A common misconception is that the First Amendment does not protect hate speech. Even among people who know that hate speech is constitutionally protected, there is a relatively popular belief that it should not be. I disagree, and in this post I’ll explain why.
In 1988, the University of Michigan, like hundreds of colleges and universities across the country, passed a hate speech code. The University created the code in response to a series of incidents in which someone hung horrifically racist fliers that used racial slurs to describe black people, said that black students ought to be lynched, and claimed they did not belong on campus. The University’s full speech policy can be found here. It’s a little lengthy, though, so for the purposes of this post I’ll use a shorter definition of hate speech that to my mind captures the spirit of hate speech codes:
Hate speech is speech that disparages a person or group of people on the basis of an immutable characteristic.
Using this definition, it is clear that the fliers that prompted the creation of Michigan’s hate speech policy qualify as hate speech. Fliers that use slurs to refer to racial groups, or argue that people should be killed because of their race, clearly disparage a person on the basis of an immutable characteristic. In fact, we don’t need a speech code to tell us that such speech is hateful, but the point is that the policy did actually cover hate speech.
The problem with hate speech codes is that they also capture speech that is not hateful. While the University of Michigan’s speech code was in effect, only one student was subjected to a full disciplinary hearing for using hate speech. He was a graduate student studying social work, who said in a class discussion that he was “developing a social work approach to move homosexuals toward heterosexuality” – what we would refer to today as conversion therapy. Under the definition I gave above, the student’s statement likely qualifies as hate speech.
Conversion therapy is premised on the idea that a person’s sexual orientation (or in some instantiations, their gender identity) can be changed, and that non-straight people are mentally ill. As it turns out, neither of these things is true. The best evidence appears to be that a person’s sexual orientation cannot be changed; it is immutable. Moreover, the DSM – the manual used by mental health practitioners to classify mental illnesses – has not included homosexuality as a mental illness since 1987, so any reference to homosexuals as mentally ill is disparaging. Saying you want to perform conversion therapy on gay people, then, is hate speech because it disparages a group of people on the basis of an immutable characteristic. Right?
I think not. What the student was saying was that he saw a group of people who were mentally ill, and he was developing a mechanism to help them. There are a number of ways to critique him for this. Was what he said unscientific? Seemingly yes, though bear in mind that the student made his statement less than two years after homosexuality was removed from the DSM, and at a time when it was still classified as a mental illness by the WHO. Was what he said condescending? Sure. Was it misguided? Absolutely. But was it hateful? No. Any definition of hate speech that covers the statement “I want to help people” – no matter how misguided that sentiment may be – stretches the definition of hate past the breaking point.
This is the fundamental problem with banning hate speech: any definition of hate speech is too vague, and ends up capturing speech that, while perhaps unsavory, is not really hateful. The reason for this is that the First Amendment prevents governments from banning specific viewpoints, so governments must phrase their hate speech policies broadly enough that they aren’t seen as targeting specific people. But general definitions of hate speech, such as the one I gave above, will inevitably be over-broad.
Skeptics of this argument might point out that I gave a very broad definition of hate speech, and that governments could try and write more detailed policies that really only covered speech that was hateful. Fair enough. In fact, though, governments have tried that. As I mentioned above, the University of Michigan’s hate speech code had a much more detailed definition of hate speech than the one I provided. When it was challenged in court, the judge still found it to be unconstitutionally vague. The same thing happened to every campus speech code that was challenged in court on First Amendment grounds.
(There is an interesting story here about Stanford University’s speech code, which was scuttled on First Amendment grounds despite the fact that Stanford is a private institution. I may write a subsequent blog post about that.)
It may be that hate speech, like hard-core pornography, fame, and religion, is something we can identify when we see, even if we can’t define it. I’m sympathetic to this viewpoint. As I said above, we don’t need a definition of hate speech to know that calling for racial groups to be killed is hateful. The “know it when we see it” approach to hate speech could probably work moderately well in our private ordering of the bounds of acceptable speech. (This is similar to the “reasonable person” approach I advocated for in my post about microaggressions.)
When we talk about the First Amendment, though, we are talking about what speech the government can sanction. We need to be very careful about the discretion we grant to the government to delimit the bounds of acceptable speech. The “know it when we see it” approach to hate speech simply gives government officials too much discretion to call speech they don’t like hateful.
To illustrate this point, think of the dispute between people who assert that “black lives matter” and those who claim that “all lives matter.” Each group sees themselves as stating a simple, undeniable truth, and that the other side is being intentionally inflammatory. The “black lives matter” crowd thinks that when people assert that “all lives matter,” they are really saying that black lives don’t matter. For their part, the “all lives matter” group think that those saying “black lives matter” sow division and discount the value of non-black people.
In reality, with the exception of disingenuous uses by a few provocateurs, both statements are probably meant as anodyne assertions of truth. To its proponents, “black lives matter” centers black people because they are the ones whose lives are in danger. “All lives matter” means to its champions that all people are worthy of equal dignity and respect. Both statements appear to come from a place of love, not hate.
Nonetheless, one can easily imagine either group, if in control of the government, banning the other group’s slogan as hate speech. They might do so out of a cynical desire to restrict the speech of people they disagree with. But importantly, they might do so out of a genuine belief that they are combatting hate.
The “know it when we see it” approach to hate speech, just like attempts to define such speech, is too broad, and will inevitably capture speech that is not actually hateful. That is why hate speech must continue to be protected by the First Amendment. To do anything else would grant the government the power to arbitrarily ban speech. There are plenty of means of combatting hate speech that we should readily employ, but government censorship is not one of them.
Chauvinisms 