Wikipedia:Wikipedia Signpost/Single/2024-03-29

A lack of technical support for interactive content on Wikimedia projects was lamented in a wide-ranging discussion on Wikimedia-l and elsewhere over the last two months. In particular, several community members expressed deep frustration about the state of the Graph MediaWiki extension, which had been disabled in April last year due to security vulnerabilities in the underlying third-party Vega framework (Signpost coverage). Back then, a Wikimedia Foundation representative had stated that "My hope is we can maybe restore some functionality in the next week or so." But eleven months later, graphs and charts remain deactivated, replaced by a prominent error message in many Wikipedia articles - despite extensive discussions about possible solutions.

Basque Wikimedian Galder Gonzalez Larrañaga (User:Theklan) opened the Wikimedia-l discussion by decrying this state of affairs:

All the solutions proposed have been dismissed, but every two months there's a proposal to make a new roadmap to solve the issue. We have plenty of roadmaps, but no vehicle to reach our destination.

He contrasted Wikimedia projects with e.g. "a place like Our World in Data [which] has been publishing data and interactive content with a compatible license for years". Several other Wikimedians likewise voiced their frustration about the lack of progress in getting graphs re-enabled.

Marshall Miller, Senior Director of Product at WMF, acknowledged these concerns on the mailing list, stating that:

to support graphs and other interactive content, we would need to take a step back and make a substantial investment in sustainable architecture to do it – so that it works well, safely, and is built to last. And because that’s a substantial investment, we need to weigh it against other important investments in order to decide whether and when to do it.

I know that it is very frustrating that the Graph extension has not been operational for many months – it means readers haven’t been seeing graphs in articles, and editors haven’t been able to use graphs to do things like monitor backlogs in WikiProjects. Over the months of trying to find a way to turn graphs back on, it has become clear that there isn’t a safe shortcut here and that the path forward will require a substantial investment – one that we have not yet started given the other priorities we’ve been working on.

How did things get to this point? Several proposals and plans had been pursued since the discovery of the XSS vulnerability in April 2023:

• One proposed solution (phab:T336595) was to re-enable the feature but treat the Vega code for graphs "like other dangerous content (Javascript, CSS) and restrict editing it to a small, trusted set of users" (similar to Interface administrators). However, this option was eventually abandoned in favor of a different approach:
• To sandbox the graphs into an iframe (a separately loaded part of a web page). This option (phab:T222807) had in fact already been proposed back in 2019 out of general security concerns, long before the discovery of the current vulnerabilities in Vega. By November 2023, some progress appears to have been made on its implementation. But serious performance problems remained due to complexities around caching. What's more, the iframe approach met principled objections by WMF engineer Timo Tijhof (User:Krinkle), who argued that it "severely limits our audience, technical choices, and assumptions going forward, and I think negatively impacts our mission and reach in a way that is in direct contradiction and violation of principles that thus far have been accepted without question" (citing the Wikimedia Foundations Guiding Principles and other WMF engineering conventions). In particular, Tijhof worried about the redistribution of content in other forms (citing examples including "Kiwix, IPFS and Apple Dictionary"). Noting that e.g. the Wayback Machine "consistently fails to archive [...] non-trivial JavaScript pipelines", he argued "there can be no JavaScript requirement for fundamental access to content". Lastly, as noted by WMF security engineer Scott Bassett, while the iframe sandbox "would reduce the risk of running potentially dangerous javascript within a user's browser, [it would] not eliminate the risk entirely." The iframe task was eventually closed as "Declined" earlier this month.
• Other options that had been proposed early on included sanitizing the input. But this approach was described as "extremely tricky to get right (after all,vega failing at it is how we landed in this situation)."
• Update the underlying library: As noted in a WMF FAQ, the Graphs extension is based on a very outdated version of the Vega library: "The last upstream release (bugfix or security) of Vega 2.x was in January 2017. Vega 5 was released in March 2019 and is still under active maintenance and development." However, by May 2023, initial hopes had been dashed that an upgrade to version 5 would suffice to resolve security concerns. The WMF's eventual plan to fix the Graphs extension (apparently now abandoned too) envisaged the version update and eventual deprecation of version 2 only in combination with the sandbox solution. What's more, as detailed in the FAQ, migrating to Vega 5 will necessitate the conversion of existing graph templates and modules, much of which needs to be done by hand (although aided by a translation tool).

Rarely, if ever, has there been a software issue that affects Wikipedia content so visibly for such a prolonged time. By January 2024, User:Sj estimated that it had "already conservatively affected 100M pageviews." According to a September 2023 analysis, over 1.3 million pages are impacted across all Wikimedia projects - the vast majority of them (1.16 million) on the Arabic Wikipedia.

Still, on English Wikipedia, only 19,160 pages were affected. (Those numbers likely already reflect the manual removal of broken graphs from many pages.) In a more detailed 2020 analysis, volunteer developer User:Bawolff had found that "the graph extension is used on 26,238 pages [on English Wikipedia]. However, most of these are in non-content namespaces, from a template that generates a graph of page views for a specific page (w:Template:PageViews graph). There are 4,140 pages on en.wikipedia.org in the main namespace that use graphs. [...] As a percentage, that's 0.07% overall, 0.2% of "Good Articles", 0.3% of Featured Articles." Another Wikipedian reported that "In ruwiki, interactive Lua-based graphs are used in more than 26000 articles about settlements and administrative units through https://ru.wikipedia.orgview_html.php?sq=&lang=&q=Module:Statistical (also, more than 8000 on ukwiki, etc.)."

Several users questioned whether the full interactive functionality of the Vega library was really needed, arguing e.g. that "Most graphs on wiki are simple bar/pie/line charts. These could be produced quite easily using even a language like Lua."

