CLIMATE CHANGE ARBITRATION BIAS AT WIKIPEDIA
by Alex HARVEY h/t GoRight
Recent events in the labyrinthine world of the Wikipedia community have prompted me to speak out about what is happening there.
Many readers will have heard of William Connolley and his Wikipedia activities. If not, I recommend three articles by Lawrence Solomon, namely The Opinionator, Wikipropaganda - spinning green, and Who am I?. Then read on here.
In Solomon's opinion, William Connolley, by virtue of his admired status in the Wikipedia community, is second only to Al Gore in terms of ability to influence the climate change debate. I think this is about right. After Facebook, Google and YouTube, Wikipedia is the most visited internet website in the world (ref). (This is partly because search engines bump up Wikipedia pages in their search results.) I imagine that climate change newcomers start at Wikipedia; I certainly did.
I believe that Solomon made a few (minor) errors, and these have been discussed elsewhere, and nowhere more so than inside Wikipedia.* In my view, though, the biggest problem is the focus on the person of William Connolley himself. This has at once made William a scapegoat and a martyr, depending on your perspective. In reality, William is just one editor, and not that different from many others. And he is not the only professional climate scientist in there either; he is simply the one using his real name.
But I referred to recent events in Wikipedia, the subject of this post. These events are the selective banning and unbanning of Wikipedia editors in line with their alignment on the climate change issue, and this is taking place right now.**
My purpose here is to re-examine the Arbitration hearing of October 2010, where 16 editors were banned, including William Connolley - and Cla68. In the case of William Connolley, the Committee is shown to be extremely lenient, compared to treatment of skeptical editors. William's ban was recently repealed despite obvious signs that nothing much had changed.*** In the case of Cla68, however, who was perceived to be a climate change skeptic, it is shown that he was banned on the basis of entirely fraudulent claims, and has just now had his ban extended by another six months on the basis of a single frank, out of context remark made in an internet forum.
This double standard - even in Wikipedia - has rarely been so stark. In my view, it challenges the image of Wikipedia as a neutral, dispassionate broker of facts. We see that Wikipedia is, in fact, run by activists who drive away the neutral, objective people who would otherwise contribute.
ARBITRATION HEARING OF 2010: WHAT REALLY HAPPENED
Back in October 2010, it was reported at WUWT and by Lawrence Solomon that William Connolley had been 'climate topic banned' in Wikipedia, which meant prohibited from editing pages related to climate change. Anthony Watts wrote, "it is finally a factual realization by Wikipedia that the sort of gatekeeping and revisioning wars ... are being ... dealt with".
It was not, however, emphasised that 15 other editors were banned as well, and about half of them were either neutral or skeptical on climate change. This meant that while an insignificant number of activist pro-AGW regulars were temporarily banned, nearly all neutral and skeptical editors were banned. It was also not emphasised that William's ban was likely to be temporary whereas the skeptic bans would probably be permanent. It was easy to predict this, because William has many friends, and his inevitable appeal would always be widely supported. And with all the skeptics banned, these banned editors would not be allowed to support the ban appeals.
As one of a tiny handful of neutrals on climate change, I was disturbed to learn of this outcome. In reality, nothing was dealt with; the ruling was like justice under the senile Tiberius, with people thrown in jail or executed at the whim of jealous informers. Wikipedia has a detailed procedure for handling difficult editors, but it was just ignored.****
As it turns out, William Connolley himself may have put it best during his ban appeal:
...at the time of [the 2010 Arbitration hearing], there were multiple wars/fights/messes all over various climate pages. No-one could be bothered to work out exactly who was to blame; the solution was to ban lots of people.
This is about right; the ruling was arbitrary and lazy.
BUT IT WORKED RIGHT?
Some will claim that the ends justified the means.
I spent some time trying to decide if anything had actually changed after this ruling. My test case was the article on Climategate, where I tried to establish some balance.
The Climategate article is one of the most important hosted by Wikipedia on the topic of climate change, after the pivotal events of November 2009. The Climategate saga afforded all of us a chance to learn. And to that end, some things were learnt, and some things did change. A host of new climate change blogs appeared, for instance: Die Klimazwiebel appeared a few weeks after Climategate, followed by Judith Curry's Climate Etc., then later Isaac Held's blog appeared, and now Tamsin Edwards and Michel Crucifix have blogs too. Mike Hulme from CRU spoke out in a new direction, and Phil Jones stepped aside as CRU's director and apologised. Steve McIntyre's efforts led to general recognition for a need for openness and better data archiving. Richard Muller and his team at Berkeley reanalysed the surface temperature record and restored confidence in the surface temperature datasets.
