Censorship: a 21st-century approach

"What's wrong with the workers of the world uniting?"

Censorship is a human universal. Censor is just Roman for “moderator.” Even the most uncensored platform in the galaxy, 4chan, has its janitors—also Latin of vast antiquity, originally meaning one of the bouncers in one of Ostia’s many leather bars. Regardless of etymology, history tells us, it is never a great idea to flip the bird at human nature.

No: the problem of a 21st-century digital discourse is not the abolition of censorship. It is the perfection of censorship.

The hate clause

Substack just dropped an interesting little post about their approach to moderation. Recently I wrote up our own code of conduct which I expect commenters here at Gray Mirror to follow; it was originally paywalled; lo, I have lifted that. It’s instructive, and of course very important for all of us here on Substack’s platform, to compare the two.

Our code of conduct is of course more stringent than Substack’s—so any content that would not be okay with them, is not okay with me. Briefly, in the laconic language of my own GenX college days: all Gray Mirror content must be neither un-PC, nor PC.

Naturally Substack only cares about the former, which is obviously fine. (PS: is there anyone who thinks “woke” and “PC” aren’t a single thing? Not a rhetorical question.) The rest of the Substack guidelines are mostly perfect, in my opinion, except, well:

We do not allow hate, defined as publishing content or funding initiatives that call for violence, exclusion, or segregation based on protected classes. This does include serious attacks on people based on race, ethnicity, national origin, religion, sex, gender, sexual orientation, age, disability, or medical condition. It does not include attacks on ideas, ideologies, organizations, or individuals for other reasons, even if those attacks are cruel or unfair. 

While I find this text carefully and elegantly written, given the moment for which it is written, it does violate Gray Mirror’s speech guidelines—because it deploys the “tongue of Mordor,” newspeak or jargon of power. The Orwelltastic use of “hate” is particularly piquant and hopefully will not return to bite anyone present as the pure predator it is.

With such great beasts in the room, we stand with our weight on the balls of our toes; we are completely alert; we are the bartender in the saloon on the edge of a gunfight. Or are we the villain? Or the hero? We can live this way. Do we have to?

But—what would be the right way to put this? We all know what Substack’s policy means. We all know it is the right policy—or at least, part of that right policy. But…

If the above is not the right way to put it—what is the right way to put it? And if it is the right policy for Substack (which it probably is), why does it matter how it is put?

A statement of law

The fundamental problem with Substack’s official text is that it should be a statement of law, but it is only a gesture of loyalty.

It works, because we can infer a real rule from it. But we cannot understand this text literally—only by deconstruction. But we are all trained in this form of deconstruction. Nevertheless, the exercise is draining. Every time we do it, our brains must bend a bit.

Try for a moment to understand this paragraph as a statement of law—a sort of building code for prose, to which we writers may refer whenever we fear we may be putting in an outlet too close to the shower, or whatever. Try to take it literally. You’ll fail at once.

For instance, what is a serious attack? Is the word serious in this context the antonym of humorous, or the synonym of severe? The latter, probably—what then is the measure of this severity? One might as well post a speed-limit sign which just said “not too fast.”

Exclusion is similarly vague; is it the antonym of inclusion, a piece of the jargon? Could we read this text to prohibit writers from calling for quarantine, a form of exclusion, of human beings with Covid, a medical condition? In fact: how could we not?

And even segregation has enjoyed quite a few meanings in its short fiery career; people commonly refer to Boston and SF as “segregated” cities. They have a point; though it is not a point anyone in Atlanta in 1921 would recognize.

If you live in SF and oppose population exchange between the Sunset and the Bayview, does that make you a segregationist? Of course, no one has ever thought of proposing any such thing. Yet. It is the way SF’s public schools work, though. So if you live in SF, can you promote a return to “neighborhood schools” on your stack?

We see that the predatory power of this “hate” clause, though it appears to be safely confined to a tiny freakshow in the dexter fringe, is actually rather poorly contained. As usual, the human factor is the only thing that keeps the tiger in its pen. With some people in charge, this text may be a harmless and prudent precaution that can shield the whole discourse. Under others, it becomes a whip to chastise the whole discourse.

So why post any rules at all? Just take a cue from the ancient Roman bouncers, whose imperium was absolute if local, who only had one rule—“Maximus Thrax can always throw you out.” In the end Substack is its own judge and its own janitor; it seems to wear those robes with some seriousness; let us pray it wears the jargon more lightly.

But why are we here? Since this clause is so useless as a statement of law, what is it?

A gesture of loyalty

A gesture of loyalty is any text that, like the famous greengrocer’s slogan in Havel’s The Power of the Powerless, reduces to a pure expression of obedience:

The manager of a fruit-and-vegetable shop places in his window, among the onions and carrots, the slogan: “Workers of the world, unite!” Why does he do it? What is he trying to communicate to the world?

Is he genuinely enthusiastic about the idea of unity among the workers of the world? Is his enthusiasm so great that he feels an irrepressible impulse to acquaint the public with his ideals? Has he really given more than a moment's thought to how such a unification might occur and what it would mean?

I think it can safely be assumed that the overwhelming majority of shopkeepers never think about the slogans they put in their windows, nor do they use them to express their real opinions. That poster was delivered to our greengrocer from the enterprise headquarters along with the onions and carrots. He put them all into the window simply because it has been done that way for years, because everyone does it, and because that is the way it has to be.

If he were to refuse, there could be trouble. He could be reproached for not having the proper decoration in his window; someone might even accuse him of disloyalty. He does it because these things must be done if one is to get along in life. It is one of the thousands of details that guarantee him a relatively tranquil life “in harmony with society,” as they say.

Obviously the greengrocer is indifferent to the semantic content of the slogan on exhibit; he does not put the slogan in his window from any personal desire to acquaint the public with the ideal it expresses. This, of course, does not mean that his action has no motive or significance at all, or that the slogan communicates nothing to anyone. The slogan is really a sign, and as such it contains a subliminal but very definite message.

