The Problem With Putting Social Media in Charge of Our Memories

The internet is very good at bringing up things from our past, but what if we don’t want to be reminded?

Photo: 3alexd/Getty Images

In late November of 2018, Facebook users who logged onto the platform found themselves flooded with old messages reappearing in Messenger as unread. The sudden surge of messages came from a bug that Facebook resolved the same day. But even after the messages vanished, the blip had a lasting effect on some users. As one journalist who saw messages brought back to life told The Atlantic, it was “yet another weird way that your past can come back to haunt you.”

Like the old Facebook messages, there are lingering reminders of the past in all our digital spaces. Odds are, you’ll scroll through Instagram and see a friend you’ve lost touch with has posted again, or reopen an old text conversation that reminds you of an older version of yourself — your jokes, stories, and slang preserved verbatim. These reminders can sometimes catch us by surprise, or even kind of hurt.

“Many [messages] were from the day my partner, Dean, passed away,” one man tweeted after the Facebook bug, “& now I’ve spent my evening in fear of what else I’m going to see.”

When the internet forces us to confront our past, how should we respond? There aren’t many options in the way of managing old chat threads and emails, posts, and photos, except to save or delete them — two pretty extreme options. Inevitably, people want to defer their decisions about which emails to consign to oblivion and save them for later when they might have the mental energy to sort through each individual message or photo and decide what it’s worth. But most people never get around to that. Most of us end up saving the majority of our past digital correspondences, despite our best efforts: one study found that we only manage to delete about 30% of our emails, another that we only scrap 17% of our photos.

Platforms like Gmail, with their ever-increasing storage capacities, may also be encouraging us to save — and incidentally, to remember. “We’re facing an intriguing inversion point in human memory,” observes Clive Thompson in his book Smarter Than You Think. “We’re moving from a period in which most of the details of our lives were forgotten to one in which many, perhaps most of them, will be captured.” This shift from ubiquitous preservation to remembering coincides with advancements in storage capacity. Until recently, limits on storage made it impossible to record our lives in such detail. A gigabyte of memory in 1981, for instance, cost roughly three hundred thousand dollars. Today, our digital tools make the passive capturing of our everyday lives cheap, and they require little maintenance to preserve these records over time. The challenge now is how we’re going to manage it all.

In his book, Thompson recalls a tale told to communications professor Jason Zalinger about a man who created a Gmail folder for all the emails he didn’t want to be reminded of anymore — he called it “Forget.” The folder, which mostly contained messages from his ex-wife, was, he said, for “the stuff I just didn’t want to think about anymore but decided I needed to archive anyway. Isn’t that a curious distinction?”

“Am I never going to be free of all this crap?”

It isn’t easy to contain such reminders in a neat folder, let alone across multiple platforms, but coming across some digital memories can be rattling. One study from researchers Corina Sas and Steve J. Whittaker found that people even struggled to move on after a breakup because of how they were constantly reminded of their past relationship online. When the researchers asked what hindered them in particular, one participant said, “Occasionally finding things that I had missed throwing out or deleting: the odd email stored in Outlook on the computer I didn’t often use or messages I missed on a social networking site I didn’t use much.”

They struggled to limit the impact of their past relationship going forward, as they had few options to contain the reminders they encountered and were resistant to deleting them. “A lack of disposal tools,” the researchers noted, “meant most participants either kept, or disposed of, everything.” And their choice impacted how they moved on. “Keepers took longer to heal, disposers often regretted their impulsiveness.”

Despite trying to distance themselves, people remained in many ways connected with their ex-partners on social networks and shared digital spaces. This has a lot to do with the robust design of social networks: They compel you to stay connected, even through circumstances, such as a breakup, in which you’re better off disconnecting.

Even if you take reasonable precautions following a breakup — like blocking your ex and untagging yourself in photos — algorithms can still dredge up things you don’t want to see. There is an unspoken etiquette around what is appropriate to bring up to someone experiencing loss which algorithms haven’t been socialized in. “People know better than to invite a friend to their ex-partner’s engagement party or suggest calling a deceased friend,” as researchers at the University of Colorado at Boulder point out, “but the algorithmic systems that underpin social media sites often lack the contextual understanding required to not make these sorts of troubling suggestions.”

The CU Boulder study builds on the previous study, finding that Facebook inadvertently confronted users with content relating to their exes while they were going through a breakup. As some users discovered, if you unfriend your ex on Facebook, you won’t see their posts anymore but you could still run into them in an untagged photo that resurfaces, or see their comments on a mutual friend’s post. Despite some people’s best efforts to tinker with settings, they were still shown things they didn’t want to see. “Am I never going to be free of all this crap?” one frustrated user said.

There’s limited awareness about how curation algorithms really work and what features like Facebook’s Take a Break (which lets users mute a person without making this decision visible to others) really do. One participant could have been spared some grief had he known there was a Take a Break feature to begin with. Still, the system first has to recognize a person is there in order to hide them, and not all photos are tagged. Algorithms can resurface this unmarked content, making suggestions that appear at odds with features like Take a Break intended to distance users. Those misfires also take their toll on users.

In an ideal world, social platforms would be able to recognize and represent the complexity of our social realities.

“Figuring out how to support people in disentangling their digital lives from each other is a particularly interesting problem,” lead researcher and PhD student Anthony Pinter told me, as previously the challenge was how to connect everyone. For one, Facebook may only be prioritizing one-to-one connections between people. So, while exes might disconnect from each other, their overlapping connections remain intact post-breakup and are still fair game for algorithms to leverage and promote content based on. This can make it even harder for users to move on when they see their ex’s relatives show up as People You May Know or an ex-mother-in-law’s status update.

Pinter and his colleagues urge designers to account for this “social periphery” that exists around former relationships and is largely responsible for users’ upsetting encounters. However, they caution against design solutions that would accomplish things en masse such as mass-blocking. Facebook could enhance existing features like unfollowing, unfriending, and blocking to help users better understand what each feature will accomplish, what will make an ex-partner stay visible, and which of their actions will be noticeable to other users. “Some actions [like blocking] should take several clicks and give people the opportunity to think through the action,” Pinter stresses.

In an ideal world, social platforms would be able to recognize and represent the complexity of our social realities. But to do that would require us to divulge even more data. “That’s certainly a Catch-22 of it,” Pinter says. “We can’t ask people to give up more data. On the other hand, we can’t make assumptions either.”

Part of the challenge too, he says, is with designing around “social norms that are fairly explicit and understood offline but get really blurry online.” In offline relationships, for instance, couples divvy up their shared possessions before separating and decide between the two of them who will get the couch or the fridge. On platforms like Facebook where couples also come to share digital possessions, there is an opportunity to design for a similar negotiation — deciding who will get which photos, statuses, and even which friends before going separate ways. For photos, Pinter envisions Facebook could make separate copies of a photo upon request and assign each copy to a separate user to be stored in their own individual archive.

More sophisticated machine learning features could streamline this process of sorting through and purging digital possessions. Implicit identification of ex-partners in untagged photos done by facial recognition, for instance, could reduce the emotional toll on users in the aftermath. “The functionality is certainly there,” Pinter adds. “It would just need to be retooled a bit.” Even so, there is an immense pressure for developers to get this sort of classification right — “it could very easily upset the user, just as much as it could save them from being upset.”

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store