- 12 Oct 06, 12:18 PM
There is no debate that Iraq is a violent place. What is fiercely contested is the level of that violence.
Yesterday the respected medical journal The Lancet published a paper suggesting that the number of people dying in the country due to violent causes is way in excess of any previous calculation. The report found that more than 600,000 people have died violently since the war in 2003 - that's 500 people a day, or one in 40 of the Iraqi population.
The report was immediately criticised as being unscientific - not least by George W Bush. It was said that extrapolating from a survey of 12,000 people was ridiculous. The people who conducted the survey countered that the work had been peer reviewed, used standard polling techniques and had been accepted as an accurate way to reflect other atrocities, such as the genocide in Rwanda.
Clearly the story was incredibly important - and the stakes were very high. That's why we decided not only to report the figures and the controversy surrounding them, but to get our science correspondent David Shukman to show the working behind them, allowing the audience to make up their own minds about whether it was an acceptable way to reach such a high figure.
Our world affairs editor, John Simpson, reached the following conclusion: "Iraq is such a violent place that it is almost impossible to tell exactly how many people are dying."
As journalists we're naturally most comfortable when we're dealing with facts - but when it's so difficult to know what the facts are, it's vital we say that too.
Craig Oliver is editor of BBC News at Six and BBC News at Ten