Raymond Richman - Jesse Richman - Howard Richman
Richmans' Trade and Taxes Blog
An (almost) censorship-free approach to the challenges of fake news and propaganda in social media
For the last several years social media and internet companies have grappled with the challenge of how to manage or contain the spread of fake news and propaganda on their sites. This remains a major problem for the big tech companies. Some recent research suggests that the degree of populism in democracies around the world is correlated to a significant degree with the extent of market power possessed by Facebook, and the degree to which internet service provision is concentrated. Facebook and related platforms have historically arguably played a magnifying role in the spread of fake news and propaganda.
More than three hundred years ago, the humorist Jonathan Swift wrote “Falsehood flies, and the Truth comes limping after it.” And so it has proved to be in social media. The attention-getting exaggeration goes viral. The fact check often fails to follow, as all too often no one finds the fact check unless they go looking for it. And the people who most need it don’t go looking at all.
The reaction to this problem began in force after the 2016 presidential election and has involved the selective identification of accounts for removal, and attaching fact checks to posts, linked articles, and tweets that have been identified as potential vectors of fake news. It has recently expanded to the wholesale removal of apps and websites believed to be vectors of propaganda, radicalization, and misinformation (e.g. the delisting of Parler by Google and Apple, followed by the cancelling of the entire Parler website by Amazon in January 2021.
This censorship strategy for coping with the problem of misinformation involves wholesale removal from the public sphere of specific platforms and voices. It often entails enormous collateral damage. A trivial example. In December 2020 I was forced to switch from Google to rival search engine DuckDuckGo in order to find fact checks of a viral story about election vote totals being changed because the search filters that seemed to be in place on Google to protect me from finding falsehoods about the election were so aggressive that they also kept me from finding a USA Today fact check of those falsehoods.
Censorship as a way to protect the integrity of political debate and thwart malign forces has ancient problems. In Federalist 10, perhaps the most profound of the Federalist Papers, James Madison wrestled with the problem of factions adverse to the rights of other citizens on the long term interests of the community. Madison noted that one possible strategy for dealing with these evil actors is to try to label and suppress them: to destroy their liberty. But he rejected this as antithetical to the entire project of liberty – the basic foundations of freedom and self government. The fundamental challenge once this road is taken is that one person’s ‘faction’ may be another person’s speaking truth to power. Like the government agents who labeled Martin Luther King a dire threat and set out to destroy him, this approach risks misidentifying threats and undermining liberty. Madison wrote
“Liberty is to faction what air is to fire, an aliment without which it instantly expires. But it could not be less folly to abolish liberty, which is essential to political life, because it nourishes faction, than it would be to wish the annihilation of air, which is essential to animal life, because it imparts to fire its destructive agency.”
Is there an alternative approach that could help thwart radicalization and better connect the public sphere without need for censorship? I think that there is. It is to find ways to reconstruct the public sphere in order to curtail the effects of factious misinformation and fake news instead of attempting to eliminate the causes. Most fundamentally, it is to give truth (or at least rebuttals) wings equal to the falsehoods they pursue. In this way, fact can balance falsehood more fully.
Most questions in politics deal with issues which are arguable rather than being capable of definitive resolution as facts. The key goal is to connect the public sphere – to get people to argue with and refute each other.
Components of the solution.
The essential element of an argument is a controversy. Someone asserts a claim about the world, about politics, about the election… If the claim is controversial, someone else will likely disagree with this assertion, and they can then frame a rejoinder. This is then responded to, and so forth. In an ideal public sphere the audience – the rest of us – observe the controversy, and come to a conclusion about which argument is stronger, and we then act on that basis.
What happens instead on social media (and in polarized political media more broadly) is that the sides of the argument often don’t really end up speaking to each other much. The argument instead gets heard only partially by most observers. If my friends tend to take one side of the issue, then I will tend to hear that side of the debate echoed back to me much more than any rebuttal of it. Around the world, as the recent Facebook leaks highlight, the results are polarization and unreality.
The starting point for a solution is a new option to tag a tweet or a post or a link shared on social media. This is the ‘rejoinder’. A rejoinder is a counter-argument. Any post or link tagged with a rejoinder will be shared with the rejoinder paired with it. When there are multiple rejoinders or links, a system of up and down voting of rejoinders could adjudicate between alternatives, potentially with an editorial role from fact checkers and others to ensure that they system isn’t abused.
There are obviously many important details underlying getting the rejoinder system to work properly. Most importantly, the key would be to have a system for upvoting and downvoting suggested rejoinders along with metrics concerning quality of sources to prioritize rejoinders from sources with better editorial quality. While the system could be mostly automated, there would need to be some management from fact-checkers and others working with the social network.
One of the strengths of this idea is that if implemented correctly, it could provide a profit motive and energy to the process of rebutting arguments. It would also give partisans of particular (often limited) viewpoints a socially productive activity to pursue – making sure that those they disagree with get appropriately rebutted across media spaces.
Let’s imagine three scenarios: Facebook-before-censorship, Facebook-after-censorship, and Facebook as it could be.
Facebook before censorship: Jones writes a blog post making a series of questionable claims. Smith writes an article criticizing the logical fallacies and lack of evidence in Jones’ argument. Jones and friends share links to Jones, and mostly ignore Smith. Smith and friends mostly share links to Smith’s article. Everyone ends up sure they are right. And they start to hate each other.
Facebook after censorship: Jones writes a blog post making a series of questionable claims. Smith writes an article criticizing the logical fallacies and lack of evidence in Jones’ argument. Jones and friends share links to Jones until they are de-platformed by Facebook fact checkers. Smith and friends mostly share links to Smith’s article. Everyone ends up sure they are right. And they start to hate each other. Jones and friends start to hate Facebook too and start looking for new platforms to share their questionable claims.
Facebook as it could be: Jones writes an article or blog post or a social media post making a series of questionable claims. Smith writes an article or a social media post criticizing the logical fallacies and lack of evidence in Jones’ argument. Smith tags Jones’ article with Smith’s rejoinder. Smith and Jones’ friends all see both articles because every time Jones’ article gets shared, the network automatically attaches a link to the leading rejoinder (by Smith). Jones writes a rebuttal to Smith’s rejoinder and tags it as a rebuttal to Smith’s article. Smith writes a rejoinder to Jones’ rebuttal. Their friends still see everything. Some people conclude that Jones is mostly right (in fairness some recognize that Smith has a few good points.) Some people conclude that Smith is mostly right. Others conclude that since neither side had very good arguments they probably had better wait for more information or better arguments. Smith and Jones eventually become rich from advertising revenue generated by their controversy but are also eventually convinced that the other did have a few good points. They become friends and go have coffee or a beer together.
Which world would you rather live in?
Comment by note, 11/24/2021:
"Justice Louis Brandeis, in his famous 1927 concurrence in Whitney v. California, hit upon the essence of robust speech as its own corrective measure. “If there be time to expose through discussion, the falsehoods and fallacies,” he wrote, “the remedy to be applied is more speech, not enforced silence.”"
Journal of Economic Literature:
Atlantic Economic Journal: