There are limits to what the First Amendment refers to as “the freedom of speech.” That freedom isn’t absolute, but it is quite broad, with certain exceptions that the courts have labored to define rigorously. For the most part, U.S. culture has taken the First Amendment and its known exceptions to establish norms of free speech and political discourse.

But the norms of free speech in American democracy were strenuously tested in 2017, with high-profile clashes driven by extremist groups. Added to the real-world confrontations is the growing perception that the digital public square (primarily through social-media platforms) has aided the spread of misinformation, hate speech and hyper-partisan rhetoric. The looming one-year anniversary of the Charlottsville “Unite The Right” rally—and the associated violent confrontations among citizens—gives us reason to reflect on how we really should think about free speech. A central dilemma: should we “tolerate intolerance” (and run the risk what philosopher Karl Popper called “the paradox of tolerance”)? Or should we be “intolerant of intolerance” (which may mean limiting fundamental rights like freedom of expression because we think “intolerance” is an existential threat to democracy)?

I think both philosophy (including Karl Popper’s great work, THE OPEN SOCIETY AND ITS ENEMIES) and American political theory provide good answers to these questions—provided that we take the time to excavate these sources of our contemporary pluralistic conventions and see where the balances should really be struck.

To be clear, these are, in fact, difficult questions. As a freedom-of-expression advocate, I viewed the run-up to the “Unite the Right” event in Charlottesville with the queasy feeling you have when the dentist approaches you with the needle and the drill. Still, I think, although freedom of speech and tolerance can, at times, be painful, the benefits of protecting that freedom to the high degree guaranteed by the First Amendment are worth the price we pay. I write this essay to explain my belief, and the reasoning that has led to it.

The First Amendment framework

Let’s start with the constitutional tradition. In American constitutional jurisprudence, the right to free speech is expansive but not absolute. We’ve taken pains to draw a line between speech that “merely advocates” violence in a nonspecific way (we say such speech is protected by the First Amendment) and speech that “incites” what we call “imminent lawless action.” We say that speech that “incites” such “imminent lawless action” (that is, it leads directly and immediately to violence) isn’t protected and therefore can be deterred, restricted, and punished. The distinction between “mere advocacy” and “incitement” is the lesson of the leading case on the issue, Brandenburg v. Ohio (1969). Although some amateur constitutionalists justify stronger censorship measures by citing the “falsely shouting fire in a crowded theater” language of the 1919 Supreme Court case Schenck v. United States, that’s not, in fact, the First Amendment standard, and it hasn’t been for half a century.

Even though the Brandenburg case sets the standard, how to draw the line between “mere advocacy” and “incitement” isn’t always clear. There’s no disputing that many—and maybe most—of the white nationalists who showed up in Charlottesville, armed to the teeth, not only hoped to intimidate counter protesters with their hateful speech, but also nourished the fantasy that they’d violently engage their opponents. They got their wish. And some militant anti-fascist activists got their wish too, recalling Woody Allen’s sentiment in “Manhattan” that “physical force is always better with Nazis,” even if just to block them from instigating more violence.

Eyewitnesses in Charlottesville, including counter protesters who stuck to their resolve to refrain from violence, have expressed gratitude that the anti-fascists were there. Did the antifascists’ presence really establish the principle that speech advocating violent intolerance must necessarily be blocked through our own willingness to resort to physical force, rather than restrict ourselves to mere “counterspeech”? (Or, as Woody Allen would put it, to “get some bricks and baseball bats and really explain things to ’em”?) Some critics have argued that Karl Popper’s analysis of “the paradox of tolerance” makes the case for resorting to force as a general remedy for the kind of hateful speech the white nationalists exemplified. Does this argument make sense?

Popper and the Paradox of Tolerance

A viral infographic invoking Karl Popper’s “paradox of intolerance”—and citing THE OPEN SOCIETY AND ITS ENEMIES—circulated widely on Twitter and other social-media sites around the time of the Charlottesville confrontations, and for quite awhile afterwards. At more than 700 pages, Popper’s 1945 book is richly provocative, and generally accessible, if at times densely reasoned. But is the infographic an accurate representation of Popper’s thought?  As it turns out, “the paradox of tolerance” isn’t itself a central thesis of THE OPEN SOCIETY. Instead, it occupies part of an essay-long endnote to Chapter 7. Researching Popper’s reasoning on tolerance requires flipping back and forward in the book to understand each section, including any endnotes, in context.

