On Tuesday night, at the fourth Democratic presidential debate, former California attorney general Kamala Harris had a question for Massachusetts senator Elizabeth Warren. Prompted by moderator Mark Lacey to address Facebook CEO Mark Zuckerburg’s resistance to regulating the tech industry, Harris switched tabs. Turning to Warren, she said,

Senator Warren, I just want to say that I was surprised to hear that you did not agree with me that on this subject of what should be the rules around corporate responsibility for these big tech companies, when I called on Twitter to suspend Donald Trump's account, that you did not agree, and I would urge you to join me.

Harris connected the question to Warren’s ongoing battle with Facebook, over their lax protocols on political disinformation. It was a matter, Harris said, “of holding big tech accountable, because they have an outsized influence on people's perceptions about issues and they actually influence behaviors.”

“So, look, I don’t just want to push Donald Trump off Twitter,” Warren responded. “I want to push him out of the White House. That’s our job.”

Harris looked wounded. She tried again.

“Well, join me,” she pleaded. “Join me in saying that his Twitter account should be shut down.”

“No,” said Warren.

“No?” asked Harris. She turned her head in disbelief. “Wow.”

Warren tried to shift the discussion to structural concerns, citing antitrust laws and campaign finance reform. But Harris had difficulty letting go of her Twitter feud. “Maybe I’m just kicking in as a former prosecutor,” she said on MSNBC after the debate. Her involvement in this issue had begun with something like a subpoena. In early October, Harris wrote an open letter to Twitter CEO Jack Dorsey, under a letterhead that reads like the first words of a courtroom opening statement: “Kamala Harris — For The People.”

“I write to call your attention to activity that President Trump has been engaged in on his Twitter account,” it read, “which appears to violate the terms of the user agreement that your company requires all users on the platform adhere to.” Harris had previously tweeted a request at Dorsey to take action against Trump, arguing that the president was engaged in targeted abuse of an administration whistleblower, in violation of Twitter’s terms of service.

Hey, @jack. Time to do something about this. https://t.co/es2fJ5PNRV

— Kamala Harris (@KamalaHarris) October 2, 2019

The White House responded in a statement to CNBC, saying:

It is not surprising that Kamala Harris, someone who believes in bigger government and more regulation, would like to silence her political opponents. In fact, it’s rather authoritarian of her. President Trump’s use of technology to communicate directly with the American people and share his Administration’s unprecedented accomplishments should be praised, not criticized.

As for Harris, she’s tweeting through it.

Trump's tweets incite violence, threaten witnesses, and obstruct justice. We can't crack down on Facebook but turn a blind eye to Twitter. Big tech companies must be held accountable for how they allow him to abuse their platforms. #DemDebate

— Kamala Harris (@KamalaHarris) October 16, 2019

Harris isn’t the first to make such a request of the Twitter CEO. In fact, personally asking Jack Dorsey to enforce bans online has pretty much become a meme. On any tweet by Dorsey, who posts from the handle @jack, you are likely to find replies demanding to know why the platform has not booted out the many Nazis who troll its furthest reaches.

Ban the Nazis. You could do it tomorrow but you choose not to.

— Kevin M. Kruse (@KevinMKruse) March 17, 2019

It’s not an unreasonable question. People are banned for far less extreme infractions than calling for genocide or making palpable threats against specific individuals. Twitter’s banning policy seems designed primarily to protect the rights of famous users to say rude things and prevent the rest of us from being rude to them in return.

“No user, regardless of their job, wealth, or stature should be exempt from abiding by Twitter’s user agreement,” Harris wrote in her letter. But Twitter sees it the other way around. A June post on Twitter’s blog included a confession. “In the past, we’ve allowed certain Tweets that violated our rules to remain on Twitter because they were in the public’s interest,” the company admitted.

A critical function of our service is providing a place where people can openly and publicly respond to their leaders and hold them accountable. With this in mind, there are certain cases where it may be in the public’s interest to have access to certain Tweets, even if they would otherwise be in violation of our rules.

