The Constitution of Knowledge: A Defense of Truth, by Jonathan Rauch (Brookings Institution Press, 280 pp., $27.99)

As the digital revolution dawned a few decades ago, commentators peddled optimism about its potential to expand humanity’s collective knowledge. Needless to say, that optimism has hit a few snags. The Internet has become a bazaar of information hawkers, where expertise is looked on with suspicion, traffic in hooey is high, and truth is harder and harder to discern.

As Jonathan Rauch puts it in his new book, The Constitution of Knowledge, consumers of online content feel “epistemic helplessness,” the “inability to know where to turn for truth.” In part, this is a product of content producers’ relentless effort to attract eyeballs, which induces readers to skip from story to story, coming away feeling none the wiser. In part, epistemic confusion is the goal of provocateurs and bad actors, for whom it’s more effective to inundate people’s attention than to censor. “If you flood the zone with distractions and deceptions and just plain garbage,” Rauch continues, “people’s attention would be diverted and exhausted and overwhelmed.”

Laments about the Internet are not new, but Rauch adroitly shows that online troubles are tainting our politics and imperiling our capacity to self-govern. “Trolls” on the right and “cancelers” on the left are kneecapping the pursuit of truth. Individual and collective knowledge—embodied in what he terms the “constitution of knowledge”—is deteriorating.

What is the constitution of knowledge? Rauch defines it as a piece of intellectual technology that generates knowledge over time, an “orderly, decentralized, and impersonal” system that depends heavily on social norms and rules to work properly. It rejects arbitrary authority and relies on institutions—journalistic, academic, scientific—to filter, organize, and elevate the most promising theories and arguments in all fields. Those involved in truth-seeking develop “accepted standards and methods and vocabularies, so that even if checkers do not agree on particular propositions, they agree on what it is they are discussing and how to discuss it productively,” Rauch writes.

Rauch deems the constitution of knowledge the crowning achievement of “liberal epistemology,” which emerged after generations in England, France, and central Europe grew weary from violent religious wars. Liberal epistemology, along with economic liberalism and political liberalism, are the three social pillars that emerged amid Europe’s upheaval, according to Rauch. All three liberalisms are social mechanisms that aim to secure cooperation, stability, and equilibrium amid deep cultural fissures.

Economic liberalism relies on the rule of law, relatively unregulated markets, and cultural respect for commercial values to produce material prosperity unimaginable to previous eras. It is “a species-transforming piece of social software, one which enables humans to function far above our designed capacity,” Rauch writes.

Similarly, political liberalism is “a social operating system which seeks to elicit cooperation and resolve differences on the basis of rules.” Its guiding principles are natural rights, rule by consent, and toleration. These political doctrines stem from theorists including John Locke and Adam Smith who seek firm theoretical grounding for governmental legitimacy.

Rauch sees the three liberalisms as interconnected features of the American regime. He contends that the U.S. Constitution and liberal epistemology share basic premises. When contemplating America’s constitutional framework, James Madison “saw that competition could be an engine converting the anarchic energy of diversity into the coordinated motion of cooperation.” The Constitution’s “core mechanism” is “forcing compromise.” Rauch continues, “To break deadlock, actors need to make new arguments, try new ideas, recruit new supporters, find new allies, and take new steps to mollify old allies.” Within and among the branches of government, ideas must have persuasive force to make it into law. Rauch observes, “The Constitution of Knowledge works the same way, except the product is not governance but reality.” To count as public knowledge, theories must be analytically rigorous, and ideas must stand up to scrutiny.

There’s much to commend in this account. Rauch persuasively outlines the many ways in which knowledge is the outcome of long-standing, resilient institutions. He offers sound strategies for combating falsehoods on college campuses and online—two places where liberal epistemology is most threatened. He praises the efforts of the Foundation for Individual Rights in Education to protect free exchange on college campuses, and he encourages similarly modeled counter-mobilization. He also applauds Facebook’s Supreme Court, Twitter’s pop-ups asking users if they want to retweet articles they haven’t read, and Wikipedia’s algorithm-less, communally driven model, with built-in “hierarchies and gatekeepers, rules and norms, due process and accountability.”

Yet the demanding epistemic methods of Rauch’s vision limit its breadth of application. In his overview of Lockean epistemology, a guiding force for the constitution of knowledge, Rauch explains: “If our claims or hypotheses cannot be reduced to particulars and then checked against the experience and reason of ourselves and others, they are outside the boundaries of what today we call ‘science.’” For something to count as “knowledge,” it must be empirically verified. Liberal epistemology as informed by Locke excludes “from intellectual respectability—from the epistemic rule book—claims which, because they are not checkable, are not adjudicable.”