But the Wikipedia article does not tell this story. At Wikipedia, there is a revisionist history of heroic scientists attacked by hackers and skeptic villains. The article is sheer propaganda. (Which is not to question the sincerity of the editors maintaining it; they doubtlessly believe that a neutral point of view means agreeing with this version of history.) And this is unfortunate, because without Wikipedia's support for this myth, I suspect that 'Climategate' would have led to more changes than it actually has. As a sometime participant in this debate, it dismays me how much has not changed.
So I tried to restore some balance to this article, and I found was that the situation in Wikipedia's climate change community had not only not improved, but in fact it had become worse. For instance, after the 2010 ruling, it became almost impossible to find uninvolved editors willing to comment on anything related to climate change. Administrators were likewise unwilling to act on anything relating to climate change. The personal attacks would fly as thick and fast as before if anyone dared suggest that an article was biased. Anyone interested can have a look at my efforts.
Another one of William's supporters, Stephan Schulz, a German computer scientist, also observed that nothing had changed. He wrote:
I think the relative calm at the climate change articles is only very partially the result of the last ArbCom intervention - we have seen such cycles of increased and decreased activity before, usually driven by external events (a new IPCC report, US elections, Hacked emails...).
WILLIAM CONNOLLEY'S BAN APPEAL
William Connolley had been banned for, among other things, 'battlefield conduct'. That is defined, I suppose, as focusing on fighting other editors rather than actually writing articles. I am not going to look at this in detail, but the evidence speaks for itself. In September 2011, though, he successfully appealed his ban and had most of his editing rights restored.
It was obvious, however, that little had changed. His characteristic battlefield conduct was in full display during the ban appeal itself. When the editor named Collect - not a climate change editor - argued against lifting the ban, William immediately accused him of lying and 'mud-flinging'. I argued that William should at least accept responsibility for what happened, as did ZuluPapa5. William responded: "by their enemies ye shall know them: I adduce [Alex Harvey's] and [ZuluPapa5]'s statements in my favour." These violations of the civility code occurred during the ban appeal itself where none of the Arbitrators could pretend to be not aware. But if one wanted evidence from outside of the Wiki, one only needs to glance at William's blog to see that his ongoing attacks against people he doesn't like or disagrees with.
The point about these blog attacks turns out to be relevant, because, as we'll see, in the case of the perceived skeptic editor Cla68, a single out of context remark made at a blog was the entire reason to reject his ban appeal.
I now turn to the parallel case of the banning of Cla68. The arbitrariness of the Committee's 2010 ruling was immediately apparent in the banning of Cla68. He had been one of the best editors in the toxic area of the climate change pages - one of only a few who managed to keep his cool at all times and stick to Wikipedia's policies. I never once saw him resort to personal attacks; I never saw him engaged in 'edit wars' to maintain a version of a page against the wishes of other editors; and I never saw him promoting a biased point of view. Indeed, he was so consistently neutral that the Wikipedia Review comment (see below) was the first time I had ever read a frank remark by Cla68. In Wikipedia, this would make him a model editor.
CLA68: THE EVIDENCE
I would now like to take a detailed look at the so-called evidence that was dug up to ban Cla68. I do this because it is important to show that the claims made against him were simply invented.
Cla68's battlefield conduct
18) Cla68 (talk · contribs) has engaged in disruptive behavior, including edit warring, inappropriate use of sources, and comments that were incivil and reinforced a battleground mentality,
Edit warring: 7 diffs
A "diff" is a URL that compares two versions of a Wikipedia page side by side. Diffs are the standard of evidence used in Arbitration Committee hearings.
In order to show that an editor is 'edit warring', it is necessary to show repeated reversions of another editor's changes. That is, one editor adds a change, and another removes it again. Then the first editor adds it again, then the second editor removes it again. This is known in Wikipedia as an 'edit war'.
These seven diffs given as evidence that Cla68 was 'edit warring', however, are just seven unrelated reverts. All editors are allowed to revert, so the diffs prove nothing. It appears that whoever found these diffs guessed correctly that the Arbitrators would not actually look at them. I asked Cla68 privately about this, and he reported that it was indeed the first time he had ever heard of anyone in Wikipedia claiming that a single revert is 'edit warring'.
