What Constitutes a Reliable Source (On Wikipedia)
My new article for Slate + Highlights from my Wikipedia Day NYC talk at the Brooklyn Public Library
Before we start, a few quick updates:
New Article Alert: I have a new article out this morning in Slate: How Wikipedia Editors are Bracing for Targeted Attacks in 2025.
Fact vs. Fiction: Ars Technica interviewed me about how fiction gave me room to explore the small but powerful Wikipedia community.
Book club: Last month, my book club read The Sun Also Rises. This month, we’re reading Fahrenheit 451. Drop me a note to join.
I was so pleased to attend Wikipedia Day NYC last month, where I had the honor of being on a panel discussing “Reliable Sources” together with journalist Emily Gertz and NYU professor Clay Shirky.
After the conversation, I had the chance to discuss these ideas further with some of New York’s most dedicated Wikipedia editors. One of the highlights was meeting Annie Rauwerda, founder of Depths of Wikipedia, whose enthusiasm for the quirky history of the website shines through.
While this might seem like a niche event, you might be surprised to learn that over 1,200 people registered to attend. There is a strong interest in contributing to the ubiquitous free internet encyclopedia.
Although the video from the event is now available, I wanted to capture some of the main ideas in writing. Note: This is not a transcript of the interview itself.
What do we mean by reliable sources?
I think it’s important to get specific because most of the people in the world are not highly-experienced Wikipedia editors.
Here’s the key principle: The articles on Wikipedia should be based on reliable, independent, published sources.
Reliability of a Source: The reliability of a source depends on context. Where possible, Wikipedia editors prefer publications with a reputation for fact-checking and accuracy. If a publication has people engaged with checking facts or scrutinizing the writing, then that is considered an indicator of reliability.
Independence of a Source: Evaluating the independence of a source means considering whether the source is self-published, not a primary source, and not closely affiliated with the subject. For instance, social media posts and company press releases generally do not qualify as reliable sources because they are not considered sufficiently independent.
Can you give me an example?
Let’s say you have a New York content creator who makes a biographical TikTok video about her life. In the video, she mentions that her favorite restaurant when she was growing up was Junior’s Diner (the original location in Downtown Brooklyn).
This TikTok video is likely not to be considered a reliable source for a biographical page about that content creator for at least two reasons:
Not From a Reputable Source. The source is not from a publication with a reputation for fact-checking and accuracy. TikTok videos are self-published and do not undergo a rigorous editorial process.
Lack of Independence. The source is not independent from the subject. It’s self-published without any third-party vetting. The creator might have a conflict of interest; her content could be biased by her desire to promote herself or to tell her story in a certain way.
This doesn’t sound that difficult.
The rules may seem simple in the abstract, but they can become more challenging when applying them or interacting with others. Because we are inundated by social media, it can be a bit jarring for people to hear that most social media posts are not considered reliable in and of themselves. For example, the TikTok creator might be rather offended that her video is not considered a reliable source about a fact that she knows very well: her favorite restaurant growing up.
It’s important that we communicate the reasoning for this policy. Wikipedia focuses on verifiability, and her self-published video has not been verified by people who are independent from herself. But again, you could imagine how the content creator could get peeved that her own firsthand account is not considered sufficient without third-party vetting.
What about the reliability of a news publication that’s not a social media post?
There are also very extensive discussions on Wikipedia about whether a publication is or is not considered reliable based upon whether the publication has a track record of fact-checking and accuracy. Some publications are deprecated after the discussion indicates that they are generally unreliable. Going forward, any user who attempts to cite a deprecated source on a Wikipedia page receives a warning and a request to cite a more reliable publication.
For instance, I wrote about how Wikipedia editors deprecated the UK newspaper The Daily Mail back in 2021 after finding that the paper had mastered the art of running false clickbait stories. The American news and opinion website The Daily Caller was deprecated in 2019 after discussion about how it consistently produces false or fabricated information.
Hold up. This sounds very controversial.
It can be. Wikipedia’s goal is to provide trustworthy information as reflected by reliable sources. That means Wikipedia editors can and do delete content when the underlying source is determined not to be trustworthy.
Keep in mind, Wikipedia has never billed itself as a free speech forum, a free-for-all where anybody can cite whatever source they like. The goal is to provide a reliable encyclopedic summary, which means rejecting and deleting poor sources.
As you might imagine, these discussions about whether a publication is or is not considered reliable can be very contentious. That’s why it’s crucial for Wikipedia editors to repeatedly remind the public that the discussions about reliable sources are all quite public and transparent. Anybody can become a volunteer Wikipedian—it’s free—and engage in these conversations. There’s not some exclusive cabal of editors (based in New York) making secret decisions behind the scenes.
I’d like to shift gears and talk about the recent news that the Heritage Foundation is planning to “identify and target” Wikipedia editors.
Last month, the Jewish-American news site Forward reported a shocking scoop: the Heritage Foundation, the conservative think tank behind Project 2025, was planning to “identify and target” Wikipedia editors. Through analyzing text patterns, usernames, technical data, and employing social-engineering tactics, Heritage aimed to reveal the identities of anonymous Wikipedia editors they believed were “abusing their position” on the platform.
What effect might these targeted attacks have on the free encyclopedia?
If the think tank succeeds in its effort to identify and target editors, the consequences could be profound. Faced with the risk of harassment or real-world retaliation, many volunteer editors—especially those covering politically sensitive topics—may simply stop contributing. Those who remain are likely to be the most ideologically-driven voices, further eroding Wikipedia’s stated goal of neutrality.
I love how you broke down the idea of “reliable sources” with clear examples. The TikTok video example is especially relatable... many content creators might be surprised that their own self-published content isn’t considered trustworthy for encyclopedia purposes. It's a reminder of how important it is to have that independent, third-party verification to ensure neutrality and accuracy.