The latest recording to emerge from Osama bin Laden's cave calls for a boycott of this month's election in Iraq. The Sunni Arab, Saudi-born zillionaire tells 25m impoverished Iraqis-mainly Shias or Kurds-that all who vote will mark themselves as "infidels." In the same tape, Bin Laden also nominates Abu Musab al-Zarqawi as his "emir" in Iraq. Zarqawi is a Jordanian, a suspected rapist, and presenter of his own home video series, in which he occasionally beheads his guest stars. His al Qaeda of Iraq group is the most visible threat to Sunni Arab participation in the poll. Thankfully, most Iraqis don't think that people like this deserve a veto over their election.

The election will happen and will happen on time. This is because Iraq's decision-makers want it to be so, and for better or worse, it is President Bush and Grand Ayatollah Ali al-Sistani-the supreme religious authority among Iraq's Shia majority-who call the shots. But sooner than the world realizes, Bush will cease to call the shots. US soldiers will still be in Iraq in February, but they will be there as the guests of an elected government chosen under UN auspices by a big majority on a large voter turnout. The new government will need US troops and dollars, but an elected government dominated by people who fought Saddam in the marshes for decades, and who battled the US marines to a negotiated draw at Najaf, will be far more independent and credible than the current Allawi administration.

The 30th January election is just the first of five steps designed to complete Iraq's constitution and government-building process by the end of 2005. The 275-member body soon to be elected is charged with proposing a new permanent constitution by 15th August. By 15th October, there must be a national referendum on this permanent constitution. By 15th December, there is to be a new national election that will follow the provisions of the new constitution. And by 31st December this newly elected government is to take over.

Iraqis, then, will go to the polls three times this year-assuming they can agree on a constitution in the autumn. Can the schedule work? For some observers, the Iraqi full-body makeover is a work of fantasy. Others say it is like a television reality show, dealing in real behaviour rendered meaningless by a fake environment. But 90 per cent of Iraqis hope it turns out to be a gritty documentary about decent people struggling to improve themselves.

The first constitution ever written from scratch for a state in the nationbuilding stage was written for Iraq by an English colonial administrator, Edward Bonham-Carter, judicial secretary for Mesopotamia, in 1924.

General Douglas MacArthur seems to have written Japan's constitution largely on his own, and it still rules the lives of 127m people. More recently, constitutions for broken, reconstituted or newly created states have been written for East Timor, the Balkans, Afghanistan, and Iraq again. In 1993, South African lawyers wrote their country's new constitution with very little help from outsiders.

In Iraq, after the US-led invasion, there were no such domestic resources: almost 30 years of Saddam had destroyed them. But half a dozen Iraqis in Baghdad, some of them lawyers who had practised in other countries, wrote the transitional administrative law which has served as the country's constitution since March 2004. It guarantees federal autonomy for minorities, shares oil money equitably, and ensures equal rights for women in language that the mullahs of Falluja and the ayatollahs of Najaf could endorse. That constitution will remain in force until the new one is ratified by the Iraqi people.

An American professor of law, Noah Feldman, fluent in classical Arabic and several local dialects, and a Rhodes scholar with an Oxford doctorate in Islamic thought, acted as a constitutional adviser to the Iraqi governing council. He led the lawyers working on the constitution through the process, debating the key issues such as the virtues of a single assembly versus bicameralism. He would brief them on the implications of various constitutional choices, and often supplied the wording that gave voice to their decisions.

Feldman came to Baghdad as an employee of the coalition in April 2003, just after the invasion. His book, After Jihad, had established him as a leading scholar at the crossroads where Islam, democracy, and constitutional law meet, but America's neo-cons were surprised at the choice. Feldman had cut his teeth clerking for two of the most influential liberal judges in Washington.

