The Senate Select Committee on Intelligence has pronounced judgment on prewar assessments of Iraqi weapons of mass destruction, and despite the continued partisan bickering, one bipartisan point of agreement seems clear: Had the intelligence been done right, the decision on whether to go to war would have been clear. It's a convenient conclusion, absolving lawmakers of responsibility for any errors in judgment they might have made. It's also naive, shortsighted and dangerous.

It's naive because even immaculate intelligence would not have produced the sort of certainty on Iraq that would have made decisions of war and peace obvious; in fact, better intelligence would have made the waters muddier. Before the war, widespread opinion properly held that the absence of evidence of Iraqi weapons of mass destruction should not be confused with evidence that they were absent. There is nothing to suggest that U.S. intelligence assessments, had they been more careful, would have found conclusive evidence of the weapons' absence.

The closest to such an offering before the war was the host of individual Iraqi accounts asserting that Iraq had no such weapons. These have frequently been cited as the overlooked silver bullet -- had they not been discounted, some say, we could have concluded that Iraq was WMD-free.

But even in hindsight, it is hard to see why they should have been seen as credible. Iraqis claiming that their state did not have the weapons ran no special risks from the Saddam Hussein regime, regardless of whether they were honest or were lying; since their statements were cost-free, why should they have been given much credit? In contrast, informants within Iraq with stories of Iraqi perfidy would have been on balance more reliable because their incriminating testimony came at greater risk to their and their families' lives -- and why put your life on the line for a lie? This logic has an exception, of course, for informants of the Ahmed Chalabi-managed variety, who had significant incentives -- the possibility of one day running a liberated Iraq -- to lie. We would apply exactly the same logic today as we did two years ago. Would -- should -- 100 individual claims from within Iran that Iran has no nuclear weapons program change U.S. opinion of that state one iota?

Improving intelligence is important, but ultimately it will never remove all -- even most -- uncertainty about weapons of mass destruction. Every day that we cling to the fiction that intelligence reform will rescue us from uncertainty is a day we delay grappling with the most important proliferation policy challenge we face: wielding U.S. power against the spread of WMD in an inherently uncertain world.

Countless critical questions need to be debated. Where should the burden of proof lie when confronting possible but uncertain proliferation of weapons of mass destruction -- should the suspect state be required to prove its innocence, or should the outside world be required to prove its guilt? Should a state obstructing inspections -- something that can be easily observed -- be assumed to be harboring banned materials -- a much more uncertain conclusion? Should that distinction even matter for purposes of policymaking? Should we regard dual-use materials -- ones whose ultimate purpose is inherently uncertain -- as illicit until proven innocuous, or the reverse? The flagship example here is the now-notorious aluminum tubes delivered to Iraq that some once argued could be used in a nuclear weapons program, but which turned out to be part of a rocket program instead. Should the standards of proof and suspicion under uncertainty be applied evenly, or should they vary with the nature of the state under investigation?

All these are critical questions, and through the summer of 2002, healthy public debate over how to handle uncertainty about weapons of mass destruction was heating up. Beginning in September, though, the foundation shifted, as officials from the president on down began speaking with more certainty about Iraqi weapons, and, in particular, about the critical matter of Iraq's nuclear arms program. Discussion of how to manage uncertainty quickly faded. Two years later, it has not yet returned to the agenda.

Yet it is the central issue we must confront and debate in crafting a responsible WMD strategy. If the soul-searching currently in vogue can admit the limits of intelligence, rather than chase an illusory holy grail, that debate will finally be underway.

The writer is the science and technology fellow in foreign policy studies at the Brookings Institution.