Map of Wikipedia articles, by Mark Graham at Oxford University

This weekend, Sarah Gilbert and I were honored to give a plenary talk at Wikipedia Day NYC, a celebration of the 25th anniversary of Wikipedia, alongside N.K. Jemisin, Bill Adair, Alexis Nikole Nelson, and other wonderful editors, archivists, and community leaders across the Wikimedia movement. We took the occasion to reflect on four lessons from the science of Wikipedia that we hope could inspire anyone to organize for the common good. Here is what we shared:

Next month when the Artemis II spacecraft launches on its journey to the moon, it’s going to be carrying Christina Koch, the first person ever to edit Wikipedia from space. She did this to correct the record about one of her spacewalks during 328 days at the International Space Station in 2019.

When we think about Wikipedia, we sometimes imagine it like the view from space – a vast resource that combines the shared light of people around the world. The English language version alone has over 7 million articles and according to economists, contributes roughly $42 billion dollars a year in consumer surplus to the US economy despite not costing anyone anything.

Maybe you like us have asked “how in the world can people build something so unimaginably grand?” If so, you’re thinking like a researcher. Together at the Citizens and Technology Lab, we work together at Cornell University to study what Wikipedia reveals about the potential for human cooperation. Our research helps Wikipedians worldwide understand how to organize more effectively – and today, we’re going to share four lessons for anyone hoping to organize people for the common good.

Every light starts with a spark, and the basic unit of collaboration on Wikipedia is a single edit to a page – something that Valentyn Nefedov discovered when he was 12 years old, a bit bored at his parent’s house in a village outside the city of Lviv in Ukraine. 

Valentyn’s family’s village doesn’t have many photos on the Internet and Google hasn’t ever sent a camera car down its roads. But this village of 235 people did have a very small Ukrainian language Wikipedia page. And over a decade ago, a 12 year old Valentyn decided to add to it. He got a pencil and paper, walked down the roads, and counted. Later that day he added a sentence to his community’s Wikipedia page: “there are 42 garden societies in the village.”

Gardening is deeply important in Ukrainian society and garden societies are a basic social unit in many communities (Keegan 2023) Without a citation, the edit was eventually removed. That didn’t stop Valentyn, who went on to add information about his local media station, his favorite shows, and other small fixes.

The Mystery of Cooperation on Wikipedia

For scientists who study cooperation, the mystery of Wikipedia is how more than a hundred million people like Valentyn Nefedov and Christina Koch, all adding edits from small towns and outer space, from a constellation of independent nonprofit community groups, have somehow created an extraordinary collection of knowledge that anyone can freely enjoy.

Economists call Wikipedia a “public good” since it’s not exclusive, and since each time a reader or contributor becomes more knowledgeable, we all benefit. It’s also “peer-produced” – people share what they know and get what they need- it’s the world’s biggest mutual aid circle.

Unique Contributions at the Head of the Long Tail 

To illustrate how that happens, consider three Wikipedia pages that mean a lot to us in Central New York that each get edited by volunteers multiple times every month:

  • The first is our bus system, which offered 2.85 million rides in 2024
  • The second is Cornell University, which has over 16 thousand students; this photo of our New York City campus was published by NYC Wikipedian Jim Henderson – thank you Jim!
  • The third is Harriet Tubman, the 19th century abolitionist, spy, and social reformer who taught Sunday School at St. James African Methodist Episcopal Church in Ithaca

Some articles like Harriet Tubman’s are viewed by hundreds of thousands each month, especially during Black History Month. Cornell University, gets tens of thousands of views per month, often going up and down with student interest in admissions. Our beloved bus system gets viewed 200 to 250 times a month, and its data inform thousands of others – when we ask chatbots a question about the bus, they link back to Wikipedia.

In statistics, we use “distributions” to illustrate the shape of participation- illustrated beautifully through dinosaurs by our friend Tamara Broderick. In this chart, you can think about the left/right axis as popularity, and the up/down axis as how many articles get that number of views.

Across Wikipedia, most edits happen when people document something they uniquely know and care about – Christina’s spacewalk, Valentyn’s village, or our local bus – those contributions are the head of the Pareto distribution – while the few, popular articles are at the tail.

This is often how human cooperation works more broadly. On a day like today, for every person on stage giving a talk, there are dozens of people whose committed, less visible work has made today a success.

to support everyone, welcome people with unique perspectives

Many of the most important people and stories aren’t yet included in Wikipedia, especially from the majority world. One project, WikiLovesAfrica, works to change that by organizing communities in countries like Sudan, Tanzania, and Côte d’Ivoire to contribute photographs to stories about Africa.