The stated goal is “to strike the right balance between enabling free expression, fostering accountability, and reducing the potential harm caused by these Tweets.” Twitter is protected here by Section 230 of the Communications Decency Act of 1996, which establishes platforms and users as separate entities. “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider," it reads.

Twitter has recently elaborated on the specifics of its moderation policy. Rude tweets by politicians will now come with a notice: “The Twitter Rules about abusive behavior apply to this Tweet. However, Twitter has determined that it may be in the public’s interest for the Tweet to remain available.”

Ironically, Twitter’s latest move helps them solve a nagging problem. According to an anonymous employee who spoke to VICE, when the company has tried to form algorithms designed to ban Nazis, there has been a side effect: their filters inadvertently include Republican politicians.

In theory, it seems clear that Nazis should not be permitted to incite violence against ethnic groups, but that the rest of us should be permitted to tell them to fuck off if they do. But in practice, this requires a byzantine process of identification and evaluation of individual tweets and their authors, undertaken by both armies of office drones and algorithmic robot overlords.

Twitter executives recently told VICE about the extent of its regulatory practices, including “content moderation centers spread across the globe, with facilities in Toronto, Tokyo, San Francisco, Manila, Budapest, Warsaw, Dublin, and Bangalore.” This is still dwarfed by Facebook’s far more extensive moderation program, but Twitter considers their different approach both a feature and a bug. "Twitter is really the only public platform in this space," Twitter executive Nick Pickles said to VICE. The company used to call itself “the free-speech wing of the free-speech party.”

But some have called on Twitter’s management to intensify their rules and regulations. A recent op-ed by New Yorker staff writer Andrew Marantz in The New York Times holds the platform responsible for real-world consequences of users’ statements. In Marantz’s view, invoking “free speech” is a “cop-out,” one that is “intellectually dishonest and just as morally bankrupt.”

Marantz points out that “the First Amendment doesn’t apply to private companies.” This fact is often invoked by liberals who support banning right-wing users, as in when InfoWars radio host Alex Jones was banned last year. Anyone can assemble in a public square, but not in a corporate lobby. This means businesses can regulate anyone on their property. “The private sector could pitch in on its own,” Marantz suggests. “Tomorrow, by fiat, Mark Zuckerberg could make Facebook slightly less profitable and enormously less immoral.”

Predictable rejoinders to Marantz from Reason, National Review, Breitbart, The Cato Institute, and New York have argued for a sharper distinction between word and deed. But there is an underlying structure that gets less attention than the more visible matters of principle. Contrary to their claims, Twitter and Facebook are not public platforms; they are businesses run for profit. Nor are they level playing fields. Only certain  users have the power to ban other users. All the rest of us can do is ask.

With freedom of speech enshrined in quite literally the first sentence of the First Amendment of America’s first legal document, it’s unavoidable that our discussion of it endlessly reverts to clichés. The earliest is a misquotation of Voltaire, putting the question in unambiguous terms: "I disapprove of what you say, but I will defend to the death your right to say it." Ambiguity is introduced by Supreme Court Jurist Oliver Wendell Holmes, with the “clear and present danger” qualification: the Constitution need not protect “falsely shouting ‘fire’ in a theater and causing a panic.” (Note that Holmes wrote this in defense of blatant political persecution, in a ruling prohibiting socialists in Philadelphia from distributing leaflets opposing the draft during World War I.) But surely the most relevant aphorism should come not from a French aristocrat or a jurist, but a journalist — and it does. As New Yorker writer A.J. Liebling put it in 1960, “freedom of the press is guaranteed only to those who own one.”

Speech is not an abstraction; it’s an act undertaken in a social setting. You need not only the right to speak, but the means to speak. The metaphor of the “marketplace of ideas,” also invoked by Oliver Wendell Holmes, is an apt one: like economic markets, discourse is determined by structures of power, with stark class divisions. The means and the power to speak are not, to say the least, evenly distributed. In the 2010 Citizens United ruling, the Supreme Court ruled that corporations can speak louder than any individual, with the legal right to spend unlimited sums on the promotion of political agendas.

