Editor’s Note: Recently, Capital New York reported that NYPD computers had been used to make edits to Wikipedia entries on police-related topics. We asked Wikipedia expert William Beutler to comment on this news story.

New Yorkers and the social web alike were stunned to learn that NYPD computers had been used to make surreptitious Wikipedia edits on entries related to police department controversies, including pages covering the “stop-and-frisk” policy and the deaths of Eric Garner and Amadou Diallo.

These edits added information favorable to the NYPD while downplaying views more critical of the department. For instance, “Garner raised both his arms in the air” became “Garner flailed his arms about as he spoke”; another edit pointed out that Garner was “considerably larger than any of the officers” — making the deceased sound more dangerous and unpredictable.

It was a scoop for reporter Kelly Weill, who went through the trouble of creating a Python script to sort through vast amounts of data to find the story. Following fast, the New York media and tech press collected denials from the department, criticism from the mayor’s office (which apparently has its own history of Wikipedia editing), and ultimately a statement from the police commissioner saying, “I don’t anticipate any punishment, quite frankly.” Meanwhile, a Twitter account called @nypdedits appeared, watching for any subsequent edits from the NYPD, and other reporters are setting up alerts for Wikipedia edits from their local police departments.

As a Wikipedia contributor of many years and the writer of a blog offering Wikipedia commentary, the thing I find most surprising about this is just how surprised people seem to be — self-interested Wikipedia edits are nothing new.

The History of Self-Interested Wikipedia Edits

Nearly eight years ago, a young web programmer named Virgil Griffith created WikiScanner, a website tracing many Wikipedia edits back to the CIA, Coca-Cola, The New York Times, ExxonMobil, and pretty much every large organization you can name. There was plenty of outrage to go around, but not enough attention to really single anyone out. There is even a Wikipedia article called “Conflict-of-interest editing on Wikipedia” that chronicles the most infamous examples.

It’s pretty much been the same story ever since, with a rotating cast of sensitive topics, intransigent authorities, and mildly sophisticated web detective-work needed to blow things wide open.

Which leads to another thing I know: The group least likely to raise a major fuss about this turns out to be … Wikipedia editors. Outwardly, Wikipedians lament the fact that this kind of thing happens all the time, avowedly disapprove of it, and then do nothing to prevent it from happening again. After all, this happens all the time.

Wikipedia’s Guidelines on Editing Entries

The relevant Wikipedia guideline is “Conflict of interest,” which strongly discourages (although does not outright prohibit) editing topics where one has a close personal or financial relationship. It’s problematic in part because most editors gravitate toward topics they are interested in. Sometimes it’s a favorite TV show, sometimes it’s an area of research, and other times it’s about one’s own industry.

So what makes the difference between perfectly acceptable edits and ones to watch out for? The guideline says, “when advancing outside interests is more important to an editor than advancing the aims of Wikipedia, that editor stands in a conflict of interest.”

The problem is that, as written, the line crossed exists only inside the editor’s head, which then invites other editors to weigh in on their own view of the subject matter. Despite the existence of a Wikipedia policy that advises to focus on “content, not on the contributor,” it’s a lofty notion that isn’t always held. Meanwhile, another policy forbids editors from posting information revealing the real-world identity of editors who have chosen to use pseudonyms. This rule is strictly enforced, which sometimes inhibits efforts to ferret out crafty offenders.

More than once in recent years, the Wikipedia community has considered an outright ban on “paid editing,” but each time it has been struck down for fears that it would go too far. In particular, any rules strong enough to be effective might also put a university professor in harm’s way for writing about their own field of research. Many Wikipedians are now employed in paid positions at libraries and universities, so drawing a line around payment is too dangerous.

A Closer Look at NYPD’s Edits

But what, exactly, did the NYPD do? It’s entirely possible that the edits were made by an officer acting on his or her own volition, as a matter of personal opinion, instead of word coming down from the NYPD brass to put the fix in on Wikipedia. That is, not to put too fine a point on it, how a very large majority of edits on Wikipedia get made.

Meanwhile, not all of the edits were made to NYPD-related cases. Weill published a Google Doc with a list of NYPD-edited articles, showing the edits covered a wide range of topics, including lots about popular culture (my favorite: NYPD Blue tough guy Andy Sipowicz). The NYPD’s editors also removed some more obvious vandalism.

To the list of not-so-bad changes one might plausibly add the specific Garner edits I quoted above. If someone without ties to the NYPD made these changes, there might be disagreement, but not necessarily a “gotcha” angle.

It’s not difficult to imagine that the articles were biased against the police to begin with, or that this was the officers’ impression. This is one reason why Wikipedia officially calls itself the “encyclopedia that anyone can edit” — through the back and forth of debate, Wikipedia will asymptotically approach the best possible version of events. This is the theory of crowdsourcing.

But this process breaks down when the involvement of “interested” parties is looked down upon. Could the NYPD have made the same changes had they brought them to discussion? Maybe. Figuring out how to fairly handle these “conflict of interest cases” is something Wikipedia needs to work on.

But it still mostly works. Having seen this happen time and again, Wikipedians are thinking this is an unfortunate but unavoidable side effect of allowing an open-editing policy. Closing off access to editing by all comers is more dangerous than leaving it open to all. Wikipedians are nervous that they aren’t able to detect these kinds of edits, but they know they can deal with them once they are uncovered.

Amid the controversy, a few Wikipedia editors will follow up on specific changes, and decide for themselves whether to keep, delete, or alter the specifics. And then they’ll steel themselves, knowing this happens all the time.

Leave a Reply