And then I listened to the post-fight commentary, and felt, as Clinto said of Lazio, that I had been “orbiting another planet.” Immediately after the debate, on MSNBC’s “Hardball,” Peggy Noonan, a notorious Hillary-hater, described the first lady’s manner as “robotic.” Even Clinton chronicler Gail Sheehy, presumably slotted as pro-Clinton, scored the fight as a draw. All agreed that Lazio had shown his mettle. In the next day’s paper, experts impaneled by the New York Daily News gave Lazio the nod 57-50; the editorial page concluded that Lazio “had exuded a more New York sense.” The New York Times editorial page, no friend to either candidate, called Lazio the “smoother, more aggressive performer.” David Broder declared that Lazio had proved himself a force to be reckoned with. What had looked to me like comic swagger and cartoonish bluster had registered as confidence and backbone with Broder. Times columnist Gail Collins did describe Lazio’s performance as “crazed,” but voices like hers were markedly in the minority.

And then the vox populi was weighed and measured. Instant polls were mixed — one for Clinton, two for Lazio. But opinion was remarkably unanimous among undecided voters. All six of the fence-sitters whom the Times had been following in recent months felt that Clinton had carried the day. In USA Today, the figure was eight for eight. Of the 10 undecided voters polled by the Buffalo News, most had moved into the Clinton column. And a Times/CBS poll conducted between one and six days after the debate provided the decisive evidence: Not only was Clinton voted the winner 47-34, but a majority of voters believed that Lazio had been too aggressive, while scarcely any felt that way about Clinton. In fact, the debate helped Clinton widen her lead in the poll from five to nine points.

The same language kept recurring in interviews: One respondent told the Times, “He came off as a much more immature person, and she came off very collected.”

The extremely wide range of judgments should remind us that the means of expression available to us — gesture, tone, word choice, facial expression — are so open-ended and indeterminate that three people scrutinizing the same set of behaviors can offer three different interpretations. But in this case the range has a pattern. What can explain the gap between expert and ordinary opinion? One hypothesis is that the public didn’t get it, and the experts did; that is, that Clinton really was robotic, and Lazio really did exude a more New York sense. But this would be odd, since the experts are essentially trying to anticipate public opinion rather than form a personal impression. A likelier explanation is that the undecided voters were reacting to their own intuitive judgments, while the experts were reacting to their own expectations. The experts, in effect, weren’t receiving their own intuitive judgments.

This is, of course, a general problem with expertise. You can’t fully experience any new performance of “Macbeth,” or “La Traviata,” because you’re conditioned by the expectations created by so many prior peformances. You hear what you already “know.” In this case, Noonan “knew” that Clinton was robotic — and of course wanted her to be robotic — while the communications mavens impaneled by the Daily News, who may have had no preference of their own, “knew” that Lazio’s likeability was bulletproof, and so must have come across as “feisty” rather than nasty. In contrast, the undecided voters, who were hoping to make a decision based on what they saw, had good reason to clear their minds of predisposition. They actually experienced the performance before them. (I, of course, may have been influenced by my dim view of Lazio, which I have expressed in print; but I don’t hold Clinton in terribly high regard either.)

But there is something Clinton-specific at work here. The great media blunder of the last few years was the widespread projection onto public opinion of the media’s own scorn for President Clinton. The media thought that it was anticipating public opinion, but in fact the public resisted the media’s own view.

Something similar may be at work with Hillary Clinton. I can’t count the number of times that I have been confidently told over the past year and a half that Clinton was not only going to lose the Senate race, but that she was going to get shellacked.

Until very recently, Clinton has been widely seen as a loser — a personal view which, as with the president, took the form of a prediction about public opinion. And scorn for Hillary, a whole subject in itself, runs even deeper than scorn for her husband, since she’s charmless, bland and self-righteous — and sometimes mechanical — while he’s a seducer and a rogue. Indeed, a shrink recently explained to me her theory that people hate Hillary as compensation for their inability to hate Bill, much though they feel he deserves their contempt.

