Remember the referendum? Last weekend the world's airwaves were full of broadcasts about the success of the voting in which millions "defied the insurgents" by turning out to cast their ballots.
Then we heard preliminary but "informed" speculation that the constitution had passed. Majorities of Kurds and Shias had given it enthusiastic support in the north and south-east.
In Sunni areas, where voters had been expected to reject it, not enough had come forward to turn it down.
The rule was that if two-thirds of voters in any three provinces rejected the constitution, it would fail. Election officials conceded that two-thirds had done so in the two fiercely anti-American provinces that include Falluja, Ramadi and Tikrit.
But Nineveh, which Sunnis share with Kurds and Christians, had not produced a big enough no vote. So the message was: "Sorry, Sunnis. Our constitution is safe."
Along comes a second big Iraqi event: the trial of Saddam Hussein. Important though it is as a catharsis for the former dictator's hundreds of thousands of surviving victims, it has little political significance since only a small minority of Iraqis still support him.
Of course, it could backfire on the Americans if Saddam is humiliated in court by unfair or high-handed treatment.
To a wider circle of Iraqis, and other Arabs, he might then become a symbol of wounded national pride, as he was briefly when Washington published pictures of his mouth being examined by a military dentist after capture.
Manipulating the trial's timing is the real story. Why suddenly this week? A fortnight ago, at Chatham House in London, Iraq's president, Jalal Talabani, said he did not know when the trial would take place.
Within days a date was fixed, conveniently diverting reporters' attention from the referendum count.
With the issue out of the spotlight, it is a fair bet that when the official result is declared - perhaps today - the announcement that the constitution has passed will be treated as pretty dull since we already "know" that from the weekend leaks by Condoleezza Rice, Jack Straw and the Iraqi government.
How could they be sure, since counting was not yet complete? Was the fact that the count would be flawed the real thing they knew? Was the trial an improvised political device to get rigging out of the headlines?
On Monday some Iraqi election officials were beginning to say they had come upon major irregularities and suspiciously high Kurdish voter turnouts, in places exceeding 95%. Below the radar of the Saddam Hussein trial, more questions have been raised.
Turnout figures in such cities as Najaf doubled from an initial figure of 45%. In Nineveh and Diyala, another province with a Sunni Arab majority, officials initially talked of startling yes votes of up to 70% in each.
Later, they changed the Nineveh figure to say the no votes had won - but the figure was only 55%, and so below the crucial 66% threshold for rejection.
In a rigorous analysis for the Inter Press Service [full article below], the American scholar Gareth Porter questions even that figure. He says it is based on an unbelievably low turnout among Sunnis.
It implies that Nineveh's Christians, who had declared their opposition to the constitution in advance, changed their mind on the day.
He quotes a US military liaison officer who used to work there as admitting that Kurdish officials, who have long vied for control over Mosul, Nineveh's main city, and inflate its population figures, stuffed ballots in January's election and may have done it again.
Does this matter? The constitution will be declared to have passed, because the Bush administration wants it passed.
It paves the way for elections in December, which will be spun as further proof of Iraq's gradual democratisation.
Yet it will have been bought at a high price. Cheating the Sunnis is not a sensible policy, especially when, out of the other side of its mouth, the Bush administration claims to be trying to get them into the political process.
The fact that large numbers of Sunnis voted last week does not mean they no longer support armed resistance. Reporters in Falluja found voters who said they backed the insurgency even as they cast their ballots. Jonathan Steele @ Guardian
REFERENDUM VOTE FGURES DON'T ADD UP
The early vote totals from Nineveh province, which suggested an overwhelming majority in favour of Iraq's draft constitution that assured its passage by national referendum, now appear to have been highly misleading.
The final official figures for the province, obtained by IPS from a U.S. official in Mosul, actually have the constitution being rejected by a fairly wide margin, but less than the two-thirds majority required to defeat it outright.
Both the initial figures and the new vote totals raise serious questions about the credibility of the reported results in Nineveh. A leading Sunni political figure has already charged that the Nineveh vote totals have been altered.
According to the widely cited preliminary figures announced by the spokesman for the Independent Electoral Commission of Iraq (IECI) in Nineveh, 326,000 people voted for the constitution and 90,000 against.
Those figures were said to be based on results from more than 90 percent of the 300 polling stations in the province.
Relying on those "unofficial" figures, the media reported that the constitution appeared to have been passed -- on the assumption that the Sunnis had failed to muster the necessary two-thirds "no" vote in Nineveh.
No further results have been released by the IECI since then, and the final tally from the national referendum is not expected until Friday at the earliest.
