Jump to content

Wikipedia:Wikipedia Signpost/2024-05-16/Arbitration report

From Wikipedia, the free encyclopedia
File:Tholos of Delphi 02.jpg
Bernard Gagnon
CC BY-SA 4.0
60
430
Arbitration report

Ruined temples for posterity to ponder over – arbitration from '22 to '24

Each vertical line indicates a single issue, with color showing how much time passed since the last (from short to long)

It has been a while! We have gone a very long time without a proper arbitration report. I would like to wish everyone ("including the haters and losers", as the saying goes) a happy holiday season, new year, and all the other stuff.

So:

At one point, the arbitration report was a regular Signpost feature. In 2005, the paper's inaugural year, 49 out of 51 issues carried a report. There were also 12 Signpost articles covering the process of the second-ever ArbCom election. In 2006, there were 59, with regular reports as well as another series following the election process. The next decade or so had a slow decline: 58 in 2007, 51 in 2008, 2009, 2010, 53 in 2011, 37 in 2012, 29 in 2013, and only 6 in 2014, a year featuring infamously high-profile cases like AP1, GGTF, and the first couple months of the GamerGate case. Since then, there was a bit of a rebound for a couple years: there were 17 in 2015 and 15 in 2016. But we went back to just five in 2017 – a year when there was no Signpost at all between February and June.

After that, The Signpost transitioned to monthly publication; there were eleven arbitration reports in 2018, and nine in 2019 and 2020. But we hit a new nadir in 2021, with just three for the whole year; in 2022 there were four, and last year there were six. This year, you are reading the first one.

In all honesty, though, there is more ground to cover than just 2024: the last report (in November) was about a single incident, and the couple of reports before that one were fairly restrained summaries of a contentious case (which had already been the center of a rather long AN/I thread about its previous Signpost coverage). There was a report in January 2023, but that didn't cover any cases either – it just went over the results of the 2022 election. The most-recent arbitration report that I can genuinely call comprehensive was from August 2022.

What's happened since then? Let's pick up where we left off, in August 2022.

We're going to be moving pretty fast. I am going to be skipping over all the tl;dr stuff like unban appeals, admonishments, reminders, clarifications, internal functionary and clerking business, announcements that the Committee has gotten a new coffee maker, and the like: if you want those, read the arbitration noticeboard archive. I will also be skipping things previously covered in The Signpost (like the controversial suspension of arbitrator Beeblebrox last December, or things that have appeared in previous arbitration reports). Here, my hope is to tell you some things that you might not already know by virtue of having a brain and an Internet connection.

Hopefully, some of these may be things that actually have some effect on your daily life as a Wikipedia editor. Indeed, maybe they can even help you keep abreast of changing social norms, so you do not end up like one of those guys who logs in, straps on his tools just like he always used to, does something you're not supposed to do anymore, and gets immediately frappéd at the nearest noticeboard. Maybe this is just wishful thinking.

It should be said, before we start, that the great majority of these things are sad and tragic: being a Wikipedia editor represents a substantial investment of time and effort, and running in the elite circles of administrators and functionaries even more so. The editors who are summarily cashiered and defrocked during these proceedings are rarely bad people – they are just people for whom circumstances have aligned to become incompatible with a hobby of editing Wikipedia. They are often people for whom the hundreds or thousands of hours they've invested in the project have cashed out to ignominy and shame. And there, but for the grace of God, go you and I.

Anyway:

Deletion business

Case closed on August 2.

A bunch of frequent AfD participants were topic-banned: 7&6=thirteen, Johnpacklambert, TenPoundHammer, and Lugnuts.
The last of those users, Lugnuts, (who had created over 93,000 Wikipedia articles over the years, most of them stubs) went berserk one day before the decision was posted, and dramatically quit the project after delivering a villainous speech:

The jury is still out on whether this was even true, or if he was just making it up on the spot to fuck with everybody on his way out. Skilled fridge-logic noticers may deduce that, for this to be true, Lugnuts would have had to prophesy his banning some decade in advance – and if you have the capacity to do that, why not just use your huge brain to not do the thing that gets you banned? At any rate, two proposals (WP:LUGSTUBS and WP:LUGSTUBS2) were eventually made to mass-draftify giant swaths of his article creations; both passed.