The status of the “paradox of tolerance” as one part of a footnote doesn’t mean it’s an afterthought—it definitely is a fully developed idea. But it does mean that the author imagined a reader who would progress through THE OPEN SOCIETY from beginning to end. Although Popper, who died in 1994, might have been flattered if he had lived to see THE OPEN SOCIETY cited widely on today’s internet, Popper was also famously persnickety about critics—or allies, or students—who oversimplified or misrepresented his views. 

This naturally raises the question of what Popper himself would have thought about “the paradox of tolerance” as the basis of this one-page infographic:

To figure out what Popper might have thought of this representation of his views, it helps to extract the words from the graphic. Start with this first part: “Should a tolerant society tolerate intolerance? … The answer is NO. It’s a paradox, but unlimited tolerance can lead to the extinction of tolerance.” Standing by itself, the sentence is certainly true, but as sometimes happens with cartoons and infographics, certain nuances are elided here. Should there be tolerance of intolerant ideas? Or is the real issue whether we should tolerate what might be called “acts of intolerance”?

If we interpret Popper’s “paradox of tolerance” the way the infographic represents it, the answer to the first question is, in fact, “NO.” From this perspective, it might seem necessary to suppress ignoble and intolerant views like those of the Charlottesville white nationalist protesters, and embrace both violent counter-protests and digital censorship. But should digital censorship of hate speech be the default expectation and the default response?

This policy question can be distinguished from the question of whether Facebook, Twitter, et al, can or should sometimes curate users’ speech and expression in furtherance of the platforms’ policies. Certainly digital platforms like Facebook are within their rights to censor white-nationalist content or other hateful speech—we crafted Section 230 of the Communications Decency Act precisely to enable such post-hoc self-policing by platforms without incurring legal liability for doing so. As a free-speech advocate, I wasn’t particularly troubled when Facebook started blocking posts with links to the The Daily Stormer — the neo-Nazi website that helped organize the Charlottesville rally — unless you were condemning them in the same post. That said, I’ve written elsewhere that almost any platform-based censorship—imposed by a private entity upon public discourse—is certain to leave some large fraction of a platform’s users unhappy, insisting that the platform has gone too far or not far enough.

But I was slightly more troubled by actions taken by other internet companies that operate on a “different level of the stack.”  It turns out that in the digital age, you don’t need bricks and baseball bats to silence people. GoDaddy and Google cancelled The Daily Stormer’s domain registrations, and Cloudflare terminated their contract. While all of these companies are well within their legal rights to terminate their services, their actions raise some disturbing questions for our norms associated with free speech in a democratic society. It’s not a perfect analogy (no analogy is, really), but it’s as if we punished hate speech by denying hate speakers the right to subscribe to a telephone service.

Leave aside the question of whether white nationalists deserve to suffer some kind of karmic justice for fomenting hate (it’s hard to argue that they don’t). The fact is, in an open society abiding by rule-of-law norms, we normally require due process before we take away the fundamental rights of even the most noxious dissident speakers. And speaking—even to large audiences, as the Klansmen in the Brandenburg case did—is a fundamental right under our system. It’s true that free speech rights on today’s internet are more typically mediated by private companies like Facebook and Twitter. (And, as a practical matter, they also can be facilitated or delimited by companies providing domain-name services or content-delivery networks.) Still, we ought to expect, and maybe require, that new limits on constitutional rights in the digital world will be the product of a deliberative democratic process, rather than the fact that some Nazis succeeded in getting under Matthew Prince’s skin. And yours, and mine.

In effect, then, if we take the “paradox of tolerance” infographic as an accurate representation of Popper’s thinking, we have conflicting paradigms. On the one hand, we have the U.S. First Amendment framework (and the norms it has given rise to), and, on the other, we have what purports to be a limits-on-intolerant-speech principle articulated by one of the 20^th century’s most prominent theorists of democracy.