WMF engineer Gergő Tisza (who appears to have done much of the technical work on the aforementioned iframe solution) observed that

Interactive animations were very much part of Yuri's vision for the Graph extension, but during the decade Graph was deployed in production the number of such animations made was approximately zero. [...] Instead, both gadgets and Graph usage are mostly focused on very basic things like showing a chess board or showing bar charts, because those are the things that can be reused across a large number of articles without manually tailoring the code to each

Concretely, Bawolff had observed in his 2020 analysis of the usage of graphs on English Wikipedia that:

Most of these are simple static graphs. Some notable exceptions is interactive time scale maps, such as the one at w:Template:Interactive COVID-19 maps and the one at w:List_of_countries_by_carbon_dioxide_emissions, which shows how geographic data evolves over time (See also w:template:Global Heat Maps by Year). Also the graph at w:Vancouver_Whitecaps_FC. Nonetheless, I have yet to see any examples where a graph based visualization makes what would otherwise be a difficult concept clear, or where the visualization stops me in my tracks, and is core to my understanding of the article.

Volunteer developer TheDJ even argued "let's be honest... the interactivity-part has been an 8 year long nightmare. Maybe its time to put that to bed and accept defeat."

On the other hand, Galder titled his post that opened the Wikimedia-l discussion "We need more interactive content: we are doing it wrong". He took a much wider view, arguing that the WMF's failure to get graphs working again was just one example of wider stagnation and lack of progress towards the goal that "By 2030, Wikimedia will become the essential infrastructure of the ecosystem of free knowledge" (quoting from a 2017 strategy document). Besides Our World in Data, Galder named several other educational websites that have surpassed Wikimedia projects on interactive content, e.g.:

• Wolfram Alpha is like a light year ahead us on giving interactive solutions to knowledge questions [...] That's also "free knowledge".
• Brilliant (https://brilliant.org/) is brilliant if you want to learn lots of things, like geometry or programming. Way better than Wikipedia. But... you need to pay for it.

(Other parts of the mailing list discussion focused on MediaWiki's shortcomings with regard to video.)

In her op-ed in this Signpost issue, Maryana Pinchuk, the Wikimedia Foundation's Principal Product Manager, pushes back against such proposals, reporting that at an event last fall, she "heard many Wikipedians express concern about where pursuing this strategy could lead us. There was fear of making Wikipedia into something it isn’t. There was also fear about the cost and risks of building big new software features and trying to compete with massive for-profit technology companies for users. I think all of these concerns are very valid."

In the Wikimedia-l discussion, Wikipedian and former WMF engineer Ori Livneh argued that direct comparisons with sites that do not contain user-generated content may severely underestimate the additional engineering work to implement such interactive features on Wikipedia. He pointed out security engineering as a bottleneck at the Foundation holding up such work:

The critical issue is *security*. Security is the reason the graph extension is not enabled. Security is the reason why interactive SVGs are not enabled. Interactive visualizations have a programmatic element that consists of code that executes in the user's browser. Such code needs to be carefully sandboxed to ensure it cannot be used to exfiltrate user data or surreptitiously perform actions on wiki.

The bar for shipping security-critical features is high. You can ship code with crummy UX [user experience] and iterate on it. But something that touches security requires a higher amount of up-front technical design work and close scrutiny in the form of peer review. And this means that it cannot progress spontaneously, through sporadic bursts of effort here and there (which is how a lot of engineering work happens) but requires a solid commitment of focused attention from multiple people with relevant expertise.

There are engineers at the Wikimedia Foundation and in the technical contributor community with the relevant expertise but as a rule they are extremely oversubscribed. My recommendation would be to engage them in crafting a job description for this role and in reviewing candidates.

On March 26, the WMF invited feedback on "the Product & Technology draft key results for next fiscal year. They aim to explain what outcomes we are working towards" as part of the 2024/25 annual plan. In reaction, Galder noted that "there's no single mention to this [Graphs outage problem], nor to improving the multimedia experience". In a discussion on the talk page, Miller said that "we are working on a possible plan for graphs, but I'm not sure yet what its scope will be or when we would resource it if we proceed with that plan".

HaeB, sawyer-mcdonell

The World Jewish Congress has published a report titled "The Bias Against Israel on Wikipedia", which has been covered by The Jerusalem Post, Jewish News Syndicate, and Spectator Australia (paywalled). The author, Dr. Shlomit Aharoni Lir, is an academic, described in a related recent publication as "a poet, essayist, lecturer, and gender studies scholar [who] holds a research fellowship at Bar-Ilan University and is a lecturer at Achva College". She had previously published a peer-reviewed paper about gender bias on Wikipedia (Signpost coverage). The present report does not seem intended to be an academic publication, although it has already been used as a citation in the article Wikipedia and the Israeli–Palestinian conflict.

The report criticizes Arabic Wikipedia's "blackout" in solidarity with Palestinians (see Signpost coverage), the English Wikipedia's coverage of the Holocaust, and its general "bias against Israel", which the author argues is exemplified through content and sourcing bias, "deletion attacks", editing restrictions, "selective enforcement" by administrators, and "anti-Israeli editors".

Lauren Dickinson of the Wikimedia Foundation's Communications Department told The Signpost that several staff had reviewed the document, and found that "the WJC report makes a number of unsubstantiated claims of bias on Wikipedia. It lacks adequate references, quotes, links, or other sources to support its purported findings. Further, the report misunderstands Wikipedia's NPOV policy, as well as the importance of anonymity for user privacy on Wikimedia projects."

But what is the deal, really? Let's take a look.

JPxG

The problem with this report is that many of its suggested improvements are things we've already been doing (publicly, and in a very prominent way) for decades; and many of the rest are intrusive threats to the personal safety of editors and administrators. This makes it hard to take most of its claims seriously; for example, if somebody thinks there's no way to see who has added text to a Wikipedia article, it seems easier to tell them where the "history" tab is at the top of the page, rather than create an unaccountable editorial council and appoint them to it.