One of the other remedies in the case was to create a couple of gigantic train-station RfCs: the type of multiple-question ordeals with multiple phases and formal moderation, in the style of RFA2021, Barkeep's bountifully-ballyhooed big beautiful baby boy of a three-phase formal debate on reforms to the adminship process, which seemed to be fairly well-received by the Wikipedian public.

  • Article creation at scale: Gigantic eighteen-part RfC, with a separate page for an eighteen-part tri-admin closing statement. Of those eighteen proposals, almost all failed or were closed as no consensus. The rest were moved to the second RfC; the only one that passed was number 2 ("Should we require (a) source(s) that plausibly contribute(s) to WP:GNG?").
  • Article deletion at scale: This one kind of sat around for a while waiting to happen, and then it never did. At the February '23 motion to rescind the remedy for the second RfC, arbitrator Izno said that "we had issues finding both moderators and closers, we didn't get the participation expected in the first one, and it's been 6 months since the motion passed and the community still hasn't had the opportunity to discuss the issues that prompted the remedy". His fellow arb CaptainEeek, contrariwise, said: "Thanks to all who participated and coordinated. In ending this, I want to clarify I don't see this as a failure. In fact I see that this process has been quite successful. Ending it now is a recognition that it has served its usefulness."

Jonathunder desysopped

Case closed on August 28.

Another "hang up the spurs" case – an administrator from the early 2000s (in this case '06) with a downwardly-sloping Xtools graph did some cowboy hip-shot and removed another user's rights without explanation, got ricocheted to ArbCom, saw the notice on his talk page, and never set foot on Wikipedia again. This all happened in February, but the case was formally closed in August. For the sake of avoiding an RSI flareup, I will just say "hung up the spurs" every time this happens in the rest of the report.

2005 was before my time. I wonder what kind of guy Jonathunder was? From his Xtools I can see that he made 129 edits to the article about the presiding bishop of the Episcopal Church, and 104 edits to the largest lake in Minneapolis (whose article he created). It seems very unlikely that he'll ever come back, so I guess I'll never know.

Discretionary Sanctions and Contentious Topics

Review concluded on December 14.

WP:DS are now WP:CT. A bunch of template language was shuffled around. People discussed it at great length in places like this. As far as I can tell, no significant changes were made to the bones of the process.

Athaenara desysopped and indeffed

Motion passed on October 16.

Hung up the spurs, with a twist: rather than a full arbitration case, this was a Level II desysop.
The series of events started with this oppose vote at a request for adminship, that read more like a jisei:

Huh?

Athaenara hadn't been an active editor in quite some time, and old-timers can be forgiven for being a little out of touch, but this was so outrageous as to utterly beggar belief. The only reasonable explanation was that her account had been broken into by some asshole who was signing her name to stupid letters, so Floquenbeam blocked her account as compromised. She was then unblocked by Lourdes (who is a whole story of her own).

Having been given a perfect ready-made opportunity to save face and blame the whole thing on an anonymous hacker and move on basically scot-free, Athaenara decided to cast it all into the flames. She not only confirmed that that really had been her, but also said a bunch of other stuff that also sucked. Well, thanks for the clarification.

There was a quite long case request, with an eye-watering 95 statements, which was eventually resolved by motion. Athaenara's block log tells the rest of the tale.

TheresNoTime decheckusered and deoversighted

Case closed on November 2.

In the wake of the Athaenara incident, there was an ancillary incident involving CheckUser and administrator TheresNoTime, who had been following along at home during the initial disaster, noticed Lourdes' unblock of Athaenara, and suspected her of being up to no good and collaborating with Athaenara offwiki. So they (TheresNoTime) ran a CheckUser on her (Lourdes). They were actually correct about Lourdes being sus – she was to be unmasked in late 2023 as having been an indefinitely-banned long-term abuse puppetmaster the whole time – but nobody knew this at the time, not even them, so a case request was made on the basis that this had been an improper use of the CheckUser tools. Primarily, due to TheresNoTime having been involved in the incident beforehand (they were the co-nominator of the RfA where the initial comment was made). The whole situation was kind of complicated, and there's a detailed timeline (with timestamps) at the case page. Eventually, TheresNoTime lost the tools but retained their adminship.

A year later, it would turn out that Lourdes had just been a permabanned LTA messing around with everybody to cause drama the whole time. See what I said about "sad and tragic"?

Stephen desysopped and then undesysopped

Desysop done on November 14; case closed on December 7.

Stephen, a quite-active administrator who was heavily involved with the Main Page's In The News section, found himself at the center of a scandal. As the finding of fact from the case says:

During the investigation of an unregistered user harassing another user, a CheckUser determined that the IP address associated with the harassment had previously been used by Stephen, an administrator. Stephen had been in disputes with the harassed editor in the past, and was the only registered account using that IP. According to CheckUser data, the harassing edits were made using a device that Stephen had not previously used. The Arbitration Committee reviewed these findings and determined that they were well founded.

Judging from the talk page section on this noticeboard entry, nobody really expected this, and most commenters were some combination of confused and dismayed to hear the news: "This is just so disappointing. Stephen was one the most dedicated admins working at ITN. I'm really shocked."
Indeed, it was shocking to hear that he did this, because he apparently didn't, per the other finding of fact:

The explanations provided by Stephen are sufficient to indicate that it was not him performing the unregistered editing and harassment. [...] The administrative permissions of Stephen are restored.

Strangely cryptic announcement about Iranian politics disruption

Announcement made on December 19.

Whatever this is:

The Arbitration Committee has been made aware by the Wikimedia Foundation's disinformation team of continued disruption in the Iranian Politics (IRANPOL) topic area, which was subject to an ArbCom case last year. Additional measures to address this disruption may be forthcoming in the year ahead from the Arbitration Committee and/or the Wikimedia Foundation. For now, the Arbitration Committee is informing the community of this disruption in the hopes that more editors and administrators may wish to begin working in the IRANPOL topic area. Uninvolved administrators are also reminded that editor restrictions and page restrictions are available for use in the topic area.

Nobody in the peanut gallery seemed to know what specific thing(s) they were talking about here either. I sure don't. Barkeep bestowed some breadcrumbs upon the beggars, giving a short list of "topics with-in the broader Iranian Politics area which are likely to have ongoing disruption and/or have had disruption in the past and in no particular order", but no further details than that.

Armenia-Azerbaijan 3

Case closed on March 18.

The last case on this conflict (WP:ARBAA2) was from 2007, and this one from January 2023 covered the then-ongoing Nagorno-Karabakh conflict between the two nations. It followed a gigantic number of threads at the arbitration enforcement noticeboard and various other locations, all of which failed to resolve the underlying issues.
The findings of fact here should be unsurprising, given that there was a real-life warzone: people were edit-warring on Wikipedia as well. Additionally, the Committee received evidence that substantial off-wiki canvassing had been carried out by editors supporting both countries, on both Reddit and Facebook (although this canvassing didn't seem to really affect the discussions that much).
At any rate, several users were subjected to various sanctions: Abrvagl, Dallavid, Olympian, and ZaniGiovanni were all given one-revert restrictions and topic bans. Golden and Grandmaster were also put on probation, and a two-way interaction ban placed between Abrvagl and ZaniGiovanni.

Dbachmann desysopped

Motion issued on April 5.

Hung up the spurs. A 2004 admin, whose editing had tapered off considerably in the last decade, controversially reversed an administrative action in which a user (AndewNguyen) was blocked as a "single purpose civil POV pusher determined to promote a fringe point of view regarding Race and intelligence that does not align with Wikipedia's long-standing consensus and reliable sources". Dbachmann, who had not used the block tool since 2011, objected to the block on the grounds of due process, undid it, and was taken to task for doing so. He said in his preliminary statement:

A variety of evidence was presented, some of it ranging back quite far; ultimately, Dbachmann's refusal to engage further with the case request (this was the last edit he ever made on en.wp), and that along with some highly questionable racial comments brought up from years previous moved the Committee to this:

For egregious misuse of an admin tool, for losing the trust or confidence of the community, and for failing to address the concerns of the community within a reasonable time period, while being aware of those concerns, contrary to the expectations of admin conduct and accountability, Dbachmann is desysopped. Dbachmann may regain the administrative tools at any time via a successful request for adminship.

Scottywong desysopped

Case closed on July 10.

This arose from a discussion at ANI, and became a true scorched-earth casepage: both parties are indefinitely blocked, one of them by his own hand.

Scottywong was an active technical editor and administrator (the latter for eleven years), who ran about a dozen Toolforge services. ಮಲ್ನಾಡಾಚ್ ಕೊಂಕ್ಣೊ (transcribed alternately as "Maldanach", "Maldanch", and "Malnadach" on the casepage) was also a technical editor, and bot operator (of the aptly-named MalnadachBot, which carried out high-volume editing to fix lint errors but occasionally introduced them).

