Hive mind

How the internet will change our coronavirus memories

"We can see through Wikipedia the evolution of our understanding of this disease.”

The first Wikipedia article about the coronavirus pandemic appeared on the internet at 11:58 a.m. Eastern Daylight Time on January 5, 2020. It was a single sentence. The disease was "commonly known as pnumia" (a misspelling of pneumonia). Its origin was unknown.

James Heilman, an emergency physician who goes by "Doc James" on Wikipedia, made his first edits to the post on at 6:37 a.m. Eastern on January 6. He tells Inverse that his first edits were relatively minor, he just moved a few references around and clarified that the outbreak was still happening at a “Huanan seafood market in Wuhan.”

Heilman is one of the hundreds of Wikipedians who voluntarily edit the website’s coronavirus pages (but he's been editing for 13 years and is now on the Wikimedia Foundation Board of Trustees). What began as a single sentence has blossomed into thousands of Wikipedia pages about the coronavirus. According to Wikipedia's internal newspaper The Signpost, there was not a single day in March when traffic to the coronavirus pandemic page didn't fall below 400,000 visitors.

If you look at the back-end of that incredibly popular page (accessible once you make an account) you'll see a minute-by-minute log of how the public began to grasp the seriousness of the coronavirus pandemic.

Wikipedia’s logging system “provides an easy method for any of us to check our own memories regarding what was known when,” Heilman tells Inverse. “Basically we can see through Wikipedia the evolution of our understanding of this disease.”

Wikipedia is only part of the online chaos surrounding the coronavirus. But it acts as a summary of six months of Twitter debates, cable news gaffes and rising, then flattening, coronavirus case counts. It’s a lot like what scientists call “collective memory” – a narrative in which individual experiences and narratives come together to form an arc in our minds.

With a global death toll of 440,000, the Covid-19 pandemic will probably be one of the 21st century's biggest tragedies. How we will remember this event 50 years from now is being shaped by Wikipedia editors right now who are the guardians of that collective memory on the internet.

The first wikipedia fragment on the coronavirus from January 5, 2020. Wikipedia

At first, our individual memories of the coronavirus will probably be unstable, like the early stages of a Wikipedia article where edits are frequent. Then we forget things, incorporate new perspectives, and adjust our own memories, explains Celia Harris, a senior cognitive neuroscience researcher at Western Sydney University.

“But they do coalesce into a stable narrative, probably because we rehearse them in conversations with others,” Harris tells Inverse. “This stable version is likely more similar across the group than our initial individual memories.”

In each of our minds, news events, tweets and our own experiences are convalescing into a collective memory. Heilman, and editors like him at Wikipedia, are just the ones writing it all down. If the research conducted by scientists who study collective memory holds up, how that memory forms and how long it lasts could have significant implications for how we approach the next pandemic and how society looks back at what went wrong with this one.

How long will we remember coronavirus? – If the pandemic were to end tomorrow, you’d be surprised how fast we’d stop talking about it, says Taha Yasseri, soon-to-be an associate professor of sociology at The University of Dublin.

“What we have observed in our research is that short-term memory is very short for tragic events. The public attention to the event fades away quite quickly, within a week,” he tells Inverse.

Because the coronavirus has endured for about six months, that obviously hasn’t happened yet.

Yasseri’s work on collective memory is based on several papers that discuss how the public reacts to plane crashes, specifically, by using Wikipedia articles (he calls them a “technology of memory”). In a 2016 paper, he showed that attention toward a massive tragic event like a plane crash follows a type of “exponential decay” – After an initial boost in readership in the first five days after a page is created, attention slowly tapers off as novelty wanes and readers become interested in other things.

The exponential decay of public attention. Traffic to Wikipedia articles after plane crashes spikes within 5 days, then hits a breaking point and declines. The Royal Society Open Science

Plane crashes are clearly different from the coronavirus, which has killed hundreds of thousands and fundamentally changed our lives. Nevertheless, in places where the pandemic has waned, like Australia, Harris says she’s starting to see the cycle of attention move on. “I can see where I am, that people have moved on quickly. As soon as restrictions eased, people were back to normal life,” she says. (Even Wikipedia traffic to the pandemic's homepage dropped by about 18 million page views between March and May).

