There once was a feller named Elon
For Twitter he just made a deal on
Is he free speech’s savior
Or is his behavior
Just something we’ll never agree on?
Social media are the best source of accurate news for Russians trying to understand what’s really going on in Ukraine.
Social media are also the best source of utterly false propaganda for Americans looking for talking points about the conflict in Ukraine they can use to justify their admiration for Vladmir Putin.
Before we get all worked up about Elon Musk’s Twitter acquisition and plans to take it private, let’s take a deep breath and de-murkify some aspects of the situation that, to my lay-person’s eyes at least, are in desperate need of de-murkification.
Does Twitter engage in censorship?
No. Censorship is something governments do. Twitter is a business, not a government.
Does Twitter currently engage in excessive editorial control?
This isn’t a yes-no question. It’s a how-much question.
The more editorial control Twitter exerts over the content published on it, the more it resembles a publisher rather than a platform. Oversimplifying, publishers are responsible for the content they publish. Platforms aren’t responsible for the content published on them. But it’s a continuum, not a binary categorization.
Twitter exerts some editorial control over the content posted on it (see The Twitter rules: safety, privacy, authenticity, and more ).
That control doesn’t extend to requiring that what’s posted on it is true, though, so to my eyes it’s still more platform than publisher. Your retinas might reach a different conclusion.
Do government efforts to regulate Twitter’s content constitute censorship?
That depends which content.
Content posted by individual human beings is, and (in my not-very-humble opinion) should be protected by the First Amendment. This also means it’s governed by the First Amendment’s well-established boundaries. Defamation, endangerment, and incitement to violence are as illegal on Twitter as when yelled out by an angry person standing on a soapbox in Central Park.
Content posted by corporations, in contrast, isn’t (or at least, shouldn’t be) protected by the First Amendment. Supreme Court rulings that “corporations are people too” notwithstanding, that’s still a legal fiction. Everything corporations do is legally subject to regulation, because forming a corporation is a privilege, not a right.
The most obvious and clear-cut example of legitimate regulation of corporate speech is false advertising. A company that publishes falsehoods about its products is violating the law.
As for content posted by ‘bots, ‘bots aren’t persons. They have no constitutional rights of any kind, whether deployed by corporations, governments, or autonomously (a terrifying thought).
Does government regulation of “the algorithm” constitute censorship?
No, and with all due deference to the late, great Frank Zappa, this, not the apostrophe, is the crux of the biscuit.
First of all, the technology social media concerns use is the polar opposite of an algorithm. Algorithms apply known rules to turn their inputs into outputs. Social media use neural networks, and as is well known, even the neural networks themselves don’t “know” how they reach their conclusions.
Social media neural networks are, that is, both autonomous and oblivious. Not a good combination.
Of greater significance, the act of deciding which content to bring to subscribers’ attention, whether it’s through the use of true algorithms, neural network “algorithms,” or an editorial committee composed of actual human beings, pushes social media like Twitter and Facebook closer to the publisher end of the platform / publisher continuum.
How about Section 230?
Section 230 was an attempt to define any and all internet services that democratize content creation as platforms. It was crafted in a simpler, algorithm-free time (1996) and needs replacing. Finding even two people who agree on what to replace it with, though, is a challenge.
Bob’s last word: If I was a predicting kind of guy, I’d go along with everyone else commenting about this subject: Musk will take a laissez faire approach to editorial control, including but not limited to removing restrictions on who is allowed to publish on Twitter.
Corrected from the original:
That would move it much closer to being publisher than platform.
And Musk will have a hard enough time keeping that from happening without tearing out the so-called algorithms Tweetsters rely on to decide what to pay attention to and what to ignore.
That would move it closer to being platform than publisher.
But Musk will have a hard time maintaining platform status without tearing out the so-called algorithms Tweetsters rely on to decide what to pay attention to and what to ignore.
If past behavior predicts future grousing, Musk won’t have much patience with the legal need to thread this metaphorical needle.
Bob’s sales pitch: Reserve May 11th, 2:40pm CST, when I engage with the estimable Roger Grimes in The Great Quantum Debate: Is There a Role in Business Yet? as part of CIO’s Future of Data Summit. You can register for the summit here. I predict you’ll find it even more informative than your average Tweet.
I don’t know whether I’m more taken with your limerick or your analysis, but I am sad to see the change that’s likely to happen to Twitter.
Elon rails about free speech. He doesn’t acknowledge that the First Amendment doesn’t obligate a corporation to allow everyone to say whatever they want using the corporation’s resources. The sophistication of propagandists and the amplification of dangerous lies that can take place on Twitter make for serious adverse consequences of some free speech in this context.
Our governments should rarely constrain free speech. When a compelling interest, e.g., preventing a deadly stampede that could result from falsely yelling “Fire!” in a theater, demands such constraint, our governments must limit themselves to applying the minimum necessary constraints.
Corporations, having no obligation to amplify dangerous speech, can and should serve the public good by applying reasonable constraints.
Elon certainly appears ready to let loose a torrent of dangerous propaganda and incendiary speech that will result in real-world harm. I dearly hope to be wrong, but it looks to me like Twitter is about to become a tool for hatred, intolerance, upheaval, and quite possibly the demise of democracy.
“… Musk will take a laissez faire approach to editorial control, including but not limited to removing restrictions on who is allowed to publish on Twitter.