Reminded recently that Iraq had had the very first "colonial" constitution for a nominally independent new state, Feldman says, "That's a nice historical irony, but there's a big difference. The current Iraqi constitution was written by Iraqis themselves, people who left comfortable lives dealmaking in London or doing medical malpractice work in the mid-western US, and returned to Baghdad." Feldman says it was a surreal experience for these prosperous exiles to be back in the bombed-out capital they had fled years before. "It was also surreal for me to think that ideas I had worked on in an academic context, about the compatibility of Islam and democracy, were now going to get a chance for actual expression in a constitutional document."

The last time Iraqis had multi-party elections was in 1953, when the British-backed monarchy was the real winner. Edward Bonham-Carter's constitution lasted 34 years, until the monarchy was overthrown in 1958. Feldman hopes the next Iraqi constitution-the one due in October-will last much longer. "A successful constitution," says Feldman, "has to reflect the country's existing power arrangements. It has to have the legitimacy of coming from representative institutions. And it must be flexible."

Perhaps the key feature of the current constitution is that it guarantees Iraq's main groups a veto over the new one. Any three of Iraq's 18 provinces can scupper the new document in October with "no" votes by a majority of two thirds or more. As the Kurds dominate three provinces, this is often called the Kurdish veto. It is in fact a Sunni veto and a Shia one too. Although Feldman fears that the majority Shias might continue to agitate against this minority The Mother of all Elections veto, the requirement of unanimity reflects the reality that Iraq cannot work if Iraqis don't want it to. Although the current constitution was written by Iraqis and ratified unanimously by representatives of a broad spectrum from every Iraqi religious, ethnic, tribal and political grouping (except the outlawed Ba'athists), the new constitution will enjoy the extra legitimacy of having been written by elected Iraqis and ratified in a referendum.

The new constitution could be very different from the current one. Any American in Washington or the green zone who thinks he is going to tell these people what to write or how to vote needs to spend more time on Haifa Street. Federalism, Islam and oil will be the three dominant issues for the new assembly as it writes the new constitution.

The veto rules will ensure that the present decentralized, federal system will remain in place or may be loosened further: the Sunni Arabs and the Kurds will veto anything that gives the Shia-dominated central government excessive power. There are already three Iraq’s: the Kurds, enjoying relative freedom and prosperity in the northeast; the Shias, dominating Baghdad and the south and co-operating with reconstruction and the elections; and 5m Sunnis, of whom about 10,000 have taken up arms to terrorise everyone else. So there will be nothing new about devolution in Iraq.

The biggest test for how Islamic the new constitution will be is family law. Early last year, the Iraqi women's movement, led by the widely respected minister of public works, Nasreen Berwari (a Kurd), forced the coalition authorities to quash a move to make Sharia the basis of family law. Will Iraq's liberal women be so successful when there is no Paul Bremer to appeal to? Even though a quarter of the seats in the new assembly are reserved for women, they may not be. Regarding oil, the Sunnis - whose region has very little - will use their veto to ensure that the central government collects oil revenues for pro rata distribution among the provinces.

There is no doubt about which group will win the elections. The United Iraqi Alliance, organized by Ahmed Chalabi, comprises all the principal currents among Iraq's 60 per cent Shia majority, plus a small but respectable selection of Iraqis with other allegiances. These include some Kurds, Turkmen, moderate and monarchist Sunnis and the important Shamar tribe, which is mostly Sunni. This coalition, plus additional representatives of Muqtada al-Sadr, should win 55-60 per cent of the votes. As a result of Sunni violence, turnout is unlikely to be greater than Afghanistan's recent 70 per cent showing. (If 85 per cent of Kurds, 80 per cent of Shias, and 30 per cent of Sunnis vote, then overall turnout will be about 70 per cent. If these predictions for the Kurds and Shias seem excessive, remember how long these people have been waiting, and how much they expect from this transitional assembly.)

Iraqis in the 15 provinces that are mostly Arab will vote for two bodies on 30th January: the 275-member transitional national assembly, plus the 15 provincial assemblies. In the three Kurdish provinces, Iraqis will also be voting for a second body: the Kurdish national assembly. Iraqi expatriates will be able to vote too, and 11 countries (including Iran and Britain) with Iraqi residents have agreed to set up polling arrangements for absentee voters.

