On December 17th, Tumblr’s ban on adult content will go into effect
, meaning that all of the pictures, GIFs, and videos that feature erotic content will be removed from the site. In the future, ongoing moderation duties will be handled by automated software and decided on by humans, although the first wave of content flagging appears to have been done by AI. The reaction to the news was immediate and distinct: porn has been tacitly allowed on Tumblr since its founding in 2007, and people were mostly angry at what they saw as an action that would alienate users and fandoms that threatened the fabric of the social network itself.
The ban on adult content will affect a truly staggering number of blogs, as
Katie Notopoulos pointed out at BuzzFeed News
. “I’m guessing there are hundreds of thousands (millions?) of blogs that will be shut down and disappear forever,” she writes. “And god knows how many millions of individual posts contain what is deemed adult content.” While Tumblr wouldn’t comment on the record about the policy change, instead referring me to the blog it posted, it’s fairly easy to figure out why the company acted so quickly. Last month, Tumblr’s official app was removed from Apple’s App Store
over the discovery of child pornography on the site.
Gathering a population together the way that social media does necessarily invites bad actors, but it’s rare that those people leave a website open to legal liability. And when they do, leadership tends to take immediate, often ill-considered action. When Apple banned Tumblr’s app, the course was set, and the path charted was an old one. What’s happening on Tumblr today happened almost exactly a decade ago on LiveJournal, in an event that would come to be known as “Strikethrough.” The executives at that company presided over site-wide suspensions over fears about child pornography, which led to an exodus of users and hurt the site’s reputation. LiveJournal was never the same after its users left. Indirectly, it’s one of the reasons for the popularity of Tumblr, which launched in early 2007 and became home to many of the users who left LiveJournal. Tumblr’s recent ban threatens to provoke another wave of user migration — although this time, it’s not clear where else they can go.
In 2007, LiveJournal began exploring its pivot to an ad-supported model, moving away from its premium, user-supported one. Google had recently IPO’d, proving that selling ads could work on a large scale. (At the time, LiveJournal was one of the few internet companies that had crossed 10 million users.) This was when Friendster was dying and Myspace was taking its place as the most influential social destination on the web. The show To Catch a Predator
had also debuted around the same time, and it seemed like every week, there was a new story about how dangerous it was to be a child online. It set the tone for a growing moral panic that eventually swept LiveJournal up and swallowed it whole. Strikethrough happened because LiveJournal reacted on instinct and without nuance.
At the time, Anil Dash was an employee at Six Apart, the company that owned LiveJournal. Though he was on a different team, Dash was involved in damage control when any of the teams screwed up, including LiveJournal. In 2006, LiveJournal had a brush with bad publicity because of the overzealous enforcement of a content policy
: a user was asked by the site to remove a picture of her breastfeeding, which apparently violated their policy against nudity in default pictures. Dash told me that after the kerfuffle, there was a “nurse in,” where women breastfed in front of Six Apart’s offices, after which the team invited them in for lunch and changed the policy. (In his apology post to the community
, Doug Bryan, the then-VP of operations, ended his note with the sign-off “Father of four breast fed children.”)
“The people making that decision, charitably, did not understand the LiveJournal community even though they were running it,” says Dash. The company had four products, and on LiveJournal, it had everything from people writing blogs about the Iraq War to Harry Potter
fan fiction. A year later, it was those Potter
fans who put LiveJournal’s content policy to the test.
That community was mostly made up of women in their teens and early 20s, which was cohesive as a demographic but wildly disparate in legal ramifications, as Dash put it. LiveJournal users in the community would create fanworks — some erotic, some not — in writing and with images. And as the films were released, the visual fan creations became more and more of a problem, as users were able to create photorealistic works based on stills. “And the catalyst that I remember was one of the particular pecadillos of fandom in Potter fandom... twincest,” Dash tells me. “I’ve never read the books, but I’ve learned a lot about the characters in the books. There are Weasley twins apparently…” (“Twincest” is exactly what it sounds like.)
Because they were based on underage actors, those photorealistic depictions were legally no different than child pornography. Six Apart took action. “This panic happened,” Dash says, “like we were going to liable for being a platform enabling child porn.” The terms of service were updated; there was the “friendly bullshit community post” that claimed the changes were good news, and then a filter was applied to flag and suspend content that the company deemed objectionable. It was a blunt-force instrument based on how users had tagged their content, and many unobjectionable posts and communities were caught up in a wave of mass suspensions. The blogs still existed, but there were strikethroughs in their names. It made “people feel like they’ve been deleted,” Dash says.
According to Casey Fiesler, an assistant professor of information science at the University of Colorado Boulder and a researcher of communities and fandoms, these LiveJournal policy changes redefined what was and wasn’t considered “obscene content” on the site. “And so the result was that they straight out deleted a lot of accounts from mostly fan artists and some fan-fiction communities. They also ended up deleting accounts for like communities of sexual assault survivors and this kind of thing as well.” The result was a mass exodus of users to places like Tumblr and the Archive of Our Own
(AO3), a fanworks archive that was started in part because of the chaos of Strikethrough. Based on her research, Fiesler thinks something similar could happen to Tumblr because, as with LiveJournal, the users she studied didn’t feel like the site cared about them. Dash, for his part, says that LiveJournal did care about its departing users. “It wasn’t actually a big number right away, but the reputational hit was a big deal and management felt it.”
According to Elizabeth Minkel — an avid Tumblr user, co-host of the
, which covers fan culture, and editor of How We Get To Next
— Tumblr’s had a growing porn problem over the last year, with spammy bots overrunning regular users. “[There’s a] constant refrain on my dash amongst people that I follow of ‘fix the porn bots, fix the porn bots,’” says Minkel. The problem, as she sees it, is that it’s impossible to implement nuanced content moderation at scale; AI is simply not yet capable of distinguishing between erotic fan art and porn. “I don’t think anyone’s doing it successfully. Lots and lots of stuff is getting flagged erroneously. I understand why they’re doing this. Is the bluntness of it good? Possibly not.”
Dash echoes this: “[Tumblr] used a very blunt instrument. I think it’s the same mistake that we made on LiveJournal back in the day. I think these problems keep recurring because it’s hard to do the nuance thing. I think the right solution long term, is prevention. But once the ship has sailed, what do you do?”
Even so, Minkel doesn’t think everyone is going to hate the purge of erotic material. “There are fans who believe that content should be censored, [who] think that it is morally wrong to depict characters engaged in, say, one character raping another. They say, ‘That’s wrong in real life, that’s wrong in fiction,’” she continues. “But then there are going to be plenty of fans who are not happy about this, even if it’s not actually something they’re creating, or even really engaging with.”
These conversations about what fans can expect from the networks hosting them aren’t new, either. As Minkel points out, they’re substantially similar to the conversations fans were having before the creation of the Organization for Transformative Works (OTW), which is the foundation that runs AO3. It was founded after Strikethrough. “That’s why the OTW was created, is that we are never going to be able to trust that a commercial platform is going to be on our side,” Minkel said. She thinks users just assumed Tumblr was fine with all of the content posted there because the company didn’t publicly condemn it. “I don’t think Tumblr necessarily disagrees with it,” she said. “I think they’re trying to protect themselves.”
And it makes sense from a corporate perspective. Yahoo famously bought Tumblr for more than a billion dollars in 2013, presumably as an infusion of youths to eventually sell ads against. Three years later, Yahoo wrote down its purchase to the tune of $482 million. This year, Verizon moved all of Yahoo’s properties into the content umbrella corporation Oath. The problem is simple: content that isn’t safe for work is hard to monetize, and Tumblr’s future has been in doubt since the day it was purchased. What’s a social network that can’t make money? Dead.
It’s telling that the same year Strikethrough happened, Six Apart sold LiveJournal to a Russian media company.
It used to be that back in the days before services like LiveJournal existed, people online would host their own blogs. (Movable Type, which was one of Six Apart’s most important products from both a technical and a social perspective, was software that let people publish their own blogs online.)
“Over the years we’ve seen this shift to all of these platforms,” says Jason Scott, who describes himself as a “free range archivist” at the Internet Archive. People bring their friends where they are, and communities begin to migrate. To him, it’s obvious that Tumblr doesn’t understand its base, and it made the decision to ban adult content in spite of what that base cares about. “Now, everybody wakes up to the nightmare that most people know,” he says. “You’re not a customer of Tumblr. You’re just staying on Tumblr’s couch, and Tumblr has now said, ‘We have a different use for that couch.’”
Tumblr’s users and creators have been caught in the middle of a series of corporate decisions that have left them out in the cold. Finally, after a decade online, it seems the quirky social network won’t be left to its own devices, and it will have to find its own way forward. “Tumblr is very important to fandom, but I have no idea how important fandom is to Tumblr,” Fiesler says. The answer to that question will soon become clear. The other question, however, remains: is this the end of an era, or the end of Tumblr itself?