Scorn keeps you from seeing straight — or hearing accurately. Clinton is never going to be popular, but she has always had a better shot at winning the race than she has been credited with. She is, after all, a pro, which is more or less what she demonstrated the other night. It’s fairly clear that the debate will come to be seen, in retrospect, as a modest step forward for Clinton, and a serious step backwards for Lazio. The Times poll found that her favorability ratings had grown slightly since June, while his unfavorables had soared. It may be cosmically unfair that the Clintons just keep on winning, but the truth is that they’re better than the competition.

It was, and is, a strikingly private and personal use of religious morality — more Baptist, to my ears, than Jewish. In the Jewish world in which I grew up in the mid-1960s, morality meant caring about poor people. Social justice was the great watchword of the temple and of the entire Jewish milieu. Because we had once been slaves in Egypt, because we had been oppressed and dispersed among strangers and because we had brought the scripture of democratic socialism with us from Eastern and Central Europe, we identified with the downtrodden.

There was no greater feather a rabbi could have in his yarmulke — if he wore one — than to have marched in Mississippi, as ours had. And of course we believed in the government, and in the state’s obligation to mobilize itself on behalf of the poor. The one sermon from the High Holy Days that I recall is a thunderous denunciation of Barry Goldwater two months before the 1964 election — to have voted for the man would have been the closest thing to a sin in our lexicon.

Now, it’s perfectly true that the species of 1960s suburban Reform Judaism in which I was raised was so thoroughly reform that it was scarcely distinguishable from Unitarianism. Mentioning God was practically bad form, and we would have cringed at Lieberman’s invocation of “Dear Lord, maker of all miracles.” Lieberman’s faith is a true and demanding one, while ours required principally that the grown-ups vote for LBJ and the kids leaflet the commuter trains to protest the Vietnam War.

But whatever its devotional deficiencies, the Judaism with which I was raised embraced the spirit of universalism, of moral obligation and moral passion, that lies at the core of the faith. Of course you were supposed to be upright, but what mattered was doing good in the world.

Moral engagement was, of course, the spirit of the moment. First the civil rights movement, then President Johnson’s Great Society and finally the protest against the war had the character of a holy crusade. Everything felt like a battle to the utmost, up to and including our high school English curriculum. It was a time when the language of public discourse was intensely moral but almost wholly secular.

President Johnson talked about justice, not divine will. It would be inconceivable for him to have announced that he asked himself, as George W. Bush recently said he did, “What would Jesus do?” And it would have been unnecessary: Secular morality did not have to be authenticated by a show of religious faith.

The coordinates of our own world are almost exactly the opposite of that vanished one. Piety has now become not just permissible but obligatory in political debate; public figures compete for proximity to God. And yet the effect of all this religious profession has not been to moralize policy debate but to politicize private behavior.

President Clinton’s moral nature has become the great consuming issue of a presidential campaign in which he is not running; Bush and Dick Cheney promise moral renewal, while we are asked to admire Lieberman’s courage because he was the first Democrat to denounce the president.

A great rift has opened up between a theologized and passionate debate over private behavior and a policy discussion that seems narrow, cool and technocratic. Can it really be that the most burning issue the nation faces — besides, of course, the question of precisely how much we should despise the president for his extramarital sex life — is whether to permit 20 percent of the Social Security trust fund to be privately invested?

I wish that Lieberman would use the words “social justice,” or speak of a religious obligation to do something about poverty, as modern Catholic theology does when it speaks of the “preferential option for the poor.” Perhaps he’ll do so Wednesday. But I’m not asking him to bless the traditional liberal agenda, nor expecting him to use the outdated vocabulary that all we naive lefties deployed in the ’60s.

Lieberman favors the use of school vouchers for inner-city children locked in bad schools, a policy that most liberals abhor — and that Gore opposes — but that answers the prayers of many desperate parents. It is also obvious now, as it wasn’t back then, that welfare shouldn’t be viewed as an inalienable right, that violent criminals should be locked up for a long time no matter how disadvantaged their background, that demanding less of black students is, as Bush has deftly put it, “soft racism.”