The 19-page report, which focuses on the English Wikipedia, "is based on research, content analysis, and interviews with Israeli Wikipedians"; its overview of challenges to Wikipedia's ideals include "The Power of the Admins and Beurocrats" [sic], as well as the gender gap (see Signpost coverage). Their citations on the gender gap include a survey taken in 2008 saying that Wikipedia editors were mostly male,^[1] and a paper from 2011 which compared Wikipedia to Encyclopedia Britannica on coverage of women in historical biography lists and concluded that Wikipedia had "significantly greater coverage" and its articles were "significantly longer than Britannica articles [...] for every source list"^[2]. It is somewhat unclear what, if any, relation the gender ratio of biography articles (or indeed of the editoriat) has to Israel and Palestine; while there is indeed a paper from 2014^[3] (cf. our coverage) that talks in greater depth about lower participation rates for female editors, and it's uncontroversially true that there exists a gender disparity among Wikipedia editors, it's hard to see what the connection is here. It is a little embarrassing, but so are the MoS arguments, and those don't have anything to do with Gaza either. The closest the report comes to making a connection between the gender gap and Gaza (apart from the alliteration) is to say:

Sure: it is true that we live in a society, and that the biases of that society pose issues for our attempt to write an encyclopedia that is both neutral and based on direct citations to sources written in that society. This is a problem apparently endemic to all encyclopedia writers, even our Britannic forebears; and it is the subject of much ongoing reflection and work, which is altogether good and proper to do. But it is not really clear how this relates to the main claim of the report, which is that Wikipedia is biased against Israel.

Regarding administrators, the report mentions issues with "concealment of decision-makers' actions, alongside the significant authority wielded by anonymous administrators who can delete entries and block participants without accountability". This may come as a surprise to readers, who may justifiably consider Wikipedia to be one of the most transparently operated major sites on the Web, and indeed in the modern history of the Web. Editorial decisions, discussions about those decisions, administrative actions, and the edits themselves are meticulously logged, to the second, in full public view, on a page that automatically generates a display of precisely which changes were made in the edit, or what actions were taken, by whom. When someone is banned from Twitter or Facebook, there is not an up-to-the-second log of which specific person pushed the button; there may be a form letter emailed to them later, but the act itself is not public, nor is it really disclosed at all. On Wikipedia it is; moreover, it can be reviewed and contested publicly. There's a public noticeboard for review of administrative actions; the Arbitration Committee regularly rules on cases of administrator misconduct, often deciding to remove admins. There is a large, highly regimented formal process for deleting articles (and a second one for formally appealing deletion decisions). Indeed, there are incidents in which administrators act beyond the bounds of propriety, but it is not really clear that Wikipedia is a "failed state" in any meaningful way with respect to their actions.

For example: this is a full public log of every formal action that I (Signpost editor and Wikipedia administrator JPxG) have ever taken.

You can see, in this log, that Ayamediainc was indefinitely blocked on March 12, 2024, at 04:45 UTC, for violation of the spam guidelines.

On their user talk page is a publicly viewable template indicating the name of the page in question, and an explanation of the specific way that it was in violation of the policy.

The log for that user account indicates it was created at 02:47 that same day, and two hours later they created a page at User:Ayamediainc/sandbox with the summary "Added sections from Ayamediainc profile and added background information". The log also specifies that this edit tripped two automated alerts, because text on the page matched patterns that are strongly associated with spam. The policies, the user warnings, the block, the reason for the block, and the identity of the blocking administrator (in this case me) are publicly viewable and can be audited by anyone. If the user appealed the block, that too would be a matter of public record, as would the response of whichever seprate administrator handled the appeal (which wouldn't be me).

One of the demands I can easily agree with: the "transparent editing history" item, which exhorts Wikipedia to "ensure that all changes to articles are transparent and traceable", which "helps in identifying editors who may consistently introduce bias into articles".

This stance is shared by the Wikipedia community, who implemented it 23 years ago; a link to the (mostly) complete history of every article has been a central element of the top of every Wikipedia page since the year 2002 (and the feature was incomplete and less prominently linked the year before that).

If there is some other website which provides greater transparency into its administrative and editorial decisions, perhaps it would provide a useful model for us to emulate. However, it is hard to come up with one. The histories viewable for every single Wikipedia article track every modification ever made to them, from major copyedits to em-dash fixes, and are permanently attributable to the editors. It's hard to come up with any way to increase the transparency of the process, except for personally doxing the administrators and editors.

This is one of the suggestions vaguely alluded to in the report, and later said explicitly: to this, I may offer the rejoinder that in that block log you can see me issuing blocks to a wide range of people encompassing bored schoolkids, scammers, vandals, and seriously disturbed and hostile individuals who carry decades-long grudges. I am a volunteer who edits for fun. These are not, generally speaking, people I would prefer to know where my family lived, especially not the guy who capped off a decade and change of Wikipedia harassment career by going to jail for making dozens of graphic death threats to the Merriam-Webster dictionary. That was not a made-up example: this guy is real and he's one of the hundred or so entries on the long-term abuse page. As for editorial integrity, I can definitely imagine what effect it would have on our article about any given scandal of corporate malfeasance or government corruption if somebody were able to instantly file vexatious lawsuits against individual editors. I just cannot imagine it being a good one.

The report says that "network bullying and discriminatory treatment increase when there is no personal responsibility and acting under cover of anonymity is possible". In general, my experience over the last twenty years of hanging out on the Internet suggests that network bullying also increases a lot when every person you ban from a website has trivial access to your home address, including this sicko.

One of the report's proposals for new features that should be implemented is to "host forums and discussions within the Wikipedia community to address concerns about neutrality and gather feedback for policy improvements"; it is not specified how this would fit into the existing directory of centralized discussions, dashboard, project-wide Request for Comment process with fifteen categories, six Village Pumps, and dozen or so noticeboards, including a neutral-point-of-view noticeboard and dispute resolution noticeboard. It is not explained why these venues are insufficient; none of them are even mentioned.

