The Durova incident

There is a lot of noise and fury about two "secret" mailing lists doing the rounds. This is largely as a result of a piece in The Register, which took at face values the opinion of one user who had decided, for whatever reason, to disbelieve every single person who had any inside knowledge of the events in question, and on top of that discount anybody who made even the slightest assumption of good faith.

Needless to say the blogosphere, which loves nothing better than "MOAR DRAMA!" has taken this up with enthusiasm.

So here is the chain of events.

Harassment on Wikipedia

In August of 2007 an email chain started on the subject of harassment on Wikipedia. It was circulated by cc list, and the list got longer and longer as more interested parties stepped forward.

The email chain was discussing the vexed issue of harassment on Wikipedia, and how to respond better to it. One specific problem is the fact that free-speech types will often revert reflexively the removal of any kind of harassment or attack as "ZOMG! CENSORSHIP! WIKIPEDIA IS NOT CENSORED!", leaving victims in the unacceptable position of either leaving the harassment or having a massive drama about it, at the end of which the consensus may well be to remove it but it will have been spread across the net.

Harassment of Wikipedia editors is real and has included:

  • Calls to work
  • Attempts to get people fired
  • Email harassment
  • Publication of private details such as home addresses
  • Speculation on real world identities of pseudonymous editors
  • Phone calls to family
  • Threats of harm, demonstrating detailed knowledge of locations
  • Police involvement
  • In one case, the imprisonment of an individual for stalking

As a result of harassment, editors have been driven from Wikipedia and in some cases have had to move house. It is obvious to most people that we cannot and must not ignore this, but to remove or downplay the repeated posting of harassment memes is close to impossible for various reasons including:

  • Hard-line free speech advocacy
  • Assertions of "no smoke without fire"
  • Obsessive promotion of harassment memes by those pursuing the harassment
  • Opposition to the individuals being harassed, many of whom are admins
  • Banned harassers being given more credit for good faith, largely due to the above
  • Good-faith users being cynically manipulated by abusers

So either we need to develop a better system of dealing with harassment or we are going to run out of admins willing to stand up against abuse from the likes of Judd Bagley (User:WordBomb). It is important to note here that Wikipedia is incredibly tolerant of criticism, allowing near-endless repeats of the same arguments. You have to make a real, special effort to get banned.

Yes, sometimes people have falsely identified harassment. But harassment has occurred, to the extent of law enforcement being called, and denial is not an acceptable way of dealing with it. Step one is acknowledging the problem. Those who refuse to acknowledge the problem exists, risk becoming part of it.

The list

The cc list of the email was unmanageable, so a private non-archived mailing list was set up on Wikia. We could have used yahoo groups or something, but Wikia is free and we knew people who would set the list up for us there and then, so it was quick and easy - and it's an associated site, of course. It has to be private because otherwise it would be impossible for victims of harassment to discuss their problems in confidence. Real names are used, and inevitably the real behaviour of other editors is discussed.

The list is, obviously, full of hurt and anger. We have even talked about banning! Yes, we have. We have talked some admins down form banning a user whose contributions to the project in the last twelve months have been almost exclusively restricted to arguing in project space. It's a place, then, where hurt and angry people can share their hurt and anger and be talked down form rash actions.

Unfortunately it's also a place where Durova chose to send her "sleuthing" email.

The other list

A second list, investigations, was set up in November to handle the noise that lists of diffs were causing on the main list (which attracts hundreds of messages on a busy day). This list has fewer than 300 messages total, and Durova did not send the email to this list. No blocks have, to my knowledge, been enacted as a result of posts to this list, but I unsubscribed a couple of days back because most of the messages seem to be a running argument between two list members.

The Durova affair

User:Durova is a past victim of harassment and attacks, and is also very keen on preventing abuse of the project. She knows (as most of us do) some of the signs of the returning user, the "sleeper sockpuppet" and other tools of the abuser. Sleeper socks register, make a few edits, lie dormant for a while, and then pitch in to cause problems. They have been extensively used by User:JB196 and user:Jon Awbrey among others, and the reason is to get in under the radar of recent changes (RC) patrol and past semi-protection on articles. Nobody seriously disputes that this goes on.

Durova enjoys the whole business of tracking down such accounts. She calls it "sleuthing". I am not a fan of that term, or of the idea of actively tracking down sockpuppets. I tend to wait for a problem edit and then ask for CheckUser. The downside of this is that it usually means a lot of reverting afterwards, and some of that will be opposed by well-intentioned people. A key part of the strategy of our most determined abusers is to exploit divisions and apparent absurdities. For example, Revert, Block, Ignore - "why revert it, it looked like a good edit to me". Yes, it might look like one, but we can't tell and we don't have the concept of "a little bit banned". RBI exists because it's easier than reviewing word by word the edit of a banned user looking for evidence of the problem that got them banned, which would require detailed records of their problem behaviour that both glorifies and immortalises the abuse. We'd rather they went away, and keeping long term abuse pages just provokes them into coming back time after time. In the view of many of us, it reduces the likelihood of reform. Some people think that "revert block ignore" is absurd. I disagree. A banned user is banned because they are unable to contribute productively and because the overhead in other volunteers' time in allowing them to do so greatly outweighs the value of their actual contributions. Those whose contributions are of greater value than the overhead, do not get banned. We did not ban SPUI despite endless aggravation.

So, to get back on track, Durova does sleuthing, but very few others join her in that.

Durova, unfortunately, made several serious mistakes. Some of these we should have seen coming, some of them completely blindsided us.

Mistake #1: Durova sent an email correctly noting that User!! was clearly not a new user. Durova decided that he was a returning banned user, not a returning user, though. There never was any credible evidence to support that. It did look as if !! was aiming for adminship, and people are a bit paranoid about banned users getting the sysop bit. We've had admins leaking IRC logs and the contents of deleted articles before now, and User:JB196 subverted one admin, very much to the regret of his friends. See also the Everyking desysop.

Mistake #2: Durova mistook silence for assent. Most of us said nothing about the email. One or two of us I think agreed that this was clearly not a new user (accurate: it wasn't). Durova appears to have mistaken this for agreement with the conclusion which was implicit and evidently much clearer in her own mind than in the message itself. It's clearly possible to agree that X, Y, Z are facts as stated without necessarily subscribing to any inference drawn from X, Y and Z. Interpretation of evidence is as important as the evidence itself. Truth to tell, I don't think many of us read the narrative; diff of first edit, second edit, third edit - not a new user, yup, obvious.

Mistake #3: Two weeks later, Durova blocked !!. She never said to the list, or to anyone else I've heard from, that she was intending to do this. Had she said so, perhaps we'd have read the "evidence" in more detail and spotted the disconnect between evidence and conclusion. As I say, mos of us did not, I think, even read the conclusion, only the evidence, which was, as represented, a clear indication that this was not a new user. I guess most of us had forgotten about the mail by this time anyway.

Mistake #4: Durova said it could not be challenged other than to ArbCom. There are some kinds of evidence which only ArbCom should see - email headers, CheckUser output and so on - but the diffs Durova provided were not of that class. There's good reason not to share with every vandal the Top Ten Tips for avoiding detection, but any admin will know these. "mail me for details" would work there. Better still, some kind of discussion before blocking, as the editor was apparently known to a few admins. There's no real problem with changing accounts and telling a few people you've done so, which is what !! did. Some editors change accounts to evade scrutiny, but there is no suggestion that !! was doing that.

Mistake #5: Durova implied that the Super Secret Cabal had approved the block. Hell no. It's private, not secret, and no approval was sought or given.

Those were Durova's mistakes. They were bad ones. There were other errors, too, but Durova did the right thing in promptly unblocking !! as soon as it became clear that there was no abuse and the previous account was in good standing.

Burn the witch!

An RfC was started on Durova. This was long on accusations of bad faith and evil intent and short on - well, anything of any merit. It went to ArbCom, whose conclusions were pretty much perfect as an assessment of the situation. A jaw-droppingly bad call by Durova, no others involved in making that call, read and learn. None of us on the list feel good about this, of course. Not only did Durova get hung out to dry, we also had the recurrent problem of people asserting what the list was about and for. I've said above what it was for, and whether you choose to believe me on this, or believe Jimbo (who is on the list and said the same thing), or choose to believe any of the other list members who've spoken out, or indeed believe nobody at all, what I've said above is the plain truth of the matter.

But there were other problems. Significant among these was the publication of Durova's email by Giano, "for your entertainment" as he admitted in one comment. Sorry, no. We don't publish private email for entertainment. The arbitrators already had the email (some of them were on the mailing list anyway), there was no need to publish that email.

Lots of discussion on wikien-l, also characterised by people on the list pointing out what it was for and that nobody approved the block, and on the other side people saying "I don't believe you, now provide the emails where the block was approved". A stale and unproductive exchange. No, not wholly unproductive: some people did accept the truth, especially once Jimbo weighed in.

