Wikipedia:Wikipedia Signpost/2020-01-27/Special report

Robert Fernandez contributed this essay to Wikipedia@20, a CC BY-NC 4.0 project collecting essays in anticipation of Wikipedia's twentieth anniversary. Robert graciously agreed to publish this piece (CC BY 4.0) on the 15th anniversary of The Signpost. Thanks to Joseph Reagle and Jackie Koerner at as well for assisting in the publishing arrangements. -S

As the largest and most successful online volunteer project, Wikipedia is rightfully celebrated as a triumph of crowdsourcing and volunteerism. However, Robert discusses the downsides of volunteerism and how volunteers can contribute to stifling innovation.-Editors of Wikipedia@20

At the turn of the millennium, the online encyclopedia, like hovercraft and jetpacks, was an unfulfilled promise of the future. Microsoft Encarta made it to the web in that year, and the venerable Encyclopedia Britannica had been on the web since 1994. Yet their subscription models, limited scope, and Web 1.0 functionality were not what people raised on Ford Prefect and Hari Seldon — touchstone characters of geek culture^[1] — had in mind. A real, universal online encyclopedia was an inevitability; the only question was what form this encyclopedia would take and how it would be created.

When Wikipedia appeared a year later on January 15, 2001, with its funny name and origins in a company otherwise dedicated to access to pictures of scantly-clad women and not open knowledge, its role as the world's encyclopedia was far from a fait accompli. In his PhD dissertation, Benjamin Mako Hill took a deep dive into the early workings of eight web encyclopedias.^[2] Wikipedia could have just as easily become a forgotten novelty of the dot com boom just as the other seven have become.

Interpedia was the oldest of the eight and was able to draw on the preexisting participant base and infrastructure of USENET. h2g2, with its famous branding from The Hitchhiker's Guide to the Galaxy and backing by the BBC, seemed an even more likely candidate for success. Everything2 was flush with money and attention from Slashdot, an early web driver of traffic and viral attention comparable to Reddit today. A teenage Aaron Swartz, now considered an open content legend and martyr, started The Info. Looking from the vantage point of 2001, who would have guessed that all these projects would be forgotten and Wikipedia would be a household name?

Hill concludes that Wikipedia succeeded in three key factors and none of the others succeeded at more than two of them. The first is the familiarity of the goal or product. Everyone knows what an encyclopedia is and should look like. (Though those born post-Wikipedia now have a different definition of encyclopedia.) But many of these projects were not quite encyclopedias, or not just an encyclopedia. Before I joined Wikipedia, I was an "editor" (the equivalent of a Wikipedia administrator) on Everything2, a project that seemed to encompass (almost) everything. This was a prospect that offered an intoxicating freedom but also served to confuse people unsure about how to contribute given a nearly unlimited choice of topic, format, and style.

This broadness of scope also served as an endless source of controversy and debate about how to define Everything. One of the site's longest-running controversies was a debate about the quality and appropriateness of a brilliantly incoherent and delightfully juvenile tone poem written about the Butterfinger McFlurry, a now-discontinued dessert product from McDonald's.^[3] More prosaically, the debate meant an ever-shifting series of goal posts as the site's possibilities narrowed, accompanied by lofty but vague rhetoric about quality and slogans like "a writer's site for writers." What could have been a rival to Wikipedia became the functional equivalent of a third-rate literary magazine.

If nothing else, Wikipedia's accidental genius is its branding. The branding is not just name recognition and puzzle globes and big Ws and stroopwafels. It's a shared goal that is specific enough and familiar enough that everyone can grasp and large enough that it can encompass everyone's individual interests and quirks. It's "the encyclopedia anyone can edit." What are we doing? Editing — a specific and familiar task, yet one that can accommodate anything from fussing over Oxford commas to writing lengthy essays. What are we editing? An encyclopedia — something specific that everyone can easily identify but something that can be literally about anything, from plants to trains to Pokémon.

The second factor cited by Hill was a relatively low barrier to entry. In order to achieve a critical mass of contributors, it needs to be easy to contribute. This includes all sorts of technological, social, psychological, and even aesthetic factors. Anyone who has watched a student wrestle with logging into a library proxy and try to search an unfamiliar database, then give up and just Google something knows that these barriers matter. Anyone who has given up halfway through the purchasing process on an e-commerce site has experienced the problem for themselves.

Wikipedia succeeded in lowering these barriers due to a number of factors. The innovative wiki software allowed essentially real-time collaboration. The markup language was simple enough to easily grasp, at least for the tech-savvy users of Web 1.0. The fact that you didn't even have to log in to edit meant you could quickly edit a typo with little learning curve and no commitment. Wikipedia had yet to develop the large thicket of policies and procedures that stymie new users today. Whether or not this is all still true has significant bearing on the future potential of Wikipedia, but it was true then, and was a large part of its explosive growth.

The final factor was the low social ownership of content. Many sites assumed that some sort of social reward in the form of recognition would be an incentive for people to participate. On Everything2, for example, this took the form of a video-game like system of experience points and levels. At the top of each article (called "writeups" on E2), the name of its contributor was prominently featured and the contributors could be rewarded with points by other users in the form of voting.

Intuitively it seems the lack of recognition and control of your contributions on Wikipedia would be a disincentive, as it was for me at first. Wikipedia records the author of every single contribution, down to punctuation, but finding that information requires some knowledge and experience with wiki software, rendering the authorship essentially opaque. It didn't hurt that Wikipedia's commitment to neutrality has caused the writing to develop an emotionless tone largely bereft of an authorial voice.

However, deemphasizing the idea of authorship has served to emphasize the existence of Wikipedia as a common, shared good. It also made many people more willing to contribute due to the absence of territoriality and implied ownership of content. People are more willing to alter and improve writing if they don't perceive it as belonging to another person. "Ironically," Hill writes, "the fact that Wikipedia made authorship less scrutable opened the door to deeper and more widespread collaboration."^[4]

Why is all of this important? If Wikipedia is able to change to meet the challenges of our jetpack-less future, it must know how to change, and to know how to change you must know how you succeeded in the first place, whether you examine the matter through Hill's framework or some other lens. However, Hill's conclusions are largely unknown even among specialists, and certainly to most of the core editor base of Wikipedia. As Hill notes, it's common to attribute Wikipedia's success to some combination of luck, timing, and technological superiority, despite the lack of empirical data for these conclusions. Even Joseph Reagle, in Good Faith Collaboration, concludes his book-length examination of Wikipedia by writing that it was "born almost as a happy accident."^[5]

Besides, you can't empirically measure luck in any case, so if you conclude that Wikipedia succeeded through sheer luck, you are basically concluding that the matter is inscrutable or unexplainable. This has more significant implications than a mere academic debate. When confronting new challenges for the encyclopedia, it's a case of the classic problem having a hammer and treating every problem as if it were a nail.^[6] The temptation amongst Wikipedia editors is to simply do the same thing all over again, even if the solution has no relationship to the problem, and wasn't even the solution in the first place.

On Wikipedia, the solution is to throw volunteers at it until it is fixed, somehow, eventually. Wikipedia is the world's most successful example of a crowdsourced volunteer project, of course, but we forget that many crowdsourcing projects simply fail. Crowdsourcing isn't the reason Wikipedia succeeded and we can't simply throw it at every problem as if it were magic. It's very much like today's technological hucksters who claim adding blockchain to everything will somehow result in a more successful and innovative product. This has even spread to include, amusingly, a recent Wikipedia competitor called Everipedia, which hired Larry Sanger as its CIO in 2017 in one of his desperate bids for post-Wikipedia relevance.

Without volunteers, of course, Wikipedia would have been impossible. The largest and most sustained volunteer mobilization in human history has resulted in the largest and most widely used information resource in human history. It has been so successful that any attempt to accurately describe the scope of what has been achieved inevitably resembles hyperbole. The entire project (or more accurately, collection of projects) has been driven entirely by those who volunteer their labor to generate the content or money to keep the servers running.

