Microsoft to Wikipedia: Just the Facts, Please

Since August, Microsoft has sought to get changed what it believes are inaccuracies contained in Wikipedia articles about desktop file formats, particularly Open XML. After months of slow responses from Wikipedia, changes were made, following word that the company offered to pay Rick Jelliffe to correct entries. The Talk:Ecma Open XML discussion page catalogs some of the requests and changes.

News stories proliferated over the last couple days, with many aghast that Microsoft would offer to pay someone to edit Wikipedia posts. Initially, I shared a similar view, which my colleague Steven J. Vaughan-Nichols expressed yesterday, about the pay-for-edit offer. But in the more than 24 hours since, my position has changed, and I'm backing Microsoft on what was an arguably risky endeavor.

I keep coming back to the same question: What other choice did Microsoft have? Given Wikipedia's current policies, which would discourage Microsoft or its representatives from correcting entries, and slow Wikipedia responsiveness to requests for corrections, the answer is none--or close to it. Microsoft had the option to publish a white paper and link back to Wikipedia discussion forums. But, c'mon, what kind of resolution is that?

Microsoft knows that Wikipedia is a widely used research tool. Even my 12-year-old daughter uses Wikipedia as a resource for school work, as do her friends. Is it surprising that Microsoft would want to remove any perceived bias on Wikipedia? Microsoft's position would be indefensible if the company approached a biased rather than fairly neutral third party. However, Jelliffe is considered to be fairly neutral.

"We believe that it is important for Wikipedia to have the most accurate content possible that reflects the true facts about Open XML," said Microsoft spokeswoman Catherine Brooker. "Since Wikipedia's editing staff had concerns about information we were calling out as inaccurate, we thought that they might be more comfortable if we enlisted the help of someone who could look at the entries with an outsider's perspective."

What Really Happened?
The pay-for-edit ruckus started on Monday, when Jelliffe posted about Microsoft's offer, over at the O'Reilly Weblog. Microsoft has denied nothing. Yesterday, Doug Mahugh blogged about the offer that he made to Jelliffe. Mahugh is a technical evangelist for Microsoft's Open XML formats. Jelliffe is CTO of Topologi.

I've read the e-mail exchange between the two men, and I simply don't see anything nefarious in Microsoft's intentions. In the initial e-mail sent last Thursday, Mahugh begins:

"As you've probably noticed, IBM has really ramped up the anti-Open XML message in recent days because we're in the ISO contradiction period. I posted on my blog about it this evening, and a bunch of folks here will probably be responding to various specific allegations in the days ahead."

Mahugh is referring to the ISO ratification process, which for Open XML is newly underway. The early process focuses on contradictions--the possibility that Open XML could contradict, or invalidate, another approved standard.

Mahugh's e-mail continues:

"We also have some situations where there is misinformation about Open XML that we'd like to hire a third party to clear up. For example, Wikipedia has an entry on Open XML that has a lot of slanted language, and we'd like for them to make it more objective but we feel that it would be best if a non-Microsoft person were the source of any corrections. We have plenty of budget to apply to this, and would rather have it done by a contractor than by an employee."

There is good reason to be sensitive to information accuracy available to ISO, particularly during the contradiction period. So, it's no surprise that Microsoft would want to correct perceived or real inaccuracies wherever possible, including on Wikipedia.

On Sunday, after Jelliffe accepted Microsoft's offer, Mahugh wrote him:

"I've received permission to move forward as I had outlined in my last email. So feel free to say anything at all on your blog about the process, about our communication with you on matters related to Open XML, or anything else. We don't need to 'approve' anything you have to say, our goal is simply to get more informed voices into the debate."

Mahugh's e-mail clearly indicates that Microsoft wanted transparency, and he did not insist on editorial control. Mahugh later reiterates that Microsoft would be willing to pay for Jelliffe's time. No strings are attached.

Mahaugh met Jelliffe just a few weeks before making the request. Expertise with Open XML and the competing OpenDocument Format, and "fairly strong relations with both Sun and Microsoft" made him the "ideal person to take on this project, as he would not have any discernible biases in favor of either format," Brooker said. "Plus he has a rich knowledge of ISO procedure, which was another point of inaccuracy in the current Wikipedia entries."

What Happens Next?
Anyone following my writing as analyst or journalist would know that I've given Microsoft a pretty hard time over its approach to file formats. Whoa ho, while researching this post, I find myself quoted in Wikipedia articles OpenDocument Adoption and WordPerfect. I'm not an easy Microsoft defender when it comes to file formats.

If there is a problem with Microsoft's efforts, it isn't offering to pay for editing but Wikipedia's resolution process. I e-mailed Wikipedia co-founder Jimmy Wales this afternoon for response to Microsoft's pay-for-edit scheme but got no response. An AP story by Brian Bergstein quotes Wales as saying that Microsoft should have posted a "white paper" and linked to Wikipedia forums.

Dare Obasanjo, a Microsoft developer, set that process to test this afternoon, in an action that's sure to stir up more controversy. He explained earlier today:

"As an experiment, I've updated the Wikipedia entry for TechCrunch with a mention of some of the claims about Mike Arrington's conflicts of interest on the site and references to negative blog posts but no link to his side of the story. TechCrunch is big enough for Mike not to care about this but what should be his course of action?

"According to [Wikipedia founder] Jimmy Wales and the pundits it seems (i) he can't edit the entry himself nor (ii) can he solicit others to do so. Instead he needs to write a white paper about his position on conflicts of and then link to it from the talkback page for his entry.Yeah, I'm sure that's going to get read as much as the Wikipedia entry."

A few hours later, Arrington, who edits TechCrunch, responded to Obasanjo: "This action would not be acceptable under any circumstances."

Obasanjo's action got me to looking at Wikipedia in a way I hadn't before. For example, there are seven entries for Microsoft, which are similar but in ways vastly different. How accurate is that?

In the all caps Microsoft entry, I am quoted: "It's almost like Microsoft is designing for geeks and Apple is designing for real people." However, the quote appears in one Microsoft entry, but not others. The context in the Wikipedia article is Microsoft's sometimes complex approach to designing user interfaces.

The quote, which I was surprised to see on Wikipedia, came from a May 2006 news story, where the context is accurate. But what if it wasn't? What recourse would I have other than sending e-mail to Wikipedia?

Journalists are accountable for inaccuracies. If I were to make a factual error in a blog post or news story, Microsoft could contact me for correction--or even go directly to my editor. There is no committee and no white papers to write. The Wikipedia process is vastly different because, according to the Web site, "Wikipedia has no editorial board."

Today, Citizendium started soliciting user accounts ahead of its public opening. The new online encyclopedia, established by Wikipedia's other co-founder, Larry Singer, will implement some editorial oversight and require participants to use their real names.

Maybe Jelliffe should sign up.