By ALAN W. BOCK
SENIOR Editorial writer

Among the issues that kept Iraqi parliamentarians from meeting the August 15 deadline to draft a new constitution (see cover article) were proposals from Shiites to create a southern autonomous region and from Kurds in the north for provisions that would make secession easy at some future time. The possibility that Iraq might not remain united makes this book - published last year and valuable for other reasons as well - more relevant than ever.

Liam Anderson, who teaches at Wright State University in Ohio and Gareth Stansfield, at the UK's Royal Institute for International Affairs, suggest that "managed partition," while short of a utopian solution, might well be the least-worst alternative for Iraq, a country that has never been governed long except by brutal force.

This provocative idea is rooted in the fact that Great Britain, when it created modern Iraq after World War I, managed to place Shiite, Sunni and Kurdish scorpions into one nation-state bottle, creating an essentially ungovernable - except by brute force, as Saddam did - entity. It is buttressed by seven chapters of Iraqi history that should be useful to anybody who wants to think about Iraq intelligently, whether they think the idea of partition is perceptive or loony.

This history demonstrates that the Kurds, who have been betrayed by just about everybody who came in contact with them through the 20th century, have never been content to be part of Iraq and have been in a state of rebellion - sometimes violent, sometimes not - since 1920. Based on experience, they have little reason to trust Arab promises to respect the semi-autonomous position they finally reached toward the end of Saddam's regime - and the Arabs have little reason to believe they will be content to submit peacefully if things don't go their way.

At a more fundamental level, the authors discuss perceptively the preconditions for successful democracy (while admitting that if anybody knew all the necessities the world might be full of functioning democracies). A shared belief in democracy is essential, but it must be buttressed by "institutions - constitutions, electoral systems, political parties and the like." Even more important is the existence of a civil society - "spontaneously created social structures (interest groups, an independent media, civil rights groups) that exist outside the realm of state control" - and a strong political culture amenable to democracy and the compromises essential to it.

The deeper levels of civil society and political culture are the most important for democratic success - and the least amenable to swift change. Iraq, like most Middle Eastern countries, has no tradition of democracy and precious few trappings of civil society. Locally or tribally controlled militias just aren't the same.

In light of all this, a partition of Iraq looks increasingly attractive. Anderson and Stansfield insist that "the governing of Iraq does not need to be enshrined immediately in a set of inviolate constitutional principles ... . Some form of loosely organized, largely powerless central government can be established to provide an umbrella over the whole. This can be made deliberately powerless and given to the Iraqi exile community to run, or can be constitutionally engineered to produce paralysis replication of the constitution of Bosnia perhaps."

The authors believe the most likely outcome, then, would be the development of a self-governing Kurdish state - it's had one de facto for more than a decade - and a Shia-dominated Arab state, perhaps to be called Mesopotamia. Both entities would have a supply of oil, and they think the Arab state, however Shia-dominated, would not succumb to domination by Iran. They recognize that Turkey, which contains about 15 million ethnic Kurds, whom the government has repressed brutally, would fear the establishment of an Iraqi Kurdistan, but they think it might even contribute to regional stability.

The authors know their proposal wouldn't usher in Nirvana. But it's thought-provoking, and their argument pulls together a lot of valuable history and cultural information in a readable way.