The Internet Platforms and Press Freedom

Our longstanding theories of press freedom prioritize decentralization and competition, under the principle “one person, one voice.” Press freedom is threatened all over the world today by rising authoritarian governments like those of Russia and China, and by corrupt oligarchs patterning themselves on Silvio Berlusconi to advance their own corrupt interests. But it has also been threatened by the large internet platforms—of which there are just three, Twitter, Google, and Facebook—who have often claimed that they are working to support freedom of the press.

It is true that the internet’s seemingly limitless bandwidth allows its users to follow an immense diversity of interests and passions, some on a microscopic scale. In this respect it is very different from severely bandwidth-limited legacy media like over-the-air broadcast media, and in this sense the press is freer than it has ever been in human history. But the large platforms also have the opposite effect: they are able to spread the same information with a speed and on a scale that no prior technology could remotely match. This phenomenon, known as virality, allows information to be targeted far more precisely than earlier forms of media. The large platforms are the beneficiaries of the same surveillance data to which the Chinese government has access; their advertising models allow them to match data to the specific preferences and characteristics of their users in unprecedented ways. But these technological capabilities have created huge problems for the practice of politics in liberal democracies, as was demonstrated in the Cambridge Analytica scandal.
___STEADY_PAYWALL___
In addition, the large platforms have become the primary channels by which individuals have access to information. If they decide to take down certain content, they have effectively silenced those voices relative to others that they choose not to block.

The infectious disease metaphor suggested by the term “virality” is a good way of understanding how the contemporary internet works. A virus replicates by coming into close contact with a susceptible individual. Memes on an internet platforms spread in a similar way, greatly facilitated by the platforms’ possession of user data which gives them knowledge of who is susceptible to certain kinds of information. The memes then replicate rapidly through sharing and likes. No one forces users to behave in this manner; it appears to them as voluntary choice, but actually is based on a sophisticated behind-the-scenes manipulation on the part of the platforms.

The internet platforms make use of a particular understanding of human cognition. The standard model underlying both modern natural science and the liberal Enlightenment more generally holds that human beings were rational: they observe an empirical reality external to themselves, made causal inferences from those observations, which allows them to act upon the world based on the theories they have developed. To the contrary, Jonathan Haidt and other social psychologists have suggested that many people follow a very different cognitive model. They do not begin with a neutral observation of empirical reality; rather, they begin with strong preferences for the reality they prefer, and use their considerable cognitive skills to select empirical data and devise theories that support that reality. In this respect, the large internet platforms’ ability to match information to preconceived preferences is unparalleled, and leads to massive reinforcement of prior assumptions.

Normally, if one’s preferred version of reality diverges from actual reality sufficiently, there will be an ultimate reckoning: one won’t get the job, or arrive at the correct destination, or protect oneself from disease. But here too, modern information technology has done many other things to interfere with people’s cognitive landscapes. Increasingly, we do not directly interact with the outside world by touching, feeling, walking, or talking with other people; these activities are mediated by screens that present us with avatars of that outside reality. Our social connections have spread far beyond the close circles of family and friends that we associated with a generation or two ago. Computer-generated simulations of reality have grown unbelievably realistic over time, and have blurred people’s sense of what is real and what is a simulacrum. This is nowhere more so than in the online gaming world, or in the fantasist world of Hollywood superheroes, which occupy a huge and growing proportion of young people’s time. In that world, one does not have to live with the body or social identity one was born with, and there is little accountability for one’s actions in this anonymous world. The fear of death, which normally forces us to limit risky behaviors like reckless driving or doing violence to other people, does not exist in the online world.

This then constitutes the technological backdrop to the present-day situation in the United States, where the opposite sides of the political polarization do not simply disagree on ideologies and policy preferences, but see different versions of reality. The normal institutional sources of validation for the factual belief like electoral administrators and courts have all indicated that Joe Biden was the clear winner of the 2020 election, but the desire to believe that Trump had won was so strong in the minds of his supporters that a large majority refuse to be persuaded. If one goes on the internet, there will be a mountain of information sources that will validate the conclusion the election was stolen. There are also plenty of sources seeking to discredit that view, but if one desperately wants to believe the Trump narrative, one is not going to carefully weigh the evidence pro and con. People have always been taken in by conspiracy theories and extremist views, but the large platforms have allowed this material to infect much broader parts of the population. As a result, a significant part of the democratic electorate is living in a fantasy world that accords with their desires rather than any kind of objective reality.

There is a final way in which digital media threatens freedom of speech, that goes under the heading of “cancel culture.” Cancel culture is not the same thing as accountability for speech; there are indeed speech acts like incitement to violence or criminal activity that have dire real-world consequences and should be limited or punished. Cancel culture refers instead to legal speech acts that offend the dignity of a particular person or identity group, and lead to real-world consequences like loss of a job or other forms of ostracism.

The internet has certain unique characteristics that distinguish it from legacy media. It operates on a scale and with a speed that is unprecedented in earlier forms of communication. It is public and easy to access, and has a long memory; an action or statement made decades ago and subsequently disavowed or atoned for lives on perpetually. It rewards instantaneous reaction rather than careful deliberation, leading to instantaneous mob-like reactions that would not occur if information cycles were slower and people forced to deliberate for longer. Voices on social media like Twitter are not necessarily representative of the views of the broader population. Its users tend to be activists of various sorts, in whose hands it concentrates a power to shame. This power is further amplified by the quasi-monopolistic control that a large platform like Amazon has over commerce, where it controls over 70 percent of all new book sales. If Amazon decides to de-list a book, which it has done with Ryan Anderson's When Harry Became Sally, it has serious consequences for the visibility of the book’s underlying ideas.

The internet provides people with an outlet for their feelings about social justice, while relieving them of the need to actually bring it about. Achieving social justice in a liberal democracy is a hard task: it begins with popular mobilization, which requires raising people’s consciousness about injustice on issues like race, gender, disability, or other conditions of discrimination. But it then requires moving from mobilization to action: someone must formulate policies and laws to remedy the situation; elections need to be contested, victories won, and governing majorities formed; legislators need to be persuaded to devote resources to solution; the policies need to be litigated through the courts; and policies then need to be implemented on a large scale. Many of these stages require persuading fellow-citizens who do not initially agree with the social justice issue at hand, which might in turn require adapting one’s objectives. The internet has allowed people to mistake initial speech acts for acts that ultimately affect outcomes in the real world.

Freedom of speech has an intrinsic value as an expression of individual autonomy, and a social value as the public sphere in which communities can understand common problems and deliberate on collective responses. It is threatened when control over speech is concentrated in the hands of a small number of actors. Classically, this has been governments, but today it is also concentrated in the hands of private parties that use speech to protect their private interests. The internet has interacted with the cognitive revolutions that have taken place in recent decades, revolutions that have undercut the authority of the institutions that previously structured popular understandings of factual reality. Legitimate criticisms of the hegemony of modern science and Enlightenment modes of thought that started on the progressive left have been taken to heart by the right, and pushed to extremes where no traditional source of authority is believed. The internet has permitted that vacuum to be filled by anyone with a computer, promoting what are often fantasist views that are completely untethered from reality. Under these conditions, speech may be nominally free, but it has lost its essential function as a mechanism for deliberation and democratic decision-making.