Never cite Wikipedia as a source. Teachers repeated these words so often in high school that they became stuck forever in the grooves of our brains, like the Pledge of Allegiance or the alphabet song. We nodded along in class, but we mined Wikipedia for information anyway, letting it shape our research by pointing us in the right direction. Yet, recent developments suggest that our teachers may have been right to distrust an online encyclopedia that anyone can edit. An increasing number of reporters and programmers have taken to Twitter to expose how government agencies regularly doctor Wikipedia’s content.
The Twitterbot government monitoring movement was pushed into the spotlight on September 22 of this year, when the Twitter page, “Gov. of Canada edits” (@gccaedits) sent out a tweet reading: “Mass Surveillance in the United States Wikipedia article edited anonymously from Canadian Department of National Defense.” This short description was followed by a link to the “Mass Surveillance in the United States” Wikipedia page’s edit history. Someone using the Canadian Department of National Defense’s network was changing Wikipedia’s information about American mass surveillance.
In spite of the rather sensational lead-in, the changes were fairly harmless. The technical phrase for describing surveillance systems, “Wide Area Motion Imagery” had been abbreviated to “WAMI.” While one could argue that the less descriptive acronym could possibly hurt a layperson’s understanding of the article, it is difficult to build much of a government censorship conspiracy theory around the minor edit. However, the question remains, just how often do government offices edit Wikipedia entries?
***
Wikipedia is a poster child for the digital age. It’s fast, reliable, easy to use, and, in theory, objective. A massive online database that boasts over 4.6 million articles, Wikipedia certainly has no shortage of information. However, in addition to being one of the most massive sources of information in the world, Wikipedia is also one of the most referenced. According to a University of Washington study, 82% of students in higher education use Wikipedia at some point during their research. What people tend to forget is that anyone can edit any article at any time, making the website a potentially powerful tool for propaganda and information control.
A small group of government watchdogs have responded to the vulnerabilities of Wikipedia editing by taking to Twitter to track just who is making what edits. The first of their kind came in the form of “Parliament WikiEdits” (@parliamentedits). The account was created on July 8 by Tom Scott, a British journalist and web developer. The Twitter page is run remotely by something called a Twitterbot. These Twitterbots are programs that are able to automatically post on Twitter in response to specific stimuli. In the case of “Parliament WikiEdits,” the stimuli are Wikipedia page edits by someone using a United Kingdom Parliament IP address.
Scott quickly shared his bot by fitting it with an open-source code. This means its program is accessible to anyone and only has to be tweaked slightly in order for others to redirect its point of focus. Within days, the “Parliament WikiEdit” format was duplicated in Canada, the United States (@congressedits) and a host of other countries. The communal spirit of the internet has continued to shine through in these online watch groups, with “Gov. of Canada edits” recently tweeting out the specific government IP addresses that were used to edit the article on surveillance.
While “Parliament WikiEdit” may have developed the idea, its successors have far surpassed its monitoring abilities. Since its first tweet, “Parliament WikiEdits” has detected 61 cases of Parliamentary editing. By comparison, @congressedits has reported 339 such cases and @gccaedits has reported a stunning 751 cases of governmental editing. People inside various political offices are editing Wikipedia with an alarming frequency, but just how much of what is being edited should be described as political censorship?
John Stoehr, managing editor of the Washington Spectator and a lecturer at Yale University, claims it would be difficult to call any editing of Wikipedia censorship, asserting that “Wikipedia has mechanisms in place to advance the cause of transparency.” The fact that government editing is so transparent proves that these mechanisms are working. According to Stoehr, as long as the government editing “is abiding by rules of law that Wikipedia enforces” it cannot truly be called censorship.“Censorship is fundamentally about power,” he notes. “Only if Wikipedia administrators were prevented from doing their jobs by government threat or force would I call that censorship.” Stoehr made it clear, though, that he is unaware of the specific edits that the Twitter bots have reported.
Indeed, the majority of these cases of governmental interference have turned out to be fairly inconsequential. Several edits from IP addresses of computers in the U.S. Capitol Building have been made to the page for the television show “It’s Always Sunny in Philadelphia,” and a series of edits coming from the Canadian government had to do with inside jokes involving maple syrup. Government employees, like any office workers, have some free time during which they use the Internet for personal interests and entertainment.
The danger of governmental editing of Wikipedia pages lies in the potential for political agenda-pushing. Twitter bots have reported a number of changes to biographical information on politicians. While impossible to track just who is making the edits, these changes seem to be biased toward improving public image. In one such example, the page of Michigan Congressional Representative Justin Amash’s underwent changes regarding his pre-political career, from describing him as a “corporate lawyer” to an “attorney.” Of course, “corporate lawyer” carries a Big Business connotation that could potentially harm Rep. Amash in election season. It is impossible to tell definitively whether or not someone in Amash’s camp made the change, for a major shortcoming of the Twitterbots is that they can only trace to government IP addresses, not specific people. Nonetheless, this change seems to be one that was politically motivated.
Other editing crusades have had a more nefarious agenda. The Wikipedia page for the “Death of Jean Charles de Menezes,” which describes the accidental killing of a misidentified bombing suspect by London police, underwent substantial edits from a Parliament computer. The user of this computer deleted an entire section containing possible issues with the Independent Police Complaints Commission’s (IPCC) investigation. Perhaps this could be excused with the reasoning that the investigation was secondary to the facts of the killing themselves. Another series of edits attempting to discredit the victim, however, are harder to justify. Suggestions that he was on heavy drugs at the time of the shooting were added, while information that he was in the United Kingdom legally at the time of the killing was entirely removed. Through a few clicks and some typing, a sober, legal immigrant had been transformed into a drug-addled miscreant.
