Is the internet worse than a brainwashing machine?

A recent article in The Atlantic argues that the internet functions as a transformative "justification engine" rather than a "brainwashing machine." The truth is more complicated.

Is the internet a justification engine?

In trying to make sense of recent political events and trends—whether polarisation, populism, radicalisation, collapsing trust in institutions, or anything else—there is no explanation more alluring to pundits than “social media”.

The explanation’s attractions are obvious.

First, it's difficult to believe that such a profound change in communication technology could fail to have enormous consequences. If you look at seemingly comparable changes throughout history—the emergence of the printing press, radio, television, and so on—they were all transformative. Why would social media be different?

Second, because social media is a primary mode of modern communication, any developments in society will inevitably correlate with changes in the digital sphere. This can make it natural to view the internet as causing those changes.

A third factor exacerbates this tendency: for most pundits, journalists, and academics, social media provides unprecedented exposure to attitudes they would otherwise be—and indeed used to be—sheltered from. Speaking personally, my offline life is dominated by highly-educated, well-mannered liberal and lefty professionals. It's only when I log onto social media that I encounter people who believe in QAnon, think vaccines contain microchips, or view George Soros as a gay Jewish space vampire. For the unreflective among us, this can make it seem like social media conjured such attitudes into existence.

Finally, there's something deeply compelling about simple explanatory stories that appeal to technological changes. In reality, sociopolitical events and trends typically have complex and counterintuitive causes. Many social, psychological, cultural, economic, technological, and institutional factors interact in loopy, non-linear, and highly path-dependent ways. Such processes are literally unimaginable; we cannot mentally simulate them. In contrast, the image of the internet as a kind of technological wrecking ball smashing into society feels intuitively compelling.

Understanding the wrecking ball

Of course, merely pointing to the internet as the cause of modern woes is unsatisfying. There is a need for some explanatory framework to make sense of its impact. And here, there is no paucity of ideas.

For example, perhaps the digital-age explosion of information has overwhelmed traditional gatekeepers, exposed the flaws of elites and institutions, and thereby caused a “revolt of the public” manifest in the surging popularity of populism worldwide. Or perhaps the engagement-maximising character of social media algorithms has exploited the bias-confirming nature of human psychology to trap people within filter bubbles and echo chambers, driving polarisation and extremism. Or perhaps the internet is best understood as a super-spreader of viral “disinformation” and “misinformation” infecting the gullible masses and leading them to reject expert knowledge and support demagogues.

In a recent article in The Atlantic titled ‘The Internet Is Worse Than a Brainwashing Machine,’ Charlie Warzel and Mike Caulfield advance a different explanatory framework. Looking at how Republicans have rewritten the history of January 6 in their collective imagination—minimising or misrepresenting it in highly misleading ways—they argue that this process provides a "powerful example of how the internet has warped our political reality."

However, rather than understanding this through the popular lens of "misinformation” brainwashing a credulous public, they argue that we should view the internet as a "justification machine"—a technology that "is powerful not because it changes minds, but because it allows people to maintain their beliefs in light of growing evidence to the contrary."

In some ways, I'm sympathetic to this argument.

In 2022, I published an article advancing a similar claim: that rather than understanding misleading communication in terms of "misinformation" and public gullibility, we should understand it in terms of what I call a "marketplace of rationalisations". In such marketplaces, people (pundits, intellectuals, journalists, media companies, etc.) compete for social and financial rewards by producing justifications—intellectual ammunition—for the decisions and narratives favoured by different elites and political tribes.

Nevertheless, I disagree with Warzel and Caulfield’s analysis.

On blaming the internet

First, they assign too much weight to the internet in driving processes like the Republican rewriting of January 6. While social media undoubtedly contains an endless torrent of content affirming and rationalising partisan narratives, including the many self-serving myths surrounding Trump’s “Big Lie”, treating this as a "powerful example of how the internet has warped our political reality" assumes such narrative creation would be impossible in a world without social media.

I’m sceptical.

The phenomenon of elites and political tribes rewriting historical events—and reality more broadly—in self-serving ways is as old as recorded human history. There’s nothing new about efficient rationalisation markets. Consider how rapidly the Bush administration built support for the Iraq War through selective and strategic framing and rationalisation, all accomplished primarily through traditional media.

More fundamentally, the marginal impact of social media on the public sphere is far less than commonly assumed. It’s not like the difference between a pre- and post-Gutenberg world. Before social media, humanity already had mass communication through television, radio, and newspapers, and it’s easy to overstate how central social media platforms are in shaping large-scale public opinion today.

For those highly interested in public affairs and extremely online—a group that disproportionately includes highly educated professionals like journalists and academics—it’s easy to forget how unrepresentative we are of the broader population, most of whom don’t engage much with online political content. For example, television is about five times more influential than social media as a news source in the US.

This isn't to deny that social media has important consequences. However, such consequences are complex, often counter-intuitive, and defy simplistic generalisation. Contrary to much conventional wisdom, people—of course, it’s always other people—are not puppets manipulated by algorithms and online content. They’re rational and vigilant agents who use communication technologies in sophisticated and surprising ways. To understand such technologies’ effects, then, you must explore their subtle interactions with people’s pre-existing interests, ideas, and allegiances and the broader cultural and political context they inhabit.

This explains why many major political trends are moving in different directions in countries with similar rates of social media use on similar platforms. If you unleash the affordances and structure of social media in a highly polarised country like the United States where the left/right divide increasingly maps onto an establishment/anti-establishment divide, you’ll get radically different outcomes than in societies with alternative political environments.