Online, that power plays out in myriad ways. While it hypoethetically could be used to prevent Nazis from agitating on various platforms, top-down regulation of the internet has also been used to obstruct sex workers, with FOSTA-SESTA, and and to foment political disinformation, as on Facebook in 2016. But the marketplace metaphor is too easily taken as a foregone conclusion. The internet was not born fully formed from the heads of gods; its economic structure was deliberately developed during the age of neoliberalism. In a review for The Nation of Margaret O’Mara’s new history of Silicon Valley, The Code, Adrian Chen traces the internet’s shift from public to private:

The Internet began as an academic resource, first as ARPANET, funded and overseen by the Department of Defense, and later as the National Science Foundation’s NSFNET. And while Al Gore didn’t invent the Internet, he did spearhead the push to privatize it: As the Clinton administration’s “technology czar,” he helped develop its landmark National Information Infrastructure (NII) plan, which emphasized the role of private industry and the importance of telecommunications deregulation in constructing America’s “information superhighway.” Not surprisingly, Gore would later do a little-known turn as a venture capitalist with the prestigious Valley firm Kleiner Perkins, becoming very wealthy in the process.

The result was a field ripe for mining by financial prospectors. As Steven Johnson has written in The New York Times Magazine, “Twenty years after the web first crested into the popular imagination, it has produced in Google, Facebook and Amazon — and indirectly, Apple — what may well be the most powerful and valuable corporations in the history of capitalism.” To understand how this economic structure is embedded into the internet, Johnson suggests thinking of “two fundamentally different kinds of systems stacked on top of each other.”

The underlying layer consists of “open protocols,” which were “defined and maintained by academic researchers and international-standards bodies, owned by no one.” Websites and emails still operate on these protocols; they are a commons. The World Wide Web, on which you are probably reading these words, is a platform so ubiquitous it has become invisible, synonymous with The Internet writ large. There is no CEO of email. But above this layer is “a second layer of web-based services — Facebook, Google, Amazon, Twitter — that largely came to power in the following decade.” Today, instead of a central public sphere, the internet is dominated by competing feudal states.

The result is, Chen concludes, the situation we find ourselves in today: “to carry out almost anything online we must subject ourselves to a hypercommodified hellscape of targeted advertising and algorithmic sorting.”

It is indeed true that, in Marantz’s words, Mark Zuckerburg could change Facebook by fiat, as Jack Dorsey could Twitter. But why should they? Trump is making them rich. Zuckerburg and Dorsey have both been holding private meetings with GOP operatives in Washington, Politico has reported. As a former Twitter employee explained to VICE, there is a financial disincentive to content regulation in general. “If we suspend more people, we lose people, which the growth team doesn’t like, and there is a lot of that,” the former employee said. A profit motive will always prove more compelling to these companies than the public trust.

“Holding big tech accountable,” as Kamala Harris has called for, is a laudable goal. But is requesting a unilateral exercise of power over users the best way to do it? Twitter and Facebook have both shown a willingness to police when it is in their interests, and turn a blind eye when it isn’t. Elizabeth Warren’s structural approach — breaking up the companies and diminishing their power — points to a different framework. In a response to Warren, Facebook declined to regulate disinformation distributed on its platform, arguing for the imperative to “let voters — not companies — decide.” Why not take them at their word?

Last year, Sarah Jaffe called for the “unrealistic demand” of nationalizing Amazon. Fortunately, unrealistic demands are Silicon Valley’s specialty. A public platform, like the World Wide Web, does indeed fall under the jurisdiction of federal law — just like public space in a city. But so do private platforms. The difference is that a public platform would not be subservient to a second master: the hoodie-clad tyrants who collect an income from every click. As Jaffe writes, “the point of calling to nationalize Amazon is to challenge the idea that Amazon should have more power than the democratically-elected governments of states or countries.”

We could ask Jack Dorsey, as Kamala Harris does, to rule with a firmer hand. But what if we could vote him out of office? That, to use the parlance of the region, could be the first actually disruptive innovation in Silicon Valley history.