Abstract
Central to the American tradition of expanding protections for controversial speech is a robust distrust of potential censors to make reasonable judgments about what speech should be suppressed. But the arguments for a more restrictive approach to speech often implicitly or explicitly evince much greater trust in the likely decisionmakers who will be entrusted with the authority to suppress speech. Whether restricting Communist speech, antiwar speech, “hate speech,” or “disinformation,” the case for empowering some authority figure-such as campus administrators, technology company employees, or government officials-builds on an assumption that those authority figures will be motivated by good intentions and be endowed with good judgment to make reasonable distinctions between the speech that should be tolerated and the speech that should not. Such confidence would often seem to be misplaced.
In designing and adopting any regulatory scheme, there are two separate but important decisions to make. First, of course, we must decide on the substantive rules or standards that will govern the behavior to be regulated. This is often the most visible and contentious decision to make. Setting out the rule to be enforced is generally viewed as tantamount to setting the policy itself. But there is a second decision that must also be made, perhaps even more consequential than the first. Once we know what rule will be enforced, we must decide who will be empowered to interpret and enforce that rule. After we design the regulation, we must design the regulator. Rules are not usually self-enforcing. Someone will have to determine whether the rule has been violated and what to do in the case of violations. Those two decisions are critical to the success and significance of any regulatory scheme.
In this regard, the regulation of speech is no different than any other regulatory scheme. Changing the context of speech regulation does not change the dilemma. When we lay down a rule about what kinds of speech should be forbidden, we must also decide who will interpret and enforce that rule. Who will decide whether the rule is violated by a particular utterance and therefore whether the speech in question should be suppressed, or the speaker punished? Moreover, such issues arise whenever we seek to regulate speech. If the government wants to prohibit some speech, it will need a process of enforcing that law or administrative regulation. If the government wants to criminalize “terroristic threats,” it will need both to specify the rule against such threats and to rely on a criminal justice process for investigating and prosecuting those who make such threats. If Congress wants to exclude from federal trademark protection marks that are “scandalous” or disparaging, it will need to articulate the exception to trademark law and empower a government official to review trademark proposals and reject those that violate the rule. If the comment section of an online journal excludes some kinds of posts, the publisher will need to specify a rule explaining what content is prohibited and designate a moderator to review and delete posts that potentially offend the rule.
A great deal of theoretical argument on speech restrictions is understandably focused on the substance of potential limitations on speech. The substantive rule is where principled distinctions are drawn and where justifications for or against tolerating some types of speech can be developed. If we want to restrict speech, we need to take great care to ensure that we are restricting the right speech and for the right reasons. Constitutional doctrine and normative theory are focused on such questions as the circumstances in which false speech should be forbidden, how to distinguish obscenity from pornography, and how to distinguish fair use from copyright infringement. Most of our arguments about whether a specific kind of speech should be restricted turn on the question of whether restricting that speech would be a good idea. Does the speech in question have a high or low social value? Does the speech in question cause harms, and if so, how substantial and of what nature? Will censorship make us worse off? Should we rely on the marketplace of ideas to winnow the true from the false, or do we need the thoughtful assistance of the censor?
Those substantive debates on speech restrictions often take the implementation and enforcement of any restrictions for granted. This is understandable but a mistake. The implementation process might pass without remark simply because, at least in broad brush strokes, we think that those decisions are already fixed. If we are debating possible exceptions to the First Amendment to the U.S. Constitution, we are effectively debating how the Supreme Court ought to interpret the First Amendment, and what kinds of legal limits on speech the justices should accept. It is tempting to think that if we can just agree on the acceptable limits on speech, then the implementation of those limits would take care of itself. The details of the enforcement process might seem irrelevant to whether we think a particular type of speech should be outlawed.
I am persuaded, to some degree, by all three of the common liberal defenses of robust speech protections. Free speech is essential to the identification of the truth and the advancement of knowledge, which is particularly relevant to thinking about the scope of speech protections in an academic context.1 The tolerance of dissent is critical to allowing democratic processes to function, which is especially important in the context of political speech. And free expression is important to respecting human dignity and autonomy, which has particular salience in the context of artistic expression.