The efficiency question

Warzel and Caulfield acknowledge that the drive to acquire rationalisations of preferred beliefs and narratives is older than the internet. However, they argue that the internet is nevertheless a game changer:

"The justification machine… didn't create this instinct, but it has made the process of erasing cognitive dissonance far more efficient. Our current, fractured media ecosystem works far faster and with less friction than past iterations, providing on-demand evidence for consumers that is more tailored than even the most frenzied cable news broadcasts can offer."

This claim about efficiency seems intuitive but confronts several objections.

The golden age

First, it misrepresents the pre-internet world. Warzel and Caulfield argue that before the internet, finding rationalisations “has not always been so easy”. Hence,

"Evidence foraging might historically have meant digging into a subject, testing arguments, or relying on genuine expertise. That was the foundation on which most of our politics, culture, and arguing was built."

As with much discourse about the internet, this analysis assumes a mythical pre-digital golden age.

Writing in 1922, Walter Lippmann observed, "If we believe that a certain thing ought to be true, we can almost always find either an instance where it is true, or someone who believes it to be true." Over a century and a half before that, Benjamin Franklin observed, “So convenient a thing it is to be a reasonable creature, since it enables one to find or make a reason for everything one has a mind to do.”

It’s always been easy to produce and acquire justifications of beliefs favoured by elites or large communities. From witchcraft accusations to millenarian cults to major world religions, political ideologies, and national myths, humanity has developed many highly efficient methods for promoting, protecting, and justifying bespoke realities.

In addition, Warzel and Caulfield’s emphasis on the internet’s allegedly unique "attention economy"—an environment where “people are rewarded with engagement and greater influence the more their audience responds to what they’re saying”—is highly misleading.

The real currency of the internet is not attention but status. However, the role of status and status competition in driving intellectual activity and governing the division of intellectual labour is ancient. Long before social media, pundits, journalists, writers, artists, intellectuals, priests, and many other content creators competed for influence and esteem by validating the preferred narratives of audiences.

Efficiency cuts both ways

Moreover, consider what's actually involved in finding useful justifications online. You’ve got to sift through vast amounts of bullshit, evaluate the credibility of unfamiliar sources, process contradictory claims and arguments, and—most importantly—confront more conflicting evidence than you would typically encounter offline.

In many ways, it’s probably easier to maintain partisan narratives in a pre-digital era when you could simply turn on Fox News or read The Guardian without encountering a torrent of opposing viewpoints. That is, the internet's efficiency at delivering rationalisations is counterbalanced by its efficiency at providing criticism and opposing perspectives.

The real difference maker

If there is something driving the ease with which partisans in America can protect and justify favoured narratives today, it has less to do with the internet per se than with the fact that Republicans and Democrats trust radically different sources of information. This means that contradictory evidence can be dismissed simply by attributing it to untrustworthy sources. To be concrete, if Republicans trusted mainstream media and establishment institutions, rewriting the history of January 6 would have been impossible.

This collapsing trust in such institutions among the American right is a significant and troubling development with many epistemological consequences. However, the vast amount of false and misleading right-wing content that flourishes on social media platforms today largely reflects this development; it doesn’t drive it.

Rationalisations for thee, evidence for me

Finally, Warzel and Caulfield’s analysis of the demand and supply of justifications focuses almost exclusively on Republicans. That is, except for a passing comment on how liberals managed to convince themselves that Trump was going to lose the recent election, their analysis of how misleading information functions to rationalise biased narratives is very one-sided. In this respect, the article is typical of the vast majority of discourse and commentary on topics like “misinformation”, “disinformation”, and the informational pathologies of social media.

When Republicans embrace narratives about election fraud or rewrite the history of January 6, this is presented as a clear case of disinformation, motivated reasoning, and rationalisation. And it is. But what about when Democrats and their preferred media outlets select, omit, frame, package, and comment on reality through their preferred narratives? Is that simply the noble pursuit of wisdom and understanding?

Consider, for example, how mainstream media selectively amplifies and interprets certain kinds of police violence whilst omitting others in ways that support narratives about systemic racism. Or how extreme weather events are covered and framed in ways that reinforce a maximally alarmist view of climate change. Or how almost the entire liberal establishment selected and amplified stories about Russian interference in ways that delegitimised Trump’s 2016 presidential victory.

For perhaps the clearest example of how a marketplace of rationalisations is not restricted to the political right, consider the creation and amplification of the “social media is destroying society” narrative within academia and mainstream media over the last several years. The coverage of this topic involves an endless stream of evidence, arguments, and interpretations designed to paint social media in the most negative light imaginable. (A recent Atlantic article on social media by Warzel is titled ‘I’m Running Out of Ways to Explain How Bad This Is’).

It’s not difficult to see why such a narrative appeals to many within the liberal establishment. It offers a simple explanation for complex social problems. It demonises the main rival to legacy media institutions. It coheres with pre-existing beliefs about the gullibility of the masses. It absolves established institutions of responsibility for declining public trust. Most importantly, it affirms and validates the epistemic superiority of traditional gatekeepers.

Reflexivity

To be clear, I’m not trying to draw a moral or epistemic equivalence between the world of MAGA and the liberal establishment. As I have argued repeatedly, my own view is that the Republican Party’s cult-like attachment to Trump and estrangement from establishment institutions mean that its errors are far more frequent and egregious than those of liberals.

Nevertheless, such differences reflect the relative epistemic quality of information sources preferred by America’s political tribes, not variation in the strength of demand for intellectual ammunition that rationalises biased narratives.

More generally, any serious analysis of the role of rationalisation markets needs to practice what sociologists call "reflexivity", applying theories of the functions and causes of ideas to those of the theorists themselves.

Everyone engages in motivated reasoning and seeks out justifications for preferred beliefs. To the extent that we treat such phenomena as if they are primarily a right-wing, internet-based phenomenon, we should ask ourselves: which narrative might we be trying to justify? And whose interests might it serve?