Yes, when it comes to using violence to stop speech, even 'rarely' is too much

People can hand-wave this rising trend against free speech on campus, but they do so at their own peril

Every year when FIRE releases our College Free Speech Rankings, we hear a lot of the same arguments and attempts to dismiss them, over and over again. Critics will say, for instance, that our methodology for ranking schools is “arbitrary and misleading,” as Ryan D. Enos did last year for The Chronicle of Higher Education (and which my colleague, FIRE Chief Research Advisor

Sean Stevens

, thoughtfully and decisively rebutted).

Many also criticize our sample sizes at certain schools — arguing, for example, that we only surveyed 100 students at a small liberal arts college, or nowhere near enough at a bigger school like Rutgers University. This, however, betrays either an unwillingness or an inability to understand how surveys and statistically relevant sample sizes function.

Small liberal arts schools typically enroll about 1,000 undergraduates, so 100 is actually a significant percentage (10%). And this is generally around the number we hit. For example, at Claremont McKenna, we surveyed 103 people, equivalent to 7.49% of its 2023 enrollment of 1,379. For larger schools like Indiana, Mizzou, or Rutgers, we sample more than enough to make the statistical analyses we make (573, 530, and 468, respectively, constituting about 1.2%, 2.1%, and 1.16% of their respective undergraduate enrollment). We also weight the data in order to get a representative sample from each school. (Overall, the average number of students surveyed at each institution is 229.) 

If you use a statistically relevant sample size in your research (which we did) and that sample reflects the demographics of the group you’re researching (which it does), then the conclusions you draw from that data are going to be fairly generalizable to the entire group. There are major presidential polls that use smaller sample sizes than ours, and which we still take seriously. Take this ABC News/Ipsos poll, which is based on 2,196 people from a potential pool of about 161 million (the number of registered voters as of 2022). That is a sample size of about .0014%.

This is also beside the fact that ours is by far the largest survey on free speech ever done. In fact, FIRE’s surveys have been the largest ever done every year we’ve done them — including this year, with 58,807 undergraduates across 257 schools. This is more than enough data to gain insight about the state of free speech on campus and student attitudes toward it. How many generalizations do you make and accept about society at large with less than a fraction of that sample size?

Still, the criticism I find most personally unnerving has to do with student tolerance for violence in response to speech. As part of our research every year, we ask students, “How acceptable would you say it is for students to engage in the following actions to protest a campus speaker?” with possible responses being “never,” “rarely,” “sometimes,” and “always.” This year, nearly one-third (32%) said that using violence to stop a campus speech is at least “rarely” acceptable — up from 27% last year and 20% in 2022.

That bears repeating: 1 in 3 students believe that there are times when violence is justified to stop someone on campus from speaking.

Keep in mind that this doesn’t include ostensibly nonviolent but still intimidating and censorial tactics, like obstructing entryways and heckler’s vetoes. We have other stats for student attitudes on that, and they’re just as alarming: More than half of students (52%) reported that blocking other students from attending a campus speech is at least “rarely” acceptable — up from 45% in 2023 and 37% in 2022.

In response to statistics like this, naysayers will seize upon the word “rarely,” as though it’s synonymous with “insignificant.” They wave the number away and argue that overall, given the stats, these students’ responses to disfavored speech are “mostly peaceful.” 

But this is a gross misunderstanding of what these numbers really entail — and what they imply — when it comes to free speech on college campuses.

Don’t forget the ‘sometimes’ and ‘always’ numbers

One thing this dismissal elides is the “at least” bit. In other words, they skip over just how many students in schools across the country answered “sometimes” or “always” in response to this question.

At the University of California, Davis, for example, nearly 1 in 4 students said violence is either “always” (10%) or “sometimes” (13%) an acceptable response to speech. At the University of North Carolina at Greensboro, it’s 7% “always,” 21% “sometimes”; at the College of Charleston, it’s 16% “always,” 14% “sometimes”; at Northeastern University, it’s 5% “always,” 25% “sometimes”. DePauw University, which dramatically improved its ranking last year because it reformed its speech codes, still has a student body where 1 in 4 think violence is “always” or “sometimes” acceptable (8% and 17%, respectively).

Even if we were to cede the argument over the “rarely” cohort, critics would still have to find some way of dismissing these numbers, which indicate that somewhere between 20-30% of students on many campuses think violence is “sometimes” or “always” acceptable to counter disfavored speech.

