Smoke billows over the northern Iraqi town of Sinjar.Bram Janssen / AP

My thanks to the authors of the three thoughtful responses to my Atlantic essay on “how ISIS spread in the Middle East,” and to the scores of people who have contacted me directly about it. Writing about this topic is painful, and properly so: It is a story of failure. Many events prove to be less consequential with the passage of time than they initially seemed. But I think the rise of the Islamic State in Iraq and Syria may be an exception—a cascade of errors whose impact and cost will deepen in coming years.

Since I wrote the essay, there have been several developments worth noting. First, Russia’s military intervention in Syria, though emphatic, has had little apparent success against either ISIS or the other groups that threaten the Assad regime (which appear to be Moscow’s principal targets). With the aid of Russian airstrikes, the Syrian Army was able to retake Kweiris air base in Aleppo, but they have not broken through most of the Islamist rebel lines. Russian action has been costly, most visibly in the apparent bombing of a Russian jetliner over the Sinai, but on a deeper level, in the indignation of Sunni groups and nations against Russian policy. It remains surprising that Russia, with its large population of Sunni militants, decided to join the Iranian-Shiite alliance to support the Sunni world’s lead villain, Bashar al-Assad. Evidently, Washington is not the only capital where leaders make stupid mistakes about using military power.

Second, partly as a consequence of Russia’s intervention, the U.S. has mobilized a renewed push for a de-escalation of the Syria conflict. Secretary of State John Kerry has once again showed himself to be the indefatigable diplomat. He managed to get Iran, Saudi Arabia, and Turkey to join the U.S. and Russia around a meeting table. That’s a start. The next step is to get the parties to agree on a common enemy (presumably, ISIS and Jabhat al-Nusra). Next, officials hope that the “non-terrorist” opposition groups could meet under the wing of UN special envoy Staffan de Mistura to frame common demands and assemble a joint list of representatives. And then, presumably, Russia would push the Assad regime to join the process. There are many “ifs” along the way, and it’s hard to imagine the Assad regime and the opposition meeting in adjoining countries, let alone rooms. But you have to applaud Kerry for trying.

Kerry’s diplomacy seems premised on a gut feeling more than a strategy. He must believe that Russian President Vladimir Putin is coming to realize that playing this conflict out on the battlefield could have ruinous consequences for Moscow. Does Putin perhaps see the Syria conflict’s disturbing resemblance to the Afghanistan war of the 1980s, which was part of the process that led to the unraveling of the Soviet Union, an outcome he regards as catastrophic? Better to find a face-saving transition for Assad now than later, Putin might decide. Standing in the way of that pragmatic outcome is Putin’s newfound ally, Iran, and its masterful covert-action chief, Qassem Suleimani, who appears to favor a greater “all in” military commitment in Syria than either President Hassan Rouhani or Supreme Leader Ali Khamenei will support.

What will Kerry and Obama do if this 11th-hour U.S. diplomatic push should fail? That may be one of the biggest tests of the Obama presidency. I’ll return to this conundrum later.

The third new development in the last few weeks is the intensification of the Turkish-Kurdish standoff. Turkish President Recep Tayyip Erdogan is stronger politically, thanks to the November 1 election victory of his Justice and Development Party. But the transnational Kurdish movement has also been bolstered, because of the Obama administration’s decision to augment support for its new best friend, the Syrian Kurdish militia known as the People’s Protection Units or YPG. Will this lead to a collision between Erdogan and the Kurdistan Workers’ Party, or PKK, with which Turkey fought a decades-long civil war and which has close links with the YPG? Perhaps, but Erdogan and the PKK could instead make an opportunistic decision (which the U.S. would like to help broker) to renew their dialogue. The U.S. hoped that it could, by now, have framed a joint military plan with Turkey to secure the roughly 90-kilometer gap in control of the Turkish-Syrian border west of the Euphrates. But that plan, on the Turkish side, has been more talk than real resources, frustrating U.S. commanders

Now to the substance of the comments on my Atlantic piece. I am grateful for Fred Hof’s contribution—in the long, ghastly story of the Syrian conflict, he is one of the few people who has consistently stayed focused on the horrors of the war for civilians and on the inadequacy of the Obama administration’s response. Hof offers a devastating summary in his comment: “The four-plus years the United States spent holding Syria at arm’s length, hoping its carnage could be contained while drawing erasable red lines and merely calling on Assad to step aside, have helped spawn horrific unanticipated consequences and narrowed policy options. Decisions America deferred in 2012 came home to roost in 2015. Everything is harder now …” Hof has been arguing for a stronger U.S. policy for the last three years, more clearly than anyone I know.

He makes several specific points that I agree with. First, he argues that protection of Syrian civilians is “the mandatory first step toward the negotiated political transition” that Kerry envisions. I share his view that such protection will be impossible without secure corridors free from Assad’s barrel bombing—let’s call them “safe zones.” One addendum I’d offer: Until Kerry’s Vienna talks collapse, they’re the right venue for discussion of this safe-zone process. If diplomacy fails, one consequence should be an immediate move by the U.S. and its allies to establish the safe zones.