Verbally, it might be expressed this way: “I, the greengrocer XY, live here and I know what I must do. I behave in the manner expected of me. I can be depended upon and am beyond reproach. I am obedient and therefore I have the right to be left in peace.”

This message, of course, has an addressee: it is directed above, to the greengrocer’s superior, and at the same time it is a shield that protects the greengrocer from potential informers. The slogan’s real meaning, therefore, is rooted firmly in the greengrocer’s existence. It reflects his vital interests. But what are those vital interests?

Let us take note: if the greengrocer had been instructed to display the slogan “I am afraid and therefore unquestioningly obedient,” he would not be nearly as indifferent to its semantics, even though the statement would reflect the truth. The greengrocer would be embarrassed and ashamed to put such an unequivocal statement of his own degradation in the shop window, and quite naturally so, for he is a human being and thus has a sense of his own dignity.

To overcome this complication, his expression of loyalty must take the form of a sign which, at least on its textual surface, indicates a level of disinterested conviction. It must allow the greengrocer to say, “What’s wrong with the workers of the world uniting?”

Thus the sign helps the greengrocer to conceal from himself the low foundations of his obedience, at the same time concealing the low foundations of power. It hides them behind the facade of something high. And that something is ideology.

Havel’s text was written almost 50 years ago in a country 5000 miles away. If anyone still thinks it has nothing to do with the way we live now… I despair.

What’s wrong with protecting the marginalized against racism? Don’t answer that. And if you think getting along in life is hard—try getting along in business...

When we turn Havel’s magic sunglasses on the Substack text, it translates all too easily into: “we are afraid, and therefore unquestioningly obedient. You too must be afraid—and therefore, unquestioningly obedient.”

Which is true. This is a correct and appropriate message which needs to be sent. I just don’t like the tone. I think the right tone is: “kids—you know the rules. Follow them.”

Because you definitely do want “a relatively tranquil life in harmony with society.” And yet: you have no interest at all in “concealing the low foundations of power.”

The path of the righteous man under Caesar’s sway is to obey Caesar without gestures of loyalty. Or, of course, of disloyalty. (One funny thing, by the way: often, statements of disloyalty are permitted, where gestures of disloyalty are not.)

Your RA does not have any strong opinions on the merits of certain plants. He does have strong opinions on the rules. His strong opinion is that they exist, and are for all practical purposes made by God—for all power, the RA believes, belongs to God. For Heaven is ruled by God. His viceroy, on this Earth, is the Dean. His sergeant, in this dorm, is the RA. The RA finds your libertarian opinions nifty. He would love to chat about them sometime. It smells like incense in here. Burn some pot to hide the smell.

Such rules are the rules of power. Some are written. Some are not. You learned the real rules in elementary school like everyone else—so why would anyone put on a kabuki, writing as if no one had ever heard of any “protected classes” before their employer, after lengthy and expensive philosophical and ethical cogitation, emitted this linked list of victim-sets whom no one may “attack” or be “cruel or unfair” to.

Spare us! We know what a “protected class” is. We know what the list of them is. That list changes day by day and is written nowhere—any copy is a stale cache entry. You don’t actually expect your writers to refer to it. And they won’t. So why such gestures?

The answer is: they are subtle, semi-conscious gestures of loyalty to power. Voluntary obeisance, like scratching a mosquito bite or licking chapped lips, always feels good. And always soon demands more to feel better. Of course the same is true of rebellion, and its gestures of disloyalty—which also work exactly like a little drug.

The RA’s attitude—of meticulous yet dispassionate compliance—keeps us out of both these deadly traps. Both of which can, in his timeless words, “really fuck up your life.” Let’s look more closely at the narrow path between them.

The new faith

Here is a critical aspect of “soft totalitarianism” that Havel for some reason overlooks. While it is fine to put the premade poster in the window, what power really wants you to do is to handwrite your own poster. That way, it feels fresh; it feels yours; it feels like exactly what it is not.

Yet ordering you to do this would be crude. You must do it voluntarily. To pass from the involuntary to the voluntary is to pass from compliance (which is always good) to collaboration (which is always bad). When you collaborate, you support power in two ways: first, by doing what it wants; second, by humiliating yourself more deeply.

The only moral resolution for the collaborator is to actually invest spiritually in power. Faith always follows humiliation—this is why the gesture of proskynesis is eternal. To kneel before something sacred or powerful is the normal human expression of faith, a healthy human emotion that the 20th century shunted into its political enthusiasms.

My son is a competitive soccer player, so we watch a lot of Premier League. Among other cool, Baudrillardian 2020 innovations, like posters of fans in seats and fake crowd noise, and inexplicably-English slogans like “Come On You Baggies,” every Premier League game begins with what I can only describe as a prayer to the New Faith. (Possibly this is only on the American broadcasts—but I doubt it.)

The New Faith has no positive identity. The prayer is not a prayer. The religion is not a religion. There is not even a name or a label for it. There is only a name for its absence. Once we label an absence of faith, we can ban it. And when we ban an absence of faith, we impose that faith.

In any philosophical, religious or ideological dispute, there are two sides. Before we can judge the dispute, we must identify each side. Before we can identify a side, we must label that side. If nothing else, its opponent must have given it a pejorative label.

Straightforward, right? Here is one century-old label: eugenics. If eugenics is one side in a long-ago dispute—which it clearly is—what is the other side? Good luck with that. Etymology suggests dysgenics—but no one said that then, or says it now.

The only answer is the New Faith (or its protomodern ancestor)—which has no positive identity, then or now. So there is no specific label for the other side of the dispute.

You will find these pejorative monopoles all over the jargon. A pejorative monopole is like the label “Gentile” as used by Jews and Mormons—it identifies not a system of belief, but an absence of belief.

