With this approach, given the direct American and other allied assistance that would be provided, one could be confident that sanctuary sites would never again have to face the prospect of rule by either Assad or ISIL. They would also constitute areas where humanitarian relief could be supplied, schools could be reopened and larger opposition forces could be recruited, trained and based. UN agencies and NGOs would help in the effort to the extent they were willing and able, focusing on health, education and basic economic recovery. Governing councils would be formed, more likely by appointment than election, to help international agencies make decisions on key matters relevant to rudimentary governance. Regardless of details, relief could certainly then be provided more effectively than today.

At least one such area should adjoin Jordan and another Turkey, and these should be created in cooperation with Amman and Ankara. These locations would allow secure transportation lines for humanitarian as well as military supplies. They would also provide bases from which to attack ISIL in its strongholds, a mission that Western forces could carry out in conjunction with local allies. The ultimate endgame for these zones would not have to be determined in advance. The interim goal would be a deconstructed Syria; the ultimate one could be some form of a confederal Syria, with several highly autonomous zones. One of those zones might be for Alawites, perhaps partly protected by Russian forces. But none of the zones could be for ISIL, al-Nusra or Assad and his inner circle.

At some point, the emergent confederation would likely require support from an international peacekeeping force, once it could be somehow codified by negotiation. The United States should be willing to commit to being part of a force, since without it, it is dubious that the conflict’s various parties will have confidence in the stability of any settlement. The challenge of creating governance structures that protect the rights of Syria’s various communities would be especially acute in the intermixed central population belt of the country. But in the short term, the ambitions of this strategy would be limited—they would be, simply, to make individual zones defensible and governable, to help provide relief for populations within them and to train and equip more recruits so that the zones could be stabilized and then gradually expanded.

As safe zones were created, over time some would eventually coalesce. For example, once appropriate understandings were reached with Turkey, a single Kurdish zone would make sense. Major sectors in the south near the Jordanian border, and in the north near Idlib and Aleppo, could be logical. Over time, if and when feasible, zones near some of the central cities such as Hama and Homs could be envisioned, though the logistical challenges and the safety challenges for Western forces (and the difficulty of collaborating safely with any Russian forces) could be greater in those cases. Prudence would have to be the watchword. In some cases, even the various members of the so-called moderate opposition might come into conflict with each other; outside parties might have to threaten to withhold support of various types to discourage such behavior.

The plan would be directed in part against Assad. But it would not have the explicit military goal of overthrowing him, at least not in the near term. American forces could concentrate on supporting opposition units fighting ISIL. Still, this plan would probably have the effect of gradually reducing the territory that Assad governs, since it would train many more opposition fighters and would not try to prevent them from liberating areas of the country currently controlled by the central government. If Assad then delayed too long in accepting a deal for exile, he could inevitably face direct dangers to his rule and even his person. The plan would still seek his removal, but over a gradual time period that allowed for a negotiated exit—with stronger moderate opposition groups part of the negotiation than is the case today—if Assad were smart enough to avail himself of the opportunity. In the short term, however, the current tacit understanding with Assad, whereby he chooses not to challenge Western airpower in Syria when it is used against ISIL, ideally would continue.

The opposition would need to accept that a peace deal that includes post-Assad Alawite elements remained Washington’s goal—and if they wished economic and other help down the road for rebuilding a new Syrian state, they would have to tolerate some influence for the United States as well as other key outside players. This approach, while not ideal for many elements of the opposition who surely seek more systematic revenge against Assad and his cronies, could nonetheless provide a more workable basis for making common cause than is the case today, since it would in fact ultimately aim for an end to Assad’s rule. For these reasons, whether they fully endorsed it or not, America’s main regional allies in the effort—Turkey, Jordan, Saudi Arabia and the Gulf states—would likely welcome such an approach since it would move significantly in the direction they have advocated. Moreover, it would be more credible than previous American strategies, stated or implied, because its means would better match ends.

This strategy might soften Iran and Russia’s opposition to the broader approach as well—perhaps reducing their inclination to escalate support for Assad and also possibly even enlisting them in an eventual negotiated deal about Syria’s ultimate future and associated peace-enforcement operations. Indeed, the strategy strikes a balance in its approach to Iran and Russia. It would grant neither a major role. But it would seek to mitigate the risks of escalating rivalry with them by holding out political hope and the prospect of an autonomous region for Alawites (even those previously associated with the Assad regime, as long as they were not from Assad’s inner circle). This approach may appeal even more to Moscow and Tehran if Assad continues to suffer battlefield setbacks. Damascus and Moscow would be much more likely to support a confederal Syria to the extent they believe that the alternative has become the complete overthrow of Assad and his government, the elimination of meaningful Alawite influence in a future government or, in a best case, civil war of indefinite duration. 

An ultimate settlement could include outright partition of the country if necessary. However, partition would not solve the question of how to address the mixed cities of the country’s center belt. As such, while it should not be taken off the table, it would hardly represent a panacea. 

Should Assad fall, the essence of this strategy would still apply, but in a modified way. Moderate insurgents would still need strongholds from which to build up capacity to challenge ISIL (the presumed main winner in such a defeat of Assad).

Ideally, the U.S. Congress would explicitly support this strategy, but existing authorities and funds are adequate to start now. Ideally, the UN Security Council would endorse the approach, too—including the near-term idea of providing relief (without Assad’s blessing) in some safe zones, and the longer-term goal of deploying a peace-implementation force to support an eventual peace deal. But again, given the emergency situation, the security stakes and the UN’s interest in the notions of the responsibility to protect and the prevention of genocide, existing authorities are sufficient to embark on this strategy.

THE BASIC logic of this ink-spot and regional strategy is not radical. Nor is it original or unique to Syria. In effect, variants of it have guided Western powers in Bosnia, in Afghanistan in the 1980s and since 1993 in Somalia. The last case is particularly relevant. Somalia, while a site of tragedy for U.S. forces in 1993, has since shown some signs of hopefulness. The Puntland and Somaliland in the north are largely self-governing and autonomous. Similar types of zones would be the interim goal for Syria as well.

We must be honest with ourselves: the interim period, including some type of American engagement in the war effort, could last a long time. For a country weary of long wars in the Middle East, this would constitute an on-the-ground role in yet another. That said, it is worth bearing in mind that while the Afghanistan war today continues to consume American resources and cost some American casualties, it is not a major source of domestic political acrimony within the United States. Perhaps Americans are more patient with long military operations than is often argued. That is especially the case if the strategy that the operations are designed to serve is responsive to a real security threat, and if it is at least moderately successful in its implementation.

There would of course be risks associated with this strategy. The most glaring would be the possibility of American casualties—either through “green on blue” insider attacks of the type that have taken dozens of American lives in Afghanistan, or through ISIL or regime elements overrunning a safe zone in which American forces are located. This is a significant risk, to be sure, and one that would have to be carefully managed, as noted above, by careful selection of where the safe zones are to be. It would also require deployment of American quick-reaction forces in the area, in more locations than they currently are found today, to improve the odds of coming to the aid of such U.S. forces in a timely fashion if their positions are brought into danger. In these ways, the operation in Syria would resemble the beginning phases of the Afghanistan campaign in 2001 and 2002, in which modest numbers of U.S. forces worked closely with the Northern Alliance and then the fledgling Afghan government, participating in raids and occasionally suffering casualties. Casualties could also be expected in any future peace-implementation mission, as spoilers use suicide bombs and other weapons to attack outside forces.