This is the first in a series of essays about Jonathan Rauch’s new book, The Constitution of Knowledge. His series is sponsored by Claremont McKenna’s Salvatori Center for the Study of
Jonathan Rauch is a senior fellow at the Brookings Institution, the author of eight books and many articles on public policy, culture, and government. He is a contributing writer for The Atlantic and recipient of the 2005 National Magazine Award, the magazine industry’s equivalent of the Pulitzer Prize.
Ask the average well-informed American where knowledge comes from, and you can be confident she will come pretty quickly to the phrase “marketplace of ideas.” And no wonder: it is a vivid and useful metaphor, dating to Justice Oliver Wendell Holmes Jr.’s dissent in the 1919 case Abrams v. U.S., where Holmes wrote, “The best test of truth is the power of the thought to get itself accepted in the competition of the market.” The notion is that competing ideas will bid for support, and the market will support the stronger ones while weeding out the weaker.
In my 1993 book Kindly Inquisitors: The New Attacks on Free Thought, I relied on a similar notion to explain how a free and peaceful society produces knowledge. According to the great 20th-century philosopher of science Karl Popper, hypotheses compete to withstand criticism in a scientific ecosystem. On any given day, the survivors of multiple falsification attempts constitute our knowledge.
Twenty-eight years later, I don’t cast aspersions on either of those models, Holmes’s or Popper’s. They are as valuable and insightful as ever. What they are not, however, is complete. In fact, by themselves, they have made liberal societies in America and elsewhere vulnerable to dangerous manipulation by propagandists, demagogues, and bullies, because they omit an essential ingredient. In very brief: it’s the institutions, stupid.
In many respects, Kindly Inquisitors (expanded and republished in 2013) holds up well as an account of what I called “liberal science,” by which I mean a liberal society’s system for producing knowledge. Liberal science includes not just “hard” sciences like physics and chemistry but the entire truth-seeking enterprise, including the social sciences and humanities and for that matter journalism and law and more: any field whose participants search systematically for truth by testing hypotheses, comparing and criticizing contending viewpoints, and hunting for their own and each other’s mistakes.
What liberal truth-seeking’s branches all have in common is their dedication to two defining rules. The first, No final say, posits that any claim might be wrong and so no person or faction can claim a warrant to end debate or control public conversation. The second, No personal authority, posits that whatever we do to check and validate a claim should work for anybody, so that persons are interchangeable, experiments are replicable, and no particular person or faction has any special privileges to decide what is true or false. No final say forces us to hunt constantly for errors, our own and each other’s, thus driving the search for truth ceaselessly forward; no personal authority forces each to persuade all, allowing us to link millions of diverse brains into a global truth-seeking network—the network which mobilized minds and resources with astounding speed to decode the Covid-19 genome, design a vaccine, and put it in my arm. No final say ensures freedom of speech and No personal authority imposes the discipline of fact, and between them they propel knowledge forward.
Liberal science, I went on to argue, is the most ambitious and successful of the three great rules-based social orders which define and distinguish liberal societies, the other two being markets and liberal democracy. All, however, share a family resemblance in their matchless ability to organize and direct social cooperation without centralized control; in their knack for self-correction; and, above all, in their capacity to manage diversity and discord peacefully and productively. Liberalism, as David French has noted, is the greatest mechanism ever invented for preventing civil war.
So far so good. But here is what I missed in 1993: humans’ cognitive limitations and vulnerabilities. And wow—are they big! Our beliefs, our guesses, and even our raw perceptions are riddled with biases and distorted by our thirst for status and self-esteem. We form communities of belief which amplify those many distortions. Instead of seeking critics to correct our beliefs, we seek allies to confirm them. We form sects which split and split again, lose touch with reality, go to war, burn dissidents, believe nonsense. That is the first 200,000 years of human history, more or less. In other words, humans’ epistemic state of nature is a lot like Twitter on a bad day. (By the way, having a high IQ gives no protection. If anything, it seems to help us rationalize our preferred falsehoods.)
