The Limits of Liberal Science

Jonathan Rauch’s important new book largely neglects the ways in which science can pose a threat to liberal politics.

The Constitution of Knowledge A Defense of Truth by Jonathan Rauch Brookings, 305 pp., $27.99

It is difficult to imagine a more timely or important subject than that of Jonathan Rauch’s new book. The “constitution of knowledge” in his book’s title refers to the complicated set of tools and rules, practices and norms that we use to determine what is true—from the replication of scientific experiments to the fact-checking procedures used by magazines. The aim of his book is to discuss how we know what we know, and to diagnose contemporary threats to the constitution of knowledge, so that it might be better protected.

Readers will come away from Rauch’s book better equipped to describe and defend the people and institutions that devote themselves to understanding reality. Rauch’s hopefulness with regards to the future integrity of knowledge and truth is refreshing and in many instances persuasive. He anticipates important objections and addresses them thoughtfully throughout. And he writes with attention and care, which is both essential and difficult given the scope of his subject.

In what follows, I want to offer a respectful critique of one specific aspect of Rauch’s argument. Throughout his book, he draws an analogy between political liberalism and epistemic liberalism—between the American constitutional order (broadly conceived) and the constitution of knowledge. He offers many insights into how they are mutually beneficial. And, far from assuming that the elective affinity between the two means they always work in concert, Rauch concedes that there are ways in which contemporary politics threatens liberal science. Indeed, half of his book is devoted to the dangers of politicized assaults on knowledge-sustaining institutions.

But Rauch largely neglects the converse problem: the ways in which institutions and persons affiliated with the production of knowledge can pose a threat to liberal politics.

Populist fears of technocratic overlords tend to be overblown, especially these days—so much so that I am wary of even wading into this discussion for fear of adding wind to their sails. But it is still the case that science, even liberal science, can thwart political life and political freedoms. Ignoring that problem creates a risk—the risk of elevating science to a primacy that it does not deserve, and ceding to it responsibilities that it cannot bear.

By not concerning himself with this problem, I worry that Rauch has written an unwitting apologia for scientism and technocracy, or at least for the domination of the public sphere by science, at a moment when we would be better off thinking about how to find common ground between genuinely clashing and contesting human outlooks.

Taking Human Fallibility Seriously

Fully fleshing out this argument would require a book in its own right, but I can at least attempt a sketch worthy of public discussion. The problem with Rauch’s book arises not so much with the core of his argument—his account of modern science and its character, and of what he calls the “reality-based community” and the truth-seeking techniques it relies on—but rather with how he contends with the outer boundaries of the constitution of knowledge. Rauch is great at describing the overlapping practices and principles of those who devote themselves sincerely to truth-finding. It’s when he makes strong claims about who does and doesn’t belong in the model that things get more questionable.

Rauch’s delineation of the bounds of liberal science is in some ways very broad and expansive; in other respects, it is pinched and monolithic. And there are some contradictions in his description that render it vulnerable.

Consider: Rauch says that liberal science is open-minded and does not privilege particular viewpoints. “In a world of conflicting certitudes, we must accept and even embrace pluralism,” he writes. “Anyone who calls for particular viewpoints to be privileged or for particular ideas to be censored is, by definition, not doing science.”

Yet at the same time, Rauch argues throughout the book that liberal science must be closed to a wide range of opinions and outlooks, and that it is the exclusive arbiter of public truth:

• “If we care about knowledge, freedom, and peace, then we need to stake a strong claim: anyone can believe anything, but liberal science—open-ended, depersonalized checking by an error-seeking social network—is the only legitimate validator of knowledge, at least in the reality-based community. Other communities, of course, can do all kinds of other things. But they cannot make social decisions about objective reality. That is a very bold, very broad, very tough claim, and it goes down very badly with lots of people and communities who feel ignored or oppressed by the Constitution of Knowledge: creationists, Christian Scientists, homeopaths, astrologists, flat-earthers, anti-vaxxers, birthers, 9/11 truthers, postmodern professors, political partisans, QAnon followers, and adherents of any number of other belief systems and religions.” (Page 87)

