In October 2010, six months after the airstrike footage, came the documents. Nearly four hundred thousand of them, each one a short report written by an American unit about something that had happened that day in Iraq.

What a field report is

Not a policy paper or an investigation. A field report is the military equivalent of a shift log. A unit records that something happened at a given time and place, who was involved, and what the outcome was.

They are terse, full of abbreviations, and written by people with no expectation that anyone outside would read them. That is exactly what makes them useful. Nobody was writing for an audience.

A field report isA field report is not
A shift log: time, place, who, outcomeAn investigation
Terse and abbreviatedWritten for an outside reader
Useful in bulk, for patternsConclusive on its own

The scale of it

~390,000Field reports published
2004–2009Period they cover
~15,000Previously unrecorded civilian deaths identified in them
Oct 2010Published, with several newspapers as partners

The finding that mattered most

For years the official position had been, in effect, that the US military did not maintain a count of Iraqi civilian deaths. The logs showed that individual deaths were in fact being recorded, one report at a time, in very large numbers.

Analysts working with the documents found roughly fifteen thousand civilian deaths that were in no existing public count. That one figure changed the historical record of the war more than anything else in the release. How casualty counting works explains why it mattered so much.

Why that one number landed so hard

The official position had been, in effect, that such deaths were not counted. The logs showed they were being recorded all along, one report at a time, in very large numbers.

Detainee treatment

The second major theme was abuse of detainees by Iraqi forces, recorded by American units in the ordinary course of reporting and, in many cases, not investigated further.

This raised a legal question with a name: whether handing detainees to a force known to mistreat them is itself a violation. The documents did not resolve it, but they made it very difficult to argue that the pattern was unknown.

How this connects to the airstrike

Same source, same disclosure. Chelsea Manning sent the video and the logs together as part of one act, inside a single extraordinary year of publication.

The logs also contain the report of 12 July 2007 itself, written in the flat institutional register these documents use. Reading that entry next to the footage is one of the most instructive exercises available in this whole archive, because it shows exactly how much of an event survives the journey into a database. Very little does.

The gap between a record and an event

Forty minutes, somewhere between twelve and eighteen deaths, two journalists, a man who stopped to help and two wounded children, compressed into a few lines of abbreviated text. Multiply that by three hundred and ninety thousand and you have a sense of what the logs are and are not.

How they were handled

The logs were handled differently from the airstrike video. Several large newspapers were given weeks of advance access. They reviewed the material, removed names that could put people at risk, and prepared their own reporting.

The redaction was better than in the earlier Afghan release, though the arguments about whether it was good enough continued for years and eventually helped split WikiLeaks from its media partners.

What it is like to actually read them

Dull, mostly, and that is the point worth making to anyone who expects revelation on every page. The overwhelming majority of entries are unremarkable: a patrol went somewhere, nothing happened, a vehicle broke down, a weapons cache was found.

The value is cumulative rather than individual. No single report proves much. Several hundred thousand of them, searchable, show patterns. Where deaths clustered. How often a kind of incident came back. What got investigated and what did not. That is a different kind of evidence from a single damning document, and it needs a different kind of reader.

It also explains why the reporting on the logs took weeks and involved teams rather than individuals. You cannot skim your way to a finding in a database this size; you have to count.

What the Iraq war logs are worth now

They remain the single largest body of contemporaneous primary material about the Iraq war available to the public, and researchers writing about that period still use them.

They are also a reminder of the strange shape of this whole story. The most complete record of a war's daily conduct reached the public because one person copied it without permission, and the most serious consequence anyone faced for its contents fell on that person. The Afghan set is covered here.

Questions people ask about this

How many documents were in the Iraq War Logs?

Roughly 390,000 US military field reports covering the period from 2004 to 2009.

What was the most important finding?

Records of roughly 15,000 civilian deaths that had not appeared in any public count, which contradicted the position that such deaths were not being tracked.

Were the documents redacted?

Yes. Newspaper partners had advance access to review them and remove names that could endanger people, though arguments about whether redaction went far enough continued for years.

Where this comes from

  1. The Guardian's Iraq war logs coverage — Guardian News and Media
  2. Iraq Body Count — Iraq Body Count project
  3. WikiLeaks — WikiLeaks

Spotted something wrong? Tell us and we will fix it.