We are rightly chastened by past failures. But where is the urgency to act, to do it right and, yes, to spend the money that needs to be spent? Wouldn’t a religiously inspired person feel that urgency?

But so would a not religiously inspired person. In the end, we don’t really need a liberal Jewish moralism to oppose to a conservative, largely Protestant moralism. In fact, if the function of piety in the public sphere is to further personalize politics, and to elevate matters of character at the expense of policy, then what we need is less faith and more secular moralism. We need to remember that politics matters not because it is the theater of public virtue but because it is the means by which we distribute public goods and adjudicate matters of public policy.

And so let us admire candidates for the personal qualities evinced by their sincere religious commitments, and then have done with the whole thing. Or better yet, let’s leave religion out of it altogether. Indeed, it would have been really brave if Gore had nominated an atheist — but that’s one minority group the American people may not be ready to accept.

Michael’s friend Duncan explained that the Chinese had built a series of walls to keep out intruders, before and linking them all together to form the Great Wall.

“Who built it?” I asked.

“Emperor Qin Hai,” Michael said. Duncan then explained the Egyptian belief in the transmigration of souls. Both boys were able to give me a pretty passable explanation of gravity. Both seemed delighted with their own knowledge.

Half the kids at Morse are poor enough to qualify for free lunches, but in 1998, five years after the school installed its new curriculum, all grades scored at or above the national norm in math and reading. The first-graders placed third in the city in reading, behind two upper-class schools.

Several months later, I was invited to participate in a round-table discussion of school reform at Harvard. In most settings I’m an educational moderate, but among professors at the Harvard Graduate School of education I qualify as a reactionary. The prevailing attitude around the table was that schools were far too preoccupied with “assuring academic skills,” that “we’ve gone berserk in this country with standardized testing,” that we need to cultivate “emotional intelligence.” Everyone talked about their favorite school; I talked about Morse. Immediately two of the speakers jumped on my comments.

Morse wasn’t really a high-poverty school, they said; it was full of Asian immigrants. It wasn’t the Core Knowledge curriculum that accounted for Morse’s success; any school with a strong “identity,” one said, would succeed. The other compared schools with specified curricula to “McDonald’s franchises.”

What’s going on here? Why the wish to deny an obvious success?

We have known for at least the last 20 years that American children generally perform far worse than their peers in much of Europe or East Asia on a wide range of measurable academic skills, and that children in inner-city schools, specifically, suffer appallingly high rates of academic failure. The knowledge has produced a sense of national alarm; and the alarm, in turn, has provoked a great deal of research and experimentation. We know a lot more than we used to about what works. It has become plain, for example, that many of the progressive, “child-centered” practices now in favor just don’t work as well as a more traditional focus on basic skills and a rigorous curriculum. In short, we know that some of the least romantic school practices work best.

But the pull of romanticism, especially on educators themselves, is very strong: We may well continue fighting the war over the classroom to the last child.

You might think progressive and traditional educational theories represent competing programs to achieve an agreed-upon goal, and that scholars would welcome signs of achievement in the same way that medical researchers welcome evidence that one cure works better than another. But you would be wrong.

Inside the educational world, these two streams of thought are viewed more as alternative moral and philosophical systems, as fighting faiths. What is mere evidence in the face of such iron certitudes?

Indeed, in “Academic Achievement Challenge,” a recently published study of effective classroom practices, Jeanne S. Chall, one of the leading figures for the last 40 years in the study of the teaching of reading, points out that studies throughout this century show that “the more traditional methods of teaching reading in the primary grades (i.e., relying on phonics) result in better achievement” than more informal methods in which the child learns to recognize words and sentences through direct contact with books (i.e., whole language). She writes: “Why, one might reasonably ask, do we not accept the research findings and base our instruction on it?” The answer, she says, is that “values, ideologies, philosophies and appealing rhetoric” weigh more heavily than “reason and common sense.”

