The majority of censorship is self-censorship

…But the majority of self-censorship is intentionally cultivated by an outside power.

Cory Doctorow
9 min readFeb 22, 2024


I’m on tour with my new novel The Bezzle! Catch me TONIGHT in SAN DIEGO (Feb 22, Mysterious Galaxy). After that, it’s LA (Saturday, with ADAM CONOVER), Seattle (Monday, with NEAL STEPHENSON), Portland, Phoenix and more!

A yellow rectangle. On the left, in blue, are the words ‘Cory Doctorow.’ On the right, in black, is ‘The Bezzle.’ Between them is the motif from the cover of *The Bezzle*: an escheresque impossible triangle. The center of the triangle is a barred, smaller triangle that imprisons a silhouetted male figure in a suit. Two other male silhouettes in suits run alongside the top edges of the triangle.

I know a lot of polymaths, but Ada Palmer takes the cake: brilliant science fiction writer, brilliant historian, brilliant librettist, brilliant singer, and then some:

Palmer is a friend and a colleague. In 2018, she, Adrian Johns and I collaborated on “Censorship, Information Control, & Information Revolutions from Printing Press to Internet,” a series of grad seminars at the U Chicago History department (where Ada is a tenured prof, specializing in the Inquisition and Renaissance forbidden knowledge):

The project had its origins in a party game that Ada and I used to play at SF conventions: Ada would describe a way that the Inquisitions’ censors attacked the printing press, and I’d find an extremely parallel maneuver from governments, the entertainment industry or other entities from the much more recent history of internet censorship battles.

With the seminars, we took it to the next level. Each 3h long session featured a roster of speakers from many disciplines, explaining everything from how encryption works to how white nationalists who were radicalized in Vietnam formed an armored-car robbery gang to finance modems and Apple ][+s to link up neo-Nazis across the USA.

We borrowed the structure of these sessions from science fiction conventions, home to a very specific kind of panel that doesn’t always work, but when it does, it’s fantastic. It was a natural choice: after all, Ada and I know each other through science fiction.

Even if you’re not an sf person, you’ve probably heard of the Hugo Awards, the most prestigious awards in the field, voted on each year by attendees of the annual World Science Fiction Convention (Worldcon). And even if you’re not an sf fan, you might have heard about a scandal involving the Hugo Awards, which were held last year in China, a first:

A little background: each year’s Worldcon is run by a committee of volunteers. These volunteers put together bids to host the Worldcon, and canvass Worldcon attendees to vote in favor of their bid. For many years, a group of Chinese fans attempted to field a successful bid to host a Worldcon, and, eventually, they won.

At the time, there were many concerns: about traveling to a country with a poor human rights record and a reputation for censorship, and about the logistics of customary Worldcon attendees getting visas. During this debate, many international fans pointed to the poor human rights record in the USA (which has hosted the vast majority of Worldcons since their inception), and the absolute ghastly rigmarole the US government subjects many foreign visitors to when they seek visas to come to the US for conventions.

Whatever side of this debate you came down on, it couldn’t be denied that the Chinese Worldcon rang a lot of alarm-bells. Communications were spotty, and then the con was unceremoniously rescheduled for months after the original scheduled date, without any good explanation. Rumors swirled of Chinese petty officials muscling their way into the con’s administration.

But the real alarm bells started clanging after the Hugo Award ceremony. Normally, after the Hugos are given out, attendees are given paper handouts tallying the nominations and votes, and those numbers are also simultaneously published online. Technically, the Hugo committee has a grace period of some weeks before this data must be published, but at every Worldcon I’ve attended over the past 30+ years, I left the Hugos with a data-sheet in my hand.

Then, in early December, at the very last moment, the Hugo committee released its data — and all hell broke loose. Numerous, acclaimed works had been unilaterally “disqualified” from the ballot. Many of these were written by writers from the Chinese diaspora, but some works — like an episode of Neil Gaiman’s Sandman — were seemingly unconnected to any national considerations.

Readers and writers erupted in outrage, demanding to know what had happened. The Hugo administrators — Americans and Canadians who’d volunteered in those roles for many years and were widely viewed as being members in good standing of the community — were either silent or responded with rude and insulting remarks. One thing they didn’t do was explain themselves.

The absence of facts left a void that rumors and speculation rushed in to fill. Stories of Chinese official censorship swirled online, and along with them, a kind of I-told-you-so: China should never have been home to a Worldcon, the country’s authoritarian national politics are fundamentally incompatible with a literary festival.

As the outrage mounted and the scandal breached from the confines of science fiction fans and writers to the wider world, more details kept emerging. A damning set of internal leaks revealed that it was those long-serving American and Canadian volunteers who decided to censor the ballot. They did so out of a vague sense that the Chinese state would visit some unspecified sanction on the con if politically unpalatable works appeared on the Hugo ballot. Incredibly, they even compiled clumsy dossiers on nominees, disqualifying one nominee out of a mistaken belief that he had once visited Tibet (it was actually Nepal).

There’s no evidence that the Chinese state asked these people to do this. Likewise, it wasn’t pressure from the Chinese state that caused them to throw out hundreds of ballots cast by Chinese fans, whom they believed were voting for a “slate” of works (it’s not clear if this is the case, but slate voting is permitted under Hugo rules).

All this has raised many questions about the future of the Hugo Awards, and the status of the awards that were given in China. There’s widespread concern that Chinese fans involved with the con may face state retaliation due to the negative press that these shenanigans stirred up.

But there’s also a lot of questions about censorship, and the nature of both state and private censorship, and the relationship between the two. These are questions that Ada is extremely well-poised to answer; indeed, they’re the subject of her book-in-progress, entitled Why We Censor: from the Inquisition to the Internet.

In a magisterial essay for Reactor, Palmer stakes out her central thesis: “The majority of censorship is self-censorship, but the majority of self-censorship is intentionally cultivated by an outside power”:

States — even very powerful states — that wish to censor lack the resources to accomplish totalizing censorship of the sort depicted in Nineteen Eighty-Four. They can’t go from house to house, searching every nook and cranny for copies of forbidden literature. The only way to kill an idea is to stop people from expressing it in the first place. Convincing people to censor themselves is, “dollar for dollar and man-hour for man-hour, much cheaper and more impactful than anything else a censorious regime can do.”

Ada invokes examples modern and ancient, including from her own area of specialty, the Inquisition and its treatment of Gailileo. The Inquistions didn’t set out to silence Galileo. If that had been its objective, it could have just assassinated him. This was cheap, easy and reliable! Instead, the Inquisition persecuted Galileo, in a very high-profile manner, making him and his ideas far more famous.

But this isn’t some early example of Inquisitorial Streisand Effect. The point of persecuting Galileo was to convince Descartes to self-censor, which he did. He took his manuscript back from the publisher and cut the sections the Inquisition was likely to find offensive. It wasn’t just Descartes: “thousands of other major thinkers of the time wrote differently, spoke differently, chose different projects, and passed different ideas on to the next century because they self-censored after the Galileo trial.”

This is direct self-censorship, where people are frightened into silencing themselves. But there’s another form of censorship, which Ada calls “middlemen censorship.” That’s when someone other than the government censors a work because they fear what the government would do if they didn’t. Think of Scholastic’s cowardly decision to pull inclusive, LGBTQ books out of its book fair selections even though no one had ordered them to do so:

This is a form of censorship outsourcing, and it “multiplies the manpower of a censorship system by the number of individuals within its power.” The censoring body doesn’t need to hire people to search everyone’s houses for offensive books — it can frighten editors, publishers, distributors, booksellers and librarians into suppressing the books in the first place.

This outsourcing blurs the line between state and private surveillance. Think about comics. After a series of high-profile Congressional hearings about the supposed danger of comics to impressionable young minds, the comics industry undertook a regime of self-censorship, through which the private Comics Code Authority would vet comings for “dangerous” content before allowing its seal of approval to appear on the comics’ covers. Distributors and retailers refused to carry books without a CCA stamp, so publishers refused to publish books unless they could get a CCA stamp.

The CCA was unaccountable, capricious — and racist. By the 60s and 70s, it became clear that comic about Black characters were subjected to much tighter scrutiny than comics featuring white heroes. The CCA would reject “a drop of sweat on the forehead of a Black astronaut as ‘too graphic’ since it ‘could be mistaken for blood.’” Every comic that got sent back by the CCA meant long, brutal reworkings by writers and illustrators to get them past the censors.

The US government never censored heroes like Black Panther, but the chain of events that created the CCA “middleman censors” made sure that Black Panther appeared in far fewer comics starring Marvel’s most prominent Black character. An analysis of censorship that tries to draw a line between private and public censorship would say that the government played no role in Black Panther’s banishment to obscurity — but without Congressional action, Black Panther would never have faced censorship.

This is why attempts to cleanly divide public and private censorship always break down. Many people will tell you that when Twitter or Facebook blocks content they disagree with, that’s not censorship, since censorship is government action, and these are private actors. What they mean is that Twitter and Facebook censorship doesn’t violate the First Amendment, but it’s perfectly possible to infringe on free speech without violating the US Constitution. What’s more, if the government fails to prevent monopolization of our speech forums — like social media — and also declines to offer its own public speech forums that are bound to respect the First Amendment, we can end up with government choices that produce an environment in which some ideas are suppressed wherever they might find an audience — all without violating the Constitution:

The great censorious regimes of the past — the USSR, the Inquisition — left behind vast troves of bureaucratic records, and these records are full of complaints about the censors’ lack of resources. They didn’t have the manpower, the office space, the money or the power to erase the ideas they were ordered to suppress. As Ada notes, “In the period that Spain’s Inquisition was wildly out of Rome’s control, the Roman Inquisition even printed manuals to guide its Inquisitors on how to bluff their way through pretending they were on top of what Spain was doing!”

Censors have always done — and still do — their work not by wielding power, but by projecting it. Even the most powerful state actors are not powerful enough to truly censor, in the sense of confiscating every work expressing an idea and punishing everyone who creates such a work. Instead, when they rely on self-censorship, both by individuals and by intermediaries. When censors act to block one work and not another, or when they punish one transgressor while another is free to speak, it’s tempting to think that they are following some arcane ruleset that defines when enforcement is strict and when it’s weak. But the truth is, they censor erratically because they are too weak to censor comprehensively.

Spectacular acts of censorship and punishment are a performance, “to change the way people act and think.” Censors “seek out actions that can cause the maximum number of people to notice and feel their presence, with a minimum of expense and manpower.”

The censor can only succeed by convincing us to do their work for them. That’s why drawing a line between state censorship and private censorship is such a misleading exercise. Censorship is, and always has been, a public-private partnership.

If you’d like an essay-formatted version of this post to read or share, here’s a link to it on, my surveillance-free, ad-free, tracker-free blog: