Shrapnel is raining on Jon Stewart, as it often does on heretics. In an appearance on CBS’ The Late Show, Stewart endorsed the theory that COVID-19 originated in a lab in Wuhan, China. That theory deserves a full inquiry. What seems to have peeved Stewart’s critics most is that he dared to question science. Make that “Science,” as in “follow the.”

“Science is incredible,” Stewart told Stephen Colbert. “But they don’t know when to stop.” And then he went, apparently, too far:

Can I say this about scientists? I love them and they do such good work but they are going to kill us all. … Here’s how I believe the world ends. … The world ends, the last words man utters are somewhere in a lab. A guy goes, ‘Huhuh, it worked.’

Dan Rather was appalled:

I cannot overemphasize how dangerous this line of thinking is. It is true that some scientists have done some bad things in the name of research—such as the Tuskegee experiments.… Science does not release us from our moral responsibilities. All of this is the case because science is a human endeavor and scientists are human, subject to the same frailties and base instincts as any member of our species. But science is also a way of thinking, where we challenge our own dogmas and beliefs, where we change our minds and approach when the data show we were wrong. 

If scientists are human, subject to human frailties, Rather might consider what frailties befall humans with unchecked authority. More on this presently. But first: It is true that science is a way of thinking. But its methods and assertions depend on philosophical resources that are external to science—something not all scientists acknowledge. Data does not capture the whole of truth. The claim that it does is itself a philosophical position that data cannot prove without the help of philosophy. Moreover, data and evidence cannot tell us what to do with them. That requires moral judgment.

Rather says we need science more than ever because of challenges like climate change. It is true that scientists are vital on a variety of issues. But these issues are ultimately matters of prudential judgment that cannot be solved by the scientific method.

There are several reasons for that, as I have argued in National Affairs. One is that scientists rarely form one voice, and the heroes among them are often the apostates who challenge the prevailing consensus. A second is that few if any political challenges can be solved scientifically. They entail value judgments and, more important, judgment calls. A third is that priorities compete with each other. Which experts should we listen to on climate change? Geographers? Oceanographers? Physicists? Economists?

Moral and political judgment must always superintend science. Science is a tool. It cannot supply its own moral suppositions or demonstrate, purely within itself, that it is the supreme way of knowing. In a republic, politics must play the role with respect to expertise that St. John Henry Newman assigned to university faculties. A university, he wrote, is “[a]n assemblage of learned men, zealous for their own sciences, and rivals of each other, are brought, by familiar intercourse and for the sake of intellectual peace, to adjust together the claims and relations of their respective subjects of investigation. They learn to respect, to consult, to aid each other.”

In politics, prudence operates on something like this synthesis. Experts are not paid to be sensitive to other forms of expertise. The politician must be aware of the whole in the same way Newman said universities must: “There is no science but tells a different tale, when viewed as a portion of a whole, from what it is likely to suggest when taken by itself, without the safeguard, as I may call it, of others.”

The slogan “follow the science” is meant to exempt politicians from the duty of judgment. If we are to follow scientists without the “political interference” we are told to dread, we should not only be wary of abuses. We should expect them. As Bertrand de Jouvenel taught, power’s intrinsic desire is to grow. That is why checks are necessary to prevent abuse.

Paul Waldman of The Washington Post, reasonably arguing that celebrities should swim in their own lanes was particularly defensive of experts:

[Stewart’s] attack on expertise reminds us why expertise is so important. … That’s not to say that experts don’t often have biases or blind spots, because they do. Sometimes, they can be catastrophic. But it’s not because experts can’t be trusted, it’s because something kept them from seeing what they should have, or — perhaps more often — they just didn’t have enough information to arrive at the best judgment.

That view of experts is built atop a romantic idea of human nature. Substitute “politicians” for “experts,” and one can see why. Are we so far removed from the 20th century that scientific and technocratic abuse is unthinkable? Waldman may be correct that most experts mean well, but that does not address the reality that people with power need to be watched. Expertise can become tyrannical when it denies the authority of politics to question it. That is not to accuse any individual of doing so, but a theory of complete deference to experts—besides entangling itself in internal tensions—abdicates political responsibility.

There is a direct line between an ethic of deferring to experts and early 20th century Progressivism, a movement whose leaders—like Woodrow Wilson—would not all survive today’s scrutiny. Wilson himself is proof that expertise can be helpful or haughty. It can inform judgment or so ensconce itself in rigidity that accommodation to circumstances becomes impossible.

Early Progressives, like Lester Frank Ward, were so enamored of expertise that they thought the scientific method could be applied to politics. Their motive bears emphasis: In Dynamic Sociology, Ward argued that expertise was necessary because citizens were ignorant. The point was to empower “the few progressive individuals by whose dynamic actions social progress is secured.” The use of experts could “place [Americans] upon the highway to a condition of intelligence which, when attained, will in turn work out the problem of inaugurating a scientific legislature and a system of scientific legislation.”

Never mind the condescension. This is a road to abuse. Especially in today’s academic climate of hyper-specialty, scientists might well not see either the potential dangers of their work or, more important, sources of knowledge beyond it. Stewart’s prediction of scientists wreaking disaster—delivered, again, as a comedic rant—should not be dismissed out of hand.

There is a populist attack on experts that consists of denying their expertise. It is part and parcel of the resistance to anything that smacks of distinctions between people, especially, perish the thought, if some are asserted to know more about certain things than others. But denying their authority and denying their expertise are different things. Total deference to experts abandons civic responsibility. So does political manipulation: Scientists and other experts should be allowed to do their work. But that work should inform the prudential judgment of politicians, who must be allowed to do their work as well.