• “You have to check your own claims and subject them to contestation from others; you have to tolerate the competing claims of others; you have to accept that your own certainty counts for nothing; you have to forswear claiming that your god, your experience, your intuition, or your group is epistemically privileged; you have to defend the exclusive legitimacy of liberal science even (in fact, especially) when you think it is wrong or unfair.” (Page 91)

• “Traders in the marketplace of persuasion . . . accept that there can be no other pathway to knowledge except by exposing their own beliefs to propositions and viewpoints which are different from their own. Provided individuals follow the fallibilist and empirical rules and hold themselves and each other accountable, and provided their viewpoints are many and diverse, the network will correct errors, even if many individuals cling to their mistakes.” (Page 94)

• “Members of the community share not just a commitment to the idea of objective reality but also an understanding that outside of that commitment lies anarchy… The Constitution of Knowledge requires a commitment to one objective reality, accessible only through the Constitution’s rules.” (Pages 103-104)

These passages contradict Rauch’s claims about pluralism and the refusal to privilege particular ideas. In essence, he is saying that liberal science must not privilege particular viewpoints while also saying that liberal science is not merely privileged, it is the exclusive arbiter of public truth. He could ease the tension by acknowledging explicitly that, of course, liberal science privileges some viewpoints over others; it is constantly testing and ranking the validity of particular notions about the world. But Rauch’s insistence on exclusivity for liberal science is a move that strikes me as both epistemically misguided and politically corrosive.

There is no reason for the average person or for politicians to give equal consideration to the views of reputable immunologists and those of anti-vaxxers, but people who are working in journalism, government, and the law must in principle be willing to hear out just about anyone, and to take on their ideas and arguments patiently and respectfully. The same goes for people working in the sciences: they ought to be willing to listen and engage with a wide range of people. And I don’t mean this merely as a matter of civility or politeness: I mean it as a real epistemic concern. As Rauch is aware, knowledge is always provisional, and it is always human—incomplete, sometimes flawed, often not fully understood. And since the scientific enterprise is a human undertaking, any particular individual in the sciences is subject to error and insularity and groupthink and corruption. It is also simply the case that we can never know ahead of time where genuine insight is going to arise. So liberal scientists and institutions have no business being dismissive of outside views and arguments a priori. That is just an excuse to shut down engagement, and most institutional incentives already point away from popular/democratic input and towards exclusion.

Rauch comes close to conceding as much when he discusses the need for a wide “social funnel” that is open to all kinds of ideas, which liberal science then sorts through and turns into knowledge. But this way of describing the process, even just in the abstract, and even with respect to the hard sciences, is still too closed. Scientists and others who work in the constitution of knowledge are trained experts within their particular fields, but they are also just ordinary people. Their expertise often means they will be especially subject to epistemic prejudice and close-mindedness, which means that they more than most need to hold tight to a principle of radical open-mindness. (Rauch admits as much at one point in the book: fallibilism has to be an “organizing principle for our lives and careers,” he writes, admitting that this is very hard.) There is plenty to learn from people in other traditions, disciplines, and from the public; sometimes unfiltered insights from the world are much better than what makes it through the institutional funnels. As Aurelian Craiutu puts it in his review of Rauch’s book, citing J. S. Mill, “truth is eclectic.”

Taking Institutional Fallibility (and Critical Thinking) Seriously

The problem of closedness is magnified tenfold with respect to more speculative areas of inquiry. As we move further from straightforward empirical matters and closer to speculative or normative or artistic thinking—everything from abstract physics, to the creative arts, to politics, ethics, and theology—then, in Rauch’s view, we start to break away from the rules of liberal science, and so too from the realm of the kinds of legitimate, “constitutional” truths that should inform public life.

Rauch is right to observe that early moderns like Bacon, Hobbes, and Locke turned away from speculative fields like politics, metaphysics, and theology—“claims which, because they are not checkable, are not adjudicable”—but they did not permanently close off those areas from intellectual inquiry. Even if they had wanted to (I don’t believe they did), they did not and they could not.

Just because a perspective or an area of inquiry is especially contentious does not mean it has no place in public life.