Generally, a persistent problem with the first part of the report is that it repeatedly claims (either explicitly or through insinuation) that Wikipedia lacks a process to deal with some issue, and gives no evidence to support that claim, when in reality Wikipedia not only has a process, but has had it for a very long time (sometimes more than 20 years) and it forms a central part of the site's administrative apparatus upon which editors spend hundreds of hours daily rigorously documenting all of their actions with direct references to policies and consensus. It seems that the main objective is to establish (or at least repeatedly assert) that Wikipedia lacks self-governance, or that it is unable to handle contentious issues, or that nobody has ever realized until now that it was possible for people to be rude about politics online; then this is used as the basis to propose all sorts of bureaucratic impositions, most of them done by external groups (perhaps including the one that made the report).

This is bad. While it is indeed the case that some people are biased towards one view or another, it seems unlikely that instituting binding top-down procedures (like an official oversight committee to dictate content at the behest of external consulting agencies and lobbyists, as is also suggested in the report) would arrive at a remotely better result. It's also not clear why we lack these: the report seems to be unaware of basic features like page history, which allow anyone to see who's written an article. Simply clicking on the history tab seems like an easier solution than doxing every editor or subjecting sitewide editorial decisions to random external think tanks.

If an unknown detective arrives at the murder scene and demands to be given authority over the investigation, of course there are questions about whether he has jurisdiction, but even before that, it seems relevant to note whether the victim is actually dead. If he's sitting at the dinner table asking what you're doing in the living room, this seems like a significant detail in the murder investigation.

After this, there are a number of specific examples given of pages concerning the conflict in Gaza. Frankly, this may be true: political articles are biased sometimes. They tend to be edited by a variety of people with various allegiances, who argue at great length over everything from trivial minutiae to the timelines of major events. It is not clear this can, or should, be fixed. Existing research seems fairly consistent on the idea that conversation and collaboration between people of various perspectives improves the overall quality of articles.

This process does involve a great deal of tedious, unpleasant argument; ask anybody who's edited (or worse, created) a contentious article on a political subject. But ultimately, if an article about an event is unduly biased towards one side, the only solution is to edit it in a way that fixes the problem. This sometimes results in contention, in which case a discussion must be carried out between the people who disagree, and if they cannot resolve it between themselves, there are a variety of ways of seeking external assistance.

There are a broad number of existing venues to which disputes can be brought, and through which disputes can be addressed. It's true that these processes often take a long time to resolve, and it's true that in the meantime an article can be grotesquely biased. It's even true that an article can stay grotesquely biased for a while. Any editor active on political topics can tell you about their personal Alamo, perhaps several of them, where they showed up armed with reason and common sense, and a dozen idiots showed up armed with idiocy, and they were crushed in ignominious defeat. In fact, maybe I am one of the idiots who ruined your article, and maybe you are one of the idiots who ruined mine. And we are both the idiots for some other third person. It is just an inescapable aspect of living in a society: sometimes people, even people collaborating on a project, have irreconcilable disagreements. This has happened thousands of times, and we have mechanisms for dealing with it; they may not always work perfectly, but it remains to be seen what other way things could possibly be run and work anywhere near as well.

Wikipedia policies and consensus processes are the worst form of collaborative encyclopedia-writing projects, except for all the other ones.

JPxG

Subsequent to the publication of this report, on March 20, the Arbitration Committee announced that a user account created that day with zero edits (Mschwartz1) would be granted extended-confirmed status "for the exclusive purpose of participating in a case request about Israel-Palestine". Extended-confirmed status, generally, is given to accounts with over 500 edits that are at least 30 days old (and is currently a prerequisite for any editing activity in the Israel–Palestine area, formally designated a Contentious Topic).

A long discussion ensued at the ArbCom noticeboard's talk page, as well as at the unmentionable BADSITE, in which it was speculated that this may have been an employee of the organization publishing the report (due to timing that closely aligned with the publishing of the report). Arbitrator Barkeep49 said that it "may or may not be a coincidence", explaining that "I can say the conversation with us that led to this grant has been going on since early February." Limited information has been made available about the nature of the editor, although the rare decision to grant EC status to a zero-edit account on the day of its creation based on private correspondence with the Committee beforehand indicates that there is something unusual about the situation.

It remains uncertain whether this account has any relation to the WJC (or to any lobbying organization); commenters at the talk page for the ArbCom noticeboard have questioned whether this unknown party has standing to request a case be opened, whether a disclosure is required per WP:COI, and other issues. A more comprehensive explanation came from arbitrator Primefac:

Mschwartz1's sole edit (on the 26th) was to add this case request against Nishidani (mistakenly putting it at Wikipedia talk:Arbitration Committee instead of Wikipedia:Arbitration/Requests/Case, after which it was closed with instructions on how to post it to the correct board).

As of press time, there seems to be no conclusive evidence either way of who or what this account belongs to, despite many fairly strong opinions and speculations being expressed on the talk page.

See also related earlier coverage: "Does Wikipedia's Gaza coverage show an anti-Israel bias?" ("In the media", November 6, 2023) and "WikiProjects Israel and Palestine ("WikiProject Report", January 10, 2024)

Maryana Pinchuk is the Wikimedia Foundation's Principal Product Manager, and has been a Wikipedian since 2011.

In March 2023, I helped organize a Zoom call to discuss external trends that may impact the Wikimedia movement, and over a hundred Wikimedians attended to talk about the opportunities and risks of generative AI. But a different concern also surfaced that really surprised and stuck with me. Longtime English Wikipedia editor Ragesoss noted that when observing how his children like to learn things online, he saw that instead of going to websites, they sought out personalities who had built large followings on platforms like YouTube, and could provide information from an authentically human point of view. In the year since that AI call, the English Wikipedia community has heatedly debated how to respond to the opportunities and threats of generative AI. But this other trend – of people (especially younger people) seeking out personalities to aggregate knowledge for them, on social platforms that make it easy and rewarding for contributors to participate in new formats like short video – hasn’t been discussed as much. And while some Wikipedians may wonder what any of this has to do with encyclopedic content, I believe that this trend deserves as much attention and conversation in our community.

