Wikipedia:Wikipedia Signpost/2015-07-15/Op-ed



"How long will this take?" This is one of the first questions new clients ask. They come to us because the Wikipedia entry about the company at which they work is wrong, incomplete, or even just outdated. The answer varies, but it often comes as a shock when we explain that fixing problems on the "encyclopedia anyone can edit" is a project measured in weeks, if not months.

Many are speaking to us because they have tried before and failed, or were spooked by headlines about others who had tried and failed. Late last month, another firm joined the list of ignominy: Sunshine Sachs, a PR agency to the stars, was busted for removing content from Naomi Campbell's entry, among others (see previous Signpost coverage). But often, even company representatives who mean well are treated as if the fault rests entirely with them, and are reprimanded simply for not understanding how Wikipedia works.

This is something we see every day. Together, we have nearly 10 years combined experience helping brands, companies, and organizations engage Wikipedia constructively. We help them understand what they can accomplish, what they should leave alone, and how to engage with Wikipedia's volunteers. Although the process has improved over time, we believe the right balance has yet to be discovered.

How the Bright Line works
The state-of-the-art in conflict of interest engagement is commonly called the "Bright Line" rule, from a quote by Jimmy Wales when he first outlined the concept in 2012. It basically goes like this: "I am opposed to allowing paid advocates to edit in article space at all, but am extremely supportive of them being given other helpful paths to assist us".

We immediately embraced this new development. After all, our greatest challenge over time was not the research and writing, nor aligning client goals with Wikipedia's mission, but rather the uncertainty involved in navigating a community that has as many views on paid advocacy as there are members. The Bright Line was an elegant solution, simplifying the process and making it more comprehensible for editors and clients alike.

It had other benefits, too: more feedback makes for better articles, and volunteer editors can help clarify things for a "lay audience". Sometimes clients are pushy, and it’s helpful to be able to use editor review as a backstop. Occasionally, it will even spark a great collaboration: identifying additional areas for improvement neither side would have found alone.

When the Bright Line does not work
The Bright Line can work, and we (and others who have embraced it) are proof. But after three years of following its prescripts, we are all too aware of the times when it does not. Jimbo's elegant solution comes with its own limitations, challenges, and even contradictions. Here are several reasons the status quo can and should be improved:


 * The Bright Line has not actually reduced uncertainty—Not all volunteer editors are familiar with the Bright Line, which creates uncertainty surrounding the community's response to COI editors. This is exacerbated by the fact that the Bright Line has never become guideline or policy, so there is nothing definitive for adherents to point to when seeking help.
 * The Bright Line places a significant burden on volunteers—With volunteer time as scarce as it has ever been, it is somewhat perverse that established community members are asked to take time away from their own projects to work on someone else's. Sometimes this is unavoidable, but in our experience it often lacks common sense. Some real-life situations where we have had to avoid making uncontroversial edits: de-orphaning a new article, fixing typos, fixing broken formatting, updating navboxes and infoboxes, repositioning images, disambiguating links, and fixing non-obvious vandalism.
 * The Bright Line has no infrastructure to support it—With no clear advice on how to best ask for help, no specific guidelines for volunteers, and no active wikiproject or noticeboards, few ask and fewer respond. The closest thing to it now is the little-known "Requested edits" category, which is backlogged in spite (and because) of its obscurity. We think more editors would help if a better process existed, especially one that turned it from the lonely responsibility of one volunteer at a time into a collaborative review effort.
 * Paid editing prohibitionists frustrate the process—Sometimes, the only responding editors are anti-paid editing, or are motivated by POV themselves. Not only do some COI editors choose not to disclose because it puts a target on their back, but some volunteers avoid offering help for the exact same reason. This can prevent edits being made that are genuine improvements, to the detriment of the encyclopedia.
 * It perpetuates the atmosphere of distrust on Wikipedia—The Bright Line was formulated in response to the Bell Pottinger scandal, and so its framing focuses on the worst actors, while giving scant consideration to the possibility that "paid advocates" might be simply looking for an accurate and fair representation of the brand, not necessarily a promotional one. It actually requires editors to assume bad faith, which is fundamentally at odds with Wikipedia's five pillars.

A proposed "First Amendment" to the Bright Line constitution


These problems raise an obvious question: what needs to be done? We have one short term suggestion that would immediately relieve some of the burden on volunteer editors and the wait times for adherents: The Bright Line should include an allowance for "maintenance edits".

Currently, the Bright Line allows exceptions for "emergency edits" that are comparatively rare: missed (obvious) vandalism and libel. A simple fix would be to allow for "maintenance edits" such as de-orphaning an article and removing the template afterward.

By applying common sense and allowing for edits that do not alter previous editorial decisions, the burden on volunteer editors can be eased, and neglected entries can be improved. To assuage concerns of potential abuse, COI editors might be required to use a standard edit summary such as "COI maintenance edit" so a filter could be created for identification and review.

Let's start over
The nuclear option, of course, would be to abandon the Bright Line altogether. While we wouldn't necessarily encourage doing so at this time, there is one (however unlikely) scenario in which doing so would make a great deal of sense: if Flagged Revisions were to make a comeback. Especially in light of the recent GamerGate debacle, the only difference really is whether the debate over GamerGate-inspired edits should have been a public game of whack-a-mole or a semi-public queue for editorial review.

In this scenario, editors with paid conflicts would receive scrutiny, and with new community infrastructure—not to mention some valuable gamification—they would be more likely to receive it in a timely manner. The uncertainty of how to participate and the absurdity of asking for help when the correction is obvious would be reduced, if not altogether eliminated.

A more elegant solution
Given our years of experience with client requests, we are comfortable explaining how Wikipedia works even to skeptical clients. Knowing how complex even "simple" requests can be, and how important it is to get things right, sometimes the months of research, writing and discussion are necessary. But the Bright Line in its current form makes no distinction between that which deserves careful scrutiny and those requiring a lighter touch. A few common sense adjustments would make the Bright Line easier to explain, more likely to be followed, and free Wikipedia volunteers to focus on more important things.

 William Beutler and 16912 Rhiannon are principals in a digital consultancy that specializes in online community content, including Wikipedia.

On Friday, July 17 at 16:45 (Mexico City time), William Beutler will be leading (with Andrew Lih) the Wikimania roundtable discussion Can Conflicts of Interest (COIs) be aligned with the Wikimedia project? Please join him to discuss this idea as well as anything else related to COI/paid editing.