The question, then, is not why humans ever engage in group-think, accept conspiracy theories, fail to challenge our biases, and believe comforting falsehoods. The question is why we ever do not do those things. The answer is that a liberal society’s collective search for truth depends not just on the judgment of individuals or tribes or sects but on institutions which took centuries of hard work to develop: universities, professional and scientific societies, journals, newspapers, scientific and professional associations, courts, think tanks, government agencies, professional schools, and countless more. It depends on norms and protocols which take years of training to inculcate: research methods, peer review, editing, fact-checking, scholarly writing, legal drafting, and, again, countless more. Defending those institutions and norms requires recognizing that they are not self-maintaining or self-defending. To the contrary, in the age of Trump and #StopTheSteal and cancel culture and anti-vax, they are under sustained attack from clever and opportunistic forces using powerful tools of social and cognitive manipulation like “firehose of falsehood” disinformation, conspiracy-mongering, consensus-falsifying, and social coercion (AKA “canceling”).
The free-market model is naïve about all of that. It imagines that individuals, left to our own devices in an unstructured environment, will act like scientists. Moreover, it encourages us to assume that epistemic liberalism is self-maintaining. Just have free speech, add information technology, and presto!—the market works its magic and out pops truth.
The trouble is that today’s enemies of epistemic liberalism are not so naïve. They understand that academia, journalism, law, and government—the Big Four of the professional truth-seeking business, or what I call the reality-based community—are institutional in nature, not individual. They realize they can weaken or even destroy the reality-based community by attacking its institutions’ credibility and corrupting their values. They know the community is vulnerable to a whole arsenal of time-tested, sophisticated cognitive weapons which divide, disorient, and disrupt: exactly what is happening right now.
A whole lot needs to be done by way of mounting a defense (see my book for much more on that), but the first step in defending liberal institutions is to see them; and this, as Yuval Levin argues in his seminal book A Time to Build, is harder than one might imagine. When social institutions like universities, news media, unions, religious organizations, and political parties work well, we tend to take them for granted. They become part of the everyday environment, shaping our associational lives while guiding us toward safe and constructive choices. Like the roads and traffic lights we rely on when we drive, they constrain us in ways which help us feel unconstrained and guide us in ways which help us feel autonomous. Only when they weaken and fail do we take much notice of them—though often to resent them rather than to rebuild them.
For someone my age, born in 1960, epistemic liberalism seemed to just work for a long time. Science delivered one breakthrough after another, American universities were the envy of the world, the evening news was authoritative. Progress seemed natural—the sure outcome of that wondrous marketplace of ideas. In 2021, when a majority of Republicans believe in the Big Lie about the 2020 election and an anti-vaccine infodemic is as viral and life-threatening as the Covid-19 epidemic, things look very different. Failure to grasp and respond to decades of attacks on the institutional and normative underpinnings of the Constitution of Knowledge is the leading reason.
The Constitution of Knowledge is my effort to bring institutions back into the picture and push them to the foreground. I show that science, journalism, law, and the other truth-seeking professions rely on rules and institutions whose function and form resemble the Madisonian Constitution in many respects. Both systems lay out regimes for governance—one in the domain of law, the other in the domain of knowledge. Both systems structure social negotiations, force compromise and persuasion, impose impersonal rules, use checks and balances, disallow final outcomes, incentivize error-correction, embrace pluralism, and deter the centralization and consolidation of power. The U.S. constitutional order has its formal institutions like Congress and the courts, its specialist corps of politicians and party professionals, its norms like lawfulness and being willing to lose an election. The Constitution of Knowledge has its formal institutions like universities and newsrooms, its specialist corps of researchers and reporters, its norms like factuality and being willing to lose an argument.
Moreover, the book seeks to show how deliberate efforts by identifiable actors created the institutions and professions and protocols which turned conjectural “natural philosophy” into scientific geology in the early 19th century, folk healing and charlatanism into professional medicine in the late 19th century, and hyperpartisan fake news into fact-based journalism in the early 20th century. Today, Facebook’s oversight board, attempting to establish norms and guidelines for social media, is the latest example of building institutions to counter chaos, charlatanism, and disinformation. How effective the board will be remains to be seen, but Facebook’s effort recognizes an essential truth: the reality-based community is a built environment, requiring conscious construction—and institutions are the building blocks.
Madisonian epistemology, as I call it, thus begins by recognizing that freedom of speech and the marketplace of ideas are necessary ingredients of a knowledgeable and peaceable society, but the secret sauce is the Constitution of Knowledge. We can’t repair humans, who will always be prone to make mistakes, indulge biases, feel too sure, jump to conclusions, suspect conspiracies, trust our friends, resent our critics, and let our tribes do our thinking. What we can do is strengthen the embattled institutions which structure our interactions, guide our behaviors, form our characters, and make us civil and civilized.