Despite the excitement (and hype) around generative AI in the press over the last year, Wikipedia is still used by more people globally than ChatGPT.^[1] But, when looking at Wikipedia usage compared to usage of popular social platforms frequented by young people, the comparison is far less favorable for us. TikTok has close to two billion monthly global users – twice as many as Wikipedia – and is still growing.^[2] And while it's tempting to dismiss TikTok and apps like it as entertainment platforms where young people only share and watch lip-syncing and viral dances, consider that a third of US people under age 30 recently reported regularly getting their news from TikTok, and one in four US TikTok users say they come to the platform for educational purposes like learning history.

Meanwhile, recent surveys by the Wikimedia Foundation Brand Health team have shown that worldwide, people aged 18–24 report being less aware of, less likely to use, and less likely to recommend Wikipedia to a friend than older age groups. Many of these younger audiences report preferring to get information via video platforms like YouTube or social knowledge sites like Quora. Traditionally, young people have always been a huge staple of Wikipedia readership. What could account for fewer young people than retirees today saying that they use or are even aware of Wikipedia's existence, and more going to social apps for educational content?

I think it's important to look outward at what's happened in the world in the last decade to better understand these data:

• There has been a proliferation of places where people can share and seek information online. Older platforms (e.g., Twitter/X, Facebook, Reddit) have been joined by newer competitors (e.g., TikTok, Bluesky, Threads, Discord) – competing not just for people's time and attention, but also for contributors who can make the most engaging, viral user-generated content. People with knowledge to share now have many choices for where and how to share their knowledge (whether that's text, long or short video, or audio). Successful creators can reach huge audiences quickly and can expect significant monetary rewards, so they are incentivized to produce highly engaging content on a variety of topics, including educational or edu-tainment content.
• How people interact with knowledge online has also changed. Millennials (my generation) were in their teens or early 20s when the first smartphones debuted, so most of us still had to use a laptop or desktop computer with a keyboard and web browser to explore and participate online. To learn or share information, we had to read or write it. In contrast, 18–24-year olds today were small children when the smartphone revolution hit, so their first online experiences likely happened on a touch screen, via mobile apps that require no interaction with web browsers or Google, and little or no typing. They also haven't known a world without user-generated video, and their first experience learning or sharing knowledge online was likely not a website but a YouTube video.
• Lastly, expectations for what "knowledge" must look like, online or offline, have changed. Millennials are the last generation to grow up routinely seeing physical encyclopedias in the home or school. I have been told by instructors in Wiki Education, who work with classrooms of US college students, that today's students have to first be told what an encyclopedia is, before being introduced to the concept of Wikipedia. These are college students in one of the wealthiest, most knowledge-privileged countries on the planet, but they've grown up in a world where traditional learning methods and resources are, like everything else, also rapidly changing in response to technology.

Given all these changes to how people expect to get information online, what opportunities does Wikipedia have to continue to meet the needs of this younger generation and generations to come?

I think there are a few strategies we could pursue.

One strategy, which I’ve heard expressed by some Wikipedians, is evolving what our content or projects looks like, to better fit some of these new expectations and preferences. There are different proposals for how to achieve this that have recently been proposed and discussed on the Wikimedia-l mailing list: for example, should we provide the ability to create and include more interactive elements (like graphs, timelines, or quizzes) on Wikipedia articles? Should we produce and include video explainers of important topics on Wikipedia? Though the specifics differ, fundamentally these ideas are about making Wikipedia more of a “knowledge destination” that younger audiences will want to visit – to read and to contribute to – instead of going to places like TikTok or YouTube. But when this conversation came up at WikiConference North America last fall, I heard many Wikipedians express concern about where pursuing this strategy could lead us. There was fear of making Wikipedia into something it isn’t. There was also fear about the cost and risks of building big new software features and trying to compete with massive for-profit technology companies for users. I think all of these concerns are very valid.

Some alternative strategies might instead focus on:

1. Making sure people continue to get reliable information from Wikipedia, and that they understand what Wikipedia is and that they can join our movement, no matter where they like to share or receive information online. This “free knowledge everywhere” strategy seems promising to me because creators on YouTube and TikTok already routinely harvest facts, ideas, and images from Wikipedia and include them in their videos (and it's possible that in this way, information from Wikipedia is already being viewed by more young people outside of Wikipedia than on it, in the formats that they prefer). However, Wikipedia is rarely credited or acknowledged in these videos, so audiences have no way of knowing about the amazing volunteer project that brings them this information for free, let alone that they can join in and contribute. One way to change this might be to make using our new sound logo a more consistent practice whenever Wikipedia knowledge is shared externally.
2. Evolving the habit that has been so standard among people of my generation when reading or hearing some claim that sounds dubious: opening up a new tab and checking what Wikipedia has to say. This strategy is about seeing if we can continue to make Wikipedia the "Internet's conscience" – i.e., the knowledge commons that provides additional context; organizes secondary sources; and gives ordinary people the ability to see, participate in, and influence the consensus on what is known – while adapting to the fact that "opening up a new tab and reading longform text" might not be the way younger audiences use the Internet anymore. One way to do this might be to provide instant verification with Wikipedia wherever people are getting information.

My team at the Wikimedia Foundation, Future Audiences, is exploring these different strategies, and our mandate is to evaluate them with quick experiments (including an experiment to build and test that last example, just-in-time third-party information verification with Wikipedia). But we are not the only ones within the Wikimedia movement who are interested in better understanding and showing up in this space: Basque Wikipedians have been turning Wikipedia articles into short videos for YouTube and TikTok as well as adding them back to Wikipedia, and the Wikimedia UK chapter has recently started their own TikTok channel.

I would love to hear from more Wikipedians: what role could Wikipedia play in this new personality-driven knowledge era? What new strategies seem most promising (or not) to you? What aren't we as a movement talking about or doing that we should be?

A monthly overview of recent academic research about Wikipedia and other Wikimedia projects, also published as the Wikimedia Research Newsletter.

How to better support new editors has long been a conundrum for Wikipedians. In 2018, the Wikimedia Foundation launched its Growth team, which tackles this issue by working on "features to encourage newcomers to make edits." A paper^[1] by four Wikimedia Foundation staff reports on the results of a long-time systematic study evaluating their impact:

"We propose the Newcomer Homepage, a central place where newcomers can learn how peer production works and find opportunities to contribute, as a solution for attracting and retaining newcomers. The homepage was built upon existing research and designed in collaboration with partner communities. Through a large-scale controlled experiment spanning 27 non-English Wikipedia wikis, we evaluate the homepage and find modest gains, and that having a positive effect on the newcomer experience depends on the newcomer’s context."

The newcomer homepage is summarized as "a central place to learn how Wikipedia works and that they can participate by editing". It offers a set of "Newcomer Tasks" to work on articles that the community has flagged as needing improvement, "with some tasks categorized as 'Easy' (e.g. copy editing, adding links), 'Medium' (e.g. adding references), and 'Hard' (e.g. expanding short articles)."

More specifically, the team conducted randomized controlled experiments, where newly registered accounts were either shown a "Get started here!" notification inviting them to visit their "Newcomer Homepage", or received the standard interface. Outcomes were tracked for four different metrics (all based on edits made to articles and article talk pages). Two different methods were used to evaluate impact: 1) An "'Intent-to-Treat' (ITT) approach, where we learn whether an invitation to the homepage results in significant differences" (combined with hierarchical regression to aggregate the results from the different wikis), and 2) a two-stage least squares approach to obtain an "estimate of the causal effect of making suggested edits conditional on being invited". The overall findings are:

• Activation A small but significant increase in overall activation [specifically, an 1% increase in the odds of making an edit within 24 hours of signing up], and that the outcome depends on newcomer context. Our intervention appears to distract newcomers who were already in the process of contributing, but seems to support those who were not, and in particular those who did not create an account with an intention to contribute.
• Retention No difference in the retention rate; we find a strong correlation with the activity level on a newcomer’s first day.
• Productivity No difference in the overall number of constructive contributions. [measured as the number of edits made within 15 days]
• Revert rate No difference in the proportion of contributions rejected by the community.

(The null results on retention and productivity contrast with the positive results that the team had earlier found in a smaller-scale experiment confined to four language Wikipedias, see our brief earlier coverage.)

The framing of these mostly null results as "modest gains" in the abstract appears a bit generous, also considering that the only metric with a significant increase (activation) seems less directly related to furthering Wikipedia's mission than some of the others. Similar A/B tests have been successfully used across the internet to greatly increase new user retention and activity on many websites, quite a few of which may be competing with Wikipedia for people's free time. However, the growth teams of commercial sites often have vastly more resources at their disposal (fueled by advertising revenues), enabling them to try out many more different features until hitting on one that has a significant impact. And in any case, in this reviewer's opinion these Wikipedia experiments should be considered a success in that they represent a major advance in helping us better understand new editors. As highlighted by the authors, there is a scarcity of existing research about what works specifically on sites like Wikipedia: "It is unclear what solutions work when it comes to attracting and retaining newcomers at scale in peer production communities." They note that in previous research (apart from an experiment that successfully used barnstar-like awards to increase long-term retention of new editors on German Wikipedia), "proposed solutions have only been available in a single community (English Wikipedia), and only two have been evaluated in controlled experiments". These are The Wikipedia Adventure (cf. our coverage: "The Wikipedia Adventure: Beloved but ineffective"), and the Teahouse, which the authors call (to their knowledge) the only "controlled experiment that has shown a significant impact on newcomer retention." (However, non-Wikimedia researchers have pointed out that "The Teahouse study might also have been a false positive" because of a statistical problem involving multiple comparisons.)

• The Wikimedia Foundation's research department invites proposals (deadline: April 29) for the "Wiki Workshop Hall", a new feature of the annual Wiki Workshop online conference consisting of two 30-minute sessions "for Wikimedia researchers and Wikimedia movement members to connect with each other."
• See the page of the monthly Wikimedia Research Showcase for videos and slides of past presentations.

Other recent publications that could not be covered in time for this issue include the items listed below. Contributions, whether reviewing or summarizing newly published research, are always welcome.

From the abstract:^[2]

[...] to understand how [climate change] was represented on English Wikipedia, we deployed a mixed-method approach on the article for “Effects of climate change” (ECC), its edit history and references, as well as hundreds of associated articles dealing with climate change in different ways. Using automated tools to scrape data from Wikipedia, we saw new articles were created as climatology-related knowledge grew and permeated into other fields, reflecting a growing body of climate research and growing public interest. Our qualitative textual analysis shows how specific descriptions of climatic phenomena became less hypothetical, reflecting the real-world public debate. The Intergovernmental Panel on Climate Change (IPCC) had a big impact on content and structure, we found using a bibliometric analysis, and what made this possible, we also discovered through a historical analysis, was the impactful work of just a few editors. This research suggests Wikipedia’s articles documented the real-world events around climate change and its wider acceptance - initially a hypothesis that soon became a regretful reality. Overall, our findings highlight the unique role IPCC reports play in making scientific knowledge about climate change actionable to the public, and underscore Wikipedia’s ability to facilitate access to research. [...]

From the abstract:^[3]

"Understanding how society reacts to climate change means understanding how different societal subsystems approach the challenge. With the help of a heuristic of systems theory two subsystems of society – science and mass media – are compared with respect to communications about climate change over the last 20 years. With text mining methods metadata of documents from two databases – OpenAlex and Wikipedia – are generated, analyzed, and visualized. We find substantial differences as well as similarities in the social, factual, and temporal dimensions. [...] This demonstrates for science a discursive shift from causes to consequences and for mass media a shift from chat to crisis. Science shows an ongoing growth process, while the attention of mass media appears cyclical."