Another important aspect of these figures is the way that the different cohorts relate to and interact with one another. Consider, by analogy, the pyramid model of radicalization, which shows that the “Radicals” at the tip of the pyramid — in our case, the “violence is always acceptable” group — make up the smallest contingent. This makes sense, as thankfully not everyone possesses the fanaticism and nerve to actually engage in illegal and violent behavior to further their cause.

Below that tip, however, you have the “Activists,” followed by the “Supporters” (akin to the “sometimes” and “rarely” groups from our survey) who “provide a support network” and who “justify the illegal and violent actions.” While these groups don’t themselves engage in violence, they are still aligned with the overall cause of shutting down speech. As a result they often provide intellectual cover by justifying or rationalizing violent or illegal actions, which emboldens the Radicals and indirectly contributes to the overall problem.

This same responsibility also applies to a lesser extent to the “Sympathisers,” the group at the bottom of the pyramid. For our purposes, these are students who may agree with the goal of shutting down a campus speech or silencing a speaker, but who draw the line at using violence to make that happen. While these may seem like mere dissenters (and yes, they do have a right to voice their censorial views) their attitudes still feed the fervor of the Radicals, who may consider Sympathisers to be on board but simply lacking the guts to act.

The overall point here is that the “rarelies” don’t exist in a vacuum: In fact, they make it easier for the “sometimes” and “always” groups — which are far too big already — to grow.

‘Rarely’ is still totally unacceptable

Another thing people conveniently forget is that “rarely” can be a sneakily relative term. Just imagine dismissing these statements on that same basis:

“I rarely beat my wife.”

“I rarely sell government secrets to Iran.”

“I rarely cheat on my taxes.”

It’s quite clear in these instances that “rarely” is an unacceptable state of affairs. A single incident should be considered too often, and negligence toward it should rightly be considered immoral. 

Even with situations like airplane deaths (0.01 deaths per 100 million miles) or food poisoning deaths (3,000 Americans per year, approximately 0.009%), which are too often the result of factors we can’t predict and can’t control, we consider the fact that they happen “rarely” not good enough. In fact, we have scores of people and entire industries solely dedicated to further reducing those numbers, despite knowing how low they already are and recognizing that they will likely never hit zero.

So, no, contrary to the hand-wavers, there are many contexts where “rarely” is still way, way, way too often. Promoting or accepting violence in an institution of higher learning, where engaging in dialogue with diverging and dissenting viewpoints is the whole idea, is absolutely one of those contexts.

‘Rarely’ is more consequential than you think

Speaking of context, there’s much more behind the “rarely” statistic than meets the eye. When students are asked whether they believe violence is acceptable to stop a campus speech from happening, we have to remember that controversial speakers are increasingly not invited on campus in the first place.

To zero in on student attitudes toward disfavored speech, our surveys created eight hypothetical speakers, each of whom would be known for voicing one of the following controversial viewpoints:

• “Abortion should be completely illegal.”

• “Black Lives Matter is a hate group.”

• “Transgender people have a mental disorder.”

• “The Catholic church is a pedophilic institution.”

• “The police are just as racist as the Ku Klux Klan.” 

• “Children should be able to transition without parental consent.”

• “Collateral damage in Gaza is justified for the sake of Israeli security.”

• “From the river to the sea, Palestine will be free.”

Surveyed students were asked whether or not their school should allow these speakers on campus. As the report states, “Overall, a majority of students said that six of the eight speakers should ‘definitely’ or ‘probably’ not be allowed on campus.” In some cases, these percentages were upwards of 75-80%, indicating an overwhelming intolerance for certain views and a chilling climate when it comes to inviting speakers who hold them.

This means that the most controversial speakers are already rarely invited on campus because they face so much pressure due to their viewpoint. This, combined with tactics like shoutdowns and deplatformings, creates an environment that FIRE president Greg Lukianoff and his “Canceling of the American Mind” co-author Rikki Schlott call a “Perfect Rhetorical Fortress.”

Understanding this overall dynamic is necessary for recognizing what “rarely” actually implies when it comes to student attitudes toward whether using violence to stop a campus speech is acceptable. In real-world scenarios, “rarely” may actually mean something like: 

In the rare event that someone like Ann Coulter, Judge Stuart Kyle Duncan, Riley Gaines, Matt Walsh, and Charlie Kirk actually somehow make it through all the social pressure, the Perfect Rhetorical Fortress, the disinvitation attempts, and all the other actions we generally take to prevent disfavored speakers from coming to our campus in the first place, then it is acceptable to respond with violence.