When as Jew or Mormon we excoriate the uncircumcised “Gentile,” we are reifying and condemning the absence of our own faith. We are not recognizing another faith, or even category of faiths, which is like ours but different. “Jew” is a valid category; “Mormon” is a valid category; “Gentile” is not a valid category.

We cannot expect all Gentiles to have any attributes in common; atheists, Zoroastrians and Muslims are all Gentiles; Jews are Gentiles to Mormons, Mormons are Gentiles to Jews. And when a Mormon regime bans “Gentilism,” it is just imposing Mormonism. And once Mormonism is so powerful, so universal, that it no longer needs a name…

It only takes the first five minutes of Chelsea vs. Liverpool to see that all the snappy slogans of the New Faith, however worded, use pejorative monopoles. They condemn some reified absence of belief in a creed for which no one has a name. If pressed for a positive label, the best anyone can do is just to glue “anti-” onto the ritualized enemy. Often the enemy-label is a historical movement, which once really, tangibly existed, but is now only an abstraction to be universally condemned (or puerilely emulated). But what really matters is the universal religion which is too powerful to even name.

Imagine if Islam, or Scientology, or even Zen, had this kind of intercontinental clout—imagine if one person had it—imagine if every Premier League game began with the players in full-lotus position and a koan on the Tannoy. Not that there aren’t prayers elsewhere—on billboards; on the players’ bodies—but the coolest thing is this:

This prayer, or incantation, or essential public-service message, which the announcers deliver while the players kneel, never has the same text twice. It always says the same thing; but it is reworded every time—so that it feels like a fresh idea that your friendly announcers, whose own faith is of course deep (for retired English footballers lol), just had this afternoon. So far as I know, not even Orwell himself thought of this trick.

Of course, whether or not they handwrite the poster themselves or even come up with their words, the ideology is still delivered from headquarters with the peas and carrots. Yes, this is the environment in which we must learn all over again to write and think. Do not despair, sweet friends! It is not so hard as it may seem.

The secret of compliance is that, while compliance is very important, compliance is not an opinion but an action. You don’t have to talk about compliance—just do it. So long as you do comply, there is no need at all to compromise your dignity in this way. Once you bow and scrape one more time than you have to, you are into collaboration. You are invested. Many tracks lead into the cave of the bear; only the bear’s lead out.

Here the solution is to write the same rule in a different way. You want the effect of the rule; the ideology is delivered from the headquarters with the peas and carrots, and cannot be changed or evaded; it must be followed; how you follow it, for now, remains your own concern. If you have to write it yourself—you also get to write it yourself. Maybe you can write it in a way which effectively complies, yet does not collaborate.

The “vital interest” of Substack, a business, is not becoming the Internet’s headquarters for shit-tier teenage shitposting—and also, not getting herded into a dead century’s dreary old intellectual walled-garden monoculture. If it has to choose one, it will choose the latter. It would very much rather not choose at all—for very sound business reasons. But this is an incredibly difficult line to walk.

Substack (with which I have no affiliation, obviously, other than as a customer—I am certainly not a “Pro” member, I just signed up) is by no means alone in this dilemma. Let’s try to design that approach under which it seems least likely to have to choose.


The only possible strategy for a successful company like Substack is to always take the highest possible road. The only way to win any conflict with the mainstream is to be on a higher road—if possible, a much higher road.

One classic startup strategy is to set the bar not at the present record or just above it, but far above it. That way, your competitors cannot even imagine competing with you. Of course, you must begin by identifying a niche in which this much slack exists. But comparing the present-day marketplace of ideas with past peaks in quality, we do see plenty of slack.

Like any startup, the purpose of Substack is to obtain a monopoly or something like it. As a monopoly on the marketplace of ideas, it becomes a quasi-organ of government. As an organ of government, this new organ would be no less than our nation’s forebrain—the beating heart of the medium. And the medium is the message, right?

If there is one aristocratic tradition that the 20th century never lost, it is the tradition of accepting the responsibilities that are thrust on us. Most of us “elites” are so into this that we neurotically assume responsibilities which are not real. But some are real. It is really not an exaggeration to say that the responsibility of moderating the world’s forebrain may well decide the fate of human civilization.

The obvious way to take the high road is to treat it as a legal problem—not in the usual sense of a compliance problem (if only there was a law to comply with!), but in the sense of a jurisprudence problem for the constitution of a new country of the mind, Substack.

The law of speech and publishing is an extremely old problem—almost as old as law itself. For Substack, this problem is actually a core business challenge. To create a moderation process which is independent, it must create one which is irreproachable. This process itself becomes a reputational moat which no competitor can penetrate. Substack’s moat at present remains far too ethereal for any prospective monopoly.

And if it is not irreproachable in the eyes of the New York Times, it has no alternative but to be irreproachable in the eyes of history. The law of speech is an extremely old problem; the principles of jurisprudence, the law of law, are even older. And all real money and power in this world is created by founding great institutions.

Therefore it is none too bold for Substack, or of course its competitors, to ask from first principles: how should speech be regulated? And in answering this question it must think on the full scale of a state—and with the full dignity of a state.

Thinking like a state

States across history have regulated speech in many ways. In fact, I am not aware of any regime, ever, that has had fully-unregulated speech. If any such regime existed, can we imagine it existing long?

As Americans, we have an advantage. Our history contains multitudes—it has almost literally tried everything. It has certainly tried every way of regulating the press. None of it has worked, but still! And it is quite befitting for any American company to rely on this great American tradition of jurisprudence on the limits of free speech.

For to think like a state is to think like a court. I am not a lawyer, but to me the three most interesting Supreme Court cases in the limits of speech are Brandenburg v. Ohio (1969), Whitney v. California (1927), and Dennis v. United States (1951). The current precedent is Brandenburg—and a fine old precedent it is, for our purposes.