She suggests that this is happening because the experience of coronavirus was “pretty monotonous and mundane” for most people.

“The days tend to blur together,” says Harris. “We remember turning points, transitions, and distinctive episodes.”

If you happen to experience a traumatic individual event thanks to coronavirus – the death of a loved one, for instance — those memories will linger on because they torched the fabric of everyday life. For the rest of us, the world only appeared to end as we watched from home.

Once the pandemic ends, we may not be discussing coronavirus anymore, but there will probably be moments that will trigger society to return to their experiences for decades to come. Yasseri argues that there’s another type of collective memory that has more sticking power: the underlying historical memory.

James Heilman, a veteran Wikipedia editor who has been working on updating the pages about coronavirus. Wikipedia

“There is an underlying long memory that is facilitated through literature, through culture, and more recently through the internet and online material,” he says.

This underlying historical memory is shaped by those “technologies of memory” like Wikipedia or news coverage, and there are other studies that show just how powerful those records of memory are.

In one 2019 study, scientists in France took brain scans of participants who were exposed to news footage from the 1980s that was part of a reckoning in French history: the complicity of the government with the atrocities of the Holocaust.

“Hopefully this will help prevent people from trying to rewrite history."

The team found that the actual content of the photos and that collective narrative were linked to different patterns of activity. They argued this collective narration is encoded differently, but “co-constructed” with the way the brain encodes our memories of actual facts and experiences. Our memories are inherently flavored by the people we speak to and the media we consume. So much so that we can't always separate the narrative arc from our own experience, Harris explains.

“Of course the narratives presented on mainstream and social media come from somewhere – individuals think of them and put them out there – but they are ‘contagious,' influencing other people’s narratives and making them more similar,” says Harris.

Over time that collective memory becomes entrenched in our history, says Yasseri.

In a 2017 study on Wikipedia traffic after plane crashes, Yasseri found that similar events tend to trigger a resurgence of interest in similar events from the past. Traffic to Wikipedia articles about past plane crashes within the last 40 years spiked every time a new plane crash happened.

He’s already seen that pattern happen in discussions about past pandemics, like the 1918 flu. In March, the Spanish Flu Wikipedia page was the third-most-popular page on Wikipedia, behind two coronavirus-related pages.

“That collective memory gets ‘triggered’ every time a similar event happens,” he says.

The technology of memory – The fact that the public attention moves on so quickly isn’t good news for scientists who study coronaviruses. Funding for projects like vet detectives who seek out new animal-borne diseases, is, in part, dependent on the attention and the will of policymakers. As Peter Daszak, the president of the EcoHealth Alliance told Inverse in February:

“I guarantee four months from now, when this outbreak is over, the programs that were set up will gradually start to wind down, the funding will dry up.”

He wasn't wrong. In May, with the pandemic still looming, there was already talk of disbanding the U.S. government's coronavirus task force.

But historically speaking, there's hope in that second kind of long-term memory doesn’t go away so easily. In fact, it resurfaces again and again. People like Heilman are charged with editing and ensuring that our collective memories of the coronavirus are accurate and include as many perspectives as possible to ensure that the collective memory actually reflects the experiences of many, not few.

That’s why Wikipedia’s Covid-19 project is broken down into task forces that specialize in case counts or translations. There’s an entire task force dedicated to women’s issues during Covid-19 called “Women in Red." But Wikipedia doesn’t have a spotless record when it comes to covering diverse perspectives, which leave holes in our collective memory on the internet. As the president of Wikipedia’s D.C. branch, James Hare told The New York Times in 2015:

“The stereotype of a Wikipedia editor is a 30-year-old white man, and so most of the articles written are about stuff that interests 30-year-old white men. So a lot of black history is left out.”

Right now, Heilman’s job is about summarizing a flood of coronavirus information. It’s a big job, because, if Yasseri’s work is right, people will return to those pages again and again if there’s ever another coronavirus or similar event.

“The issue is more how to summarize what is available and give appropriate weight to different perspectives. A lot of effort now also goes into making sure hypotheses do not get presented as fact and that we present prevalent falsehoods as such,” says Heilman.

If anything goes wrong, though, there’s always the backlog of edits. It’s our collective memory in real-time, but it’s also his safeguard.

“Hopefully this will help prevent people from trying to rewrite history,” Heilman says.