Second, Hof argues that ISIS is an easier target in Syria (where it’s a transplant) than Iraq, and for this reason, Syria “should be the top battlefield priority.” It’s bizarre that after spending more than a trillion dollars in Iraq, and building a new Iraqi army from scratch, America has less leverage against ISIS there than it does with 25,000 Kurdish fighters in Syria. But that appears to be so. As this week’s news of the peshmerga offensive on Sinjar showed, America’s only reliable allies in Iraq, too, are Kurdish forces.

The new Obama rubric seems to be “Do what works,” rather than the old formula, “Don’t do stupid shit.” The logic of supporting the Syrian Kurds is that they actually fight and seize ground. The YPG is allied with a smaller Arab tribal force drawn from Raqqa and Hassaka. In theory, this Arab force will take the Islamic State’s capital. But the group is small (less than 5,000 fighters) and largely untested. What will Obama do if it turns out that this “clear and hold” force can’t actually clear anything?

For his part, Dominic Tierney rightly raises questions about how to create the Syrian “safe zones” that I advocated. As noted earlier, I think the safe-zone action shifts for now to a UN-led process that would be part of Kerry’s diplomacy, assuming that it makes any headway. But if the diplomacy collapses, civilian protection becomes even more important. I would argue for a declaratory zone, rather than a preemptive strike against the Syrian air force and air-defense system. That is, the United States and its allies should announce that they will protect civilians in a defined zone, and that any attack on them by Assad’s forces will trigger a devastating reprisal. A private warning should first go to Russia, to reduce the possibility of a U.S.-Russia confrontation.

The reality is that if Kerry’s diplomacy fails, Syria will be effectively partitioned. The Assad regime will control part of the territory, backed by Hezbollah and the Russian military. Various rebel groups will control the rest. In this situation, safe zones will be another name for “cantons.” The goal of the U.S.-led coalition (an aim that ought to be shared by Russia) should be to establish stable cease-fire lines and degrade the power of ISIS.

NATO members must be full partners in creating the safe zones. Because of refugee flows from Syria, European countries have a greater stake in the success of this humanitarian effort than does the U.S. Tierney correctly stresses that this is a big commitment: “If the U.S. and its partners create a safe zone, they are responsible for the consequences.” True enough, and they are also responsible for the consequences of not taking action.

Tierney’s basic argument against safe zones is that they won’t really be safe—that they could be exploited by the jihadist groups and that “the ark could be a trap.” That’s a good caution, especially given the failure of the U.S. “train and equip” program, which was supposed to provide a stabilization force in areas cleared of ISIS extremists but got rolled by Jabhat al-Nusra. To make the zones secure, the U.S.-led coalition will probably have to work with Islamist groups (not Nusra) that have been backed by Saudi Arabia, Turkey, and Qatar. That will be tricky, but a version of it is already taking place through the CIA’s covert training program.

Let me turn finally to the comment from Lisa Blaydes and Martha Crenshaw, “Is There a Sunni Solution to ISIS?” The piece goes to the heart of the issue facing the region. ISIS is a product of disorientation and the yearning for political, cultural, and religious definition throughout the Sunni world. The rot infects all the weight-bearing beams—Egypt and Saudi Arabia, most especially. Perhaps the demolition we are now witnessing must precede reconstruction.

The Arab Spring showed what a new generation of Arab citizens would look like—we saw them in Tahrir Square, and heard their insistence that “the barrier of fear is broken.” Their disorganized, bottom-up movement sadly couldn’t create leaders; the opposition in Syria has been so chaotic, and the Sunni resistance to ISIS so weak, for some of the same reasons. The default U.S. answer to Sunni empowerment, of late, has been tribal politics. Blaydes and Crenshaw note the durability of tribal disaffection among the al-Dulaimi and al-Duri tribes in Anbar Province, and “their willingness to rebel against a ruthless regime,” whether it’s led by Saddam Hussein or ISIS. As a journalist, I have attempted for several decades to meet and understand Sunni tribal leaders, whose power was often overlooked by the U.S. But they are overlooked no longer. The sheiks know the Washington game so well now that they hire lobbyists to dicker for their share of the spoils.

In fighting ISIS, the U.S. had a “Field of Dreams” strategy for the Sunni tribes: Build training camps at Al-Assad and Al-Taqqadum airbases, and the tribal fighters would come. But it hasn’t happened. It’s too cynical a game, on both sides.

Perhaps I’m naïve, but going forward, I would be more inclined to bet on Sunni citizens than Sunni sheiks. Civil-society groups can have a world-changing aspect. That’s the lesson of the “color revolutions” that Putin fears so much. Iraq and Syria are full of angry young people who are appalled by ISIS, and hate dictators, too. That’s the sweet spot, if we can find it.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.