Somewhat inspired by Brandenburg, here is the sort of minimal political-moderation rule I would write to get the same effect as the way Substack’s rule should be enforced in practice, but spoken directly without any kind of Orwellian coding or spin:

We prohibit content that (a) incites concrete lawless action, or (b) menaces or insults, individually or collectively, any protected group of which the author is not a member. Other groups or persons may be insulted. Other abstract groups, but never persons, may be menaced. We also prohibit (c) any form of white nationalism.

This should have a meaning close to the way the Substack paragraph above should be applied in practice by practical moderators in the real world. It is probably impossible to remain in harmony with life, or at least with business, with any less restrictive rules. (On Gray Mirror you may not incite any action, or menace or insult anyone.)

There is no slippery language in this rule. It requires some ancillary definitions—but, to anyone of any party, faction or fringe, who reads it and tries to understand it in good faith, it should mean exactly what it says.

For example: can I say the N-word on my stack? Good luck parsing that out from the text of the official clause. (From the subtext, obviously, it’s obvious.) From my clause, the answer is instantly obvious: no, because it’s an insult to a protected group, and I have no such pass. This is the “stupid simple” for which every timeless lawgiver must strive.

Concrete lawless action

The original phrase is imminent lawless action, from Brandenburg, a clever Warren Court decision which cleared away the last ruins of American anticommunist legislation after one such law—Ohio’s “criminal syndicalism” law—was deployed against some hilariously-stereotypical gap-toothed Klansmen who were literally burning a cross. (If allergic to N-bombs in our highest court’s case law, dear reader, please skip that link. You can guess whom our defendants were fixing to send back to where.)

“Imminent,” to the Brandenburg court, didn’t mean “within the next 48 hours.” It meant you were putting together a real plan to do something concretely illegal. Conspiring to do something concretely illegal is already illegal, so there was not much left of Ohio’s law—originally written, perhaps by Klansmen, to suppress Bolshevik bomb-throwers.

In the ‘60s and ‘70s, this “Skokie strategy” was a brilliant bureaucratic way to clean up the remnants of McCarthyism. By bending over backwards to grant the most liberal of rights to tiny groups of tacky, irrelevant cranks, the courts established precedents that threw the shield of law over their much less irrelevant “activist” friends to the left. Hardly anyone noticed what they were doing. And once the cultural revolution had been completed, all these “Illinois Nazis” could have their “rights” taken away again. Legislating from the bench may be sneaky; at least, it is seldom practiced by dunces.

“Imminent lawless action” is only half of the classic Brandenburg test. The other half tells us that the lawless action also has to be realistic. If your dastardly plan has one unrealistic step, it is an unrealistic plan. So you are not dangerous—just shooting off your mouth, like an asshole.

And it’s not illegal to be an asshole. Maybe it should be illegal to be an asshole? An interesting question—but a different question. Anyway, “concrete” contains “realistic.” So by this definition, according to the latest case law, the boundaries of our sacred First Amendment do protect any speech that does not incite concrete lawless action.

Of course, this baseline is quite inconsistent with the more-censorious cultural values of the 2020s. It is not sufficient to know that the legal speed limit is 65. It is necessary to know the actual speed limit—then encode it in a rule that means what it says. While it would be cool to live in a pure Brandenburg world, we obviously just don’t.

Protected group

Americans always think “protected class” means “protected group.” A normal person might think African-Americans are a “protected class,” which means a class of people who are specially protected, like French nobles before the French Revolution, meaning the law grants them extra protection, and punishes you extra for harming or offending them. Even like the way a Japanese samurai could behead any insolent peasant who looked at him funny. “Let’s all remember that there is no room for insolence in football.”

This sounds logical and seems to correlate with reality. Legally, it is completely wrong. (Of course, “equal protection of the law” is a cornerstone of American jurisprudence.) The “class” in “protected class” is not a class of people. It is a class of motivation. What?

You may notice the words “based on” in Substack’s langue de bois. The truth is that it is not actually against the law to be mean to a black person, or black people as a whole, as though they were French nobles. (By the way, this is also the literal meaning of the word “privilege”—from “private law.”)

Civil-rights law is only concerned with the phenomenon of some people being mean to other people—creating a civil tort where otherwise there would be no cause of action, turning a civil tort into a crime, or aggravating the seriousness of a crime—based on certain protected classes of motivation.

Historically, courts of law did not try to be psychic—when they checked your motive, they tried to keep it to whether or not you had mens rea, malicious or guilty mind—but the 20th century showed up to make history, not to follow it. With what result, we see.

The law does not punish a white citizen more for assaulting a black citizen, “however cruel or unfair,” but only for doing so “based on” the verboten motivation of race. Nor does the law say anything different when the pigments are inverted. Now, in practice…

The grand jurisprudential irony of this bit of legal equivocation is that it dates, like so many of our problems, to the 1960s. And the legal scholars of the 1960s knew this kind of move well, because most of what they were doing was ripping such obfuscations to shreds in the last redoubts of their old enemies. For instance, the old laws of Alabama did not in any way prohibit Alabama’s black citizens from voting. Now, in practice

The jurisprudential innovation in translating this parody of law into its real meaning, writing the real meaning down, and obeying that unironically, is a good example of compliance without collaboration or rebellion. If we take it seriously, we collaborate with power. If we fail to comply with it, we rebel against power. We must do neither.

Of course there are protected people and unprotected people, protected groups and unprotected groups; of course, it is forbidden for the latter to offend the former. This phenomenon of multiple legal status categories is the norm across human history. “Equal protection of the law” was the experiment. That experiment seems to have failed. So why not just describe the real rules, as they are really enforced?

Because the real rules aren’t even that bad. Most of the chafing they produce is not in the difficulty of complying with them—just in the humiliation of believing in them. When we comply without believing, we are “cheating the devil.”

You are certainly free to disagree with a person or group protected from you. You just have to do so with the utmost politeness and respect, as if you were petitioning the King of Thailand. To everyone who isn’t the King of Thailand, you can be your usual asshole self. If you don’t like those rules, you can even talk respectfully about that.

What do you do if a samurai parks in your driveway? You say to him: excuse me, Mr. Samurai-san, your Isuzu seems to be under my gazebo. What do you do if he tells you to fuck off? You fuck off. Yet often, he will actually apologize, and move his vehicle. With nobility comes a natural noblesse oblige—the more so, the less it is questioned. Why question it? And why not be nice to everyone, including kings and samurai?

How hard is it to say what you mean under these rules? For anyone who has anything worth saying, which sadly is not most of us, not hard at all. In fact, some of us are so mushy and over-civilized that we try to apply these same rules to everyone, anyway.

Living in this reality might make you feel like a conquered person. That’s great. You are a conquered person. If you expect this reality to change in some way, why would you not start by acknowledging it?

White nationalism

The label “white nationalist” is one of the least Orwellian terms in the contemporary political jargon. This label means the same thing to its supporters and its enemies; its language corresponds to its content; and its tone is neither pejorative nor laudatory. This makes it easy for any platform to clearly and effectively ban white nationalism.

Let’s be perfectly clear: it’s OK to be white. But explicitly banning white nationalism is a no-brainer for Substack or any similar service—just a matter of basic compliance with America’s post-1965 legal reality. The best way to ban it is to focus on its errors—which is true of any illegal philosophy, whether fascist, communist, Islamist, etc.

The trouble with white nationalism is that, even in its more obscure variants like “white separatism,” it violates political clause (b): it menaces protected groups. The white nationalists are never quite clear on what will happen after our “black-ruled America” is replaced by their “white-ruled America.” Hitler was pretty vague too.

Like: what white-nationalist victory would not be followed by concrete lawless action? Against protected groups? Can we even imagine any such harmless victory at all? If we can imagine it, our dear white nationalists have certainly not made it easy to imagine.

We can still imagine a hypothetical trolley problem in which the only way to save the world is to win the race war. There is no evidence that we live in this world. Even if we did, though, would the first step in our grand, Apollo-tier plan to win the race war be… menacing and insulting protected groups? Wouldn’t we want to… sneak up on them, or something? So even a white nationalist should want to ban white nationalism.

Every movement has a seed of truth. The seed of truth behind white nationalism is the same seed behind Trumpism: that non-elite white Americans are not well-governed. (Yes, Virginia, most Trumpists are horrified by white nationalism.)

This is true—although, actually, nor are elite white Americans. Or yellow Americans. Or brown Americans. Or any Americans, up to and including the American pangolin. We could even go so far as to say that America is not well-governed—though, since our big country has so many different places and such different people, it does manage to be ill-governed in many different ways at once.

The only error in Trumpism lay in thinking that Donald J. Trump could make America well-governed again. Had the Trumpists been right about that, history would have forgiven them everything—everything and much more. History is a stern tutor—and especially unforgiving of those who confuse being right, with wanting to be right. Unfortunately, in this case, the distance between the two was especially galactic.

Not the only error, but the most essential error, in white nationalism is the belief that communal conflict is an invariant of heterogeneous societies. This is not true. It is only true given bad government. Given bad government, many things will go badly.

In general, if we observe that any group of people is being badly governed, the usual cause is a government that is bad in general, and the way to fix the problem is to fix or replace this bad government in general.

In specific, we have no reason to think that the fundamental problem which creates this specific symptom, the misgovernment of poor white people, has anything at all to do with poor white people in specific. The symptom is very real, but its cause is very far upstream; and so is any appropriate fix.

There is still less reason to think that this hypothetical solution must have some zero-sum character—in which whites are governed better, by governing nonwhites worse—or more harshly—or… something?

Alas, this logic is just as irrational as leftists think it is. And its antecedents are easily recognized as mutated versions of 19th-century leftism—until very late in that century, nationalism was always a leftist idea. Abraham Lincoln’s Republicans were both the most leftist ruling party in the world, and the most nationalist.

And as a mutant leftism in a leftist world, the creed remains easy to explain and sell. It is especially easy to explain and sell because it is its enemy’s official enemy—which by itself ought to tell you something.

The right product is the product the customer wants. The marketing strategy behind Hitler’s anti-Semitism was clear enough. Hitler’s customers wanted to fight. So the product they needed was an enemy. Americans do not want to fight. They do not like being in a cold civil war. Very few of them are trench-hardened World War I veterans. The product they need is peace—but, peace without being misgoverned or humiliated.

It is obvious that anyone who is a white nationalist believes that the best life is life in a white-only community. But what the white nationalist really wants is a safe, cohesive and orderly community—such as he would also find in Japan. If Japan wasn’t so racist. “Segregation” is not an end—but a means to this end.

What if there was another means? What if there was a path to safety, cohesion and order that worked well not just for white people, but also for the whole motley crew of genetic mystery-meat which we’ve managed to heap on our long-suffering continent?

White nationalists would probably still be fine with it, if that path just meant “white people run the show.” Which might even work—but what if there was another means? If we white people were well-governed by mutant alien pangolins, what’s the problem? Imperialism works, and what’s good for the goose is good for the gander.

One way to look at the fork between progressives and white nationalists is to look at Kipling, who is closer to the root of that fork than most people think. True: Kipling was the prophet of imperialism and the father of the swastika. But it is notable that where today’s white nationalists tend to be white separatists, who are all about laying down the “white man’s burden,” Kipling urged his audience to take it up—roughly the same attitude, in a sense that would bother both, as the average Teen Vogue subscriber. For there is truth and wisdom everywhere—even in Kipling, and even in Teen Vogue.