Those arguments are important, but they are ultimately not decisive for me. At the very core of my own skepticism about speech restrictions is distrust of those who would wield the power to suppress speech. Even if I were completely convinced that some particular type of speech is of low value and generally harmful, I would be extremely reluctant to agree to a rule prohibiting that speech because I have little faith that speech restrictions would be applied in a manner that did not have serious social costs. Censors would likely be overly aggressive in enforcing speech restrictions and biased in what they judge to be intolerable speech. It is precisely in the context of controversial speech that we will find it difficult to reach uncontroversial conclusions about whether a particular example of speech is beyond the pale. As James Madison pointed out, “if angels were to govern men, neither external nor internal controls on government would be necessary”; but the great problem with “framing a government which is to be administered by men over men” is that “you must first enable the government to control the governed; and in the next place oblige it to control itself.”2 Obliging the government to control itself has been particularly challenging in the context of freedom of speech. Even if we could design the ideal speech code, we should not have much faith that it would be implemented in an ideal way.
For me, those concerns about who will watch the watchmen create a very strong presumption against any significant restriction on speech. The long struggle to expand freedom of speech has been to an important degree the result of a dawning realization that censors cannot be trusted and thus the scope of their authority had to be significantly narrowed. I have often found that those who favor more restrictions on the freedom of speech also tend to have more confidence about how those rules will be implemented. If we do not need to worry about the second problem, the problem of implementation, then it becomes easier to imagine that desirable rules might be developed. Those who have faith in administrators tend also to be more willing to endorse speech codes than I am. Even when I can agree that a given example of speech is a net loss for society, I am much more reluctant to take the further step of empowering someone to limit such speech. If I am asked whether we must tolerate the speech of Nazis, I am not overly concerned about the possibility that Nazis might have interesting or illuminating things to say, but I am quite concerned that building the machinery of censorship to suppress the speech of Nazis will prove threatening to speech that is valuable. I would share the view that it would be unfortunate if my fellow citizens found Nazis to be persuasive, but I have trouble imagining who I might trust to make determinations as to which ideas my fellow citizens should be allowed to hear and assess.
The interlinked problem of substance and procedure in designing regulatory policy is pervasive. Some general examples might clarify and emphasize the importance of thinking about process as well as substance when assessing the wisdom of a potentially regulatory initiative.
The procedural question is particularly explicit in the context of administrative law. Administrative law is concerned with the rules governing how judges ought to review the actions of administrative agencies to secure agency compliance with statutory authorizations and mandates. A legislature might set down a variety of substantive statutory regulations, whether narrowly or broadly drawn, and create an administrative agency to enforce those regulations and develop subsidiary regulations to implement the statutory directives. Creating administrative agencies with substantial policymaking authority runs the risk, however, that the agency might not do what the legislature had expected or desired. The legislature faces a principal-agent problem when delegating such discretion to administrative agents.
Appreciating that political actors struggle over bureaucratic structure as well as over regulatory policy clarifies why substance and process might seem like they are in tension with one another and why administrative decision-making is so cumbersome. As bureaucracy scholar Terry Moe has pointed out:
Political compromise ushers the fox into the chicken coop. Opposing groups are dedicated to crippling the bureaucracy and gaining control over its decisions, and they will pressure for fragmented authority, labyrinthine procedures, mechanisms of political intervention, and other structures that subvert the bureaucracy's performance and open it up to attack. In the politics of structural choice, the inevitability of compromise means that agencies will be burdened with structures fully intended to cause their failure.3
The political game is not over when the substantive policy has been determined. Sophisticated political actors will also try to influence the means and procedures by which implementation decisions will be made. The second part of the political fight might be as consequential as the first. Alternatively, if you know that your political opponents will be the ones controlling how a policy is interpreted and implemented, then a sophisticated political actor will have to approach the decision about the substance of policy differently. Narrowing the scope of an agency's discretion will be quite pressing if one does not trust those who are likely to sit at the controls of the agency.