From the abstract:^[4]

"[...] The aim of this paper is to [... analyze] whether the research topics of greatest academic interest align with those that attract the most social attention. To this end, the OpenAlex concepts are explored by comparing their works count with the page views of their respective Wikipedia articles. As a result, a correlation analysis between the two metrics reveals a lack of connection between the two realms.

See also a presentation at the November 2023 Wikimedia Research Showcase, and earlier coverage of related publications involving the first author

From the abstract:^[5]

[...] I trace how the global warming-related articles in Wikipedia changed over time, particularly in the wake of the publication of the 2007 International Panel on Climate Change Fourth Assessment Report. [...] I trace how Wikipedians enact genre in an unstable environment by analyzing how arguments unfold in Wikipedia talk pages, how the article text and citations change, as well as the larger network of global warming-related articles. [...] In chapter 2, I find that Wikipedians’ arguments create boundaries around the discursive spheres that can be cited within different articles, which suggests the significance of arguments not only about the topic but about genre as a deliberative resource in networked discourse. In chapter 3, I find that editors’ work in enacting genre results in facts becoming more at issue, or destabilized, within articles through the course of 2007. This analysis suggests that arguments about genre, and the easy availability of circulating texts online, may challenge consensus about controversial issues. In chapter 4, I use argument and network analysis to trace both Article for Deletion discussions and also the larger ecosystem of articles about global warming. This analysis shows how the talk page and article editing practices that I trace in earlier chapters become sedimented within the site’s information architecture, shaping what Internet users may learn about the issue. [...]

From the abstract:^[6]

"Wikipedia articles are categorized into different levels of quality, so we analyzed all 7,048 environmental articles in the Environment Assessment project on English-language Wikipedia. Based on a review of literature, we selected indicators of information quality (number of editors, number of edits, article length, number of references, and the ratio of references to words) and tested the correlation between these indicators and quality perception in the Wikipedia Assessment project. We found that articles perceived as higher quality typically have more editors and edits, are longer, and contain more references, as well as a higher ratio of references to words"

From the abstract:^[7]

"[...] This study examines global engagement with climate change and related concepts through an analysis of around 517 Million Wikipedia pageviews of 3965 items from WikiProject Climate Change across 213 countries in the years 2017 to 2022. We take advantage of Wikimedia Foundation's differentially-private daily pageview dataset, which makes it possible to study Wikipedia viewing behavior in a language edition agnostic way and on a per-country basis. Temporal analysis reveals a stagnant engagement with climate change articles, contrary to societal trends, possibly due to the attitude-behavior gap. We also found substantial regional differences, with countries from the global north displaying greater traffic compared to the global south. Specific events, notably Greta Thunberg's speech at the UN climate summit in 2019, drive peaks in climate change engagement [...]. However, causal time series analyses show that events like these do not lead to long-lasting increased traffic."

The Charter for the Universal Code of Conduct Coordinating Committee (holy alliteration!) has been ratified, with nearly 75% support from 1,746 voters. Some voters were excited about the potential benefits of the charter regarding administrator accountability and representation, while others were concerned about local-project autonomy, potential abuse of the policies, increasing bureaucracy, and even political bias. Still others commented on the voting process itself, including the availability of translations into other languages. Full voter comments are available on Meta-Wiki. – s, AK

The winners of the 2023 Wiki Loves Monuments contest were announced earlier this week. The announcements were staggered and spanned two days, starting with honorable mentions (places 25 to 16) on Monday 25 March, and concluding with the 15 winners on Tuesday 26 March. To see the 15 winning entries, go to the Wiki Love Monuments website. Wikimedia Commons has a complete overview of all nominees, winners and runners-up. – AK

• 
  2nd place: Marksburg Castle, Germany, in winter
• 
  3rd place: Panoramic view of Premana, Lombardy, Italy

• Wikimania 2024 CfP: A call for session proposals for Wikimania 2024 (to be held in Katowice, Poland from August 7-10) will close on April 3.
• Annual reports: Wikimedia Ukraine (in Ukrainian), Diff post (in English); West Bengal Wikimedians User Group; Dagbani Wikimedians User Group; Wikimedia Community Ireland.
• New administrators: None to report this issue.
• Record low active administrators: 442 was the low point for active admins in 2023 (see prior Signpost coverage, and followup), and 2023's low point was just undercut on 28 March 2024, when 441 were counted by the bot (see page history for the tally). Just as 442 was last year, 441 is a record for nearly the last 20 years of English Wikipedia, i.e. the period encompassing the current Requests for Adminship (RfA) procedure. Currently, reform of the RfA process is being discussed at the 2024 RfA Review.
• Articles for Improvement: This week's Article for Improvement is Abdalá Bucaram, the former president of Ecuador. Please be bold in helping improve this article!

• AARoads, take me home: A long-running conflict over Wikipedia's articles about US roads led to the forking of many roads articles back in September. Gizmodo recently interviewed a number of AARoads editors and Wikipedians (Imzadi1979, BMACS1002, and onel5969) about the split, calling it "a promised land where the editors hope, at last, that they can find peace". See also previous Signpost coverage: "A fork in the Roads WikiProject" ("News and notes", September 16, 2023)

• Getting away from it all?: AFAR magazine explains "Why Google Flights and Wikipedia (Yep!) Are a Traveler’s Best Friends", giving some tips on finding airports connecting aspiring travelers to their destination. The article calls the customary "Airlines and destinations" sections in Wikipedia articles about airports (example) "a great resource for those wanting to know options for nonstop flights and possibilities for mileage redemptions."

