KENSINGTON, Calif. — On a tape recording made in the Oval Office on June 14, 1971, H. R. Haldeman, Richard Nixon's chief of staff, can be heard citing Donald Rumsfeld, then a White House aide, on the effect of the Pentagon Papers, news of which had been published on the front page of that morning's newspaper:
"Rumsfeld was making this point this morning,'' Haldeman says. "To the ordinary guy, all this is a bunch of gobbledygook. But out of the gobbledygook comes a very clear thing: you can't trust the government; you can't believe what they say, and you can't rely on their judgment. And the implicit infallibility of presidents, which has been an accepted thing in America, is badly hurt by this, because it shows that people do things the president wants to do even though it's wrong, and the president can be wrong."
He got it exactly right. But it's a lesson that each generation of voters and each new set of leaders have to learn for themselves. Perhaps Mr. Rumsfeld - now secretary of defense, of course - has reflected on this truth recently as he has contemplated the deteriorating conditions in Iraq. According to the government's own reporting, the situation there is far bleaker than Mr. Rumsfeld has recognized or President Bush has acknowledged on the campaign trail.
Understandably, the American people are reluctant to believe that their president has made errors of judgment that have cost American lives. To convince them otherwise, there is no substitute for hard evidence: documents, photographs, transcripts. Often the only way for the public to get such evidence is if a dedicated public servant decides to release it without permission.
Such a leak occurred recently with the National Intelligence Estimate on Iraq, which was prepared in July. Reports of the estimate's existence and overall pessimism - but not its actual conclusions - have prompted a long-overdue debate on the realities and prospects of the war. But its judgments of the relative likelihood and the strength of evidence pointing to the worst possibilities remain undisclosed. Since the White House has refused to release the full report, someone else should do so.
Leakers are often accused of being partisan, and undoubtedly many of them are. But the measure of their patriotism should be the accuracy and the importance of the information they reveal. It would be a great public service to reveal a true picture of the administration's plans for Iraq - especially before this week's debate on foreign policy between Mr. Bush and Senator John Kerry.
The military's real estimates of the projected costs - in manpower, money and casualties - of various long-term plans for Iraq should be made public, in addition to the more immediate costs in American and Iraqi lives of the planned offensive against resistant cities in Iraq that appears scheduled for November. If military or intelligence experts within the government predict disastrous political consequences in Iraq from such urban attacks, these judgments should not remain secret.
Leaks on the timing of this offensive - and on possible call-up of reserves just after the election - take me back to Election Day 1964, which I spent in an interagency working group in the State Department. The purpose of our meeting was to examine plans to expand the war - precisely the policy that voters soundly rejected at the polls that day.
We couldn't wait until the next day to hold our meeting because the plan for the bombing of North Vietnam had to be ready as soon as possible. But we couldn't have held our meeting the day before because news of it might have been leaked - not by me, I'm sorry to say. And President Lyndon Johnson might not have won in a landslide had voters known he was lying when he said that his administration sought "no wider war."
Seven years and almost 50,000 American deaths later, after I had leaked the Pentagon Papers, I had a conversation with Senator Wayne Morse of Oregon, one of the two senators who had voted against the Tonkin Gulf resolution in August 1964. If I had leaked the documents then, he said, the resolution never would have passed.
That was hard to hear. But in 1964 it hadn't occurred to me to break my vow of secrecy. Though I knew that the war was a mistake, my loyalties then were to the secretary of defense and the president. It took five years of war before I recognized the higher loyalty all officials owe to the Constitution, the rule of law, the soldiers in harm's way or their fellow citizens.
Like Robert McNamara, under whom I served, Mr. Rumsfeld appears to inspire great loyalty among his aides. As the scandal at Abu Ghraib shows, however, there are more important principles. Mr. Rumsfeld might not have seen the damning photographs and the report of Maj. Gen. Antonio M. Taguba as soon as he did - just as he would never have seen the Pentagon Papers 33 years ago - if some anonymous people in his own department had not bypassed the chain of command and disclosed them, without authorization, to the news media. And without public awareness of the scandal, reforms would be less likely.
A federal judge has ordered the administration to issue a list of all documents relating to the scandal by Oct. 15. Will Mr. Rumsfeld release the remaining photos, which depict treatment that he has described as even worse? It's highly unlikely, especially before Nov. 2. Meanwhile, the full Taguba report remains classified, and the findings of several other inquiries into military interrogation and detention practices have yet to be released.
All administrations classify far more information than is justifiable in a democracy - and the Bush administration has been especially secretive. Information should never be classified as secret merely because it is embarrassing or incriminating. But in practice, in this as in any administration, no information is guarded more closely.
Surely there are officials in the present administration who recognize that the United States has been misled into a war in Iraq, but who have so far kept their silence - as I long did about the war in Vietnam. To them I have a personal message: don't repeat my mistakes. Don't wait until more troops are sent, and thousands more have died, before telling truths that could end a war and save lives. Do what I wish I had done in 1964: go to the press, to Congress, and document your claims.
Technology may make it easier to tell your story, but the decision to do so will be no less difficult. The personal risks of making disclosures embarrassing to your superiors are real. If you are identified as the source, your career will be over; friendships will be lost; you may even be prosecuted. But some 140,000 Americans are risking their lives every day in Iraq. Our nation is in urgent need of comparable moral courage from its public officials.
Daniel Ellsberg is the author of "Secrets: A Memoir of Vietnam and the Pentagon Papers."