One recurrent issue was: "give us the evidence of who approved the block". "There is no evidence, because nobody did". "I don't believe that, now give us the evidence" - repeated ad nauseam.

And now we have MOAR DRAMA! - especially thanks to those who want to promote some agenda. You know how it is with witches: you duck them in the pond, if they drown they are innocent and their souls go to Heaven, if they survive they are witches and you burn them at the stake.

Which brings us right back to square one: Wikipedia has a serous problem. If an outright lie is posted on Wikipedia to harass someone, the only options at present appear to be to let it stand, or to remove it and suffer a month of drama.

Giano

Giano is a long-standing Wikipedia editor with chronic civility issues but whose commitment to the project is beyond doubt. Giano posted, repeatedly, the content of Durova's email, leading to a short block and consideration of a 90 day ban. This was represented in some quarters as a principled stand; there is a userbox doing the rounds supporting him in the arbotration committee elections for posting "the evidence that exonerated user !!" but that does not stand up to scrutiny on a number of levels. Specifically:

  • User !! was already exonerated and unblocked before Giano posted the email
  • The exonerating evidence was not the email, but an admin who knew the previous account of !!
  • The arbitration committee already had the email anyway
  • Giano stated that he could no longer post the email "for your entertainment"

This is, then, a classic piece of Wikidrama, a bit of silliness escalated to absurd proportions and misrepresented even after the truth was noted. There should be nobody who is unaware of the four facts above, at least nobody who paid any attention at all to the case. And yet Cla68 and others continue to state their side of the story as if it were true.

Enter Bagley

Enter Judd Bagley, a banned user whose job is attacking those who are less than laudatory towards overstock.com (OSTK), a company whose poor share performance is due to naked short selling Gary weiss Wikipedia. Yes, definitely, must be Wikipedia, can't possibly be poor management, consistent failure to show a profit, or earnings per share of -3.68 (yes, that's minus, as in loss). The Register picked up on Bagley's story, ably assisted by Cla68 and long-time gadfly Dan Tobias, who never could resist an opportunity to push his BADSITES meme, wiht the result that Bagley's version of events was reported entirely uncritically, represented as Wikipedia doing down his poor beleagured company.

Here's some stuff about Bagley, offered as-is, no warranty expressed or implied. His earliest edits as an IP were to skew the naked short selling article and push a website which WHOIS shows was registered to... overstock.com. There is no evidence of any significant good-faith activity by Bagley on Wikipedia.

Bagley has himself and via sockpuppets and proxies tried for a long time to achieve three things:

  • Making the article on overstock.com to more positive, i.e. downplaying its poor financial performance
  • Making the article on Gary Weiss less positive, and preferably an outright hatchet job
  • Making the article on naked short selling reflect his boss' POV, i.e. that naked shorting is fraud (which it isn't, not legally at least, although I for one find it as reprehensible as hedge funds and derivatives).

In pursuit of that, he spawned a meme that Weiss was editing Wikipedia as a sockpuppet, and tried to blackmail an administrator into allowing him to get what he wanted, else he'd publish Bad Things about her - which turn out to be the same bad things that have already been promoted by Daniel Brandt but which have never had a shred of supporting evidence.

So once again we are doomed to go round the same loop of the various smears Bagley has tried against Weiss in the past, weighing them up against WP:BLP and finding them wanting, because in the end neutrality is not the average between what Weiss thinks of himself and what Bagley thinks of him.

The memes

Every dispute is escalated to the maximum extent possible, and then talked about outside, and this is sometimes abused by one or other side in the dispute as a way to get their version of events written into Wikipedia as The Truth™. Every Colbert fan knows that Wikipedia is The Truth, right? And let's be clear, these memes are very similar to the elephant meme in The Colbert Report: they inspire rapid, determined attempts to insert the meme into Wikipedia, and they then write about the fuss this creates as a way of rpeserving the meme for posterity. It is no different with WordBomb's naked shorting meme.