If you are a technical editor, you may be able to figure out what the point of contention was: Scotty thought MalnadachBot was a waste of resources that was breaking pages for no reason, and that Malnadach was unnecessarily changing stuff. Malnadach thought that Scotty was unnecessarily raising objections to important technical fixes.

Scotty demanded that Malnadach "stop with the annoying useless edits already". He also made a derogatory reference to Malnadach's username: "Hello, user with non-English characters on the English Wikipedia. I don't even know what to call you. In my head, I just think of you as 'Mr. Squiggles' because your username just looks like a bunch of squiggly lines to me." This last part was a bridge too far for mostly everyone, who universally condemned him for this strange and ostensibly racist comment.

For failure to meet the conduct standards expected of an administrator, Scottywong's administrative user rights are removed. Scottywong may regain them at any time via a successful request for adminship.

Like the TheresNoTime case, this was another stupid tragedy: Malnadach was revealed to have been some permabanned sockmaster troll the entire time, and his account globally locked – during the case, no less. Nonetheless, Scotty's damage had already been done. He was desysopped by the Committee's decision, and as his final admin action he blocked himself with an expiry of forever.

AlisonW

Case closed on July 16.

Hung up the spurs: arbitration case brought on by a AN/I thread brought on by an involved block. This block came from Alison W, a 2004 admin who was an attendee of the first-ever WikiMeetup and involved in the earliest days of the Wikimedia Foundation. Her most recent month with more than a hundred edits was December 2009. Unlike most of these cases, she took an active role in responding to the case, and she's even edited in the months since.

Ultimately, the Committee found that her interpretations of the relevant guidelines no longer comported with what is now expected of administrators. Arbitrator GeneralNotability said: "AlisonW, your years of service are appreciated, but I think the most honorable course of action here is to recognize that you are no longer up on community norms, voluntarily hand in the bit, take the time to familiarize yourself with those new norms, and then re-RfA in the future." Ultimately, she was desysopped.

SmallCats

Case closed on August 25.

This is one of those cases that resist all attempt to make sense of what's going on. Everything feels like a reference to some older, more ancient conflict; it's possible to just directly read what's being said, but this feels like missing the point in some important way.

Ultimately, the outcome of this case was that two editors were banned: one with sixty thousand edits over thirteen years, and the other with three million edits over the course of nearly twenty. Perhaps, given more time, it would be possible to say more about it than that, in a way that gave some genuine understanding of the underlying issues; in a way that explained what was going on here; but today is not that time.

When I mouse over BrownHairedGirl's username, a little box pops up that says "2,942,733 edits since: 2006-01-04". The talk pages and precedents and twenty-year-old diffs being discussed in this case feel like ruined temples with inscriptions I can't read: maybe some day I'll be able to.

Mark Ironie and CorbieVreccan

Motion issued on September 17.

This case request was brought about by this AN/I thread. During the course of the case request, both of them maintained that their situation had been disclosed to the Committee for many years, and was not considered to be a problem. Nonetheless, the main allegation was of potential conflicts of interest, misuse of administrative powers, and patterns of support in discussions; for example, Mark often showed up to agree with Corbie in arguments. It didn't result in the Committee opening a case, but it did result in both users requesting a desysop, and a rather unusual motion being resolved:

Mark Ironie and CorbieVreccan will be considered a single user for Wikipedia's purposes. When editing the same articles, participating in the same community discussion, or supporting each other in any sort of dispute, these editors must disclose their connection and observe relevant policies such as edit warring as if they were a single account.

Subsequent events

This recap brings us up from the middle of '22 up through almost the end of '23. The astute observer will note that it's currently May of '24, meaning that even after a marathon report covering eighteen months (!), we remain a couple months behind. There are some additional cases that I feel deserve more attention than being one brief list item in a report that's already nearing the limits of the human attention span.

Contrary to some claims made in the badsite threads "Is the Signpost ignoring an ArbCom case?" (no) and "Jacob Gotts aka JPxG: liar or braindead?" (no), the absence of an arbitration report covering the Nihonjoe case in the last couple months was not a deliberate act of censorship (there wasn't anything to censor), nor was it a declaration of war against the badsite (about which I remain ambivalent). Rather, it was an issue of insufficient resources to report on the case thoroughly. We will cover that case, and some others, in the next report.

In the meantime, please enjoy the encyclopedia with understanding and compassion in the search for truth, and love your neighbor as you do yourself.