One thing we have learned from this presidential campaign is that an honest debate on the terrorist threat is not possible. This is unfortunate, since no debate would be more important.

When President Bush lapsed briefly into candor before the Republican convention, pointing out the self-evident truth that the war on terrorism can never be won in the traditional sense, the Democratic camp lit into him for alleged defeatism.

When Sen. John F. Kerry recently pointed out the equally obvious truth that it would be a great thing if we could ever beat terrorists back to the point that they represented only a nuisance, the Republicans pretended that he had said fighting terrorism was the same as fighting prostitution.

Presumably neither candidate will again commit honesty on this subject, unless by mistake, late at night when they are very tired. In some ways, of course, their campaigns remain focused on terrorism. Many of the words and phrases spinning out of the debates -- Iraq, shipping containers, border guards, North Korea -- in the end come back to how best to keep the country safe. On Kerry's Web site, you can find a serious strategy for corralling loose nukes from unsafe locations.

But the discussions mostly take place one or two steps removed from the horrifying central questions: How likely is a catastrophic attack? How do we know? Have U.S. policies made an attack less likely? If they eventually do make an attack less likely, will we know that, or are we "doomed to a perpetual state of panic?"

Those words are Ashton B. Carter's, writing in Foreign Affairs, and his answer is, no, we are not so doomed, if we would work far more energetically to prevent an attack (Carter, now at Harvard, served in the Clinton Pentagon). But you don't finish his article feeling reassured.

There are many possibilities to worry about, but two more than all others: nuclear and biological. Both are within the imaginable technical competence of terrorist groups, and both could cause death and (in the nuclear case) destruction on a scale that would make the World Trade Center attack seem like a misdemeanor. Both also could so convulse society that the current debate on FBI agents snooping around library records would seem quaint.

So if we knew such an attack were pending, it would be worth doing almost anything -- including preemptively forfeiting some civil liberties -- to prevent it. But how can a citizen measure the likelihood?

Graham Allison, another Harvard expert and veteran of the Clinton Pentagon, says that a nuclear attack "is more likely than not in the decade ahead." But almost a decade ago, he said nearly the same thing, in a Post op-ed: "In the absence of a determined program of action, we have every reason to anticipate acts of nuclear terrorism against American targets before this decade is out."

I don't point that out to undermine Allison, who I think is performing a valuable service, as is his new book, "Nuclear Terrorism: The Ultimate Preventable Catastrophe," but to point out the perils of prediction and the dilemma for officials and experts who would like to steer the country somewhere between complacency and nervous breakdown.

The experts themselves, even officials with access to all the secrets of government, chasing down thousands of elusive threat reports every day, must be working in the dark more than we would like to know. The rest of us, rightfully excluded from much intelligence, mystified by reports of waxing and waning "chatter," will be all the more so. This administration has made things worse by being secretive even when openness would have been appropriate.

Now politics exacerbate the dilemma. Democrats suspect the administration of hyping the threat, to justify the reelection of a "war president," and simultaneously of hyping how much safer the country has become, to tout the administration's efforts. Republicans have the same contradictory impulses: to insist that Bush has accomplished more than Kerry alleges and that the nation is too much in danger to risk Kerry's untested leadership.

The politicians also may assume that voters don't want to think about terrible things happening. It's true that few of us would choose to dwell on the possibility of central Washington being vaporized, or of a mystery virus felling hundreds of thousands of Americans in some ugly and contagious way.

But surveys suggest that most Americans have a fairly realistic attitude about all this. Two-thirds of adults, one recent poll showed, believe a major terrorist attack is likely in the next year -- and expectations haven't changed all that much as Sept. 11 has receded. Most could probably handle an honest debate about living with uncertainty, about how to prevent the worst -- and how to prepare for it.