A different version of this dynamic can be seen in the U.S. Supreme Court's doctrine on administrative law. The adoption and implementation of regulations by administrative agencies, in furtherance of legislative mandates, raise a compliance issue of whether the administrative regulations are genuinely consistent with the statutory authority upon which they depend. The courts might insist that they should make the final call on how statutes should be interpreted and whether agency actions are consistent with the statute. Alternatively, the courts might defer to the statutory interpretations made by the executive branch itself, perhaps under presidential and legislative oversight. The Chevron doctrine, the policy of judicial deference given to administrative actions, instructs lower courts to defer to agencies’ interpretation of ambiguous statutory directives so long as those interpretations are not arbitrary or capricious.4 Recently, the Court has become more skeptical of the executive branch, and thus less willing to defer to controversial administrative interpretations of statutes, culminating in the Court's recent decision to overrule Chevron.5 A trusted agency may be given more leeway to interpret and implement substantive regulatory policies. Unlike Congress, the Court does not have the authority to alter the substantive policy embedded in the statute if it is distrustful of executive officers, but it can shift more interpretive authority from the executive to the judiciary.
A quite different example of the interaction of substantive rules and decision-making procedures can be found in the context of the impeachment power. The U.S. Constitution gives little detail about the power of impeachment, but it does specify both a substantive policy and a decision rule for implementing that policy. The substantive policy is that officers can be impeached and removed for treason, bribery, and high crimes and misdemeanors. Of course, that last category of impeachable offenses leaves a great deal of discretion in the hands of the implementing authority: in this case, Congress. But the Constitution establishes a very important decision rule. The House of Representatives can impeach an officer with a simple majority vote. The Senate, however, can only convict and remove an officer on impeachment charges with the agreement of two-thirds of the senators present. The impeachment power might operate very differently if officers could be removed on the basis of a majority vote in the House alone, even if the substantive rule regarding impeachable offenses were exactly the same. If one were worried that the House might abuse the impeachment power, the worry could be alleviated by either tightening the standards for impeachable offenses or raising the hurdle on making decisions, or both.6
A final example comes from a realm closer to the free speech context. Imagine a new communication technology that made it possible to produce, distribute, and exhibit to public audiences strips of film with moving pictures and sound. Such a technology might make community leaders and politicians very nervous about what the youth of America might encounter. They might think that it would be a good idea to restrict the access of minors to some content that might appear on film. To do so, they would need both to determine what substantive rule should be applied to distinguish films that could and could not be shown to children and they would need to identify someone who could screen films and categorize them as appropriate for general audiences or not in accordance with that substantive rule.
The first effort to do so involved a set of local governmental film censorship boards. That the town fathers of New York City thought a film was fit to exhibit to the public did not mean that the film would not be banned in Boston. Even if the rule on paper for banning films was the same in both cities, neither would have been content to trust the judgment of the other's censor board. Subsequently, the film industry, and the courts, convinced local governments to get out of the film censorship business and to rely instead on the motion picture industry itself to rate films and restrict access to them. What one might think would be a good rule to separate restricted from unrestricted films might depend heavily on how much one trusted the body empowered to apply the rule to specific films. Conservative parents might want a far more restrictive and clearly defined set of rules if they think the Motion Picture Association is run by a bunch of libertines; alternatively, they might turn to outside rating services, like Common Sense Media, that would more closely mirror their own preferences and social mores.
Trust engenders discretionary leeway for regulatory bodies. Substantive rules that could safely be placed in the hands of a trusted administrator would need to be reconsidered if the administrator is not trusted. Faith in authority, which often amounts to shared preferences and conviction, facilitates delegating power to those in authority. The same basic political logic operates in the realm of speech regulations.
The question of trust in authorities empowered to restrict speech is most pressing in the context of the coercive power of the state. The government is not the only institution that can restrict some forms of speech, but the government has the most sweeping power to do so and can deploy coercive force to gain compliance.
The United States is an outlier among democratic countries for both its relative lack of trust in government and its relatively libertarian policies regarding freedom of speech. That probably has particular significance relative to debates about the best approach to hate speech. Many democratic countries impose some legal restrictions on what might broadly be called hate speech. Hate speech seems like an easy target for government regulation. The costs of tolerating such speech are evident. The benefits are negligible. Why not regulate it? The rise of social media has spurred many countries to specifically restrict hateful content on digital platforms.7 Those same countries often had criminal laws against various forms of hate speech already in place.8 Constitutional scholars from those countries are often quite comfortable with the idea that free speech guarantees can be balanced against a desire to protect the dignity of individuals or communities and with empowering law enforcement officials with the authority to punish speech that incites hatred. From an American civil libertarian perspective, by contrast, such laws not only suppress speech that should be tolerated but also risk empowering police and prosecutors with a discretionary authority that can be turned against some of the same groups that hate speech laws are intended to protect. Can we trust government officials - or even our fellow citizens - to know hate speech when they hear it? Where there is a broad social (or at least a relevant elite) consensus on the contours of what counts as hate speech, then regulation becomes more viable. But in global surveys, Americans say they trust the government at rates more comparable with Colombia and Slovakia than with Switzerland, Germany, and Japan.9 It is hard to be eager to empower the government to suppress hateful speech if you fear that the government is likely to abuse that power or turn it against you.
The fate of the somewhat related “fighting words” exception to the First Amendment is telling of the difficulties in the American context. The U.S. Supreme Court has traditionally recognized a set of substantive “exceptions” to the protections of the First Amendment, such as obscenity and true threats of violence. The latter gave little protection for individuals uttering so-called fighting words. In 1942, the Court observed that “there are certain well-defined and narrowly limited classes of speech, the prevention and punishment of which have never been thought to raise any Constitutional problem.” Among those classes of speech were “insulting or ‘fighting’ words,” which had no redeeming social value and ran contrary to a “social interest in order and morality.” Such words could be forbidden because they were thought to tend “to incite an immediate breach of the peace.”10
The “fighting words” doctrine has never been formally overruled, but after articulating the doctrine in the midst of the Second World War, the Court has steadily narrowed its scope. The blossoming of the civil liberties revolution on the Court in the mid-twentieth century shifted the judicial orientation on the proper balance between law and order on the one hand and individualistic expression on the other. Just a few years later, the majority of the Court would instead emphasize that the right to speak freely is “one of the chief distinctions that sets us apart from totalitarian regimes.” Included in that right was speech that “induces a condition of unrest” and “even stirs people to anger.”11 Whether delivering racist diatribes, burning national flags, or shouting obscenities at a funeral, the justices increasingly came to think that it was the onlookers who had a responsibility to refrain from turning to violence rather than speakers who had a responsibility to refrain from stirring people to it. The person who threw the first punch breached the peace, not the one who hurled the first insult. Whatever social consensus that had once held that you could not wear a jacket emblazoned with profane slogans in a public place or yell insults at a police officer had washed away. As the country rebelled against the man in the gray flannel suit, the Court concluded “that one man's vulgarity is another's lyric.”12
The breakdown of a once buttoned-down social consensus was accompanied by greater skepticism of government authority and the officials who might be tasked with determining which speech had no social value or could stir people to anger. In the very case in which the Court first laid out the fighting words exception, the speech in question involved a sidewalk speaker telling a police officer that he was “a damned fascist,” among other “offensive, derisive and annoying words and names.”13 Later cases before the Court similarly involved opprobrious language aimed at police officers, and not infrequently “direct conflict of testimony as to ‘who said what.’”14 Calling a police officer a “white son of a bitch,” or yelling at them “you god damn motherfucking police,” or saying to them “why don’t you pick on somebody your own size?” could get one arrested for creating a disturbance.15 The justices began to think it unwise to “provide the police with unfettered discretion to arrest individuals for words or conduct that annoy or offend them.”16 The kind of people who were repeatedly arrested for “interrupting” the police with their unwelcome language-that is, people like Raymond Hill, an advocate of the Gay Political Caucus in Houston, Texas-might have already distrusted the authorities.17 In the turbulence of the 1960s, even federal judges came to share some of that distrust and decided that First Amendment protections needed to be stronger.