“That would move it much closer to being publisher than platform.”
Maybe I don’t understand, but wouldn’t less control over content make him a platform?
A little confused by the “last word.” In my mind, reducing the amount of editorial control would push Twitter more in the direction of a platform rather than in the direction of a publisher. If you want to publish an op-ed in The New York Times, you’ll have to navigate the editorial maze to do so. Hence, a publisher. If you want to post your op-ed on the bulletin board at the local diner along with business cards and tear-off phone numbers, all you need is a thumbtack. Hence, a platform. Do I have this backward?
No, regrettably I stated my position backward, the result of multiple tries at rearranging paragraphs to try to make the thoughts flow better.
Taking a more laissez faire approach moves Twitter closer to being a platform. Not tearing out the “algorithm” leaves aspects of publisher-dom in place.
Let’s see. If Musk’s Twitter will be “laissez faire” and “removing restrictions”, how is that “closer to a publisher than a platform”? It would seem the opposite is true.
The problem with social media is that they use left-leaning “fact checkers” to justify their limiting or removal of views and users who disagree with left-leaning views. This bias is the major argument that social media platforms are actually publishers. Musk doing less of that is more like a platform than a publisher.
perhaps this reference will be of use.
publisher/platform differentiation is not very relevant with respect to 230.
https://www.techdirt.com/2020/06/23/hello-youve-been-referred-here-because-youre-wrong-about-section-230-communications-decency-act/
Hmmm. The linked article states, “Really, this is the simplest, most basic understanding of Section 230: it is about placing the liability for content online on whoever created that content, and not on whoever is hosting it. If you understand that one thing, you’ll understand most of the most important things about Section 230.”
That’s exactly the distinction I made, only I said “platform” where the article says “hosting.”
Good reasons I have not yet created a twitter account.
Thought experiment:
I make a provocative flyer and place it on the public announcement bulletin board in a Starbucks. Someone else took copies of that flyer and placed it in every other Starbucks in the area.
** Am I responsible for someone at the Starbucks across town getting riled up, or is the person who took my flyer and duplicated it?
** Does the person who duplicated it have any responsibility for good or bad outcomes?
It isn’t one or the other. Both had a hand in it. Presume you’re drawing a parallel between making copies and the dreaded algorithm. The person to works hard to bring the flyer to everyone’s attention does, I’d think, share responsibility for its contents.
“Does Twitter engage in censorship?
No. Censorship is something governments do. Twitter is a business, not a government.”
This is misleading; and the misleadingness is far from trivial. Better — at the cost of brevity — to rearrange and expand it something like this:
Does Twitter engage in censorship?
Yes, Twitter’s ownership/management can, and does, engage in censoring speakers on the Twitter platform. However, Twitter’s ownership/management cannot, and does not, engage in censoring speakers on platforms and publishers and venues that ARE NOT TWITTER – such as Facebook, the New York Times, and soapboxes in public parks. Because Twitter is not the government, but is instead a privately-owned enterprise, it cannot enforce its censorship desires on platforms/publishers/venues that are owned by entities that are NOT owned by Twitter; it also cannot enforce its censorship desires against individual humans who are not on the Twitter platform, though it CAN, and sometimes does, retaliate against entities and individual humans who ARE on the Twitter platform, for their speech or behavior that took place OUTSIDE of the Twitter platform. This retaliation can take the form of censoring individual speech items by the speaker on the Twitter platform, or even can take the form of banning a speaker entirely and forever from the Twitter platform.
Censorship by a non-governmental entity, such as Twitter, is less extreme than censorship by the government, because the government can censor speech on ALL platforms and ALL publishers and ALL venues simultaneously, whereas a non-governmental entity can ONLY censor speech on platforms/publishers/venues that it owns. A speaker who is censored by Twitter can turn to other platforms/publishers/venues, and attempt to speak THERE, in the hope of NOT being censored THERE; but a speaker who is censored by the government is censored on ALL platforms and ALL publishers and ALL venues simultaneously, and has nowhere else in which to attempt to speak. A speaker who is censored by the government is censored EVERYWHERE; a speaker who is censored by Twitter is censored only on Twitter, and nowhere else.
It appears we define censorship differently. As I use the term, it’s about preventing the public from encountering the content. Governments have some semblance of the power to do this. Twitter does not. It can prevent the public from encountering it on Twitter, but that’s the extent of its power.
To me, the difference between “you won’t read it here” and “you won’t find it anywhere” is profound, more a matter of kind than of degree.
With respect to Section 230, there is no difference between a platform and a publisher. As far as I can tell, the law doesn’t recognize a platform as a distinct from a publisher (the difference only seems to exist when law makers are talking about social media in a attempt to control what social media sites are doing). All social media platforms are publishers. The distinction here is that social media platforms are mainly filled with user generated content. Other publishers (News sites, TV sites, bloggers, etc) are mainly filled with content being created directly for the site. Everyone can be sued, but section 230 says that publishers can’t be sued for user generated content. And before any says, what about editorial control? That is one of main purposes of section 230, to allow publishers (like Facebook, Twitter, YouTube) to delete user generated content without becoming libel for the user generated content that was left. This happened in the 1990’s. That is why section 230 was passed, to allow for user generated content to be deleted without the web sites (publishers) to then be libel for anything left on the web site.