WHEN I WAS EIGHT OR NINE, my father bought an encyclopedia. To him, maybe because there had been one in his childhood home—a prized possession his parents had bothered to box up and ship when they immigrated in the 1970s—owning an encyclopedia was some sort of milestone, a marker of adulthood. I had trouble grasping the potential utility. Why do you need that, I asked, when you can use Wikipedia? This resulted in a game: we would come up with an arbitrary topic or question (What are the names of Jupiter’s moons? What was Kublai Khan’s love life like?), and see who could find the answer first—me using Wikipedia, him using the encyclopedia. After several rounds, I was declared the winner.
Wikipedia, at that time, was a relatively recent development, two or three years old and only a fraction of its present size. It was widely mocked for its potential to promulgate mistruths, because any anonymous jokester or propagandist could edit it. “Wikipedia is not a valid source” appeared as a standard proviso on school assignments. Just a few years later, in 2005, Nature would conduct an informal study comparing the accuracy of Wikipedia and Britannica, and judge it nearly a tie. By 2012, Britannica was out of print, and today Wikipedia is one of the ten most popular sites in existence. In study after study, it is found to be more reliable and more exhaustive than older, traditional encyclopedias. In 2018, The Atlantic called it “the last bastion of shared reality” online. The site features more than six million articles in English alone, and more than fifty million articles in its three hundred independently produced language editions combined.
A new collection, Wikipedia @ 20: Stories of an Incomplete Revolution, looks back on the first two decades of the free encyclopedia. Edited by Joseph Reagle and Jackie Koerner, the volume includes short essays by thirty-four contributors—scholars, activists, close observers, and dedicated Wikipedians. It was produced “in the wiki-spirit of open collaboration,” through an online platform called PubPub. Would-be contributors submitted abstracts and drafts, which passed through public and then editorial review, with some subset selected for inclusion.
Ostensibly mimetic of the Wikipedia process, this method did not yield Wiki-quality results: the consequent essays might have used a more thoroughly crowdsourced edit. Several are repetitive or overlap awkwardly. Two feature the same lengthy block quote from Denis Diderot, chief editor of the eighteenth-century Encyclopédie, but date it twenty years apart. Even more frequently quoted in the collection is the official Wikimedia “vision”: “Imagine a world in which every single human being can freely share in the sum of all knowledge.” Formulated in 2004 by Wikipedia cofounder Jimmy Wales, the sentence is rearticulated and parsed, phrase by phrase, in the capstone essay by current Wikimedia executive director Katherine Maher. It’s “an impossible, asymptotic vision,” she writes, but it’s also a vision for “a better world.”
Wikipedia is far from the only website with aspirational corporate rhetoric. But where social-media companies’ statements read like hokum, all but entirely disconnected from their actual monetized functions, Wikipedia does seem to do what it says it does. There is no hustle, no money-making operation running parallel to the user’s experience. Wikipedia does not sell advertisements. It has no personalized feeds or viral content or manipulative algorithms. My entry on Donald Trump reads exactly the same as yours. In 2020, you’d be hard-pressed to find scholars willing to parrot the pablum of a Facebook or a Google in a book for a university press. The essays collected in Wikipedia @ 20, meanwhile, are unabashedly idealistic. Contributors write of Wikipedia as “a great human endeavor, the largest collective project ever,” a “utopian project,” a step “on the road to an imagined utopia,” and even “a useful utopia for conceiving how people could cooperate productively without market relations.” You get the sense, reading through the various contributions, of stumbling on the last corner of the internet where early Silicon Valley optimism is substantiated.
As more and more of the internet is consolidated, discredited, and co-opted by capital, Wikipedia begins to look like a vestige of a bygone era. With its volunteer-run editing process and its open-source ethos, the site may be the one success of an early-internet ethos (crowdsourced, democratized information-sharing, with little centralized control) that otherwise has come to look like empty rhetoric. Wikipedia is produced independent of capital, and is not only competitive, but has fully demolished its competitors: encyclopedia companies that recruit and pay experts to write authoritative entries. Wikipedia functions on unpaid labor without being exploitative, because no one stands to gain much from the addition of minutiae to the entries for “Meta-joke” or “List of sandwiches” or “Cultural history of the buttocks.”
WHAT MAKES WIKIPEDIA’S success as a nonexploitative collective endeavor even more surprising is that it began with different intentions. At the turn of the millennium, Wales was already a cofounder of the start-up Bomis (home of “The Babe Engine,” which allowed users to search for X-rated images of “Bomis Babes”). He met Larry Sanger on an online discussion group devoted to Ayn Rand’s philosophy of Objectivism and hired him to head up Bomis’s newest project, a for-profit internet encyclopedia called Nupedia. By the end of 2000, due to a lengthy process whereby experts were recruited and their entries subjected to rigorous review, Nupedia only had two finished articles. Frustrated with slow progress, in January 2001, Sanger and Wales created a wiki where nonexperts could edit one another online.