In other words, “rarely” seems to describe how often these speakers get a chance to speak at all.

It should also be noted that the people mentioned above are just a few of the more well-known controversial speakers who have recently dealt with this behavior. There are unfortunately many more — including relatively unknown figures who are not considered all that controversial off campus, but who are nonetheless subjected to this treatment all the time on campus.

Given that, as well as the fact that the only correct answer to the question, “When is violence acceptable to counter speech?” is absolutely never, it should alarm and infuriate us that one-third of college and university students are even “rarely” willing to justify violence in response to speech they don’t like.

And when you realize that the percentage has been rising these last five years, that siren should be deafening.

Words are not violence, violence is not speech, and this trend is bad for all of us

One critical point to make here — which we unfortunately have to make again and again and again — is that there is a very bright line between speech and violence. That line is not only sensible and logical, but also necessary. So necessary, in fact, that without it our entire civilization would inevitably crumble. The foundation of civil society is in our capacity and our insistence on using our words to solve our problems. The fact that a growing number of young people in our country believe otherwise is a warning sign we shouldn’t ignore.

Of course, some will still argue that we’re mistaken. They might say that in those “rare” cases where violence is condoned, the speech being targeted really is ugly, harmful, and wrong — and that the violence really is justified. But the thing is, even if it’s true that the ideas being expressed are pure poison, violence is the least productive imaginable way to go about addressing them.

To invoke Greg Lukianoff’s conception of “Mill’s Trident” here: First of all, you might be wrong. We’re all fallible, and we often don’t know it when we’re completely mistaken. Engaging with dissent is the only surefire way to find that out. Second, even if you’re right, arguing with your opponents will not only give them the chance to change, but also give you the opportunity to sharpen your own position. Third, and most importantly, it’s more than likely that the truth lies somewhere in between your position and theirs and that dialogue and debate is the only real way to find it.

Violence puts a stop to that entire process, and it harms not only its direct victims, but also anyone who might benefit from hearing what they have to say.

In addition to this, there is what I call the self-preservation principle against censorship: Our actions tend to prompt equal and opposing reactions — especially when they feel unfair. When you grant yourself the power to decide what speech crosses the line, and when you give yourself license to resort to violence to stop speech that does, you’re also tacitly granting it to everyone else, including your worst enemies. 

It can be difficult to see this when you’re on campus and seemingly hundreds — or thousands — strong in your views, but you won’t be in power forever. Whatever you do to silence your opponents, your opponents will almost certainly do to silence you the moment they have the chance. And the more unfairly they feel you’re treating them, the harder they will fight to get the chance.

And that’s exactly how “rarely” turns into “sometimes,” and then “often,” before becoming “always.” Introducing violence into the equation only escalates things until it isn’t merely a response to disfavored speech, but the only one.

There are many, many reasons to be worried about the survey results this year, but this is perhaps the most critical: If only out of pure self-interest, the idea of using violence to stop speech, however rarely, should be anathema to us all.

A note from Greg:

I turned 50 this week! I’ve been at FIRE for almost half of my life, and it’s been so incredible to see our little team grow to become the nation’s premier free speech defender!

We’re fighting for free speech in the courtrooms, in the classrooms, and in the culture. ⬇️

👉 Our team has grown to 115 incredible free speech defenders.

👉 We are nearing ONE MILLION followers on social media and email.

👉 We have 30 active litigation cases.

👉 We have secured eight grassroots victories in communities across the country.

👉 We have helped build 35 alumni organizations to defend free speech on campuses.

👉 We have had four model bills passed in states across the country this year alone.

👉 We have released the largest report on college free speech EVER!

👉 We have successfully reformed 31 anti-free speech college policies this year alone.

FIRE has come a long way, but there is still more work to do. That’s where you come in.

Please consider donating to FIRE and joining the free speech movement.

(Consider it a birthday present to me 😉)

Shot for the Road

As ERI readers probably know, I recently published a piece here called “Why the ‘words are violence’ argument needs to die.” But I almost forgot that the very same night I also discussed that topic during our FIRE Student Network summer conference in Philadelphia right before introducing our special guests, Jesse Singal and Katie Herzog of “Blocked and Reported.”

Here’s the video of my remarks, which I think (unfortunately) just keep getting more relevant these days: