Let’s begin by updating Justice Oliver Wendell Holmes for the digital era:

Persecution for Suppression of the expression of opinions seems to me perfectly logical. If you have no doubt of your premises or your power and want a certain result with all your heart you naturally express your wishes in law content moderation and sweep away all opposition.

(The original quotation may be found on page 888 of this worthwhile essay).

Today many people have no doubt about one premise: they know what a lie is. Naturally, they demand that online platforms refuse to run political ads containing lies. Put another way, the platforms should make sure that no one hears false speech. Democratic citizens should not be burdened by the task of making up their own minds about the truth or falsity of speech in political ads online. After all, the people demanding suppression of speech in ads have no doubts that they know what is true and what is false in politics.

Here are a couple of practical problems with all this. In the politics today (and probably yesterday too), the answer to the question: “what is a falsehood?” will be “what the other side says.” But in politics there will be at least two sides to various controversies and maybe many sides. And the proponents of those many sides will all be convinced that the “other side” is lying. In an illiberal society, they will deem their opponents’ lies worthy of suppression. The First Amendment will prevent the government from acting on everyone’s censorious desires. But the First Amendment does not apply to online platforms. What should the platforms do about “lies”?

Let’s consider a hypothetical but realistic case. Imagine the Elizabeth Warren for president campaign wants to buy an ad from Facebook that says, “Two prominent economists—Emmanuel Saez and Gabriel Zucman—have found that inequality has grown rapidly. We need a wealth tax now. Vote for Elizabeth Warren for president.” Seeing the ad, some demand it be taken down because critics of Saez-Zucman have argued that their conclusions involve manipulating data; it is, the critics say, filled with falsehoods. Other people associated with the Justin Amash for president campaign buy a Facebook ad that says, “The studies of inequality come from the Data Fudging School of Taxation Advocacy. Let’s have prosperity not a wealth tax. Vote for Justin Amash for president.” Seeing this ad, Saez-Zucman demand that it be taken down. The ad is, they say, based on lies about their research.

What should Mark Zuckerberg do? I see three choices. He would be well within his rights to refuse to run either ad. Jack Dorsey at Twitter has decided to go down this path. Second, Zuckerberg could take sides and run the Warren ad or the ad of its critics. In short, he (or a fact-checking organization) could decide which one is true and protect Facebook’s users from the other one, the lie. Third, Zuckerberg could do what he has done: he could run both ads and leave it to those who see the ads on Facebook to decide their truth or falsity and the implications of either answer. Zuckerberg has selected the most liberal choice. He has refused to decide whether Piketty-Saez-Zucman or their critics have the best of the argument. Facebook has placed that power and responsibility to decide with its users.

That trust in individuals to read, think, and decide marks much of classical liberalism and its philosophical descendants over the past couple of centuries. That trust is one foundation of our First Amendment and our politics. Yet Zuckerberg has been roundly criticized for refusing to run ads said to include falsehoods. We live at a moment when liberalism, the philosophical foundation of our society, is unpopular.

From this point things get more complicated. Remove the candidates from the above examples and Facebook’s policies change. In theory, Facebook could send Saez-Zucman or their critics to fact-checkers and refuse to sell ads to whoever fact-checkers deemed the liar. In practice, neither Saez-Zucman or their critics are likely to end up at the fact-checker since the latter already has a full agenda dealing with obvious hoaxes and conspiracy theories. As a matter of free speech theory, however, the logic of Zuckerberg’s decision about political ads applies to almost all political speech, paid or unpaid. So users should be left to decide the truth or falsity of all speech, including Saez-Zucman­­­, their critics, obvious hoaxes, and conspiracy theories.

But leaving the obvious hoaxes and conspiracy theories up may alienate many users and spook advertisers, driving shareholder value down. What appeared at first to be a conflict between liberalism and its critics is actually a tension within liberalism which values both political and economic liberty. The challenge of digital liberalism will be to accommodate both kinds of liberty while keeping government and politicians at bay. Perhaps those most given to denouncing Facebook would do better to keep in mind the complexities of the challenge they are struggling with.

Meanwhile, everyone should keep in mind that the facts at issue in free speech controversies are not like “Stanford University is (is not) north of San Francisco.” They are more commonly like the facts at issue between Saez-Zucman and their critics. The facts at issue in such struggles may become established after “more speech,” but are, for now, “essentially contested.” One thing is clear though. We should never allow such controversies to be settled by people who never doubt their premises and want a certain result with all their heart.