Naturally and appropriately, much of the discussion about Wikipedia has revolved around the dedication, ingenuity, intelligence, and other positive traits of volunteers. What this discussion omits is that, often, volunteers are terrible. They act out of their own self-interest, their social and psychological needs, and their biases and grudges. Some of this is perfectly appropriate, as no act can be purely selfless, and many volunteer projects are successful precisely because they have found a way to harness volunteer energy and satisfy the healthy needs of volunteers, and these needs can serve as incentives to continued volunteer participation. In practice, however, many volunteers, consciously or not, put their own needs ahead of the mission they purport to be serving.

The most widely cited example of a discussion of this problem is probably Yochai Benkler's seminal paper "Coase's Penguin, or, Linux and the Nature of the Firm."^[7] Benkler coined the word jalt to describe the combination of jealousy and altruism affecting those working on volunteer projects. Though it is seldom discussed in that context, jalt has become one of the most powerful factors shaping the behavior of a number of the most active participants on Wikipedia. Jalt is a natural, understandable impulse. After all, if you're a pizza delivery guy or IT help desk employee who has contributed thousands of unpaid hours to Wikipedia, it's natural to get a little annoyed when Jimmy Wales gets to honeymoon on Richard Branson's private island.

A little green-eyed jealousy is harmless, but real problems begin when that altruism blossoms into a sense of entitlement. In 2011, an episode of This American Life focused on Daniel H. Weiss, the former manager of the gift shops at the Kennedy Center for the Performing Arts in Washington DC.^[8] (Weiss is now president and COO of The Metropolitan Museum of Art in New York City). The gift shops were staffed entirely with volunteers, and the Kennedy Center had some three hundred of them. "Many of them," Weiss noted, "simply wanted to be in the Kennedy Center to support the mission," an admirable, altruistic impulse.

The well-stocked and well-staffed gift shops were somehow not generating that much money for the Kennedy Center, and it was estimated that the shortfall was about 150 thousand dollars. Weiss was charged with getting to the bottom of the matter, taking him from investigating the books to undercover stakeouts with a United States Park Police detective. The cause was not one or two thieves or a ring of crooks, it was everyone. Despite (or as we will see, precisely because of) the altruism of all involved, theft of both cash and merchandize was rampant amongst the hundreds of volunteers who staffed the gift shops.

Weiss recalled that:

There were some volunteers who were taking money. There were some young employees who were taking money. There were lots of people who were taking merchandise, at every level. People were all stealing from this wonderful, uplifting organization because they could. Because it was easy and it was available.
— This American Life, 431, See no Evil, Act Three

According to This American Life:

...people weren't just grabbing t-shirts. Some had been taking cash-- mostly small amounts, like a cab fare home. After all, they'd just worked three hours for free. It's just a few dollars. What's the harm? But still, they put their hands in the cash box, took some bills out, and put the money in their pockets.
— This American Life, 431, See no Evil, Act Three

The altruism of these volunteers has grown beyond petty jealousies and grown into a sense of entitlement that provides a self-justification for acts that cause direct harm to the mission that they profess to be working to support. The harm need not be physical or monetary to directly impact the mission, it can be a harm to morale or harm that prompts volunteers to leave a project, or never engage with it in the first place.

Likely hundreds of examples of this kind of harm can be found a day on Wikipedia. The most common is what's referred to as "biting the newcomers," or hostile engagement with editors new to Wikipedia.^[9] This has been cautioned against since the earliest days of Wikipedia, but if anything the practice is accelerating as Wikipedia's policies, procedures, and the use of jargon and acronyms for both grow like kudzu. It's an easy enough practice to fall into even for well-meaning editors, who tire at having to deal with the same issue a hundredth time, and even easier for the many editors who proudly proclaim their abrasiveness and, to paraphrase Tennessee Williams, imagine themselves paragons of frankness.^[10]

This is, of course, another example of entitlement promoting volunteers to place their own personal needs over the health of the project. For all the triumphant citing of metrics by proponents of Wikimedia projects, one unmeasurable metric that is never cited is the loss of the volunteers who depart or never engage in the first place because of the barriers to entry, one of the most cited of which is hostile editors.

As we've seen from Hill's framework, a low barrier to entry was a key factor in the explosive growth and success of Wikipedia. The editors needed to create, maintain, and update millions of articles will not continue to join the project if the existing ones are hostile. This is all the more critical as Wikipedia grows into a part of the online and information infrastructure and fewer people see it as a distinct project they can participate in themselves. Elsewhere I've called this problem the "conceptual barrier," the inability or resistance of people to seeing themselves as having the ability and capability to edit Wikipedia.^[11]

Studies have shown that many readers see Wikipedia as one of many search engine results, not as a distinct resource they could search on their own.^[12] This problem will only grow as more people interact with information from Wikipedia in different ways: on mobile devices, through the Google Knowledge graph, read to them by an Alexa or other smart device. New editors cannot be recruited to volunteer on Wikipedia if they cannot overcome the barriers to entry, or even conceive of Wikipedia in the same way as its existing editors do.

Yet most of those existing editors, at least the most vocal ones, refuse to acknowledge dwindling volunteer recruitment as a problem. Much of this refusal is likely a result of the culture of Wikipedia itself. It's been well established that the culture of an organization is derived from its original participants and persists long after those individuals depart.^[13] Wikipedia grew out of early online and open-source communities, and undoubtedly this has contributed to its flexibility and dynamism. However it has also contributed to its exclusionism. Paradoxically, the encyclopedia "anyone can edit" grew out of a community hostile to outsiders, where new participants were unwelcome because they were unaware of or thought not to share community norms and values. Recall the concept of "eternal September" from USENET, when new users were perceived to swamp and disrupt online communities at the beginning of the fall semester when they regained internet access after returning to school.^[14]

This inherited culture has also influenced how decisions are made and who gets to make them on Wikipedia. Decision-making on Wikipedia revolves around "consensus," where those participating collaboratively come to a conclusion, through a process of voting and discussion. This is justifiably celebrated as transparent and democratic, but there's also a deliberate process to exclude people from the discussion. New editors arriving to participate in a discussion are rejected as "meat puppets" and even the opinions of established editors are rejected if they've been asked to participate or become aware of the discussion through certain channels, such as a social media post.^[15]

The principle of decisions being made by those who are doing the work and supporting the mission grew out of those online communities that gave birth to Wikipedia's organizational culture, and the principle seems logical enough for an open-source software project. Can this kind of exclusion be justified by an encyclopedia that "anyone can edit" and is the world's largest information resource? Why should decisions that affect how information is disseminated around the entire world be made by a handful of participants who are almost certainly English-speaking white males privileged by their access to computer literacy, free time, and awareness of a particular discussion?

In practice, to have a voice in a Wikipedia discussion requires a combination of stubbornness and privilege. It requires the knowledge and experience of how to monitor these discussions and the free time to participate in them before they conclude. Even well-established Wikipedia veteran editors don't have a voice in these discussions, because they don't constantly keep tabs on particular discussion forums. Effectively, Wikipedia decision-making has been ceded to a small group of obsessives whose hobby is Wikipedia governance.

In 2015, an article about a Wikipedia editor named Bryan Henderson was widely discussed in American media outlets,^[16] and Henderson was even interviewed for a brief profile on CBS News. Henderson, who edits Wikipedia under the user name Giraffedata, is obsessively focused on one thing, removing the words "comprised of" – a phrase he considers grammatically incorrect – from all Wikipedia articles. In the media and on Wikipedia, Henderson's quest has been largely treated with bemusement.

It's more serious when you realize this is the basic dynamic for Wikipedia decision making and control. The logical, sane response to disagreeing with Giraffedata is to shrug and move on. Since decisions are by those who participate in a localized discussion, leaving cedes the decision-making power to those willing to engage in the least logical and sane response. This incentivizes not just obsessive but also belligerent behavior and even harassment, and empowers those privileged with the time and resources to engage in this behavior. Minor quibbles about grammar is one thing, but these techniques are frequently used by political ideologues, ethnic nationalists, and conspiracy theorists. Professor Bryce Peake called this the "hegemony of the asshole consensus."^[17]

