This is a crosspost for When to Suppress Speech by Michael Huemer, which was originally published on Fake Noûs on 9 August 2025.
Here, I address when private agents should try to suppress undesirable speech.*
[ *Based on: “When to Suppress Speech,” Georgetown Journal of Law and Public Policy 20 (2022): 825-43. ]
1. The Question of Private Suppression
My question: When is it desirable for a private agent to suppress speech due to disagreement with the content of that speech?
Examples:
- Google firing James Damore for suggesting that men and women might choose different occupations due to different preferences, and that Google had an ideological echo chamber that prevented open discussion of the issue.
- College students shouting down Charles Murray when he is invited to speak on campus.
- Georgetown University firing adjunct professor Sandra Sellers for mentioning to another professor that black students often do poorly in class.
- Twitter and Facebook banning President Trump from their platforms in January, 2021.
Not my question: when the government should suppress speech; when private parties are or should be legally allowed to suppress speech.
2. The Case for Suppression
Two reasons why you might want to suppress speech:
A. Psychological Harm
Some people today claim that words are “violence” and that people are harmed by hearing the wrong political views. One theory is that certain kinds of bad political speech cause stress, and stress causes actual medical harms; hence, bad political speech is kind of violence.
This would be a convenient rationale for suppressing dissent. Naturally, my political views wouldn’t cause any stress since they’re all true and beneficial, and nobody would be stressed out by hearing such wonderful, sensible ideas. But other people’s political ideas, if they disagree with me, are obviously wrong, and a sensible person could easily get stressed out by hearing such horrible, oppressive ideas. So people who disagree with me should probably be silenced. What could possibly go wrong? There’s no way that dogmatic ideologues would abuse this rationale by claiming to be stressed out every time an opponent makes arguments that they have no good answer to.
In response: There is indeed evidence of medical harms from severe stress, such as that caused by child abuse or other serious trauma. However, there is no evidence of any harms resulting from hearing political views that one disagrees with. That’s just complete bullshit. In fact, hearing views you disagree with is probably good for you. It probably makes you more tolerant and resilient and helps you better hone your own views. Telling people that they need to be protected from hearing ideas they disagree with is more likely to cause stress by creating groundless fears. People need to learn to live with disagreement in a liberal society.
That said, I do agree that personal insults and threats lack cognitive value and are usually suppression-worthy.
B. Political Harm
The other problem with incorrect political speech is that it may persuade people, which could result in worse public policy.
Traditional liberal reply: We need a free interchange of ideas to figure out what the best policies are in the first place. The best ideas will tend to emerge from open competition.
Counter: Sure, maybe this works most of the time. But sometimes, lies have unfair advantages over truth. E.g., lies can be optimized for entertainment value or to appeal to people’s prejudices, and it can be harder to debunk the lies than it is to spread them. Perhaps, therefore, we should suppress lies about objective matters of fact but not sincere mistakes or expressions of opinion.
3. The Case for Freedom
A. The Risk of Error
If you’re trying to suppress controversial political speech, there’s a high chance that your side is wrong and you are suppressing truth. Consider:
- The fact that the issue is controversial already means people are unreliable about it.
- Political beliefs are especially irrational. There is empirical evidence that they are partly genetic and tied to general personality traits, that people evaluate political information in a highly biased way, and that this class of beliefs functions more as a group affiliation signal than an attempt to describe reality.
- The church suppressed Galileo; the Nazis suppressed dissent; the communists suppressed dissent; Nixon tried to suppress Daniel Ellsberg; etc. In history, how often are the people suppressing speech the good guys? The answer is almost never. If you’re suppressing speech, you’re probably the bad guys.
B. Entrenched Opposition
The result of suppressing ideas you disagree with is not that everyone then agrees with you. That might work if your side has control of all information sources, like the government of North Korea. But it doesn’t work in a liberal society with many information sources. In our society, the result of suppressing ideas in certain forums is that (a) the people who disagree with you increasingly get their information from other sources, so they get siloed into separate ideological subcultures, (b) they grow to distrust and resent the forums that suppress their views. E.g., say that right-wing views are suppressed in academia. Right-wing people don’t then say, “Oh, I guess right-wing views are wrong.” Rather, they say, “Oh, I guess we can’t trust academia. We need to get our information from right-wing think tanks. Or Fox News. Or the manosphere.”
Both sides wind up worse off because they stop interacting with each other, and both sides become more extreme.
C. Intellectual Corruption
If we have a widespread practice of speech suppression, people who have social power in a given organization will increasingly use it to suppress those with less social influence. The range of suppressible opinions will expand to include any idea that the dominant group cannot or does not want to debate. The dominant ideology, even if it started out generally correct, will become increasingly extreme, simplistic, and dumb.
D. Penumbral Suppression
Another problem with speech suppression is that you wind up suppressing a lot of speech that you weren’t trying to suppress. The rules for what we suppress are often super-vague (like: we suppress “offensive” ideas), and many people (esp. academics) are highly risk averse. So many will just decide to avoid all discussion of certain broad areas, such as race, gender, or IQ.
E. The Foundations of Liberal Order
People disagree, a lot, and about important things. This is a threat to social order—many societies descend into violence due to normative disagreement. The way we address this problem in a liberal society is through discussion. That’s the alternative to violence. Speech-suppressors are trying to end discussion of important matters of controversy, thus undermining the foundations of liberal social order.