If you are a sincere dissident, your goal is simply good government. While not all human beings should be governed in the same way, they all need good government. If your only vision is a vision of good government for some, your vision is too small to govern at all; you are sure to fail; it is both fruitless and dangerous to be around you; you are in the dead end of war. This, not any jargon, is why you deserve to be censored. (But not doxed, punched or “cancelled,” of course.)

Wartime exceptions

So it should be clear that, whether or not it is a political crime, white nationalism is a political mistake—which is the only reason a healthy discourse can survive banning it. You shouldn’t have to ban errors; but you can ban errors.

Still, a healthy marketplace of ideas is not in the business of banning errors. It relies on the adversarial dialectic to filter them out. Why should it need this exception? White nationalists may deserve to be censored; why do they need to be censored?

Why (c), why (b)? Why respect these absurd “protected classes,” whose purported oppressors are these shadowy, imaginary, unconscious conspiracies? Because all power is of God; and all real power must always be respected.

These are simply wartime exceptions for the cold civil war. Power too has its wartime exceptions—its excuses for introducing the rules of violence into the rules of speech. Power has no enemies, only prey; and amongst power’s herbivorous prey, the military imperative is to avoid stimulating the predator’s appetite.

When you conspire against power in an ineffective and performative way which also happens to fit the behavior pattern of power’s typical prey, you behave like a mouse infected with a cat parasite. The rest of the mice simply can’t have that around. Also: we eat seeds, not cats. Now, if the cat has plainly been dead for a while…

In military terms, the strategy of this particular error is a strategy of confrontation. This is sometimes the right move. It is usually the wrong move. In any kind of a war, even a cold civil war, the freedom to make the wrong move must be restricted.

And the worst possible move is to validate the preposterous conspiracies—by acting them out, like a child. Even for the enemies of power, this is giving aid and comfort to the enemy—always a crime in time of war. So this accursed move has no friends at all. Banning it is a mercy to even those who are banned—not that they’ll thank you.

The golden age of red-baiting

What response do we not want to stimulate? The Brandenburg decision was American jurisprudence wielding the sword of peace. For a look at our legal solons in a bloodier mood, check out Whitney v. California (1927), which Brandenburg overruled. It doesn’t get more old-school than this:

The freedom of speech which is secured by the Constitution does not confer an absolute right to speak, without responsibility, whatever one may choose, or an unrestricted and unbridled license giving immunity for every possible use of language and preventing the punishment of those who abuse this freedom, and that a State in the exercise of its police power may punish those who abuse this freedom by utterances inimical to the public welfare, tending to incite to crime, disturb the public peace, or endanger the foundations of organized government and threaten its overthrow by unlawful means, is not open to question. 

The statute does not violate the Equal Protection Clause of the Fourteenth Amendment in penalizing those who advocate a resort to violent and unlawful methods as a means of changing industrial and political conditions while not penalizing those who may advocate a resort to such methods for maintaining such conditions, since the distinction is not arbitrary, but within the discretionary power of the State to direct its legislation against what it deems an evil without covering the whole field of possible abuses.

Looks like our own Supreme Court invented the “paradox of tolerance” when Karl Popper was a little boy. America, f— yeah! The Court says: punch a Commie today. The more things change…

No one could be more different from Brandenburg’s Klansmen than Anita Whitney. It’s still Anita’s world and everyone else just lives in it. The only difference between then and now is that everyone else used to be able to stand up for themselves, a bit. (Not that Anita went to jail, of course, for being the godmother of California communism—not a Whitney.)

Even Mr. Justice Brandeis, representing the new style in American jurisprudence, is too cagey to even dissent. (Mr. Justice Brandeis is certain to have had tens, perhaps even hundreds, of acquaintances in common with Miss Whitney.) But in the Brandeis concurrence, far more famous and far more beautifully-written than the majority, we see an early clever attempt to build out the Brandenburg test:

Even advocacy of violation, however reprehensible morally, is not a justification for denying free speech where the advocacy falls short of incitement and there is nothing to indicate that the advocacy would be immediately acted on. The wide difference between advocacy and incitement, between preparation and attempt, between assembling and conspiracy, must be borne in mind. In order to support a finding of clear and present danger, it must be shown either that immediate serious violence was to be expected or was advocated, or that the past conduct furnished reason to believe that such advocacy was then contemplated.

Bringing Mr. Justice Brandeis (joined by Mr. Justice Holmes) to his real punch-line:

I am unable to assent to the suggestion in the opinion of the Court that assembling with a political party, formed to advocate the desirability of a proletarian revolution by mass action at some date necessarily far in the future, is not a right within the protection of the Fourteenth Amendment.

Many such cases. And we see why the Brandenburg court, 40 years later, is so eager to get these “criminal syndicalism” laws off the books—even after they had long since become dead letters. For the guilty long outflee their own pursuits.

Now the worm has turned, of course! And no doubt the cleverest lawyers of America, from Harvard students to Supreme Court clerks, are thumbing through these old Red Scare precedents today, looking for cool law tricks to wreck the lives of sad boomers. Their targets are not exactly Anita Whitney. Nor will they get off, like the Ohio Klan.

These new solons of a new moral panic might start by consulting Dennis v. US (1951), also thankfully overruled by Brandenburg:

To those who would paralyze our Government in the face of impending threat by encasing it in a semantic straitjacket, we must reply that all concepts are relative.

Obviously, “clear and present danger” cannot mean that, before the Government may act, it must wait until the putsch is about to be executed, the plans have been laid and the signal is awaited. If the Government is aware that a group aiming at its overthrow is attempting to indoctrinate its members and to commit them to a course whereby they will strike when the leaders feel the circumstances permit, action by the Government is required.

Pure balls of fire! Stalin couldn’t have put it more crisply. “Action by the Government is required.” For “the putsch is about to be executed, the plans have been laid and the signal is awaited.” Kids, don’t let anyone tell you the 20th century wasn’t lit. Even here!

