The situation in a nutshell: Syria's dictator is powerful in his domain. His forces are well armed. The rebels who oppose him are relatively weak, and any successes they experience, or the provision of assistance to them from external players, can incentivise the kind of clampdown on the people of Syria which the international community wishes to avoid. Real success on their part might result in the lynching of their oppressor, as the examples of Ceausescu, Hussein and Gaddafi make clear, so Assad might think this quite the existential fight - and if he gives way, trial in an international court awaits, Milosevic-style. So why do anything other than carry on?

The dilemma, mutatis mutandis, is a familiar one for the West. What to do? How to help the rebels without effectively encouraging more repressive tactics? How to encourage the dictator to yield the levers of power?

My suggestion is simple: a twin "safe harbour" policy. One set of safe harbours is for the rebels, in the form of areas of Syria along the Turkish and Jordanian borders, with their protection guaranteed by the West. The other is for Assad. 

Follow @TheCommentator

Protection for rebels is tried and tested policy (for better or worse). The offer of immunity for dictators who give up power – an unappetising deal done in the ultimate best interests of the people concerned – is a long-mooted idea, albeit one yet properly to be tested (certainly not in the modern age of international courts). The dual new aspects of my suggestion are these.

First, that both are tried in tandem. Secondly, that both safe harbours should shift in size over time, to a stated and declared timetable; as the first set of harbours, the rebels' protected areas, get bigger, the second, Assad's deal, gets smaller.

In the beginning - say, over the next three months - the rebels are accorded very small safe territories which would offer no real offensive assistance, and Assad is offered total immunity, for him and his family, and gets to retain whatever goods he takes with him if he walks away. From mid-summer on, the rebels get more significant territory protected by the West, and Assad's deal only protects him from prosecution - no bank accounts, no booty. In the late autumn, the rebels get more serious territory - and Assad's protection for prosecution is only for himself, only pertains to pre-2012 acts, and he still gets no stuff. From the end of 2012, if still in power, the rebels get full Western support and Assad gets no deal.

You can (reasonably) dispute the morality of dealing with such a man. You can argue (no doubt with force and merit) about the right places to draw the various points of expansion of harbour one and diminution of harbour two. But, as a starter for 10, I humbly suggest it's worth discussing.

Alex Deane is Head of Public Affairs at Weber Shandwick. You can follow Alex on Twitter at @ajcdeane