4. For Limited Censorship
We’ve seen reasons both for and against suppressing bad speech. The reasons for suppression apply most strongly to lies about matters of objective fact. The reasons for supporting free speech apply most strongly to sincere opinions. So we should probably suppress the former but not the latter.
Example: Facebook, Twitter, and other forums have banned QAnon content. This is appropriate since QAnon content is just lies about objective facts: the authors know that they are not in fact high-ranking government officials with top secret information about a ring of satanic pedophiles. Banning this type of content does not incur large costs, since (referring to the considerations from sec. 3 above):
a) There is no significant chance that we’re wrong and that QAnon is actually telling the truth;
b) QAnon is fringe content that can actually be pretty effectively suppressed, unlike, say, mainstream Republican views;
c) There is no real intellectual value gained by actively discussing QAnon content, and there is relatively low risk that banning QAnon-type content will lead us to banning things that are worth discussing;
d) Banning this type of lie about objective facts is relatively unlikely to lead to accidental suppression of adjacent ideas that are worth discussing;
e) It’s also unlikely to undermine liberal social order, since the QAnon lies are not things that needed to be discussed or that people were thinking about prior to QAnon’s inventing them.
Notice that matters are quite different when it comes to suppressing mainstream or moderate conservative ideas, such as that women must be biologically female, or that men and women have biologically-based psychological differences, or that different races have different average abilities. These are sincerely held opinions, discussion of which has cognitive value.
Not to be naive, there is no guarantee that open discussion of ideas will lead to the truth’s winning out. False, even disastrous ideas can triumph for decades (see the case of Marxism). The problem of bad political speech leading to political harm is real.
It’s just that speech suppression doesn’t fix the problem; suppression tends to make the problem even worse, for the reasons discussed in sec. 3 above. You’re more likely to get dumb, extremist ideologies winning out.
As flawed as the marketplace of ideas is, it is the best thing we have.
5. What About Completely Unreasonable Ideas?
Many people would accept the case for free speech in general but make an exception for very harmful and completely unreasonable ideas, such as white nationalism. Surely, one might think, there is little to be lost by suppressing such ideas, and little progress to be gained from engaging with such completely unreasonable people. Is there?
Consider the case of Adrianne Black (formerly Derek Black), who was once considered a rising star in the white nationalist movement. She was the child of Stormfront founder Don Black and the godchild of former KKK Grand Wizard David Duke. She went to college with firm white supremacist beliefs, then encountered some Jewish students who invited her to Shabbat dinners and talked with her about her beliefs. Over a period of one to two years, they converted Black away from white nationalism through reasoned discourse. Black wound up renouncing her former beliefs and apologizing for her role in the white nationalist movement.
Or consider the example of Daryl Davis, a black musician who has had many conversations and even friendships with KKK members. Davis reports having directly convinced 40 to 60 people to leave the Klan and indirectly caused about 200 to leave. He converted the leader of the KKK in Maryland away from racism through patient dialogue. For more, see the documentary Accidental Courtesy.
There are many cases like these—people who have been converted away from hate through dialogue. These conversions do not happen through someone’s berating the person, silencing them, or trying to ruin their life. They usually happen through personal interaction with a member of the hated group and patient, reasoned dialogue. Hate is not defeated by more hate; hate is defeated by love. Bad ideas are not defeated by silence; bad ideas are defeated by better ideas.

I think if someone is running a blog, it should be socially acceptable to ban people from commenting for almost any reason, including just finding someone annoying. According to the definition used in this article, this counts as "suppression of speech". Maybe it is in the literal sense, but I don't think smuggling in the bad feelings associated with government censorship is fair.
Or say you are s run a fish and chips shop, and it turns out the person you hired at the front is an open racist who drives customers away by telling them how much he despises albanian people. Are you meant to sacrifice your own money and livelihood for the sake of "protecting the man's speech"?
People have a right to curate their spaces for their actual needs. The questions become thornier in a case like college campuses, because academic debate and discussion is part of the needs of such an institution. Organisations have to determine the pros and cons of what they allow people to say on their platforms.
Thanks for comment, titotal. Agreed.
Could you say more about the relevance you perceive to the theory and/or practice of effective altruism?
Part 1 frames this significantly in terms of actions by major corporations (that includes big universities) exercising their power, especially tech companies. On the one side, I'm less worried about the risk of individual people and relatively small social movements effectively suppressing speech than I am about big tech companies or even "ordinary" major corporations doing so. To the extent that an EA actor is powerful enough to achieve some suppression, the actual effect on the suppressed speech's ability to obtain a hearing should be pretty minimal.
On the other side, I find the interests of individuals and social movements in distancing themselves from speech they find odious to be much greater than for large corporations. Some of that is about association -- if you're trying to create a certain sort of community, tolerating problematic speech is going to impede that.
So the balance implied by the linked material -- which to be fair, wasn't written with the EA community in mind -- doesn't strike me as particularly helpful for the types of "suppression" decisions that individual EAs, EA actors, and the EA community are likely to face.
Thanks for the comment, Jason.
I was thinking the post could be helpful to make people reflect about whether they are someway elevating/suppressing content based too much on agreement/disagreement, and too little based on whether it could update views.