The goal, at the outset, was to work up to 100,000 entries (more than most print encyclopedias)—and to make money. In fact, the site didn’t switch over from a dot-com to a dot-org until 2002, after Wales floated the idea of ad sales over the Wikipedia email list, prompting an insurrection among the contributor base. Spanish Wikipedians went so far as to splinter off, copying the entire contents of their language version to an external server. While Wales seemed only too happy to adopt contributors’ altruistic values ex post facto, Sanger grew increasingly disillusioned. Objecting strenuously to the site’s ethos of “disrespect toward expertise,” he founded a new encyclopedia, Citizendium. Like Nupedia, and its corporate competitor Microsoft Encarta, it was not a success: Citizendium still exists, but to date only 166 articles have passed editorial muster.
Meanwhile, Wikipedia ballooned—passing the two-million-article mark in the English edition in 2007—and evolved a culture and an administrative structure. The kinds of people who liked to edit Wikipedia entries, it turned out, were also the kinds of people who liked to design elaborate systems for keeping those entries error-free. Today, to remain on Wikipedia, an edit must pass through neural network machine-learning bots on the lookout for vandalism, then an army of “‘recent change’ patrollers,” at which point regular editors are alerted that an article in their areas of interest has been updated. All that before ordinary readers have a chance to flag, dispute, challenge, undo.
WIKIPEDIANS JUDGE ONE ANOTHER primarily on the basis of “edit count”—how many times they’ve successfully landed changes to the public site. Some have a tendency to get obsessive: one single user, for instance, deleted the term “comprised of” from over 40,000 articles. There are now over 40 million registered English Wikipedia editors, 131,000 of whom are considered active contributors. The sheer volume of rules—45,000 words of regulations about proper conduct, and millions of words in auxiliary documents—is a testament to the community’s zest for self-governance. While it makes for a highly consistent, accurate product, this “labyrinth of rules and guidelines,” Ian Ramjohn and LiAnna Davis explain, “keeps all but the most dedicated newcomers out.”
Complexity is far from the only barrier to entry. As Siân Evans, Jacqueline Mabey, Michael Mandiberg, and Melissa Tamani point out, “Wikipedia’s idealistic community guidelines—‘be bold’ and ‘assume good faith’—do not take into account the pervasiveness of online harassment.” They describe “a veritable hellscape of microaggressions within the Wikipedia community,” a closed-off and homogenous group—mostly male (80 to 90 percent, according to several studies), mostly white (77 percent, per less reliable data) millennials in Europe and North America. Wikipedia has more contributors from the Netherlands than from the entire continent of Africa.
There is a corresponding disparity in content: more articles on North American and European cities than others, disproportionately few on women scientists and Black artists, too many on military history. Only 4 percent of the geotagged articles on English Wikipedia are devoted to Africa. Fewer than a quarter of biographies are of women. Critiques of Wikipedia’s gender gap have been especially persistent. Many Wikipedia @ 20 contributors are activists (from groups like the Working Wikipedia Collaborative, Art+Feminism, the Black Lunch Table project) attempting to change Wikipedia from within, both by lobbying for better site-wide policies and by hosting “edit-a-thons” to bring in new contributors and diversify content. But even the critics, at least those included in the book, are loyalists; they characterize their work as “tough love.”
That some voices are louder than others, that they create arcane rules to disenfranchise outsiders, that representation is not equally distributed: these are the problems of any encyclopedia, any library, any news outlet, any democracy. If Wikipedia is a test case for techno-utopianism, it is also a test case for an older ideology similarly unfashionable these days—your garden-variety Enlightenment-era liberalism. As Jake Orlowitz writes, Wikipedia “created what the Enlightenment philosophers only dreamed of”; it is Diderot’s “Encyclopédie reborn in a digital age.” The encyclopedia is the quintessential Enlightenment project, embodying the aspiration to compile all human knowledge for the benefit of a more educated public. And though skepticism about liberal notions of political progress is justified, it does seem true that progress on Wikipedia is real, if slow: edit-a-thons are beginning the uphill battle of rebalancing and un-biasing the site, and the official Wikimedia 2030 project aims to “advance our world by collecting knowledge that fully represents human diversity.” If the trajectory of the moral universe is currently uncertain, it seems fair to say that the long arc of Wikipedia edit histories bends toward justice—or at least toward accuracy.
Rebecca Panovka is a writer and the coeditor of The Drift.