And so I see no upside to dismissive lists like the one Rauch supplies (see the first bullet above) to delineate whom he believes de facto to be excluded from the circles of liberal knowledge—creationists, homeopaths, 9/11 truthers, QAnoners, and so on. The list might well be unobjectionable to some, but it reveals some serious limits to Rauch’s approach. The fact, for one, that he would be so disdainful of “postmodern professors and political partisans,” and vaguely contemptuous of “adherents of any number of other belief systems and religions” is remarkable.

Regardless of where you come down on questions of ultimate truth and meaning, or on the whole world of critical theory, and even granting that plenty of academic work translates very badly to the so-called “real world,” the idea that postmodern professors should be epistemically shunned—or that they do not contribute to the constitution of knowledge in important ways—is simply untrue. The same goes for all kinds of work in theology, and politics, and literature, and the arts, and philosophy, and the humanities more generally. I would venture that Rauch’s list in fact encompasses the vast majority of Americans: most of us have at least a streak of countercultural skepticism, or partisanship, or mysticism, or conspiracism lurking in our souls.

At some points, Rauch claims that his version of liberal science is open to at least some non-empirical disciplines. He describes how liberal science is meant to be expansive, and so to include “the softer sciences and even humanities such as literary criticism and moral philosophy.” And he contends that “science, as conventionally defined, makes up only a fraction of the reality-based community.” But Rauch’s descriptions of how liberal science operates fails to capture a lot of what actually goes in the more humanistic disciplines. This is partly because he is devoted to the principle of objectivity, or “no personal authority,” and so is resistant to anything that smacks of subjectivity or perspectivalism. But some forms of genius or intelligence amount to little more than idiosyncratic expressions of personal authority. And clarity about subjectivism is a major part of much social science and humanities research. Taking subjectivism seriously isn’t a rejection of truth or knowledge, it’s a rejection of the simplistic idea that truth, to be true, must always in principle be broadly shared or communicable.

Unfortunately, Rauch is overly dismissive of precisely those species of inquiry that are often the most thoughtful about political power, in addition to being the most rigorous and self-aware regarding the problem of intellectual overreach.

Rauch’s rhetorical choices similarly reflect a mode of thinking that excludes much philosophy, theology, and humanistic thinking. My own humanist’s ear is fine with the idea of the “constitution of knowledge,” but pretty displeased by the presumptuous notion of a “reality-based community” in the singular, even one as broadly conceived as Rauch’s. I believe strongly in a collective liberal process of knowledge-accretion, but I do not care for Rauch’s notion of “outsourcing reality to a social network,” nor for the use of the metaphor of the “operating system” to describe a human community.

So what to do? Let’s revisit Rauch’s bold claim:

You have to check your own claims and subject them to contestation from others; you have to tolerate the competing claims of others; you have to accept that your own certainty counts for nothing; you have to forswear claiming that your god, your experience, your intuition, or your group is epistemically privileged; you have to defend the exclusive legitimacy of liberal science even (in fact, especially) when you think it is wrong or unfair. (Page 91)

Here is how I would revise it to make it more liberal, and more consistent with genuine pluralism:

You have to check your own claims and subject them to contestation from others; you have to tolerate the competing claims of others; you have to accept that even your own feelings of certainty are fallible; you have to honestly admit that you do believe your mode of engagement to be epistemically privileged (and be able to give reasons why), while at the same time sustaining a radical openness to counterarguments from any and all quarters, even (in fact, especially) when you think it is wrong or unfair.

One of the reasons that it is so important to stay rigidly open-minded about potential sources of knowledge and truth, even those that are highly contestable or seem dead wrong, is that the liberal sciences (which, again, for Rauch also include jurisprudence, government research, and journalism), for all of their considerable virtues, are hardly foolproof. The constitution of knowledge, being a human endeavor, is not merely fallible, but also is ridden with internal tensions, is often ineffective, is prone to insularity, and is subject to abuses, in ways that Rauch is too blithe about.