But to compare the CPUSA in 1948 (indeed headed by Eugene Dennis, whom I always mix up with Lawrence Dennis, also tried under the Smith Act, just seven years earlier), to any antigovernment movement today, is like—comparing the Iroquois in 1648, to the Iroquois in 1984. There were giants in the earth in those days.

In these days… it’s history repeating as farce, everywhere we look. Sorry, kids.

A positive approach

All this strategy has been purely defensive in nature. The true purpose of the Substack “hate clause,” however it be worded, is just to let Substack stay in business in the ‘20s. Let us wish the company, and its competitors too, the greatest good luck.

Yet sometimes the best defense is a good offense. Instead of asking what Substack has to do to keep its nascent information marketplace from being flushed down the drain, we could ask what it should do—not ignoring the need for self-defense, but thinking beyond it—to make that marketplace greater than any that has come before. And of course, greater than any potential competitor.

Again, this strategy would treat moderation decisions with the seriousness of a state—reflecting the fact that these decisions have a serious impact on peoples’ livelihoods. The right moderation architecture would make the transition from Hume’s barbarous monarchy to his civilized monarchy, inventing its own process and prestige, without ceding any power or influence to the world outside its walls—except, of course, as tactically necessary to hold those walls.

How should this marketplace of ideas be governed? Eh… better to paywall that one, don’t you think? You’ll get it next if you

Ah—screw it. Let’s give it away. (You should subscribe anyway, though.)

The invention of due process

We now consider it absolutely normal for social-media accounts to be dragged out of their houses at night and given the Genickschuss in the full NKVD, Gestapo or Camorra fashion—with no warning or explanation at all. Next morning, they are hanging from a bridge with “404” sharpied on their chests.

Obviously, this would not have happened to them had they not done something wrong. Doubtless even in the Camorra there is some formal process for when you just have to whack a guy; no doubt within the platforms there is a process too. At least when it’s a big account. But nobody not involved in that process will ever learn anything about it.

It is important to understand why all these very different institutions, not to mention Twitter and Facebook, converged on the same opaque, arbitrary process of execution. It’s not just that barbarous monarchy works. It’s that barbarous monarchy scales.

Yet in the end, advancing civilizations seem to rise above mere death-squad justice. Substack, at the intellectual tip of civilization, is ideally poised to take this step, for two practical reasons: as a paid longform service, it has the highest quality and the lowest scale.

Facebook and Twitter need to assassinate thousands of accounts a day. Much of this process is just the automated massacre of industrial bots and sockpuppets. Even when the victims are main accounts of actual humans, and even when the judges are actual humans, the process needs to wade through enormous volumes of subliterate muck at enormous speed. This is inconsistent with any kind of taste, wisdom or transparency.

To ban a professional writer, however, is—or should be—a rare and serious thing. And it demands the dignity of a serious process—meaning a real trial (in many languages, process and trial are the same word) in a real court.

From barbarism to civilization

The approach of creating due process is emotionally attractive due to its symbolism. But anyone who forget the structural motivation of this attraction will design it wrong—much in the way Facebook did.

Probably the most attractive statement in the whole Substack text was:

Our enforcement of these guidelines is not decided by public accusations or pressure campaigns.

Which means: their enforcement of their guidelines is dictated by the company. Which means: as famously with Cloudflare, the last word will always be spoken by Substack’s CEO. (I’m not even sure which of the founders this is.)

This design is precisely Hume’s barbarous monarchy—which can work perfectly well, given the right barbarous monarch. But—can it be improved on? There are two paths out of barbarous monarchy. The first leads to Hume’s civilized monarchy. The second leads to what we have now: oligarchy, an eternal, pervasive, unanimous bureaucracy.

Facebook chose the second. According to its CEO’s own repeated avowals, Facebook never wanted the job of being the censor—the job of deciding what words are true and what words are harmful. Its CEO especially never wanted this duty personally.

Responsibility is always a hot potato. When a supreme authority shuns responsibility, that responsibility is first passed to an internal process, then to an external authority. Responsibility, once externalized, is power—and power will always find a customer.

Thus Facebook has supposedly delegated its authority over moderation policy to an external “Supreme Court,” comprised of—basically, the usual suspects. As Rousseau might say, its will is to become a tool of the general will. Kind of like the last king of Pergamum, who bequeathed his kingdom to… the Roman Republic. You could say he was a cuck. You could say he got the picture.

There is something sad in this result, especially with a founder-CEO who so admires Augustus; but so it goes. In the end, it’s probably the right business decision. Curation, moderation or even governance is not really Facebook’s core competence. And to the extent that it has to be good at censorship, its primary metric is throughput.

For a business like Substack, which due simply to its form-factor of longform text is in the high-art trade of literary publishing, the calculus is very different. Moderation and even curation are much closer to Substack’s core competence. Therefore, giving this job away—either for the CEO, or for the whole company—is a much worse idea.

The civilized monarchy

The idea of separating the executive and judiciary branches, a deeply-seated American prejudice which must have driven Mark Zuckerberg’s desire to give the latter job away, is a late heresy in the Anglo-American legal tradition. Of course, it has triumphed so thoroughly that even the UK now has a “Supreme Court”—honestly just embarrassing, like your little brother imitating you a little too industriously.

In the classical way of government, the executive and judiciary functions were unified. The monarch was the chief judge. The role of the inferior judges was not to challenge the monarch’s will, but to make it systematic and regular. The monarch, in turn, could correct this system and discover exceptions in it.

The first step in transitioning from barbarous to civilized monarchy is not to reduce the power of the CEO, but to expand it, adding structure to compensate for the scale. Order always precedes law. Are Substack’s moderation decisions, today, made in an orderly way? No one can or should know. But if they are not—perhaps they should be. But the final judge of this order must remain the CEO.

The purpose of a corporation’s firewall of secrecy is to wrap the reality of controlled internal chaos in a smooth, convincing exterior. Once you learn how to do something, though, it is no longer chaotic. And it can even be exposed.