• Balkan edit wars: In an interview for Portuguese newspaper Público (in Portuguese, paywalled), University of Novi Sad professor and fact-checker Stefan Janjić broke down the influence of Russian disinformation in his native country of Serbia, including a recent dispute over the content and tone of the article about the Srebrenica massacre on the Serbian Wikipedia.

• Around the world in 292 ways: Wikipedian Lucy Moore has created an article for a woman from every country in the world (as well as nearly every partially-recognized state, United Kingdom subdivision, autonomous region, special administrative region, overseas territory and disputed ethnic republic!) Moore participates in Women in Red, a WikiProject dedicated to closing the gender gap on Wikipedia. An article in The Guardian and an interview with CNN cover Moore's work and highlight women from lesser-known corners of the world. In honor of International Women's Day, an editorial in The Guardian analyzes Wikipedia's coverage of women.

• Covering Pakistan's elections on Wikipedia: The Friday Times covers how Pakistan's turbulent politics and recent general election has caused problems for Wikipedians; accusations of political bias fly as editors try to figure out who to list as the leader of Pakistan Tehreek-e-Insaf, whether to list the party on election maps at all, and perhaps most importantly, who won the election.

• French Wikipedia discusses transgender names: Le Monde (in English, paywalled) reports that "Wikipedia's French-speaking community is torn apart over 'deadnaming' trans people": "The source of the conflict lies in a debate over the 'deadnames' – or morinoms in French – of trans people. Should these names that are no longer in use, such as birthnames, be mentioned on Wikipedia? If so, in which cases and under what conditions? On the English-language version, contributors settled the matter back in 2015, without too much controversy: They agreed to only mention the deadnames of people who became famous under that name. However, among French-speaking Wikipedians, the question has remained unanswered, and even resurfaces regularly."

Dear fellow Wikipedian:
I have heard your concerns, and I understand that this is a contentious discussion that requires careful deliberation. Upon consideration, I have determined that more deliberation will mean better results. It must also be recognized that I am right, hence all deliberation must be for the purpose of stating my argument. Therefore, I will continue to reply to every comment that disagrees with me in this 7,000-word-long discussion, even if I am not being addressed in any way. I do not recognize that this may crowd the discussion, make it difficult for other editors to speak, or fail to accomplish anything that I could not have done with a single comment. I have added the page to my watchlist, and I have enabled notifications for this discussion. I look forward to engaging with you further. Extensively.
Kind regards,
Your persistent friend

Dear fellow Wikipedian:
Thank you for filing your report. You have accurately identified 750 instances of incivility, point of view editing, ignoring of consensus, ownership of content, and insertion of original research by this user in the last 7 days, and you are correct that they have chased away three productive editors. We have determined that this is a content dispute. None of these individual infractions rises to the level of sanctions, and you are chastised for wasting our time. Furthermore, you used a slightly brusque tone in one of your interactions with this user. You are cautioned to improve your conduct in the future. Please hesitate to file additional reports here if you encounter further disruption.
Sincerely,
The Administrator's Noticeboard

Dear fellow Wikipedian:
It has come to my attention that you have nominated my article for deletion. What you fail to understand is that you are a buffoon, because this thing was mentioned in seven different newspapers, only twelve years ago. Therefore any encyclopedia would be incomplete without a thorough repetition of every complete thought that was reported in regard to this subject. I reject any idea that more than a brief acknowledgement of its existence is necessary for this to be a notable subject. You don't need to read any notability guidelines to know that three reliable sources means automatic victory.
Thanks,
AfD arguer

Dear fellow Wikipedian:
I resent your comments about my participation in the Palestine/Israel topic area. I assure you, I take the utmost care in following policies and guidelines while editing and engaging in discussion. I carefully deliberate on every issue before developing an opinion. Yes, you are correct that my actions in this area have been beneficial to the same side and detrimental to the other every single time I have ever engaged with the topic in my ten years on Wikipedia. This is not because I am incapable of a neutral point of view. It is because my side has been correct in 100% of the disputes that have emerged. If you accuse me of editing with a POV again, I will disparage you on every noticeboard I can find. You have only proven that you and your people cannot edit on Wikipedia neutrally.
Best wishes,
An account that edits almost exclusively in the Palestine/Israel area (swap the topic as needed)

Dear fellow Wikipedian:
I fail to understand your concerns about the quality of my sources. I happen to be an avid reader of Alternative Facts Today, The Extremist Times, and The Government is Lying to You Weekly. I can attest to their accuracy in reporting. I expect that you would prefer coverage from the lying mainstream media, like Reuters or The New York Times. My friend, has it occurred to you that it is actually the reliable sources that are the biased ones, while my sources are entirely neutral and correct? Pull the wool from your eyes, and wake up to the lies that keep them in power.
Until the revolution proves me right,
A neutral Wikipedian

Dear fellow Wikipedian:
It appears you have taken an interest in our favorite subject. What you don't realize is that this subject is governed by our WikiProject. The five of us who engage at the WikiProject have total ownership over all 12,000 articles under our domain, and any edits to these articles must follow the rules we have imposed. Please see our WikiProject style guide essay for what you are and are not allowed to do when editing on this subject. You are welcome to join, but we will all vote against your ideas because they are not part of our established paradigm. If you attempt to make an edit that complies with Wikipedia best practices, we will continue to inform you that we don't do that here and that our articles operate on different rules from all of the others... because we said so.
Cordially,
Your local WikiProject

Dear fellow Wikipedian:
I have noticed that you attempted to make a change on Wikipedia. Have you asked permission first? Wikipedia is not some place where anyone can edit, you must first seek approval before doing anything beyond fixing a typo. I object to your change because you have not been given permission, and I will continue to object on account of the fact that it has been objected to (by me). I have no suggestions for an alternative or a compromise, so we'd better just leave everything exactly as it was without fixing whatever you were trying to improve.
Sending my love,
The neighborhood watch

All characters and letters on this page are fictitious. The concepts expressed herein are representative of broad trends that exist throughout Wikipedia. Any resemblance to specific editors is purely coincidental.