Science is our best process for understanding empirical facts, but ideas wrapped in the mantle of science have often had pernicious effects—most notoriously when scientific arguments were offered on behalf of slavery, racist policies, eugenics, and other affronts to human dignity. It took moral reasoning and political action to challenge and overcome those arguments, usually long before they were decisively discarded as pseudoscientific. And students of history will have little difficulty thinking of many other cases when attempts to apply what was understood to be scientific rationalism to human affairs resulted in what would eventually come to be recognized as bad policy, as when highways were designed to maximize speed and convenience by gutting American cities; as when the great boon of powerful painkillers combined with perverse financial incentives to create the opioid crisis; as when the effort to create more precise, targeted weapons so that we could spare lives contributed (as Samuel Moyn argues in his new book) to new, dehumanizing modes of warfare. And the scientistic, technocratic impulse in our politics—the inclination to defer to those who claim their innovations represent not just technical progress but human, social progress—deforms how we govern emerging technologies, from social media to pharmaceuticals to biotechnology to tools of surveillance to, any day now, artificial intelligence and virtual reality.

Rauch’s model supposes that the various branches and institutions of the reality-based community will provide a check on one another. I don’t think liberal science can afford to close off any avenues of counter-critique or caution a priori.

Politics Matters

Rauch makes another effort to safeguard genuine epistemic pluralism at the end of Chapter 4, in a short section called “Reality Is a Part-Time Job.” Here he once again argues that liberal science, as he defines it, should have exclusive epistemic authority, but he adds a caveat: only when it comes to public knowing (“supremacy in the realm of public knowledge, but not in the realm of private belief”; liberal science cannot “run your life, rule the world, or control your brain”). This effort to save other forms of belief falls a bit flat insofar as they have already been deemed not “reality-based.” But beyond that, while I agree with Rauch that public decision-making bodies should privilege knowledge and insights that are generated by accredited scientists, lawyers, and researchers, when it comes to creating laws and other overtly political work, we enter a different realm. In my view, many forms of knowledge can and should come to bat in political life. My sense is that Rauch would give liberal science the starring role.

In other words, for Rauch, those involved in politics should rely, perhaps exclusively, on the knowledge that comes from liberal science. The idea that science should dominate in public life—even just as a theoretical proposition, and even just as an epistemic matter—is necessarily to diminish the importance of politics and contestation, which in some ways must always be primary (since liberal science depends on political institutions more than political institutions depend on liberal science).

I want liberal science to have a lot of authority in public life, too, but I want that to be the result of high-quality public education, strong advocacy and persuasion on the part of literate citizenry, politicians who understand and care about reality, and scientists who are willing to speak for science (and are thoughtful about how that can be done and about its limitations)—not because anyone said it should in principle be so. The older image of the clash of ultimate values—where science is one voice in a broader cacophony, as Max Weber laid out in his famous lecture Science as a Vocation, among other places—is truer to America’s liberal constitutional order than the idea of a streamlined liberal science that reigns supreme.

Absent a clearer articulation of the ways in which knowledge can threaten politics, I foresee two distinctive long-term dangers for American democracy:

On the one hand, I worry about the emergence of a technocratic liberalism that is overconfident to the point of inhumanity—one that is oppressive to dissidents of all stripes, that regards artists, critics, and the faithful, as well as whistleblowers, radicals, and other renegades, as existential threats to the reigning order, rather than as unruly citizens of a messy pluralistic reality.

On the other hand, I worry about a future reactionary politics that uses all the trappings of objectivity and liberal science to usurp political authority from the people. I worry about right-wing authoritarians who will use the climate crisis to justify illiberalism, and about conservative intellectuals who would abandon democracy in a second to secure their own vision of the “objective” good (backed by plenty of statistics).

Again, Rauch has admirably labored to build public confidence in a set of important institutions and practices, and I hope that the book is successful in helping Americans advocate for truth and knowledge. But his approach can easily tip into the kind of ‘rah rah science’ that is already a distinct flavor of liberal hubris and that does not easily win over new friends.

If I had my druthers, liberals would talk less about how #ScienceIsReal and #TruthMatters and more about science’s complexities and limits and all that we don’t know. As I read Rauch’s book, the famous line from Hamlet kept popping up in my mind: “There are more things in heaven and Earth, Horatio, / Than are dreamt of in your philosophy.”

It’s a good line: An appropriate mantra for an epistemic liberal at any time, but maybe especially in times of heady polarization, strife, and uncertainty. In the long term, humility makes for a better epistemic balm than Rauch’s bolder prescription.