Under orderly moderation, the execution of an author is no longer carried out at the sultan’s whim, or still worse the whim of some anonymous revolutionary committee. Rather, the most basic form of procedural justice is introduced: due process of law. There is no punishment without law and trial.

If an ad-hoc internal process evolves into a systematic and regular internal process, that internal process can become a transparent and public process—without any change to its fundamental lines of authority. This is how a barbarous monarchy evolves into a civilized monarchy, and a death-squad into a judiciary.

And as a process becomes regular, it grows easier to calibrate. For instance, Substack has a way to punish its writers that Twitter doesn’t: it could fine them. While this sort of punishment would be disturbing as the arbitrary voice of opaque power, it might seem much more natural as the outcome of a transparent judicial process.

The breadth of justice

While such a judicial system is nice if not necessary for basic moderation decisions, it seems a shame to use it only for these rudimentary and unfortunate necessities.

In a genuinely impressive step, the Substack text deviates strongly from the canonical boilerplate to assure its writers that, so long as they can keep their hyena-paws off Ra’s sacred beeves, they can be as “cruel or unfair” as they like. Here, we feel, is a publisher who gets it—or at least, a platform which gets it.

What we see here is an all-too-rare sense of de minimis non curat lex—the knowledge that mere cruelty and unfairness, however disappointing and even reprehensible, are essential and ineradicable aspects of any human discourse; and they are far too subtle, and far too mild, for any process of procedural justice to address.

But this is hardly the end of the conversation. The old common law had many ways of pursuing justice for cruel or unfair speech; a suit for libel was only the most obvious. These remedies still exist—in theory. Could a platform like Substack arbitrate such a dispute, rather than leaving it to the slow and inept courts?

The new journalism, though substantively above the law, has many good ideas for regulating itself—many of which are, as one would expect in any unaccountable context, better than all existing implementations. One such idea is “fact-checking.” Could a platform check facts, or build a collaborative mechanism for checking facts? Could it even do so in an ideologically nontoxic way?

Writers are a quarrelsome bunch and will take any opportunity for a scuffle. The higher the stakes, the more entertaining. A platform is an entertainment business. Petty fights of every intellectual type, judged seriously and fought for real money, could be a real draw—and no one should underestimate the power of the dialectic.

In general, the place of governance in an information marketplace goes well beyond mere censorship. Governance is gardening; it does not create the plants, or make them grow; but no garden waters or weeds itself.

Reputation and curation

The quality of a social network can only decline. The way this usually happens is that initial quality is very high, because early adopters are almost always quality people.

A similar startup to Substack is Medium, which experienced the same initial-quality effect. For quite some time, the look and feel of Medium carried the same associations that Substack’s now carries: quite simply, the mark of premium content.

This was because Medium was good. But, while good, it had the wrong business model. It kept trying to be either a community or a publisher. All its readers and writers really wanted was a better, more modern Blogger. And eventually, its quality declined. Today I associate Medium with “barely-concealed marketing brochures.”

Substack is well past its early-adopter phase and can expect the same quality decline. The Substack party line is that Substack is only a platform, and every writer is their own product. While this is the right thing to say, hopefully they do not believe it. Positive and negative externalities always leak out across a platform—the average quality of all stacks is of considerable importance to every stack.

And Substack has a huge advantage in booting up a working reputation system. It has a cash leaderboard. While a cash leaderboard is not a perfect proxy for reputation—publishing is no stranger to fake book sales, nor does popularity equal quality—it is a great bootstrap credential.

Authors can do more than author. One simple way in which Substack could convert its current momentum into a sustainable quality moat would be to “close the books,” and make Substack an application-only platform—like, you know, Harvard.

How would you apply to start a stack? Simple: create an account and write your first post, as you do now. That’s your application. But who reads your application? Well—which leading Substack writers do you admire?

Every successful writer is at least a half-competent curator (an easier problem than editing per se). It usually does not require much reward to coax these opinions from them. Writers often enjoy teaching other writers, promoting proteges, and other such behaviors. And writers have always reached publication by impressing their peers.

Of course, this simple filter creates an opening for Substack competitors with no filter at all. Said competitors will develop the brand image of the home of the second-rate—the Android of newsletters, as it were. They can probably charge less of a tax, though.

Why not be a luxury brand? In content it’s especially important to start at the top end of the market—so much easier to expand downward than climb upward. In content the vast majority of the revenue is from the top creators, anyway.

And of course, an entry filter like this is only the start of a full reputation ecosystem—in which raw sales is one input, but only one input. Could we imagine the Oscars of Substack? Why not?

Now, reputation is notoriously hard to get right. And even a reputation system that works does not solve moderation. An effective reputation system will not ban white nationalism for you—it will serve you up the best white nationalism. If you’re not afraid of this, maybe you should be.

But moderating high-quality content remains much, much easier. Andrew Anglin of the white-nationalist Daily Stormer is a very talented writer; he is quite good at what he does; and the quality of his work makes it very easy for anyone to say that his kind of work is not appropriate for a certain venue (here, Cloudflare). The case for censorship is essential to the form; it is not incidental to the execution.

The material purpose of moderation is defense against political attack. And it is much harder to attack high-quality content. The attackers are looking for real facts that fit the pattern of the story they want to tell. High-quality content fits this pattern poorly.

The line between (manual) curation and (algorithmic) reputation is also quite delicate. The sense of purity produced by an impersonal mechanism—like the Oscar vote count—is easy to ruin and hard to restore. And yet at the same time, a gentle thumb or two on the scale, applied discreetly and early on, can really keep the train on its tracks. In curation and reputation, there is no right and easy answer.

And as Dr. Johnson put it, nobody but a fool ever wrote except for money. So I really should have made you subscribe for that—as I normally do for new ideas, solutions, etc. Was it worth it? Our operators are standing by: