Anatomy of a Hatchet Job

The Atlantic's War on Substack

Dear Friends,

I was really hoping to not write too much more about the Substack/Atlantic controversy, but a recent Atlantic piece set me off. I thought it would be useful to have most of the major points of this controversy in one place and on something like a timeline.

Best,

Sam

ANATOMY OF A HATCHET JOB

Let’s go through this in order, shall we. 

Back story: 

- Substack launches in 2017. At that time (between Trump’s election, Charlottesville, etc) there is considerable rethinking of the then-prevailing First Amendment-driven standards for speech on the web. Given the mood of the time, Facebook and Twitter rapidly reverse-course on their existing moderation policies, hiring many thousands of (off-shore, low-paid) content moderators — which, within a few years, morphs into simply using AI for content decisions. 

- Substack from the beginning charts a different course. The Terms of Use, in the ‘Hate’ section, specify a standard prohibiting content that “incites violence based on protected classes” — a high standard that parallels First Amendment jurisprudence and the so-called Brandenburg Test, in which the bar for inadmissible speech is speech that is “likely to incite…imminent lawless action.” In a 2020 post, the Substack founders acknowledge that the trend of the moment is for “a more controlled community that can guarantee safe spaces to all involved” to which they respectfully disagree. “We think the principles of free speech can not only survive the internet, but that they can help us survive as a society that now must live with all the good and bad that the internet brings,” they write. 

- There are controversies from early on. In 2021, Jude Ellison Doyle announces his “intent to abandon Substack” — on the grounds that Substack hosts writers including Graham Linehan and Jesse Singal who are “dedicated to spreading transphobic rhetoric through their work…and use their platforms in such a way as to operate as ‘harassment influencers’ [against trans people].” Linehan had had his Twitter account permanently suspended the year before. Ironically enough, given what is to come, the charge against Singal seemed mostly to stem from a cover story he wrote for The Atlantic called “When Children Say They’re Trans.” A year later, the Center for Countering Digital Hate complains that “Substack does not explicitly prohibit users from spreading misinformation….about vaccines or Covid.” In both cases, a flurry of stories from major publications followed, concerned, as The Washington Post for instance put it, that “Conspiracy theorists, banned on major social networks, connect with audiences on newsletters.” 

Timeline: 

1.Jonathan M. Katz, somewhat like Doyle, becomes increasingly concerned that Substack has departed from his politics. Unfortunately, I can’t quote Katz directly, since he has closed his Substack and thrown up a paywall around his posts, but, paraphrasing from memory, he describes Substack in the old days, c.2019, as being predominantly left-of-center — an ethos that was undermined when a set of “canceled” or contrarian figures (Matt Taibbi, Bari Weiss, Alex Berenson, Robert Malone, Michael Shellenberger, etc) found a home on Substack in the contentious period c.2020-2021. 

2.Katz is enraged by Hamish McKenzie’s decision to host Richard Hanania on The Active Voice in June. Hanania — who is maybe the single most controversial and contrarian of the new sort on Substack — describes himself in the interview as “seeking enlightened centrism” and argues vociferously against “wokeism,” including that “civil rights law….has undermined the integrity of public institutions.” Within six weeks of the interview’s airing, Hanania is exposed for having, under a pseudonym, written racist posts for white supremacist websites. Hanania is widely denounced and there is a sort of guilt-by-association levied against anybody who had ever praised or promoted Hanania. In addition to McKenzie, that includes Elon Musk and J.D. Vance, who had both spoken positively of him, as well as Steven Pinker and Marc Andreeseen, who had been interviewed on Hanania’s podcast. (Ironically again, Hanania had also been published in both The New York Times and The Atlantic.) 

3.Katz pitches an article on what he describes as a “trend” of far-right content on Substack, and The Atlantic is interested in a piece. The piece does not, however, lead with Hanania or with concerns about heretics like Taibbi or Weiss (surprisingly enough, Weiss is now cast in a different role, as an established figure who might share Katz’s misgivings about the free speech-based moderation standards). Instead, Katz goes through the real fringes of Substack and produces a trickle of content that is obviously noxious but is, let’s say, less-than-influential. He names: 

a) The Tribalist, a “pro-white think tank,” which has two posts and a grand total of four likes.

b) Turning Point Stocks, which addresses “the stock market and the Jewish Question,” and which, when I looked a month ago, had zero likes and a single comment on years’ worth of material. (At the moment there are no likes or comments on it at all, but I’m guessing that’s because comments have been disabled.) 

c) Andkon’s Reich Press, which provides a history of World War II from a pro-Nazi perspective. It has since been terminated, but when I looked at it shortly after the Katz piece, it had likes in the single digits. 

d) The People’s Initiative of New England, which as its sole piece of content posted a manifesto calling for secession from the United States out of stated concern for the importation of “racial and cultural strangers.” The manifesto, which Rolling Stone in a separate piece described as a “coming out” for a neo-Nazi organization called NSC-131, garnered a total of six likes. 

e) White-Papers, which offers “pro-white analysis” and does have a following of over 1,000, promotes the racist Great Replacement Theory but does not, it has to be admitted, come close to “inciting” violence.

f) Richard B. Spencer, who was a featured speaker at Charlottesville as well as other neo-Nazi rallies but who claims to have disavowed white nationalism (leading to the uncomfortable question of whether anyone who has even previously been associated with far-right or Nazi views should be considered irredeemable).

g) Patrick Casey, who had been a leader of Identity Evropa, a defunct neo-Nazi organization, and had been banned from Stripe, Substack’s payment processor. Katz’s most damning charge is that Substack allowed Casey to “get back on his feet…getting around Stripe’s ban.” But Jesse Singal (apparently taking time off from “harassment influencing”) fully debunks that assessment here — showing that Casey’s quote “I’m able to live comfortably doing something I find enjoyable and fulfilling” was taken out of context by Katz, that Casey was still heavily restricted by the Stripe ban, that Casey had a modest following of free subscribers on Substack, and that Casey in the quote in question was referring to another source of income. 

4.In its victory lap of an article, written by Jacob Stern, The Atlantic calls Katz’s piece “an investigation.” But even by Katz’s own standards, that would be putting it strongly. He describes his piece as “an informal search of Substack” while also, perplexingly, lumping in “Telegram channels that circulate Substack posts.” In terms of the gross aggregate problem, he produces the odd number of “at least 16” newsletters with “overt Nazi symbols” but leaves it unstated whether some of those might have been mainstream historical sites writing on World War II; and, for his grand tally, proffers the even-less scientific number of “scores” of extremist sites, which could mean just about anything but implies that he found less than one hundred out of the tens of thousands of Substack newsletters. Shalom Auslander’s back-of-the-envelope calculation is that .00145% of the paid subscribers on Substack “subscribe to Nazis.” 

5.Glossing over the scale of what it is actually depicting, The Atlantic’s headline (which I assume was written by an editor rather than Katz) is: “Substack Has A Nazi Problem.” With that, The Atlantic had spoken, and Substack was thrown into great consternation, with the issuing of the Substackers Against Nazis open letter and with the mainstream media somewhat giddily piling in after Substack management decided to stick by its reigning moderation policies. The New York Times reported: “Substack Says It Will Not Ban Nazis or Extremist Speech,” and the tech magazine The Verge, which has been vitriolic in its previous Substack coverage, wrote, “Substack says it will not remove or demonetize Nazi content.” 

6.In all headlines, the words “Substack” and “Nazi” are side-by-side — as terrible a bit of publicity as any organization can ask for. What’s elided over though, in addition to the scale of the problem, is the nature of Substack’s offense. Katz’s Atlantic piece describes Substack as having “terms of service that formally proscribe ‘hate,’” but that’s what a lawyer would call a “term of art” that drastically misstates Substack’s actual policy. In the section on “hate,” Substack makes it clear that the proscribed speech is “incitement to violence,” while beliefs and views, however noxious, are tolerated. To be simplistic, Katz’s accusation would be like accusing someone of jaywalking in front of a traffic light, without stating whether or not the traffic light was green or red.

7.The tech journalist Casey Newton, who is sort of the new-and-improved Katz, leaps into the fray by arguing that “the correct number of newsletters using Nazi symbols…on your platform is zero” and “until Substack, I was not aware of any major US consumer internet platform that stated it would not remove or even demonetize Nazi accounts.” Newton and a “team” from his Substack “Platformer” also submit to management a number of accounts believed to be in violation of the existing incitement standard (which, according to Substack, totals five accounts and less than a hundred subscribers). Newton, like all other critics, seems to be confused about what they’re actually asking for. Is it that Substack had been lax in enforcing its existing Terms of Service, as Katz tried to imply (“We are not—as lazy rumors have it—trying to “deplatform Substack contributors [we don’t] like,” Katz wrote on his Substack)? Is it a concern about Substack’s writer recruitment and promotion, as the Substackers Against Nazis letter stated (“there is a difference between a hands-off [moderation] approach and putting your thumb on the scale,” the letter read)? Or is it, as Newton is contending, Substack’s First Amendment-based moderation policies are untenable given the proliferation of hate on the web?

8.Newton’s argument echoes a point of view that’s prevalent within academia and legacy media — that the inherent nature of the internet renders free speech principles “immature.” In a 2019 New York Times op-ed, Brittan Heller, of the Anti-Defamation League, wrote, “The idea that platforms like Twitter, Facebook and Instagram should remove hate speech is relatively uncontroversial” — which must have been a surprise to two centuries’ worth of free speech advocates within the American First Amendment tradition. In a study on the ascendance of robust content moderation in the period 2016-2023, The Journal of Communication writes: “The goal is not necessarily adjudicating content as more or less acceptable, but moderating it on the basis of evolving and ever contingent public conceptions of accountability.” In other words, robust content moderation was nothing more nor less than profit-driven companies responding to political pressure. But somehow, as Newton’s argument would have it, since other social media companies employ intensive content moderation, with a threshold of “hate,” Substack is atavistic or naive or worse in clinging to the tradition inherited from the First Amendment and Brandenburg v. Ohio. 

9.It’s highly debatable whether the content moderation standards adopted c.2016-17 by Facebook, Twitter, and others actually work — they do, incidentally, make amusing reading, with Facebook adopting such Solomonic rulings as prohibiting images of “fully-nude close-ups of buttocks” unless they are “photoshopped on a public figure” — but what’s been evident enough from the experience of Substack over the past month is that the mere specter of moderation deeply distorts dialogue. In the year-and-a-half that I’ve been on Substack, I have had two trolls in my comments and have hardly ever come across an uncivil exchange. Since I’ve written on Notes to defend Substack’s existing content moderation, I’ve been called “a brownshirt,” “a true warrior for the alt-right,” and a “George Santos scale sociopath” — and this by the people who are concerned about “hate” on the web. The writer of an open letter defending Substack’s position is called, in Jonathan Katz’s comment thread, “a dangerously naive human tool who is being strung along by nefarious techno-zealots….to do their cult recruitment work for them.” Katz, in one of his now-paywalled posts, accuses Substack of having a “pro-Nazi policy,” which goes far beyond any of the “evidence” he produces in his Atlantic piece. And Jude Ellison Doyle, who started the campaign against Substack for being transphobic, currently calls his Substack “Hamish McKenzie Killed Someone” — which is also the title of every one of his posts. 

10.The point here isn’t that these comments are hateful or defamatory (although they are), it’s that they are protected by the same content moderation rules that Katz & co are trying so assiduously to criticize. Once robust content moderation is installed, then, presumably, a whole bunch of other content comes under scrutiny (what about the “Stalinists” on Substack, what about the “Maoists,” and then, inevitably, what about the “anti-vaxxers” or those who generate “unsafe spaces”?) and a tattling culture comes into being, as has happened over the past month, with Substackers incessantly reporting on each other to management.

Incidentally, speaking as a Jewish person, the anti-Semitism that’s concerning in the the world right now is not coming from Andkon’s Reich Press or even from The People’s Initiative with their attempt to “turn New England into a white ethnostate.” It’s from pro-Hamas and over-charged “Free Palestine” rhetoric, which has resulted in at least one attempted pogrom and includes such openly genocidal calls as “From the river to the sea,” with its implied destruction of a Jewish state, as well of the Jews within it. I wouldn’t wish to ban “Free Palestine” rhetoric on Substack, even when it’s inflammatory, but I would be curious to know if those who are at the moment so concerned about “hate” would be as staunch in their views if the same criteria were turned on a left-wing cause. 

11.The Atlantic celebrated its own reportage with a piece yesterday, referring to Substack in the past tense, with the headline “Substack Was A Ticking Time Bomb” and the surprisingly mortuarial line, “This shift….to a no-question-about-it social-media company may have sealed Substack’s fate.” Substack had it coming, is The Atlantic’s claim — “Substack seeded its own moderation crisis” — and what Katz braggadociously refers to as a “mass exodus” of users is what Substack deserves for having naively thought that it could stick to the free speech-principle for content moderation. For The Atlantic — and, here, the publication tips its hand a bit — the basic issue is that Substack “was” always hard to define. “When the site launched in 2017, there was considerable ambiguity about what it even was,” Stern writes. “A media company trying to pioneer a new model of journalism? A social-media company trying to correct the ills and excesses of its predecessors? A modest software for sending out email newsletters?” 

And here, finally, we’re getting close to what this whole issue is really about. There is room for honest debate about online content moderation — “we appreciate that there are reasonable arguments to be made on all sides of these questions,” wrote Substack management in 2020 — but what is dishonest is to claim, as Katz, Heller, and Newton all do, that these questions are somehow already settled and anyone adhering to the free speech standard is therefore, implicitly, condoning extremism. And what is deeply dishonest is to try to depict a platform like Substack as being extremist or far-right simply for trying to do things differently. 

“There is considerable ambiguity about what Substack even was,” Stern writes. Well, yes. Because Substack is something new in the world — and not everybody gets to have been around since 1857. It is part self-publishing, part blogosphere, part social media, genuinely trying to make use of new technology to both monetize writing and to foster a more thoughtful discourse. The legacy media has treated it with foreboding and suspicion, as if it were some strange beast encountered in the jungle — “Why We’re Freaking Out About Substack,” one New York Times piece was headlined — and then, whenever possible, with outright hostility. 

I am not as critical of legacy media as somebody like Matt Taibbi is. Publications like The New York Times or The Atlantic still have immense resources and wherewithal. They attract talent, they cover stories that nobody else is able to cover, and they are among the few places on earth that can make writing professionally viable. But they are publisher and editor-driven. They have their slants and their politics. They have deeply-ingrained institutional limits on what their writers can say and how they can write — and they have disproportionate influence over ‘respectable’ public discourse. 

They are not neutral and they are not arbiters. They are a river, and there is a sea¹ of free-thinking, independent-minded writers, of which Substack is, in a way, only the tip of the iceberg. A Trojan Horse campaign like the “Nazis on Substack” smear is never about what it seems to be — it’s an attempt to discredit an uncomfortable rival bidding for a surprisingly large share of the public discourse. 

On the one hand, there is “authority,” an old and cozy way of doing things, with long-toothed publications calling balls and strikes (and hiding their stake in the argument). And, on the other hand, there is a new way — which rests on a certain looseness —and allows writers to take charge of their own creative lives, to write in their own voices and grow their own following, and to have as much freedom as they can handle. 

Which side are you on? 

1. That rhetorical point is best accompanied by this track (1:13).