At the same time that the Court was discovering that one man's vulgarity could be another man's lyric, it was also wrestling with the scope of another exception to the First Amendment: obscenity. An American judge in the nineteenth century gave the conventional wisdom in noting, “while happily we have outlived the epoch of censors and licensors of the press, to whom the publisher must submit his material in advance, responsibility yet attaches to him when he transcends the boundary line where he outrages the common sense of decency.”18 As long as we thought we could identify a “common sense of decency” and trust government officials with enforcing it, the suppression of obscene material did not seem to cause much of a problem, and American courts were quite happy to cast those materials into the darkness beyond First Amendment protection. The Court thought it “apparent that the unconditional phrasing of the First Amendment was not intended to protect every utterance.”19 It was not intended to protect those items “utterly without redeeming social importance.”20 But how to distinguish the obscene from the merely lewd? Perhaps the courts could divine and rely upon “the average person, applying contemporary community standards.”21 Such standards proved to be slippery. Judges soon found themselves second-guessing customs officials and local prosecutors on whether works of modern literature by authors such as James Joyce, D. H. Lawrence, Henry Miller, and William S. Burroughs outraged the common sense of decency. The Supreme Court justices and their clerks found themselves crowding into the basement on “movie day” to decide which films had been properly declared obscene by local judges and juries. After watching Louis Malle's 1958 film The Lovers, which had led to a theater manager in Ohio being fined for exhibiting an obscene film, Justice Potter Stewart threw up his hands and declared he could not “intelligibly” define obscenity: “But I know it when I see it, and the motion picture involved in this case is not that.”22 So long as the Court trusted judges and juries to distinguish art from “hard-core pornography,” then an obscenity exception to the First Amendment could be left ill-defined and capacious. As that trust was lost, censors were put on a shorter leash.
The First Amendment itself was born out of distrust of the traditional censors. It enshrines the command that Congress shall make no law “abridging the freedom of speech, or of the press,” but what does that mean? As the Court later pointed out, the “unconditional phrasing” of the constitutional text was understood by almost everyone to be more restrictive than grammar alone might suggest. Justice William O. Douglas skipped movie night at the Court since, in his view, the First Amendment laid down “its prohibition in terms absolute.”23 Few others found the First Amendment to be so easy.
James Wilson, who would later become one of our nation's first Supreme Court justices, gave one fairly conventional answer to the question of what “freedom of speech” means during the ratification debates for the U.S. Constitution. Although the First Amendment had not yet been written, and Wilson thought such an amendment was unnecessary, he found himself defending the Constitution against the charge that it would open the door to abridgments of the freedom of speech. To such an accusation, Wilson thought it important to define terms. Of course, Congress was not barred from punishing literally every type of speech for
the idea of the liberty of the press is not carried so far as this in any country…. What is meant by the liberty of the press is, that there should be no antecedent restraint upon it; but that every author is responsible when he attacks the security or welfare of the government, or the safety, character, and property of the individual.24
The constitutional framers did not trust the system of “censors and licensors” that the British king had established to control what texts could be disseminated in his realm, but that did not mean that they did not trust anyone to identify dangerous and harmful speech. Indeed, by Wilson's reasoning, the “liberty of the press” protected in both English law and American law was the same. What had changed in the United States was the assessment of who could be trusted to respect the boundaries of the liberty of the press. The Americans did not trust executive licensors to play that role, but perhaps they could trust an independent judiciary to do so. If the true threat to free speech came from monarchs, then eliminating that threat might do most of the work that was necessary to protect freedom of speech. For good measure, judges could be insulated from control by the quasi-monarchical executive branch, and then perhaps they could be trusted to defend the people's interest in free speech. But even if that failed, there was an ultimate trustworthy body to determine whether an author “attacks the security or welfare of the government”: a jury of the author's peers.
It took more experience with republican government to demonstrate that kings were not the only threat to freedom of speech. The emergence of partisanship and the passage of the Sedition Act of 1798 showed that not even judges and juries could be entrusted to distinguish protected from unprotected speech. Federalist editors preached that “it is Patriotism to write in favor of our government - it is Sedition to write against it.”25 Not all judges and juries, they thought, could be entirely trusted to uphold the authority of the government. What was needed was someone like Justice Samuel Chase, who was “a sworn enemy of free democrats” and could be counted on “to terrify democratic printers from insolently avowing opinions contrary to the ruling powers.”26 Chase did not have much faith that juries would help him in that task, and he had to instruct one federal marshal to be sure to remove “any of those creatures called democrats” from the jury pool.27 Meanwhile, the Jeffersonians were learning that they could not trust Federalist judges and juries to uphold the liberty of the press and properly distinguish protected political opinion from unprotected seditious speech. With that declining trust came a reassessment of the wisdom of prohibiting “seditious” speech in a republic.
A century ago, philosopher Zechariah Chafee helped transform American constitutional law by hammering home the broader lesson of that experience. “The essential question,” he thought, “is not, who is judge of the criminality of an utterance, but what is the test of its criminality.” Lack of trust in potential censors meant that we had to adopt more robust protections for freedom of speech.
The transference of that censorship from the judge to the jury is indeed important when the attack on the government which is prosecuted expresses a widespread popular sentiment, but the right to jury trial is of much less value in times of war or threatened disorder when the herd instinct runs strong, if the opinion of the defendant is highly objectionable to the majority of the population, or even to the particular class of men from whom and by whom the jury are drawn.28
There was value in shifting power to a more trustworthy authority, but unpopular speech could only be reliably protected if stringent rules were adopted to tie the hands of those in charge.
The same interplay of trust and rules arises in the context of protecting speech on college campuses. The stakes may not be as high as a sedition prosecution by the government; however, preserving a free intellectual environment in American universities is nonetheless consequential. Free inquiry stagnates if it is not adequately protected, and there are myriad pressures to curtail free inquiry in the present moment.
Those developing rules, practices, and institutions to regulate speech in the academic environment have often evinced a great deal of trust in the authorities who would be entrusted with patrolling the boundaries of acceptable and unacceptable speech. Some degree of trust might be unavoidable if universities are to function, but the more we have faith in those who will be exercising authority, the more comfortable we tend to be with empowering them with broad discretion to evaluate speech. Let me note two quite different contexts in which this plays out.
Universities have always had codes of conduct, and those codes of conduct have often applied to speech, especially student speech. In the age of in loco parentis, the freedom of the students was limited, and the discretion of the dean was vast. The same cultural sea change that weakened the power of the censors in other parts of American life also loosened the grip of the dean on student speech. When the dean of students at the University of Missouri decided to expel Barbara Papish, a thirty-two-year-old graduate student and member of the Students for a Democratic Society, for distributing on campus a newspaper featuring on its cover a cartoon depicting a policeman raping the Statue of Liberty and articles containing obscenities, she made a federal case out of it. Papish had previously been put on probation for disseminating literature containing “pornographic, indecent and obscene words” while high school seniors and their parents were visiting campus in 1967. The Court made it plain that “no matter how offensive to good taste,” the dean could not enforce “conventions of decency” in student publications. State universities were “not enclaves immune from the sweep of the First Amendment.”29
The campus speech codes of the 1980s, like the campus speech codes of the 1960s, were top-down. It is hardly surprising that when college administrators get together to draft regulations restricting the speech of college students, the administrators will give themselves substantial leeway to crack down on harmful speech. Drafters of such policies often “intended that speech need only be offensive to be sanctionable.”30 When the power to make law and the power to enforce the law are united in the same hands, the lawmaker tends to trust the law enforcer to know what to do. The administrators thought the administrators would recognize offensive speech when they saw it.
More remarkably, the campus speech codes of the twenty-first century are often bottom-up rather than top-down. The current generation of students has a striking degree of trust in college authorities and consequently often seeks to invest them with substantial authority to regulate speech on campus. I have frequently found that current students are shocked to discover that students had ever bristled against the dean's authority or that university officials ever exercised their authority to suppress speech because it was offensive to good taste or embarrassing to the university. Students now assume that campus administrators will share their values and commitments and think like they do. As a consequence, students are trusting that campus administrators will suppress only the right kind of speech. The campus activists now issuing these demands might well be making a safe bet. When a majority of the student body and the campus administrators are ideologically aligned, it is only the few dissenting students who have to worry that their speech might be ruled out as harmful.
The events of October 7th in Israel might have fundamentally altered that calculation. As the war in Gaza continued into the spring of 2024, student protests broke out on campuses across the globe, demanding their universities’ disclosure of and divestment from military contracts and deals with private arms companies. Suddenly, members of the campus community found themselves divided on what kinds of political speech and political activism were tolerable. Community members who were on the same side of many political disputes were now at odds with one another, and speech regulations that had gone unnoticed began to pinch. There was evident dissensus over which examples of political rhetoric were hateful or made the campus feel unsafe. Administrators’ responses to the protests have included quiet tolerance, on one end of the spectrum, and the expulsion, eviction, and violent arrest of protesters, including some faculty, on the other. Whether universities further respond by narrowing or broadening speech restrictions on campus through formal policy remains to be seen, but a new generation of students might have learned that they can no longer trust that college administrators will always see things as they do.
The speech of professors is regulated as well. At this moment, there are very serious threats to professorial speech coming from outside the university, but for now I want to call attention to how speech is regulated inside the university.31 Of course, the professoriate has little faith in how conservative politicians would seek to regulate academic discourse. Less visibly, the professoriate relies on a more trusted authority to regulate academic discourse: themselves.
Scholarship is routinely subjected to speech regulation. Literary scholar Stanley Fish likes, provocatively, to call this “censorship.”32 One need not be so provocative to see the point. Legal scholar Robert C. Post has emphasized that “academic freedom has always been conceived as a barrier to ‘the pressure in a democracy of a concentrated multitudinous public opinion.’”33 Academic freedom does so, however, to protect the development of disciplinary knowledge. But here he would emphasize the disciplinary. Academic freedom is ultimately the freedom of a scholarly community to generate and discuss ideas. We allow that scholarly community to discount purported knowledge that is not regarded as consistent with disciplinary norms. As academics, we trust the discipline, not the individual scholar. That trust in the discipline is rendered concrete in the boundaries that are raised up and enforced around academic freedom. Speech that the discipline judges to be unworthy is unapologetically disfavored. Such speech is inexpert and incompetent. Academic freedom is designed to protect only the speech that the scholarly community regards as competent.
The system works well so long as scholars trust their colleagues, peers, and disciplinary communities. Just as judges looked to community norms to determine what expression could be suppressed as obscene, universities look to disciplinary norms to determine what expression can be suppressed as irresponsible or unwise. That trust can fray. Peer-reviewed scholarly journals are one site where discipline is enforced. The purpose of peer review is to provide a disciplinary quality control of what gains the imprimatur of scholarship. However, the familiar concept of “peer-review cartels,” by which a small subset of academics use the peer-review process to exclude scholarly work that might challenge their own preeminence in the field, lays bare how trust in disciplinary communities can break down. The relative authority of peer review hinges on the trustworthiness of the peers who are doing the reviewing. If the gatekeeping process is seen as abusive, scholars might develop alternative channels for disseminating their work, whether through upstart journals with more trusted editors and reviewers, non-peer-reviewed paper repositories, or non-peer-reviewed journals.
While peer-review cartels can damage scholarly careers and pervert the shape of scholarly discourse, other processes can expel professors entirely from the scholarly community. A central goal of academic freedom advocates in the United States in the early twentieth century was to free faculty hiring and promotion decisions from the control of the nonexperts. University presidents and trustees could not be trusted to be reliable gatekeepers of expert knowledge. Academic freedom could only exist if the scholarly realm inside the university was controlled by the scholars. But again, that gatekeeping function can endure and thereby protect intellectual freedom only if those exercising the gatekeeping function retain trust in their capabilities and integrity. On one front, self-governing professors must be able to demonstrate to outside stakeholders that they are acting in good faith and to the ultimate benefit of the university. Donors and politicians who become convinced that the self-governing faculty has created a cartel of its own that operates to the detriment of the university are likely to want to seize back the power over personnel.
On another front, if academic freedom and free inquiry are to thrive, then it is crucial that faculty governance not be weaponized against heterodox thinkers. One recent proposal shows how academic freedom, built on the logic of a disciplinary community, can be turned into a sword rather than a shield. Literature and media scholars Michael Bérubé and Jennifer Ruth chose the title It's Not Free Speech for their recent book precisely to emphasize that academic freedom is not truly an individual right. Scholars enjoy academic freedom only by the grace of their colleagues. Thus, Bérubé and Ruth suggest that universities should create “academic freedom committees,” not to help ensure that professors are protected from threats to their work, but to help universities identify and expel professors who express the wrong normative commitments.34 The thicker the commitments of a scholarly community, the less room there will be for the dissident. Free inquiry can flourish only where the gatekeepers can be trusted to allow it.
If one trusts the censors, or expects to control the censors, or expects to be the censors, then one is likely to have more confidence designing rules that allow for restricting speech. When contemplating a set of ideal speech restrictions, however, one should think carefully about how those rules might be applied if one's antagonists were operating the system. For me, distrust of the potential censors dictates robust free speech protections. I might trust myself to exercise good judgment about which speech ought to be suppressed, but I do not trust anyone else- and I would not advise anyone else to trust me.
ENDNOTES
Keith E. Whittington, Speak Freely: Why Universities Must Defend Free Speech (Princeton, N.J.: Princeton University Press, 2018).
James Madison, “The Structure of the Government Must Furnish the Proper Checks and Balances between the Different Departments,” Federalist No. 51, in Alexander Hamilton, James Madison, and John Jay, The Federalist Papers, ed. Clinton Rossiter (New York: New American Library, 1961), 322.
Terry M. Moe, “The Politics of Bureaucratic Structure,” in Can the Government Govern? ed. John E. Chubb and Paul E. Peterson (Washington, D.C.: Brookings Institution Press, 1989), 276. For related points, see also Matthew D. McCubbins, Roger G. Noll, and Barry R. Weingast, “Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies,” Virginia Law Review 75 (1989): 431-482.
Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
Loper Bright Enterprises v. Raimondo, 603 U.S.___ (2024).
On the impeachment power, see Keith E. Whittington, The Impeachment Power: The Law, Politics, and Purpose of an Extraordinary Constitutional Tool (Princeton, N.J.: Princeton University Press, forthcoming 2024).
David Bromell, Regulating Free Speech in a Digital Age: Hate, Harm and the Limits of Censorship (Berlin: Springer International Publishing, 2022); and Timothy Garton Ash, Free Speech: Ten Principles for a Connected World (New Haven, Conn.: Yale University Press, 2016).
Caitlin Ring Carlson, Hate Speech (Cambridge, Mass.: The MIT Press, 2021); Chris Demaske, Free Speech and Hate Speech in the United States (London: Routledge, 2021); and Nadine Strossen, Hate: Why We Should Resist It with Free Speech, Not Censorship (Oxford: Oxford University Press, 2018).
OECD (Organisation for Economic Co-operation and Development) iLibrary, “Trust in Government,” http://doi.org/10.1787/1de9675e-en (accessed June 5, 2024).
Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942).
Terminiello v. City of Chicago, 337 U.S. 1, 4 (1949).
Cohen v. California, 403 U.S. 15, 25 (1971).
Chaplinsky v. New Hampshire, 574, 568.
Lewis v. New Orleans, 415 U.S. 136 (1974).
Gooding v. Wilson, 405 U.S. 528 (1972); Lewis v. New Orleans, 415 U.S. 134 (1974); and Houston v. Hill, 482 U.S. 454 (1987).
Houston v. Hill, 482 U.S. 465 (1987).
Ibid., 454.
United States v. Harmon, 45 F. 414, 416 (Dist. Kansas 1891).
Roth v. United States, 354 U.S. 476 (1957).
Ibid.
Ibid., 489.
Jacobellis v. Ohio, 378 U.S. 184, 197 (1964).
Roth v. United States, 354 U.S. 476, 514 (1957).
Jonathan Elliot, ed., The Debates in the Several State Conventions on the Adoption of the Federal Constitution, Volume2 (Philadelphia: J. B. Lippincott Company, 1836), 449.
Gazette of the United States, October 10, 1798.
New London Bee, September 3, 1800.
The Debates and Proceedings in the Congress of the United States: With an Appendix Containing Important State Papers and Public Documents, and All the Laws of a Public Nature, Volume 14 (Washington, D.C.: Gales and Seaton, 1852), 218. See also Keith E. Whittington, Constitutional Construction: Divided Powers and Constitutional Meaning (Cambridge, Mass.: Harvard University Press, 1999), 42, 43, 46.
Zechariah Chafee Jr., Freedom of Speech (New York: Harcourt, Brace, and Howe, 1920), 25, 26.
Papish v. Board of Curators of the University of Missouri, 410 U.S. 667, 669 (1973).
Doe v. University of Michigan, 721 F. Supp. 852, 860 (E.D. Michigan 1989).
On external threats, see Keith E. Whittington, You Can’t Teach That! The Battle over University Classrooms (Cambridge: Polity, 2024).
Stanley Fish, The First: How to Think About Hate Speech, Campus Speech, Religious Speech, Fake News, Post-Truth, and Donald Trump (New York: Simon and Schuster, 2019).
Robert C. Post, Democracy, Expertise, Academic Freedom: A First Amendment Jurisprudence for the Modern State (New Haven, Conn.: Yale University Press, 2012), 67.
Michael Bérubé and Jennifer Ruth, It's Not Free Speech: Race, Democracy, and the Future of Academic Freedom (Baltimore: Johns Hopkins University Press, 2022).