Inappropriate use of sources: 3 diffs
If the edit warring accusation was fake, the second claim is strange. Three diffs were given to suggest a pattern of behaviour, but two of them were just talk page comments. So we see immediately that this evidence is also bogus. But there is one relevant diff provided; it may show an innocent mistake, or it may show an addition that deserved to remain - this is a matter of opinion. What is certain, though, is that it doesn't show an editor intentionally misusing sources. I say it's 'strange', because Cla68 had pointed out that the term 'alarmist' had also been used to describe scientists who believed in global cooling in the 1970s. This is correct. As evidence, he added a link to William Connolley's peer reviewed paper on the myth of global cooling. Inappropriate use of sources indeed? Certainly William's paper was relevant, but it is arguable that it did not establish Cla68's wording. In any case, whatever the problem, it is clear that Cla68 did nothing wrong by adding the link.
Comments that were incivil and reinforced a battleground mentality: 4 diffs
Although four diffs are given as evidence, once again only one of them is relevant. Three were comments made during the hearing itself, when the usual rules of etiquette are relaxed somewhat, and the diffs show at best borderline lapses of politeness that barely warranted a rebuke - not to mention a topic ban.
And what did he say? Cla68 wrote,
Trying to introduce [certain contrary viewpoints] into an AGW article in Wikipedia is often extremely difficult because of [point of view]-warring by a group of editors who mainly edits those articles.
I might add that never were truer words spoken, although you are not allowed to say so in Wikipedia. To be sure, it is a minor infringement of civility. Some will ask why? Well, while practically no one strictly adheres to Wikipedia's civility code, comments like this one are discouraged because, although no one is named, other editors in the discussion might wonder if they are the point-of-view-warring editors referred to.
But let us be clear: if this is the best evidence that those vying to remove Cla68 could come up with, it shows that Cla68 was a model editor, as many of us observed. If they had wanted to ban me, for instance, they could have found diffs showing far more significant lapses of civility at one time or another than what is shown for Cla68.
So, there was absolutely no evidence given to show an editor conduct problem. Yet ArbCom banned him anyway.
CLA68'S BAN APPEAL
Given that there was no case for Cla68 to be banned in the first place, and given the parallel case of William Connolley's ban that was removed despite continued evidence of bad behaviour, Cla68's ban appeal should have been straightforward. However, William Connolley and his supporters appeared with more evidence and arguments in favour of rejecting the ban appeal. (Recall, freed of his own ban now, William himself was allowed to give evidence - and he did.)
Connolley supporter MastCell appeared with the following:
On 9 October 2011, Cla68 (talk · contribs) posted on Wikipedia Review:
Fortunately for [William Connolley], Wikipedia doesn't have a "Child of Privilege-big-ego, artificially affected misanthropic, jaded, high-falooting" activist rule, or he would have been sent on his way long ago.
Here's the thing, I truly don't believe that [William Connolley], [and others redacted] are really scientists, because I can't belive that true scientists would act as deceitfully, dishonestly, or as insecurely and cowardly as they act. If they are really scientists, I would like to know which universities they teach at to ensure that I don't send my kids to those bush league institutions. To be clear, I respect scientists who truly believe in man-made global warming but recognize that they might be wrong. The ones who don't are the ones who try to use Wikipedia to artificially socialize their positions. 
Statements like this one, arguably, are inappropriate - even off-wiki. But this comment was made nearly six months ago, and was a single off-hand remark. Meanwhile, William attacks people at his blog almost daily. This is no secret.
But according to William,
I oppose Cla68's request for a blanket lifting of his ban. I would have been prepared to argue for a partial lifting, but I think that the quote MastCell provides is powerful evidence of the disruptive nature of Cla68, and that that his problematic attitudes continue. It is also evidence of his two-faced-ness: on-wiki, he strives for smoothness, but off-wiki the truth emerges. Cla68 attempts to dismiss this as a mere ad hominem argument and fails in any way to address the obvious problems that it demonstrates; I suggest that means any relaxation is inappropriate.
Now for the responses by the Arbitrators, the same ones who removed William Connolley's ban, and who are presently happily removing the bans of all the other activists requesting their bans be overturned.
I then presented the same evidence as I did above that showed that Cla68 had been banned on the basis of faked evidence. The Arbitrators showed no interested at all in my inconvenient submission. But after six weeks, the case was silently closed without the Arbitrators even voting.
