Last year, in Matal v. Tam, the Supreme Court considered a constitutional challenge to provisions of the Lanham Act that prohibit the Patent and Trademark Office (PTO) from registering any trademarks that may “disparage … or bring into contempt or disrepute” any “persons, living or dead.” Tam, the leader of an Asian rock band called “The Slants,” wanted to register that name as a trademark. “Slants,” however, is a derogatory term for Asians (Tam wants to appropriate the term and so drain it of pejorative force), and so the PTO refused to register the mark. Tam sued, claiming that the relevant provisions of the act violated the First Amendment. A unanimous Supreme Court agreed, holding that giving offense is a particular viewpoint, and a law restricting expression on the basis of viewpoint violates the First Amendment.
The Supreme Court has often held that the government may not ban speech merely because it is offensive, and so the result in Matal was not surprising. Nevertheless, the case is important because, it makes clear that, although many European nations ban “hate speech” and other forms of speech deemed offensive, the United States will continue its long tradition of protecting more speech than other nations do.
Now, most Americans think this is a good thing, but my friend Professor Hadley Arkes, writing in the Claremont Review of Books, disagrees. Like the Europeans, he thinks that offensive speech should not be protected. On his view, the First Amendment should protect arguments but not epithets, reasoned discourse but not insults. As to Matal, he thinks that the Court’s opinion can be explained only by supposing that the justices are falling into moral relativism.
I have the greatest respect for Professor Arkes. I also agree completely with the position with which he is universally identified and for which he is rightly regarded as a hero of modern conservative jurisprudence—that almost all legal reasoning inevitably involves some natural-law reasoning. I cannot, however, follow him down this road about Matal. I think there is a simpler explanation of Matal and the Supreme Court’s other free speech cases. In my view, the Court is not saying, or even implying, that there is no objective standard about which speech is offensive and which is not. Like other people, the justices can recognize the difference between an argument and an epithet. Rather, the justices are saying that, although there is an objective standard about which speech is offensive, the government cannot be trusted to apply this standard in particular cases, regardless of whether the application is done by the legislature, the courts, jurors, or any other government actor. To see why, it helps to consider some other legal problems that may at first seem rather different but that in fact involve quite similar principles.
The Murder Statute and the Twenty-Sixth Amendment
Consider the murder statute. If anything is wrong, certainly murder is, and so we make it illegal. When it comes to enforcing this law, however, we do not punish everyone who commits a murder. On the contrary, we punish only those people whom the state can prove beyond a reasonable doubt committed a murder. Undoubtedly, some people who violate the law escape punishment because the state cannot prove their guilt.
Why do we have such a system? Because we recognize that, in applying the rule against murder, we fallible human beings will make mistakes—both acquitting the guilty (false negatives) and convicting the innocent (false positives). Since we believe, as Blackstone said long ago, that “it is better that ten guilty persons escape than one innocent suffer,” we create a rule of enforcement that greatly reduces the number of false positives at the expense of increasing the number of false negatives.
Or consider the Twenty-Sixth Amendment, which gives eighteen-year-olds the right to vote. The idea behind a minimum voting age is the eminently sensible one that, before someone be allowed to vote, he or she must have a certain level of maturity. But does everyone younger than eighteen lack this level of maturity and everyone over eighteen have it? Of course not. The reality is that the rule is both under-inclusive and over-inclusive: it excludes some people mature in mind but young in years, and it includes some people old in years but immature in mind.
We could have a rule that provides that only people mentally mature enough to vote be allowed to do so. That rule would admit to the franchise precisely the right group of people, if it were perfectly enforced. We know, however, that the enforcement would be very far from perfect. Some government agency or court would have to decide whether a person was mature enough to vote. Even if the officials charged with making this determination had the best of intentions, they would have to conduct extensive factual investigations about the applicant, and they would inevitably make a great many mistakes. Indeed, they might well make more mistakes than would result from applying the bright-line rule based on mere chronological age.
Moreover, the rule would be an invitation to mischief. For the officials charged with administering the rule would of course belong to one political party or another, and they would thus have a strong partisan incentive to find against applicants who seem likely to vote for the other party. In Democratic strongholds, young would-be Trump voters would often be found too immature to vote, while in Republican strongholds the same treatment would be meted out to likely liberal voters. Given all this, it is easy to see that we are much better off with a bright-line rule based on chronological age, even though that rule will result in a large number of errors.
The lesson is that, in framing a rule of law, we need to take account not only of the content of the rule but also of the likely problems involved in its enforcement. Some rules that would be just the thing if they could be enforced perfectly will be terrible in practice precisely because, when administered and enforced by fallible human beings, they generate huge problems. In such cases, we are often better off with other rules that, while imperfect by their terms, present fewer problems of enforcement and so produce better results overall.
Why Do We Need Constitutional Protections for Speech?
And this turns out to be the case with protections for speech. Let us begin from first principles. Why do we need constitutional limits on the power of the legislature to regulate speech? Why not trust the legislature to apply the objective standard about which speech is offensive and which is inoffensive, banning the former and permitting the latter?
The answer, of course, is that history shows that legislatures tend to make an especially large number of mistakes in regulating speech. Moreover, these mistakes tend to run all in the same direction: they tend to be false positives (which ban speech that is not objectively offensive) rather than false negatives (which fail to ban objectively offensive speech). Even worse, these false negatives are not randomly distributed across intellectual or ideological viewpoints but tend to be concentrated on viewpoints with which the majority faction in society disagrees. In a system in which there were no constitutional limits on the legislature’s power to regulate speech, the majority’s views would always be permitted by law, but many minority views would be systematically suppressed—not because those views are objectively offensive but because the majority mistakenly believes them so.
What kind of constitutional limit on the legislature’s power will curb this tendency? One possibility would be a constitutional provision stating that only objectively offensive speech may be prohibited. That would be an improvement over a system that included no constitutional protection for speech, but only a very modest one. In this system, we get two levels of protection: first, the honesty and wisdom of the legislators, and then the honesty and wisdom of the judges, each applying the same standard.
But legislators and the judges tend to be similar people, with similar educational and socioeconomic backgrounds and holding similar values. Ultimately, they all owe their positions to democratic majorities. Their biases and prejudices in determining which speech is objectively offensive and which is not will thus tend to coincide. Hence, judges are likely to make the very same mistakes as legislators in applying the objective standard to determining which speech should be banned and which should not, and this seriously undermines the value of the proposed constitutional limit on the power of the legislature.
To get a system that meaningfully reduces the legislative error rate, we need a constitutional limit on the power of the legislature that, when applied by the courts, strikes down more laws and leaves legal more speech. We thus get doctrines like those relied upon in Matal: the government may not discriminate among viewpoints in regulating speech and may not ban speech merely because it is offensive. In this way, we get a system that almost never suppresses good speech, even though it lets through a great deal of bad speech.
How It Really Works Out
But is it really true that, if we allowed government to ban just the speech that is objectively offensive, we would end up suppressing inoffensive minority viewpoints? Professor Arkes thinks not. In my terms, he thinks that the error rate, while not zero, will not be very high. Indeed, he argues that average people have no trouble sorting out insulting from non-insulting terms. His examples of the latter include urologist, meter maid, and saint, and his examples of the former include kike and wop. Even these examples are not as clear as Professor Arkes thinks (some women would find meter maid offensive), but the real problem is that Professor Arkes has not considered the harder cases that will inevitably crop up.
Take the term homophobe. Some people will say that, while homophobia is offensive, the word homophobe is not; it merely refers to people who hate or fear homosexuals. But the case is not so simple. Professor Arkes himself says—truly, to be sure—that sometimes people making arguments against same-sex marriage are “branded as ‘homophobes,’ as though they were carriers of disease rather than bearers of arguments that deserve to be respected and addressed as arguments.” In other words, homophobe, like a great many other words (racist, Communist, traitor), can be used as an argument-ending insult. Or again, is Holocaust-denier an insult, or a factually correct description of some people with certain odious beliefs? What about climate-change denier? Before we can figure out whether that term is an insult, we would need to settle a bunch of empirical questions about the environment and the likely future effects of certain current trends. Suddenly, figuring out which terms are insults is not so easy.
For that matter, we should remember that insults actually have a legitimate place in rational discourse. When Winston Churchill said of Clement Atlee, “He was a modest man with much to be modest about,” he was making a serious point. So too was Theodore Roosevelt, when, after Justice Holmes voted against the government in the Northern Securities antitrust case, Roosevelt said of him, “I could carve out of a banana a judge with more backbone than that.” The Apostle Paul called the Galatians stupid, and Our Blessed Lord called the Pharisees, “You brood of vipers!” It is just not true, in my view, that insults are always worthless and offensive speech that we can do without. Will courts be able reliably to distinguish the insults that make a serious point from the ones that merely insult?
But, quite apart from all that, we need not speculate about what is likely to happen if we start allowing the government to ban offensive speech. We need look at what is actually happening in those jurisdictions that already have such bans. Floyd Abrams, the prominent First Amendment lawyer, collects examples:
In Belgium, a member of Parliament and leader of a right-wing political party was convicted in 2009 for distributing leaflets calling for a “Belgians and Europeans First” policy and that said, “Stop the sham integration policy,” “Send non-European job-seekers home,” and “Stand up against the Islamification of Belgium.” He was convicted of incitement to racial discrimination, disqualified from holding office for ten years, and sentenced to community service. His conviction was upheld by the European Court of Human Rights on the ground that such literature “sought to make fun of the immigrants,” leading to hatred of foreigners, particularly by “less knowledgeable members of the public.” In England, an individual was tried and convicted for displaying a poster that showed the World Trade Center ablaze with the caption “Islam out of Britain—Protect the British People.” The European Court of Human Rights permitted that conviction, as well, to stand, concluding that the poster constituted a “public expression of attack on all Muslims in the United Kingdom.”
Similarly, in France, legendary actress Brigitte Bardot has been convicted five times for hate speech against Muslims, most recently for saying that Muslims in France “lead us around by the nose” and “destroy our country.” In Switzerland, a judge issued an arrest warrant for Italian author Oriana Falacci for allegedly racist passages in her book The Rage and the Pride (Italy refused to extradite her). In the Netherlands, when politician Geert Wilders asked a crowd of his supporters whether they wanted “more or fewer Moroccans” in the country, the crowd shouted, “Fewer!” Wilders replied, “Well, we’ll take care of that.” For that, he was later convicted of hate speech. Germany has recently enacted a new law banning hate speech on the internet and used it to take down tweets from members of parliament when these were deemed offensive to Muslims.
There are hundreds and hundreds of more examples. If we allowed the government to ban offensive speech, exactly the same thing would happen in the United States.
The American Experiment
These examples show how quickly the noble impulse to purge the public square of offensive and insulting language degenerates into censorship of unpopular viewpoints. But if we want to protect such viewpoints, we need rules severely limiting the power of government to regulate speech. Such rules come at a cost, which is that they allow more objectively offensive speech to go about in society, and that speech, we must admit, does real damage.
But all solutions have costs as well as benefits, and in the area of speech we have available basically two solutions. We can have weak protections for speech, in which case much offensive speech will be suppressed, but so will many minority viewpoints that are not objectively offensive. Or we can have strong protections for speech, in which case there will be more offensive speech in society, but the minority viewpoints that are not objectively offensive will be heard too. The one solution that is not available to fallible human beings in the real world is to suppress all the offensive speech and none of the inoffensive speech. No group of human beings can apply the relevant standard, objective though it be, without making a very large number of errors.
Faced with this dilemma, the characteristically American solution is to err on the side of liberty, which means trusting in the wisdom and virtue of the American people. That is, by allowing virtually all speech, we are trusting that the great majority of the American people will not abuse their freedom to engage in bad speech. We trust that, when a small minority does abuse that freedom, the great majority will not be deceived by their lies, corrupted by their seductions, or provoked to violence by their incitements. It is no accident that the European nations, which have never believed in the ability of their people to govern themselves as we Americans have, think they need more restrictions on free speech. The American experiment is founded on the opposite view: the view that a people capable of governing themselves are worthy of the trust that the First Amendment places in them.