Considering both perspectives in the light of Northern Ireland and Iraq yields some additional insight into the viability of partitioning as a means of resolving ethnic conflict. In Northern Ireland, partitioning was attempted and physically enforced to a degree, though there were always ethnic Irish living in the North and even some British/Protestants living in Ireland proper. That partitioning failed to end the ethnic conflict on this island is painfully clear with even a cursory glance at the last decades of the twentieth century (violence tapered off dramatically in the first decade of the twenty-first century due in large part to a changing geopolitical scene): bombings of civilian areas, assassinations, and open resentment and street violence were major problems in Belfast and throughout Northern Ireland, with parts of Ireland proper affects as well. This situation highlights one of the key problems with partitioning as a means of addressing ethnic violence -- creating a new nation and separating the populations in conflict creates a new conflict in the division of land. The Irish still had (and have) a claim to the territory of Northern Ireland, and thus while the conflict changed from intrastate to interstate, it was not fundamentally changed by the partitioning.
In Iraq, a nation formed basically by outside...
While Shiite and Sunni groups might come into territorial conflicts in regards to certain cities (along with the Kurds, though to a lesser degree proportionally speaking), the nation itself is already largely divided into ethnic regions through historical happenstance. Oil riches and other factors complicate the matter of breaking up the nation, as even a people with no traditional claim to an area would be loathe to give up the resources it contains to an "enemy" people, but without the history of conquest and subjugation that exists between the Irish and the Northern Irish it is actually possible that the creation of physical political partitions could be effective here.
History has not shown that nations know how to be at peace for long periods of time regardless of ho much or how little partitioning has taken place. It also seems to demonstrate that violence and conflict is also much more commonly and directly related to resource access and control than it is to religious beliefs or ethnic identities. Tackling these problems form this perspective, rather than from a purely ethnic perspective, would lead to a greater chance of peace.