IS WIKIPEDIA NEUTRAL?
I have shown that Cla68, a neutral editor on climate change, was banned from Wikipedia's climate change pages on the basis of faked evidence; that William Connolley, who was also banned, was allowed back despite significant question marks hanging over his commitment to change his ways; and that Cla68, on the basis of a single frank remark made outside of Wikipedia, had his ban extended for at least another six months.
Can the above be reconciled with Wikipedia's advertised neutrality? Note that Cla68 is not a climate change skeptic. In private conversations with Cla68 he has always maintained belief that the world is warming and that humans are probably causing it. There is not a single remark on record anywhere to support the contention that he is a skeptic. Cla68 was banned because he is neutral; not because he is a skeptic.
I am not sure how this can be fixed, but I certainly think that the public needs to be more aware of and more involved in what is going on in there. Wikipedia gets away with it because journalists don't care, and the public don't understand how it works and are generally not interested. This needs to change if Wikipedia is to live up to its noble promises.
Others will argue that Wikipedia's neutrality somehow exists despite actions from Wikipedia's highest panel of appeal that appear so flagrantly to contradict the notion of neutrality. If a real court was found to disregard all evidence, laws and precedents, it would be an unbelievable scandal and all judges would be forced to resign. It is my opinion that this is what needs to happen for Wikipedia to retain any credibility. The Arbitrators need to resign immediately, and if they refuse to do, then Jimmy Wales or the Wikipedia Foundation needs to remove them forcibly. This will not happen, because Jimmy Wales already knows about this and either doesn't care, or feels powerless to act.
In fairness, I am sure he cares. He doesn't have the power to stand up to these activists.
SUMMARY AND CONCLUSION
To summarise, the October 2010 banning of 16 editors from Wikipedia's climate change pages was arbitrary, and had the effect of significantly tilting the balance to the climate change activist editors who mostly control the pages, although William Connolley himself was made to take a break. It was not a decision forced on the Arbitrators by evidence, but an arbitrary decision made for convenience that violated Wikipedia's principle of treating others with dignity and respect, and its own processes for dispute resolution. The decision to repeal the ban of William Connolley despite his continued display of battlefield conduct showed that the ban appeals were also arbitrary, and the decision to then apply a double standard in the case of the neutral Cla68 shows that Wikipedia's own claim to neutrality is fiction.
* It should be noted that it takes at least a year for outsiders to make sense of what is really going on in a controversial discussion inside Wikipedia.
** Since I wrote this the case of another AGW activist appealing his ban is now underway. We see the same pattern. Being an AGW editor, the same Arbitrators are shrugging off compelling evidence that this frequently banned editor hasn't changed or even, in this case, complied with the ban. Cla68 has attempted to make a huge submission of evidence, but it has been deleted because Cla68 remains a topic banned editor. I recommend visiting GoRight's blog and reading the whole article.
*** This is not to say that William's behaviour on-wiki hasn't changed; it might have; I am not spending time there at present. The point is, at the time of the hearing, there was no reason to believe that it had.
**** I contacted Jimmy Wales and offered him the chance to dispute any factual errors, and he disputed this particular point. I then pointed out that the DR procedure states that Arbitration is a last resort, to be applied only after all avenues to resolve the problematic behaviour have failed. In the case of Cla68, though, nothing else was tried (and of course nothing was tried because there never a problem in the first place). I did not hear back from Jimmy Wales after this. Perhaps his view is that the provisions I referred to are designed to best deploy ArbCom's time, and not limit their power. If so, it hardly changes the arbitrary nature of the outcome.
Update per 25. May 2012 by Alex Harvey
Since writing this article a while ago I discovered that I made a mistake in one part. I claimed that the seven diffs given as evidence that Cla68 was edit warring were 'seven unrelated reverts'. However, only three of the diffs are completely unrelated as I first thought. Of the remaining four, two were reversions of the same content but on separate days and the other two were reversions of the same content on a single day. Still, none of these violate the "three revert rule" which is documented http://en.wikipedia.org/wiki/Wikipedia:3RR#The_three-revert_rule here. The take home point is the same: there is still no evidence of a violation of Wikipedia's rules, and I still maintain that it is impossible to explain this presentation of invalid evidence as human error. Everyone in Wikipedia knows what the 3-revert-rule is and everyone can count to three. - AH