Similarly, numerous articles about potential misconduct by the American government were heavily cut. Of particular note is the deletion of information about the FBI’s Counterintelligence Program, which ran from 1956 to 1971 and involved spying on several domestic political groups like the Black Panthers and the Klu Klux Klan. While the program engaged in illegal and unconstitutional activities at several junctures, all references to any illegality were deleted from the opening paragraph.
The editing process can also be used to bolster a government figure’s image through omission. The Canadian Twitterbots in particular encountered multiple removals of information on controversial missteps. On Patrick Brown’s page, a conservative member of parliament (MP) , the lengthy “Controversies” section was deleted. The section detailed his highly contentious stance on reproductive rights and abortion, a position that openly broke with members of his own party. Edits on MP Shelly Glover’s page also raised concerns over the expungement of several campaign overspending allegations and a full-blown call for her suspension from Parliament.
But far from simply deleting potential controversies in politicians’ careers, government Wikipedia editing has occasionally crossed over into covering up criminal indiscretion. Pamela Wallin, a member of the Canadian Senate, had a checkered history wiped from the record. She had been accused of using government money for personal travel expenses, at one point charging an expense account for traveling to a conference and arriving at her destination hours after the conference had ended. She was also found to have several inconsistencies in her stated place of residence. Wallin was ultimately suspended from the senate and ordered to reimburse the government $120,000. After government editing, the entire section devoted to covering Wallin’s expense scandal was deleted. The anonymous editor gave no reason for the deletion on the editing comments section.
Again, the weakness of the Twitterbots is that they are only able to track edits back to IP addresses, not to individual employees or offices. As such, it seems extremely unlikely that those attempting to skew public information will ever be held accountable in a tangible way. The public, however, does have an advantage in the relative transparency of the Wikipedia editing system.
The various Twitterbots are by no means the first systems to monitor government editing of Wikipedia articles. Scott himself had been tracking these manually for several years by using Wikipedia’s system of storing editors’ IP addresses. Since almost all web traffic coming out of the Houses of Parliament was limited to two proxy servers, tracking where the edits were coming from was not difficult. Surprisingly, the majority of references to these pre-Twitter Wikipedia editing campaigns have been confined to tech and alternative blogs. This was the revolutionary coup of the Twitter feeds. Despite their relatively small following, they are getting mainstream attention in a way that previous attempts at exposure have been unable to get. Since all previous edit monitoring had to be done manually, the automated Twitterbots provide a new perspective on just how widespread and frequent this type of government editing of public information has become.
The governments now under surveillance have responded to the Twitter feeds in various ways. Canada’s parliament has mostly responded on a case-by-case basis, giving stock answers to questions about widespread interference. Congress initially attempted to take a similarly passive course of action until Wikipedia itself put a 10-day ban on Congress IP addresses editing Wikipedia articles. Since that initially promising stand against government editing, the ban has been lifted and widespread modification has continued as before.
The only government to tackle the issue head-on has been the British government, although their method has been to attempt to censor all coverage of the issue. After the debut of Parliament WikiEdits, the IP addresses used by the Internet proxies in Parliament were all changed. A Freedom of Information request for the new IP addresses was turned down, with the reasoning that “releasing the information would be likely to prejudice the prevention or detection of crime”—language from the Freedom of Information Act of 2000. This statement was made in spite of the fact that Parliament IP addresses had been previously disclosed with no apparent concern for detection of crime. The fact that nothing was done about a government’s attempt to silence watchdogs is unfortunately symptomatic of a bureaucracy whose primary goal is self-preservation, and may actually be more troubling than the edits themselves.
Wikipedia page editing could be considered in line with a recent authoritarian bent in the behaviors of democratic governments across the globe. From NSA wiretapping to CCTV surveillance, the gulf between the people and the governments that are supposed to represent them has continued to grow. Recent accusations of governmental mishandling and police brutality in a series of high profile protests have further cemented the us-versus-them mentality that has come to define government-civilian relations. Growing mistrust could have potentially dire consequences for the public. As Sterling Professor of Political Science and Director of the MacMillan Center for International and Area Studies Ian Shapiro states, “it is a failure of government if people lose their trust in it. In the limiting case it makes a country ungovernable, since compliance cannot be compelled indefinitely.”
Despite the apparent surplus of mistrust in government, the Wikipedia edit Twitter watchdogs remain painfully overlooked. “Gov. of Canada edits,” the feed covering the most widespread, or at least most poorly covered up, censorship has a paltry 5,672 followers. For a point of comparison, Canadian pop singer Justin Bieber has 55.9 million followers.
Perhaps part of the reason that government editing is receiving so little attention is that people are actually aware of the very serious limitations of Wikipedia. As Andrew Sable, UCLA professor of public policy and political science put it, “the first thing this reveals is how foolish it is for anyone to rely on Wikipedia in the first place, precisely because anyone can edit the entries.” Further, he believes that “the capacity of private actors to alter and distort information to their hearts’ content completely overwhelms the cuts of the government.” If the government is changing articles that it considers slanted, that is tantamount to a “press conference or a press release,” simply attempting to control the tone of a story. The furor the edits have raised in certain corners of the internet also raises some uncomfortable questions about why some are so troubled by Wikipedia editing, while wars and drone strikes and arguably more consequential overreaches of power seem to go unnoticed.
Ultimately, the question of the Wikipedia editing wars comes down to a bit of a stalemate. On one hand, people within the government are changing public information. On the other, the source they are editing is, by its nature, unreliable and open to manipulation. Indeed, Wikipedia only exists because people are so willing to edit and alter information to constantly keeps its pages updated. While the Twitterbots may not have exposed an earthshattering conspiracy like their creators had hoped, they have served as another reminder that everyone should be aware of not just the content of what they read, but of who wrote it and why. Especially when it comes to Wikipedia.