Raw Data Rarely Produce Certainty

As the pressures of impending war mounted, the facility was moved onto the target list. When it was hit, Iraqi officials claimed a propaganda victory. And the DIA could offer nothing certain, only strong suspicion. The plant was "apparently never used" to produce infant formula, it concluded. In a Feb. 6, 1991, position paper, the agency said the plant "was correctly identified as a suspect facility."

Powell did not back down. "There was a body of evidence to suggest we knew what we were doing," he told USA Today in March 1991. "Some of the so-called baby powder that was around could not have been made there. We saw the packages and read the labels. It was made by a company that was not, to the best of our knowledge, doing business in Iraq."

Powell may well be right that the baby-milk story was an elaborate Iraqi hoax. On the other hand, anyone who has spent time in Hussein's Iraq knows that industrial facilities are commonly surrounded by security fences. As for the camouflage, in trips to Iraq after Desert Storm, I saw countless facilities painted like the baby-milk plant: food storage warehouses, oil tanks, flour mills. Facility managers all said the same thing: Orders from Baghdad had said the facilities should be painted, and painted they were.

In the end, of course, Iraq was proven to have an enormous biological weapons program. And U.N. inspectors who visited the baby-milk site after the war found evidence that three engineers from Iraq's bioweapons program had been assigned to the facility in 1989.The inspectors placed the plant on a monitoring list because the dual-use equipment meant it still could be turned to development and production of biological agents.

What is the lesson to be learned here? What do my photograph of the bombed-out factory and the Bush administration's desire to republish it now have to do with the secretary of State's appearance at the U.N. or the question of whether the United States should go to war with Iraq? Simply this: The photograph and the story of the "baby-milk factory" illustrate the tension that almost always exists between intelligence and decision-making. It arises from the difference between interpretation and certainty.

Gathering, collating and assessing information is what the intelligence community does. As Powell's presentation demonstrated, the information normally comes in bits and pieces from all sorts of sources: an account from an agent, an extracted confession, captured documents, images from spy satellites, snippets of intercepted communications. Intelligence analysts spend their careers studying how best to assemble those fragments, assess their reliability and judge what they mean.