Ideological and visionary school literature has flourished in the absence of crisp evidence. But that era is drawing to an end; education looks more and more like a real social science; specific instructional practices have been isolated and studied as never before. And the outcome is pretty clear: One study after another has shown that traditional instructional methods, which Chall calls “teacher-centered,” produce better academic results than progressive, “student-centered,” ones, and not only in reading.

A recent study of 24 models of schoolwide reform found that only three could furnish “strong evidence of positive effects on student achievement.” All three were teacher-centered. Two of them, called Success For All and Direct Instruction, focused on imparting basic skills to children in the early grades; teachers read from a script, and students were often called upon to respond in chorus. The third, High Schools That Work, offered a rigorous academic curriculum combined with a strong vocational program. Core Knowledge offered “promising evidence,” and a recent study found that children in Oklahoma City schools who had used the Core Knowledge curriculum for only one year significantly outperformed comparable children in reading, vocabulary and social studies. The more progressive models, in which learning is seen as a natural outcome of a child’s curiosity and teaching as an act of guidance, registered “marginal evidence” or “evidence of mixed, weak or no effects on student achievement.”

Some progressive educators have begun to acknowledge the virtues of a traditional curriculum and even the need for measurable results (just as some traditionalists have come to see the merits of such progressive innovations as small schools with a strong sense of community). Nevertheless, it would be naive to think that the new evidence will carry the day.

Traditionalism may have the results, but progressivism still occupies the philosophical high ground. Progressives see themselves as children’s champions, their advocates in the face of adult society’s preoccupations — social conformity, conventional success and “standards.” The core of the progressive faith, writes Alfie Kohn, the author of “The Schools Our Children Deserve,” is: “Kids should be taken seriously. Because learning is regarded as an active process, learners are given an active role.” In what Kohn calls the “Old School of education,” on the other hand, “The idea is to have students memorize facts and definitions, to make sure that skills are ‘drilled into’ them.”

It’s no coincidence that this process sounds physically painful. Kohn writes as if traditionalists derived sadistic satisfaction from their mistreatment of the young. The two educational views have thus become proxies for political views — progressivism for a liberal concern about social justice, and traditionalism for a defense of the existing order. And so it has required an act of courage, at least in the world of the bien-pensant, to stand up for standardized tests or a fact-based curriculum.

Jim Coady, the principal of the Morse School, found himself denounced as a racist and a reactionary when he adopted the Core Knowledge program. E.D. Hirsch, the literary scholar who is the theoretician and creator of Core Knowledge, writes that he is a “political liberal” who believes that “the only practical way to achieve liberalism’s aim of greater social justice is to pursue conservative educational policies;” but Hirsch has been vilified in the graduate schools of education.

You have to wonder how long this can last. There are simply too many people, like Hirsch, who are politically liberal but educationally conservative for this formulation to continue to seem credible. And even political conservatives have stopped writing off the public schools in favor of vouchers or other privatization schemes, and started writing studies like “No Excuses: Lessons From 21 High-Performing Schools,” a very useful guide from the impeccably right-wing Heritage Foundation. And the evidence is just too clear-cut. Educational progressivism, whose origins go back to Rousseau and Emerson and John Dewey, is profoundly appealing and very much in the grain of American liberalism; but its intellectual monopoly is fading.

The smug certainty that one is speaking for children, or for the poor, precludes debate and makes self-reflection unnecessary. This is precisely why so many years had to pass before liberals could accept the idea that welfare was having a harmful effect on many of its own beneficiaries. Welfare was a right; protecting that right against the encroachment of conservatives became a defining issue for liberals — until it became impossible to deny that welfare dependency was an obstacle to self-sufficiency. Finally it became acceptable to favor welfare reform.

The same process occurred with bilingual education, which was a benefit given to largely impoverished Hispanic immigrants. Only after mountains of evidence accumulated that bilingual instruction did not help children make the transition to English –and after it became plain that the majority of beneficiaries did not want or believe in it — did liberals even begin to examine the merits of the issue. In education, as elsewhere, liberalism must be rescued from itself.