In 2019, our lab supported WikiLovesAfrica to test reminder messages that helped recruit 70 experienced former contributors across numerous countries to add photos, as well as support and mentor first time photographers. Our experiments led to over two thousand more photographs that wouldn’t have been published that year- yielding recruitment insights used by other projects around the world.

So the first lesson from Wikipedia about cooperation is: to support everyone, welcome people with unique perspectives.

Managing Trust During Conflict

Sometimes something you care about becomes popular whether you like it or not. This is what Valentyn and millions of other Ukranians experienced on February 24 2022 when Russia invaded Ukraine in a war that is still ongoing. One thing Valentyn remembers about that dreadful day is that he was the person who created the Ukranian Wikipedia article for the invasion, a page that hundreds of thousands of people continue to access for accurate information about the war.

Here’s what the day of the invasion looks like in the data. Views to Ukrainian Wikipedia increased substantially, but active contributors declined as their lives were disrupted and people were drafted into military service. We are currently working with Valentyn to study and write about how the Ukrainian Wikipedia organized to meet this demand.

This brings us to a second question: can you trust an encyclopedia that anyone can edit during a conflict? Can you trust a resource that opposing sides have a chance to shape?

Valentyn told me about a debate in 2023 about how to describe a Petro Poroshenko who was president of Ukraine before the war. Some people called him an “oligarch” – and other people disagreed. On Wikipedia, anyone can click the “Talk” link to debate what should go onto the page. And they did, with great enthusiasm. 

Eventually Valentyn offered a suggestion: the community’s goal is a neutral page with the most reliable information from published sources. So he suggested that they cite all of the arguments next to each other: what the media say, what scholars say, what Poroshenko claims, and what Ukranian’s report in opinion polls. The community accepted this solution and the article has kept this structure in the years that followed.

Scientists have studied diversity and disagreement on Wikipedia by looking at how specialized individual editors are. In one study of English language editors, researchers classified Wikipedia contributors by how much they contributed to a wide range of topics or just ones about certain political issues (Shi et al 2019). They found that:

  • Many editors have diverse expertise
  • And other editors focus on one area of knowledge, sometimes on opposing sides of a political disagreement- for example only editing pages for politicians from a single party

Using this information, it’s possible to describe an article by whether it is edited by people just on one side, by people in the middle, or by people with different focuses.

The scientists then applied this analysis to articles in science, politics, and social issues across the English Wikipedia. For each article, they looked at whether the editors specialized in writing about the right or the left. In those cases, the dot for the article would be more on the right or the left of the chart. If contributors to the article had more diverse or even opposing expertise, the dot would be placed in the middle.

diversity and disagreement can make something better

This chart shows plotted the alignment of editors against the number of people who contributed to the article (the vertical axis). The researchers found that popular articles with more editors attract more diverse, polarized contributors that balance each other out, while less popular articles tend to attract more specialists.

Next, they studied the relationship between editor diversity and article quality. In this chart, articles further to the right are higher quality, and articles higher on the chart have more diverse, polarized editors. On average, the researchers found that articles with more editor diversity are considered higher quality by the English language Wikipedia community.

And that’s the second lesson from the science of Wikipedia – diversity and disagreement can make something better, even if it sometimes takes some debate to get there.

The Care Work of Community

Managing disagreements involves a lot of work, and it’s just one of the kinds of tasks Valentyn takes on in his role as a leader. 

As the war in Ukraine continued, people who lost loved ones looked to Wikipedia as a place to share their memories. But as people share personal stories, these memorials sometimes put people at risk and fell out of line with Wikipedia’s norms of citing secondary sources. That means someone has the unpleasant task of telling an editor that their article about a loved one is going to be deleted. This task falls on Wikipedia’s leaders, a role held by Valentyn. But instead of just deleting the pages, Valentyn wanted to remove them in the kindest way possible. So, he developed a protocol. When removing memorial pages he explained Wikipedia’s rules and why the memorial was in violation of them, then, he provided information about where they could share memorials, and importantly he passed along his condolences. 

Valentyn also contributed to an article with tips for people who might consider adding memorials. Guides like these can help reduce the risk of people experiencing the pain of losing their loved one to war followed by the removal of a memorial article on Wikipedia.

