After two-and-a-half years of steadily metastasizing violence in Syria, the harsh reality is that the country isn't going to become a stable, unified state again in the foreseeable future, let alone a remotely democratic one. It may be time to start thinking about alternatives.

Syria has already fragmented into increasingly well-defined enclaves. A multiplicity of Sunni Arab rebels control large swaths of the north and east, while the regime is dominant in the capital and major cities, the largely non-Sunni coastal provinces, and a corridor connecting them.  Kurds control small border areas in the far northeast. The Druze, heavily concentrated in southwest Syria, have formed militias to fend off rebel incursions, while tending otherwise toward neutrality.

Although the rebels are likely to eventually bring down the minoritarian regime of President Bashar Assad, as Michael O'Hanlon helpfully analogizes, the fall of Damascus will "no more end the war in Syria than the overthrow of Saddam in 2003 brought stability to Iraq."  Pro-regime forces will fall back toward the coast, Kurds and Druze will fortify their enclaves, and the Sunni rebels will fight each other to reclaim what's left of the Syrian capital.

It could be many years before any coalition in Syria is sufficiently powerful, willing, and united to impose a state monopoly of force. By that time, the country will have lost hundreds of thousands of lives, most of its urban infrastructure, and a generation's worth of economic growth and development.  The collateral damage of a grinding fight to the finish could render the dominion of whoever wins all but ungovernable and destabilize the rest of the Levant.

Keenly aware of this, Obama administration officials continue to place their hopes on brokering a negotiated settlement providing for a peaceful transition to majoritarian rule in Syria. The problem is that powerful jihadist rebel factions and their wealthy donors in the Arab Gulf will never accept a political accord that curtails their pursuit of an Islamic state in Syria, while most Alawites and other minorities will reject any transition plan that doesn't.  There's no way to bridge the gap until someone reins in the jihadists, and that clearly isn't going to happen before pro-regime forces are decisively defeated (if then).

The most viable alternative to the violent restoration of Sunni Arab hegemony in Syria is partition – either "hard," resulting in two or more independent states (e.g. Sudan, 2011), or "soft," as O'Hanlon proposes,[1] resulting in autonomous centralized cantons under a weak federal government (e.g. Bosnia, 1995).

Of course, Syria is not neatly divided into geographic ethno-sectarian constituencies.  Apart from Jabal Druze and the coastal hinterland of Latakia and Tartous provinces, there are few sizable blocs of Syrian territory where any one minority is a majority (though there are plenty where Sunnis predominate). All of the major cities are confessionally mixed.[2] However, insofar as refugees from government-controlled areas are disproportionately Sunni and those fleeing rebel-held areas are disproportionately non-Sunni, territories under the control of both the regime and the opposition are becoming steadily more homogenous.

As in Lebanon during its 1975-1990 civil war, de facto partition is happening every day.  The question at hand is whether the international community should encourage a settlement that reifies and institutionalizes this fragmentation, rather than seeking to propel one side or the other to victory.

Though explicit advocacy of either soft or hard partition is deeply taboo in Syria (an unfortunate legacy of French colonialism), it appears that most Alawites, nearly all Kurds, and many Christians and Druze would prefer some form of secession or extreme decentralization over the uncertain outcome of majoritarian (and almost certainly Islamist) rule, particularly if it means an earlier end to the fighting. 

Most Sunni Arabs vehemently oppose partition, not so much because they see Syria's sovereignty and current borders as sacrosanct (in fact, the more deeply religious see both as largely irrelevant), but because they expect to win out under majority rule in a unitary Syria.  However, Sunnis may come to support partition if it means escaping Assad's grip without paying the catastrophic costs of defeating him.  They may also come to prefer living in a more homogenous polity less susceptible to outside (particularly Iranian) influence.  Jihadists will unquestionably reject partition, but no more than they have rejected all other conceivable political outcomes acceptable to a critical mass of minorities (and perhaps less so, insofar as a rump Sunni Arab state with fewer minorities has fewer obstacles to their political ascendancy).

At the regional and international level, partition is no one's first choice of outcome.  However, with the exception of Turkey (always a stickler about Kurdish autonomy anywhere), neither is it anyone's last choice.  Jordan and perhaps Israel would find a friend in a Druze statelet, while a coastal Alawite-dominated statelet would be sure to align with Tehran and Moscow (indeed, partition could be Russia's best hope of holding onto its naval facility at Tartous long-term).  The Kurdish zone would likely form a close relationship with its counterpart in Iraq.  The Arab Gulf states would own the center (literally, in many places). Partition could prove to be a stable equilibrium, an arrangement from which regional and international players have little incentive to unilaterally depart once it is in place. 

Partition obviously won't solve all of Syria's problems, least of all its authoritarian political culture.  Assad and his minions will surely dominate the coast, while Islamists will reign supreme in the bulk of what's left. The largest and most militarily powerful Kurdish group, the Democratic Union Party (PYD), is deeply autocratic.  Those not belonging to the dominant ethno-sectarian group in their statelet are going to have a hard time of it.  Such a solution would be a travesty if prospects for peaceful reunification and democratization weren't so abysmal. Under the present circumstances, according to New York Times columnist (and former Beirut bureau chief) Thomas Friedman, it "might actually be the most stable and humanitarian long-term option."[3]

Nor will geopolitical fragmentation of the Levant solve all of America's problems.  Partition would mean a less-than-complete Iranian strategic defeat in Syria.  However, this outcome can be mitigated by ensuring that the coastal state does not encompass territory contiguous either with Iraq or with the Lebanese Shiite Hezbollah movement's fiefdom in Lebanon, and it must be weighed against the hefty price tag of dealing Iran a complete defeat – a Sunni Islamist capture of power in Damascus.  While an interior Sunni Arab state will surely have a decidedly Islamist bent, it will be less fractious than its multiconfessional predecessor, and therefore a more reliable buffer against Iranian encroachment in the Arab world.

This doesn't mean the Obama administration should propose partition as the answer in Syria, at least not at a time when most domestic and regional actors continue to loudly repudiate the idea.  But neither should it insist on a negotiating framework that explicitly disallows partition (by mandating a transitional government of "mutual consent" as the cornerstone), as those who dare not speak its name today are likely to gravitate toward it in practice as the war drags on. Even Turkey is now losing enthusiasm for the rebel effort.  Down the road, it's not inconceivable that Assad's inner circle can be persuaded to abandon the interior without a bloodbath in exchange for Western recognition of an Alawite-dominated mini-state.  If that is what it takes to avert the calamities yet to come in Syria and the region, Washington should keep an open mind.