Any individual, political party or group of parties that had gathered 500 signatures by 15th December has a place on the national ballot. Voters will select one from over 109 of these choices. They will appear on the ballot in an order selected by lottery (recently broadcast on national television). Any ballot choice that receives 1/275th of votes or more will be guaranteed representation in the new assembly. A party that receives 10 per cent of the vote will receive 27 of the 275 seats, and so on. As every third person on the party lists must be a woman, such a party would return nine women to the assembly.

Despite reports to the contrary, and the pronouncements of Bin Laden, every Sunni tendency, including former Ba'athists, is still on the ballot. The leading radical Islamist group, the Iraqi Islamic party, is still on. Two Sunni monarchist parties are on. President Qhazi al-Yawar and his tribe, the biggest in Iraq, are still on. Smaller Sunni tribes are on. There are Sunnis on the big Shia list. There are Sunnis on the Socialist and Communist lists. There are Sunni secular lists, and there are Sunnis on the non-sectarian secular lists. Sunnis of every stripe will find friendly choices on 30th January.

The real question is, will the Sunnis vote? They will be substantially, maybe massively, under-represented when the final count is made. The Shias must address this. They will have to bring Sunnis into the government, heed Sunni interests, and signal to an already unpopular insurgency that this is an Iraqi government-that the meaning of revolt has changed.

Sunni under-representation could provide the Shias with the constitutional opportunity to behave badly. Will they? Will Sistani, Muqtada, Chalabi et al bust open the parental liquor cabinet and take the Bimmer for a joyride? Probably not. Already they have shown great political maturity in forming and maintaining the alliance that will win the election. In Ahmed Chalabi they have a very practical man at the nexus of their politics. And they have been waiting for this chance for five centuries. They don't have the muscle to repress Sunnis (or Kurds), and they have every interest in isolating the insurgency and enjoying the fruits of reconstruction and self-rule.

Iraq under a Shia government will be the test case for the prime challenge in international politics: can Islamic democracy work in the Arab world? The only elected government in the region, it will be democratic-at least at the start. But it will have for the most part a sectarian identity, its biggest factions run by members of an ancient clerical aristocracy and the whole united under the mantle of a grand ayatollah.

Many people believe that Islamic democracy is impossible and that trying to create it in Iraq is just a dangerous adventure for unsophisticated cowboys. But Islamic democracy already exists, and often thrives. The world's second most populous country, India, is home to 150m Muslims who participate in a well-functioning democracy. The fourth most populous nation, Indonesia, is a democratic home to around 211m Muslims. We should also remember that Muslims are far more likely to vote women into power than are many other countries with illiberal histories, such as Germany, France and Japan.

A leading American theorist of Islamic democracy, Noah Feld- manperhaps surprisingly -is a political enemy of the neo-cons who hope to shake up the middle east with their vision of democratic regime change. Although he first went to Iraq as constitutional consultant to the Bush coalition, he had earlier been a volunteer legal adviser to the Democrats in the Florida dispute in 2000. In Iraq, he disagreed with the Bush administration's fears of elected Islamists, and chose to work for the Iraqis instead, doing more than anyone else to write the constitution. Who is he? The most accomplished Islam scholar of his age is a blond and glamorously dressed 34-year old nice Jewish boy from the Bolshevik hotbed of Cambridge, Massachusetts.

Does this all seem contradictory? It shouldn't. Feldman's career and books, and the Iraqi constitution that he helped to write, prove how useless our traditional labels can be in a part of the world where it now seems only the "conservatives" want change. A product of the most elegant colleges and best wine cellars at Harvard and Oxford (Eliot House and Christ Church), Feldman has done the most sophisticated and influential work of any legal scholar of his generation while sleeping on a kitchen floor in the Republican Palace in Baghdad.

