, but frequently looks like it. He does not believe in the very concept of allies, because an alliance is a form of a positive-sum relationship. Trump can only wrap his mind around zero-sum relationships. His is an adolescent two-dimensional kind of realism that feels heroic when it is merely immature. We saw his attitude in full form in Canada this past weekend.

So he never sees the cooperative aspect of human nature inherent in these kinds of relationships, but only the competitive aspects. That is why the very existence of such institutions as the European Union and the United Nations constitutes existential irritants to him (and to John Bolton). That is also why he cares much more about strategically trivial trade imbalances than he cares about, or can even seem to understand, the benefits to U.S. security of our supplying common global security goods. He sees only price tags, never what we get for our money, because the former is expressed in a simple number and the latter cannot be. He looks only at the bottom line of a calculation, without understanding in the slightest the equation that produces it.

It is possible to summarize how the President thinks about such matters by recourse to an old musical comedy lyric. Thus Yul Brynner, singing Oscar Hammerstein’s rhymes, in the 1956 movie “The King and I”:

Shall I join with other nations in alliance?
If allies are weak, am I not best alone?
If allies are strong with power to protect me
Might they not protect me out of all I own?

Is a danger to be trusting one another
One will seldom want to do what other wishes
But unless someday somebody trust somebody
There’ll be nothing left on earth excepting fishes.

We may hope, at least, that one day the President will take the final couplet as much to heart as he does the dour spirit of the first three. But the G7 event in Canada tells us that he has not yet done so. That also explains the difference between the sound logic of adjusting the U.S.-ROK relationship in a careful and cooperative manner so as to obviate the danger of decoupling, and the underlying objectively pro-decoupling attitude that seems to be driving Trump’s North Korea strategy.

We are thus at a key hinge point in a process predictable even before the inauguration: the movement from “same world” to “lonely world” and, ultimately, to “cold world.” Widespread predictions that the arc of Trump’s foreign policy would in time regress toward the mean have proven quite wrong. As of this past week, we have definitively moved from “same world” to “lonely world”— and unless the Trumpian trampling of what benign order remains in the world is not stopped, it is only a matter of time before “lonely world” turns into “cold world.”⁴

The Arabs have a saying, equal parts witty and wise, that “everything starts small except calamity.” We’ll see. If the good reasons for talking to North Korea (and for ending U.S.-ROK military exercises as a step toward getting U.S. troops out of there) prevail in time—and these things do take time, generally more time than the duration of a single presidential term—then good outcomes will have been produced from not-so-good motives. And there is a good outcome to be had here: a deal in which the Norks get a no-regime-change pledge, the end of the Korean War, and some economic goodies, and we and our allies get a diminished prospect of a war crisis, and a Nork nuclear weapons program surge frozen and possibly reversed—even if full denuclearization as we define it remains, in my view, extremely unlikely.

That possible good outcome is why the idea of a U.S.-DPRK summit did not deserve the widespread ridicule to which it was subjected when it first surfaced just over a year ago.⁵ It wasn’t as though the status quo—gussied up by the Obama Administration as “strategic patience”—was working, or that it was even a status quo with the North Korean program advancing apace. Calamity is still possible, although futility is ultimately more likely when the long-time North Korean interpretation of “full denuclearization” again becomes too clear for the Administration to ignore. But it is at least possible that something good can come out of all this, though it probably will need to wait for consummation in the post-Trump era.

If it happens at all, this will require patience, luck, and a tolerance for some interim white knuckles. These days a situation thusly described has to count as a small mercy.