Even without a particular agenda, a potent sense of volunteer entitlement is a key driver of the asshole consensus. This is directed not just at new editors; in many cases the worst of it is directed at volunteers in key positions and employees of the Wikimedia Foundation. An exemplar is a comment unironically directed at a volunteer MediaWiki software developer: "We, the community are the souvereign [sic]."^[18]

If it were only a matter of it being an unpleasant or even hostile experience for many contributors, that unpleasant reality could be arguably dismissed if it resulted in a successful encyclopedia for both volunteers and readers. This is exactly what the asshole consensus argues. A quote frequently deployed in this context is this: "We are here to build an encyclopedia, not to sing kumbaya, and this is a shop floor."^[19] The intellectual labor of encyclopedia creation is likened to a fetishized male, blue collar workplace, diminishing and dismissing other modes of contribution, styles of communication, and types of volunteers.

The framing of the asshole consensus rests on a priori assumptions that this behavior is necessary for a successful project, and actually results in one. By their standards, it is successful, successful for them. When measured against the publicly stated mission, norms, and principles of the project — "be bold," innovate, treat one another with civility — it is an utter failure.

Innovation in particular suffers from the established consensus. One infamous example is a notorious controversy in Wikipedia's recent history that revolved around something called "superprotect," briefly instituted in 2014. Wikipedia administrators have always been able to protect pages, everything from articles to pages containing programming code, from changes due to vandalism and other mischief. When the Wikimedia Foundation wanted to prevent administrators on the German Wikipedia from removing a new feature they had developed called Media Viewer, they created a new type of protection that even administrators could not override, only some WMF staffers.

As you might imagine, this created a gigantic uproar from not just the German Wikipedia, but the entire Wikimedia community. There's no question this was a serious misstep by the WMF. They acknowledged as much when they removed it a year later, with the then-executive director stating that it "set up a precedent of mistrust."^[20] To this day, superprotect is constantly invoked in response to a perceived threat of WMF overreach or to assert the privileges of established editors and administrators over the Foundation.

It's worth examining what this event illustrates about Wikipedia community dynamics. While it is framed as a triumph of the community over oppression, the reaction is part of a pattern that constantly reoccurs in situations that lack any oppressive acts by the WMF or other forces: An improvement that provides new functionality is introduced and it immediately faces hostile resistance from entrenched volunteers.

Not coincidentally, the innovations that bring the most vociferous resistance are ones designed to improve the reading and editing experience for those outside the established community: Media Viewer was designed to make viewing images and media on Wikimedia projects easier; Visual Editor introduced a WYSIWYG editor so new contributors didn't have to wrestle with confusing wiki markup text; the content translation tool was designed to make translating articles between different language Wikipedias far easier. It is no accident that the text-heavy interface of today's Wikipedia looks much like it did in 2001 and is not geared towards today's internet users; it is a direct result of this resistance to change and innovation.

Wikipedia's community has inherited an antipathy to new contributors from the early internet projects it grew out of. Sometimes that antipathy is expressed in the form of direct hostility towards the contributors themselves, other times it is in resistance to changes and improvements that improve the recruitment or experience of new contributors.

Even in the form of indifference the antipathy is not benign, as established contributors center their own experience over that of the new contributor or reader. The actual mission of the project is secondary, the sense of entitlement of the volunteer is paramount. The entitled volunteer resists improvement because it upsets their comfortable vision of how things should be done in a way that accommodates them, regardless of the harm this does to the project and mission they are supposedly volunteering on behalf of. They are indifferent to examinations like Benjamin Hill's of how and why Wikipedia succeeded; they prefer an explanation involving serendipity or inevitability because what matters is not the success of Wikipedia's mission, but their own gratification and entitlement.

Wikipedia was able to surpass projects like Microsoft Encarta and the Encyclopedia Britannica because it brushed aside the gatekeepers in favor of flexibility and innovation. Over the last two decades, however, Wikipedia's institutional culture has exchanged flexibility for bureaucracy and established volunteers have set themselves up as the new gatekeepers. Collectively they form a Team Encarta stifling innovation in favor of ossification.

Footnotes