What Popper’s OPEN SOCIETY Really Preaches

Is the cartoonized representation of Popper’s thoughts about intolerance accurate? Despite the appealing simplicity of the “paradox of intolerance” infographic, Popper framed that “paradox” with more nuance and more subtlety in THE OPEN SOCIETY. Because he is a thorough, discursive thinker and writer, Popper includes the following argument in the course of a longer essay on paradoxes (in that long footnote!) to explain tolerance in an open society does need some limits. To avoid the risk of oversimplifying, I include the full paragraph below. (Please bear with me and read it through.)

“Less well known is the paradox of tolerance: unlimited tolerance must lead to the disappearance of tolerance. If we extend unlimited tolerance even to those who are intolerant, if we are not prepared to defend a tolerant society against the onslaught of the intolerant, then the tolerant will be destroyed, and tolerance with them.—In this formulation, I do not imply, for instance, that we should always suppress the utterance of intolerant philosophies; as long as we can counter them by rational argument and keep them in check by public opinion, suppression would certainly be most unwise. But we should claim the right to suppress them if necessary even by force; for it may easily turn out that they are not prepared to meet us on the level of rational argument, but begin by denouncing all argument; they may forbid their followers to listen to rational argument, because it is deceptive, and teach them to answer arguments by the use of their fists or pistols. We should therefore claim, in the name of tolerance, the right not to tolerate the intolerant. We should claim that any movement preaching intolerance places itself outside the law, and we should consider incitement to intolerance and persecution as criminal, in the same way as we should consider incitement to murder, or to kidnapping, or to the revival of the slave trade, as criminal.”

That’s a big, complex paragraph, and Popper’s reasoning in it doesn’t lend itself to being abbreviated, ellipsized, or summarized as a social-media meme or as a cartoon (a mischievous adaptation of which has also been used to promote intolerance of Muslims). Which is to say, his line of reasoning isn’t easily tweetable (even at a full 280 characters). Popper’s “paradox,” as simplified in social media and on blogs, doesn’t establish a blanket justification for violent responses to repugnant intolerance, or the mass censorship of hateful ideas.

Proving this requires a close textual reading, which means we have to note Popper’s precise language. At the outset, the philosopher makes clear that he is focusing not on tolerance—which is generally regarded as a virtue in open societies—but specifically on “unlimited tolerance.” [The emphasis is mine.] As he puts it, “I do not imply […] that we should always suppress the utterance of intolerant philosophies.” Only when “incitement to intolerance and persecution” happens should we treat the utterance as criminal, he explains. And it’s Popper’s use of “incitement” here that links his thought helpfully to First Amendment theory.

To be clear, Popper is not using “incitement” in the specific narrow way that First Amendment free-speech cases use it—as shorthand for the Brandenburg case’s “incitement to imminent lawless action that is likely to produce such action.” It seems possible, even likely, that he is using “incitement” to encompass at least some of what the Supreme Court classified in Brandenburg as “mere advocacy” when he writes that any open society should have the right to suppress advocacy of intolerance.

But Popper’s expressed preference for “rational argument” suggests that he doesn’t want that right to be used normally as the first resort in response to such advocacy. And his enumeration of other kinds of “incitement”—to specific criminal acts like murder, kidnapping, and human trafficking—suggests that he has other kinds of speech in mind (like criminal conspiracy) that under our system of laws can be punished consistent with the general protections of the First Amendment. It’s well-established that one could argue for the revival of the slave trade, as an abstract notion, and be protected under the First Amendment, but one can’t conspire to enslave actual people without breaking the law.  And it’s well-established that one can’t defend such a criminal conspiracy as protected speech by invoking the First Amendment.

In Popper’s view, tolerance even of intolerant philosophies should be the rule as long as we can keep them in check through other means. As the Cato Institute’s Jason Kuznicki has explained, the preferred first response is through “rational argument” or democratic consensus (the latter is what Popper means by “public opinion”). 

Of course, shaping public opinion may present new challenges in the social media age. “Filter bubbles,” increasing polarization, “fake news,” and viral conspiracy theories have all been cited as reasons to conscript platforms to take a heavier hand in censoring certain kinds of speech, or groups of speakers. I’ve written about those demands a lot lately, notably here and here. I’m on the record as being more skeptical about “filter bubbles” than most. On the one hand, I think the problem of cherrypicking facts to support your views predates the internet. On the other, I think the predisposition of human beings to get into arguments on the internet undercuts the notion, that we just want to hear what confirms our opinions. (A version of this notion, advanced by law professor Cass Sunstein among others, argues that we’re predisposed to hide in “information cocoons.”) To be clear, I agree that polarization and fake news—on social-media plaforms as well as in traditional media—are problems. But these problems don’t entail a solution that imposes censorship obligations on social-media platforms (or elsewhere in the stack) and then simply trusts those private companies to censor content for us. If you believe freedom of speech and the rule of law are central values of an open society, you shouldn’t choose a cure that kills the disease by crippling the patient.