  • Use of Wikia for hosting the lists was used to support a harassment meme proposed by banned user:MyWikiBiz. This user was banned for editing for pay, as a result of which the community clarified and codified its conflict of interest guidelines. His off-wiki campaign has included attacking named editors for supposed conflicts of interest and calling the IRS to try and get the Wikimedia Foundation fined (i.e. shut down) for this supposed conflict of interest. My take on this is pretty simple: we stopped him making a buck out of our volunteer-run project, so he wants to take the whole thing down out of spite. He's a charming and intelligent fellow (no, really) and very plausible, but as soon as you scratch the surface you find determination, ruthlessness and a complete failure to get the whole idea of Wikipedia the free content encyclopaedia. He genuinely does not see what is wrong with allowing companies editorial control over their articles. Well, his own project allows that control, but any such idea has been comprehensively rejected here. As a project, Wikipedia did not ban him for being evil, we banned him for completely failing to internalise the Wikipedia ideal (with supporting evidence in the form of sockpuppets, sometimes with names being anagrams of "Jimbo Wales sucks"). Not everybody is cut out to be a Wikipedian, those who are not should be thanked for their interest but asked to leave.
  • ZOMG! MI5! A recurrent meme pursued by Daniel Brandt and now by WordBomb, a conspiracy theory with two massive logical disconnects and for which absolutely no credible evidence has ever been advanced. In fact, it's a classic conspiracy theory, because the fact that Wikipedia won't publish it is the principal evidence offered that it is true. The reason we won't publish it is because no reliable source has ever been provided.
  • ZOMG! Secret cabal! Debunked comprehensively by those who were on the mailing list, but pursued by a small group of people who seem, for reasons known only to themselves, intent on the most damaging possible interpretation of the situation. I know why the list exists, and what went on there, and what they told The Register is as close to outright fabrication as makes no difference. It's real black helicopters stuff.
  • ZOMG! Naked short selling! Ah yes, the WordBomb story. Cade, the author of the Register piece, has clearly been contacted by Judd Bagley (WordBomb) of overstock.com to push his agenda. Cade accepted his story at face value, something that pretty much nobody else has done, and now people want to use Cade's story to insert Bagley's attacks into other articles. No thanks. We debated this extensively when Bagley was first outed as the creator of antisocialmedia.net, due to his attempts to blackmail an admin into allowing him to use Wikipedia to promote his agenda. That's pretty much the main reason he's banned. Part of the WordBomb meme is the statement that various Wikipedia editors (and one in particular) are sockpuppets of Gary Weiss, an assertion that Bagley made so cleverly that even some long-standing editors believed it. This has been extensively investigated and as far as I can see those who originally believed it, no longer do. I've seen some of the rebuttal evidence, it is unquestionably the case that the editor is in a different state from Weiss and has provided some credible evidence of real world identity that shows he is not Weiss. He should not have had to do this, but assume bad faith and extrapolate from there does appear to have become the default position in some cases.

The role of Wikipedia Review

Wikipedia Review is the place where Bagley / WordBomb, MyWikiBiz (user Thekohser there), JB196 (user Looch there) Jon Awbrey (user Jonny Cache there) and others have set up shop. They are very keen to undermine those who pushed back against their abuse of the project. They are intelligent people and delight in finding and exploiting every tiny mistake made against them, as a way of drawing attention away form the fact that the vast majority of actions taken against them are 100% spot on. Their role is aking to that of Grima Wormtongue. Most especially, they are the key source of the idea that refusal to republish their harassment memes is evidence that the memes are true.

Until such time as WR drops this crap, I personally consider that active participation there is fundamentally incompatible with being a good Wikipedian. They have declared us their enemies, the site's agenda is dominated by people justly banned for abuse. I have in the past actively participated in sites that seek to offer informed critique and improve the project - this is not one of them. It is, at present, a cesspit, and it is actively damaging Wikipedia day by day. In other words, they are succeeding in their aim, and it is time for those Wikipedians who help them in this to stop doing so.

Wikipedia Review is currently dominated by a clique who despise Wikipedia - or rather, they despise the fact that it manages to do pretty well without them. They want to exploit every possible difference between various groups who are involved in Wikipedia, with the explicit aim of bringing down those they do not like (which in this case is equivalent to: any admin who will stand up against their abuse of the project). Most of them are frustrated vanity spammers, but Bagley is worse.

Why no link to this site? Because they have not yet banned the worst offenders, most notably Judd Bagley. When they do, I will believe they have some worth. Right now they fulfil one of the definitions of a sewer - full of shit and stinks to high heaven - without bringing the collateral benefit of taking effluent away. This is wikien-l's job, one that forum does well, and the fact that several WR regulars are banned from said list speaks volumes.

WR is an absolutely great place for those who are constitutionally unable to edit Wikipedia, to go and find people who are just as dysfunctional, and thus continue to pretend that the problem is everybody else.