In other words, claims must be “fallabilist,” which means capable in principle of being proven wrong. This was the scientific methodology championed by philosopher Karl Popper, who argued that, while no claim can receive definitive confirmation, empirical claims “can achieve disconfirmation.” Therefore, fallibilism must “search for error,” since “error is something we can find.” Rauch explains, “By testing ideas experimentally, logically, and many other ways, science shoots down thousands, even millions, of hypotheses every day.” In the constitution of knowledge, only claims that can be falsified and tested against others’ experience are worthy of respect.

Plainly, as Rauch points out, this way of knowing cannot be applied to all of life. Love, friendship, and faith are separate realms that operate according to their own logics.

But even within the realm of hard science, truth-seekers must proceed on the basis of sub-scientific, or a-factual, functions of reason. AEI fellow Tony Mills writes in his recent paper, “The Role of Judgment and Deliberation in Science-based Policy,” that “expert judgement,” an unconscious mental process born from years of practice, guides every step of scientific experimentation. Only those who have spent years in their field can anticipate, account for, and properly classify the various contingencies of their chosen subject. “Some amount of error in [the scientist’s] data is of course inevitable, since neither he nor his instrument is perfect,” Mills writes. “But how does he decide which errors are tolerable and which are problematic? How does he differentiate, for instance, between systematic error and those introduced by unknown sources (say, magnetic interference)?” The scientist’s judgment about the source of error is not “subjective or arbitrary.” But scientific judgment is sometimes so closely bound to its object of study that it eludes explicit rational formulation.

Judgment—a peculiar faculty that is neither mere intuition nor formal logic—is indispensable in the path to knowledge, though it is rarely falsifiable or fact-checkable. Scientific communities inevitably rely on claims born from judgment rather than strict empirical methods. Yet liberal science professes to accept only what can in principle be disproven.

One study Rauch cites illustrates a key oversight of the empiricist mindset. Study participants were shown data about severe property damage from deer overpopulation, compared with the milder property damage resulting from crime. Even in groups for whom the damage caused by deer far exceeded that of crime, participants chose to allocate more public resources toward addressing the crime problem. Rauch attributes these results to the “high emotional valence” associated with crime. The authors of the study wrote, “It was as though people simply couldn’t get past the emotional tug of the problem—crime—no matter what empirical data we presented about it.”

Another explanation is that people, furnished with judgment, tend to evaluate a multitude of factors when making decisions. Sure, crashing one’s car into a deer is jarring. But there’s something more disturbing and therefore more urgent about fellow human beings equipped with a conscience willfully harming us—even if the harms are less than others. This claim isn’t quite checkable in an empirical sense, but it is nonetheless a judgment compatible with reason.

This leads to the matter of the place of the constitution of knowledge in public policy. What role is there for knowledge?

Debates surrounding Covid-19 vividly illustrate this challenge. With rising case numbers and a substantial unvaccinated population, leaders face difficult choices. Should the government mandate vaccinations? Should the unvaccinated get paid to receive the vaccine? Should they be left alone, leaving everyone else to hope for the best? Data and studies certainly should inform responses to these questions, but no amount of data can give us final answers. Profound complexity often plagues social debates, and final decisions involve non-adjudicable moral deliberation.

In a 1977 essay, “Policy Science as Metaphysical Madness,” Edward C. Banfield wrote: “Social problems are at bottom political; they arise from differences of opinion and interest and, except in trivial instances, are difficulties to be coped with (ignored, got around, put up with, exorcised by the arts of rhetoric, etc.) rather than puzzles to be solved.” Formal modeling, regression analysis, and quantitative evaluation sometimes attempt to fix social problems that are impervious to scientific exactitude. The multidimensional nature of politics, and human beings more generally, often renders policy science too narrow to be useful. To apply empirical logic to morals and politics is to commit a category error.

Rauch is right to stress the importance of facts in public debate. He cites many disturbing examples of the truth being pilloried for political gain, and he presents a perceptive account of knowledge as a product of community, dialogue, and intellectual organization. But when, by sleight of hand, contestable moral judgments masquerade as scientific facts, skepticism and mistrust will follow. In public deliberations about what we know, we cannot dispense with distinctions between the types of claims to knowledge we’re making: scientific, moral, or prudential.

To address epistemic helplessness, we must rely on institutions to guide debate, but these institutions need a capacious definition of reason—one that understands not just the power of facts but also of judgment and moral deliberation. Embracing reason in its fullness results in a messier constitution of knowledge, to be sure, but also a more abundant one.

Photo: MicrovOne/iStock


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next