Open debate remains the best antidote to noxious doctrines promoting intolerance. But this conviction doesn’t depend on any notion that democracy is necessarily robust—it isn’t. Both the rise of populist antidemocratic movements around the world and various assaults on public institutions in the United States suggest that even the most hardy democratic institutions are often more fragile than we realize. Even in this country, where the First Amendment has sometimes seemed to be privileged just because of its “firstness,” free speech and freedom of the press may have fallen, for the moment, out of favor.  

And Popper himself was painfully aware of the fragility of democratic, open societies. In the late 1930s, Popper responded to the growing threat of Nazism in his native Austria by emigrating to New Zealand. He knew that democracies can be overcome by the rise of intolerant philosophies. Plato, the “enemy” of open societies with whom Popper “argues” through the first half of THE OPEN SOCIETY, knew this too—on that particular issue, the two opposing philosophers necessarily had to agree.

Nevertheless, Popper argues throughout THE OPEN SOCIETY that because we can’t know for certain that our given policy solutions to social problems are the best, censorship is generally at odds with a free society. He insists that the central pillar of an open society is not authority as such, but on what he calls “the spirit of criticism.” Today we’d call it the spirit of debate, and while we should maintain the right of social-media platforms to curate speech consistent with (ideally) a pluralistic, open-society policy framework, we should also support the platforms’ right to choose to tolerate even a range of intolerant speech. But only so long as it’s not “unlimited”—Popper quite rightly criticizes “unlimited tolerance” of anti-democratic, authoritarian speech. At the same time he also implicitly supports what might be called “limited tolerance” – a tolerance that relies first and foremost on the expectation that the other citizens of an open society will show up and challenge intolerance.

This brings us back to our grim first anniversary of the “Unite the Right” violence in Charlottesville. The alt-right protestors arrived in Charlottesville equipped for a street fight. They carried shields and shared tips for disguising cudgels as signs. They brought pepper spray and practiced moving as a phalanx.It seems clear that they were planning for something more than the “symbolic speech” that might be expressed by simply carrying weapons they believed they had a right to carry. They wanted to be threatening, and to seem threatening to their nonviolent opponents. (For that particular reason I’m sympathetic to the ACLU’s decision not to support arms-bearing protests going forward).

The most effective responses to violent ideologies, when it comes to marginalizing or defeating doctrines of hate, still has to be in “the public square”—sometimes the literal square and sometimes the public square that is public debate on social media. The literal public square is still sometimes the right place to congregate in opposition to intolerance. As David Cole, the ACLU’s national legal director, pointed out in The New York Review of Books, the white supremacists who called a rally in Boston the week after Charlottesville were “vastly outnumbered by tens of thousands of counter protesters who peacefully marched through the streets” to express their opposition to the white supremacists’ views. Per Cole, “Free speech, in short, is exposing white supremacists’ ideas to the condemnation they deserve.”

Sometimes the public square will be Facebook or Twitter or some other current platform, or some new platform that hasn’t even been rolled out yet.

Cole’s right about how free speech works, which means our first takeaway, from a close reading of both our First Amendment framework and the philosopher who personified commitment to open societies, is to begin by taking the risk of being tolerant. This doesn’t have to be our only response or our ultimate response. It just has to be our first response to intolerant speech. If we begin by tolerating the expression of intolerant views, we have the advantage. Intolerance, articulated openly and precisely, puts itself right in our critical crosshairs.

We know that when intolerance is out there, making itself known and earning (appropriate) condemnation, its weaknesses and shallowness and illogic becomes self-evident rather than covert. And when that intolerance is openly expressed, we can’t and mustn’t passively accept the First Amendment’s protection of intolerant thought. Instead, we know our first duty as members of an open society is to exercise our own First Amendment rights to challenge intolerance and refute it.