Feldman's first book, After Jihad, published in April 2003 just after the invasion of Iraq, posed the question, "Can democracy be made to flourish in the lands where Islam prevails?" His short answer is, "Yes. And that would be good for Muslims and good for the rest of us." He argues that it would even be all right if more extreme Islamists won elections-which history says they are unlikely to do. He politely describes the neo-cons' caustic response to this opinion as, "Feldman, you're naive."

Feldman argues that the virulence of popularly elected Islamists would be moderated by the exigencies of popular rule and global living within one or two electoral cycles. Moreover, he says, continuing to deny popular aspirations in places like Egypt or Tunisia or the Gulf will create worse tensions and more anti-western feelings than allowing Muslims to elect Islamic governments. Indeed, by supporting oppression in these places, our current policies are forging an alliance between democratic hopes and anti-western radicalism. Arguing for Islam's democratic potential, however, Feldman does not rely on the truism that "people just want to be free." He perceives deep historical roots for the main tenets of liberal democracy growing in Islamic soil. A basic egalitarianism underpins the entire religious community, based on the notion that all Muslims are equal before God. The idea of justice, a natural restraint on tyranny, pervades Islamic thought. Most important, says Feldman, is that traditional Islamic rule has always acknowledged the primacy of law.

"The caliphs never had absolute authority. It was limited government, government according to the rule of law. I'm not saying that medieval Islam was a modern democracy, any more than medieval England was. What I am saying is that the historical roots are there for modern Muslims who want to draw on a historical narrative. In the Islamic world it won't do for people to say that democracy is Greek or Roman. They need to identify origins for democratic practice that resonate with their own communities. And there is ample material for that in Islamic tradition."

The generous old-fashioned liberalism of Feldman's intellect and ethics-so different from the dyspeptic politicking of a Michael Moore-pervades his books. In one section of After Jihad, he praises the influence of Al- Jazeera. The channel, with its anti-American polemics, he says, might be bad for America's short-term interests in the middle-east. But it operates free of government interference, Feldman points out, and thus, "In the long run, Al-Jazeera sets the precedent that Arab governments will not be able to control the flow of information to all their citizens. On the whole that is going to be good for the truth. That is the basic theory of free speech and hence of democracy." A Jew who thinks Al-Jazeera is good for America? No wonder the neocons don't like this guy.

In his latest book, What We Owe Iraq, Feldman argues that an occupation such as the current one in Iraq can be legitimate if it acts in a role of "trusteeship" that sees it handing rule in a timely way back to a domestic arrangement that is substantially more legitimate than its predecessor. There is a crucial condition: the first role of the nationbuilder is "to impose security by constituting a power large enough to prevent civil war or anarchy."

In his New York office recently, savouring a recent offer of tenure from Yale and reflecting on those days living on kitchen floors in Baghdad, Feldman assessed the performance of the occupation in Iraq. "We have done a pretty good job of providing free speech and Iraqi participation in politics, and in directing everything towards Iraqi self-government. But in terms of providing security, we just haven't fulfilled our duty."

Despite the occupation's failure to deliver security, and Sunni threats to the 9,000 polling stations, Iraq's Shias and Kurds will vote in huge numbers. Many Sunnis will boycott an election that will ratify their new status as a wait-your-turn minority. Others will be too frightened to vote, or will find their polling stations vaporized by Chechens and Yemenis.

Enough flexibility has been built into Iraq's constitutional and electoral process for the country to have a good chance of surviving this first shock. First, the 30th January national election is simply for an 11-month assembly whose main job is to write the permanent constitution. The Sunnis will have two more chances to go to the polls this year: once to ratify the constitution, once to elect the new government. In other words, if the new assembly is imperfect, that is OK: it is not permanent. If it wants to write an oppressive constitution, it knows the Sunnis and Kurds will veto it. A flawed election this month-and it will certainly be marred by Sunni violence and abstention-is not the end of the world. It is not even the end of the election process. It is just the beginning.