This brings us to our third question, what kind of leadership is needed to sustain trustworthy cooperation? 

This is an urgent question for Wikipedia. In the largest Wikipedias, like in the English language Wikipedia, new pages keep getting added but the number of editors has not remained at the high levels Wikipedia experienced when it was new, 20 years ago. 

Researchers have looked into why that might be. One possibility is that when people care deeply about a topic, the conversations can get heated- and people can sometimes be rude, disrespectful, needlessly aggressive, or make personal accusations.  

A group of researchers found that roughly 1 in 100 of the Wikipedians who participate in public discussions have ever left a toxic comment toward another person at some point in their past. While it’s rare, it’s common enough to make a difference (Smirnov et al 2023).

These researchers produced a chart that shows the impact of toxic comments on editor’s activity. It’s just a snapshot, but it does a nice job of visualizing what happens after people receive a toxic comment, which are represented by the red triangles. Not everyone responds the same. At the top, one person gets a toxic comment but keeps editing after. At the bottom, someone who was editing pretty consistently needed to break after receiving a toxic comment. And then in the middle, someone didn’t edit again after receiving a second toxic comment. When the authors looked at how this plays out overall, they found that toxic comments consistently reduce activity—between one half to two active days. 

These impacts are not distributed equally. Studies have found the women editors engage in what’s known as emotional labour—this is the work it takes to manage your own emotions and those of others. So for example, women editors have described avoiding work that could result in conflict and describe being the targets of online abuse, but they persevere because contributing to Wikipedia is important to them (Menking & Erickson 2015).

effective cooperation requires care

While these problems aren’t fully resolved, there is a group of people who take on special leadership responsibilities to help manage these issues—people like Valentyn. These leaders play a variety of roles.  

They wield the banhammer, which allows them to ban or unban users, delete and undelete pages, and apply page protections. With the banhammer, these leaders fight trolls. 

They’re also builders. They develop tools that can automate some of the work required to quickly tackle vandalism and help make Wikipedia more trustworthy. 

But importantly, they’re gardners. They help Wikipedia grow by giving editors feedback and helping them resolve conflicts. This often requires a lot of empathy and emotional labour—just like we saw with Valentyn. 

So our third lesson from Wikipedia is that that effective cooperation requires care. And that can be hard work. 

Sustaining Participation in volunteering over time

When we talked with Valentyn, he remarked that if people only hear about the hard work, they might not want to volunteer. So we asked him what makes him feel proud and fulfilled. Valentyn told us about the community – of times when people trusted him with questions, moments when he could mentor younger people to get started, and all the people he’s met around the world who care about free knowledge.

And this is a question not just for Wikipedia but for us all: how can we sustain participation in volunteer work?

Several years ago, the CAT Lab team met with a group of community leaders to imagine interventions to sustain participation on Wikipedia. We decided together to study the practice of gratitude.

When Wikipedians created a feature for thanking people for contributions to knowledge, they hoped to encourage a welcoming culture and also provide a way to encourage contributors to keep volunteering. By the time we started our research, Wikipedians were sending nearly 70 thousand of these private messages per month.

To study the effects of receiving thanks, we organized a group of hundreds of volunteers across four language Wikipedias and built a list of people whose edits were especially worthy of thanks. In a randomized trial, volunteers used our app to send thanks to over 2,700 newcomers and experienced editors alike. We then waited to see what happened over the following months. 

The study is still under review. But in our preliminary results, across all languages, for both newcomers and experienced accounts, people who received thanks remained active Wikipedians for longer periods of time and also spent more time making edits.

This study also discovered something powerful about gratitude that applies beyond Wikipedia: Gratitude spreads. People assigned to receive thanks went on to send 61% more thanks to others, usually to someone new. Sending someone appreciation causes them to share gratitude further to others.

That’s our final lesson from the science of Wikipedia: even when things are are hard, relationships, community, and gratitude sustain cooperation over time

Cooperation for the Common Good in a Complicated World

To scientists, Wikipedia is more than an encyclopedia- it’s a way for humanity to learn more about our collective potential. 

And what we learn from the science of Wikipedia is that cooperation depends on unique perspectives, (b) diversity and disagreement may be hard, but they can make things better, (c) cooperation is sustained through care, and (d) sparked and tended through the gratitude we share with each other. 

So when you see photos from the Artemis mission next month, read news about Ukraine, or look up your local bus route on Wikipedia, we hope these lessons inspire you to organize something that benefits everyone.

References