However, according to the U.S. military liaison with the IECI in Nineveh, Maj. Jeffrey Houston, the final totals for the province were 424,491 "no" votes and 353,348 "yes" votes.
This means that the earlier figures actually represented only 54 percent of the official vote total -- not 90 percent, as the media had been led to believe. And the votes which had not been revealed earlier went against the constitution by a ratio more than 12 to 1.
These ballots could only have come from the Sunni sections of Mosul, a city of 1.7 million people.
Although the votes from polling centres in those densely populated urban areas would take longer to count than those from more sparsely populated towns and cities outside Mosul, they should not have taken much longer than those for the Kurdish sections of Mosul.
Thus there seems to be no logistical reason for failing to announce the results for the 340,000 votes that went overwhelmingly against the constitution.
Rather, the evidence suggests that it was a deliberate effort to mislead the media by Kurdish and Shiite political leaders who were intent on ensuring that the constitution would pass.
They knew that all eyes would be on Nineveh as the province where the referendum would be decided.
By issuing figures that appeared to show that the vote in Nineveh was a runaway victory for the constitution, they not only shaped the main story line in the media that the constitution had already passed, but effectively discouraged any further media curiosity about the vote in that province.
The final figures revealed by the U.S. military liaison with the IECI suggest a voter turnout in Nineveh that strains credibility.
On a day when Sunni turnout reached 88 percent in Salahuddin province and 90 percent in Fallujah, a total of only 778,000 votes -- about 60 percent of the eligible voters -- in Nineveh appears anomalous.
Even if the turnout in the province had only been 70 percent, the total would have been 930,000.
The final vote totals suggest that the Sunnis, who clearly voted with near unanimity against the constitution, are a minority in the province.
It is generally acknowledged that Sunnis constitute a hefty majority of the population of Nineveh, although Kurdish leaders have never conceded that fact.
A total of 350,000 votes for the constitution in the province is questionable based on the area's ethnic-religious composition.
The final vote breakdown for the January election reveals that the Kurds and Shiites in Nineveh had mustered a combined total of only 130,000 votes for Kurdish and Shiite candidates, despite high rates of turnout for both groups.
To have amassed 350,000 votes for the constitution, they would have had to obtain overwhelming support from the non-Kurdish, non-Arab minorities in the province.
According to official census data, before the invasion of Iraq in 2003, Assyrian Christians and Sunni Arabs accounted 46 percent of the more than 350,000 people on the Nineveh plain. Most of the others are Shabaks and Yezidis. Kurds represented just 6 percent of the population.
But the Kurds have asserted political control over the towns and villages of the plains, with a heavy Kurdish paramilitary and Kurdish Democratic Party (KDP) presence.
That Kurdish presence provoked widespread opposition and some public protests among non-Kurdish communities on the plains, especially Christians and Shabaks.
Assyrian Christians are particularly afraid the constitution's article 135, which divides the Christian community into Chaldeans and Assyrians, will be used by Kurds to expropriate their lands and villages in North Iraq.
Michael Youash, director of the Iraq Sustainable Democracy Project in Washington, has spoken with Assyrian Christian leaders in two district towns, Bakhdeda and BarTilla, on the Nineveh plain where Christians represent roughly half the combined total population of more than 100,000 people.
He says Assyrian Christian political organisations mounted big demonstrations against the constitution in both towns, and that their local leaders are sure that very high percentages in both towns voted against the constitution.
In response to an e-mail query, Maj. Houston, the U.S. military liaison with the IECI, said, "It was my understanding that the Christian communities would be opposed to the constitution," but he dismissed the suspicions of vote fraud in the province.
Saleh al-Mutlek, one of the Sunni negotiators on the constitution last summer and now a leading opponent of the constitution, told reporters, "There is a scheme to alter the results" of the vote.
He alleged that members of the Iraqi National Guard had seized ballot boxes from a polling station in Mosul and transferred them to a governorate office controlled by Kurds.
A former U.S. military liaison with the Nineveh province IECI has confirmed a similar incident of seizure of ballot boxes from a polling station during the January elections.
According to Maj. Anthony Cruz, Kurdish militiamen tried to bribe local electoral commission staff to accept ballots that had obviously been tampered with. Cruz also confirmed a much larger ballot-stuffing scheme by Kurdish officials in the province, as reported by IPS in September.
On Monday, the Electoral Commission announced that it would conduct an audit to examine the high "yes" vote, but it is not clear that it will include the results in Nineveh.
Gareth Porter @ IPS