Since I last wrote about editing Wikimedia projects, quite a bit has changed — I’m not entirely sure if writing in detail about what happened is a good idea, nor am I sure why I’m even writing this at all, but I suppose we’ll just see how this post goes… but first, a crash-course in some relevant user groups, policies, and procedures of the world of Wikimedia. If you’re already familiar with the checkuser user group, requests for adminship and related policies, you may wish to skip the explainer.
These user groups grant permissions such as:
- Viewing recent IP addresses a given user has logged in/edited from
- Viewing user-agent information for a given user’s actions
- Hiding revision content/log content from the public and administrators
As such, these user groups are considered highly sensitive positions of trust, only granted after considerable deliberation either by direct community election (i.e. the steward elections) or by a mixture of community consultation and internal deliberation by committee (CUOS call for applications).
The use of both checkuser and oversight is governed in a number of different policies; often defined as local policy, global policy and legal policy.
In addition to these policies, as an English Wikipedia administrator (and at the time, checkuser and oversighter) I was expected to abide by the local administrator policy in the actions I mention later.
To save you reading and digesting all of those policies and trying to figure out how they apply below, I’ll summarise the relevant parts here. Of course, my own bias will be evident in how I read and summarise these policies, and I provide a link to each, so you can read them verbatim.
Involved admins (link)
In general, editors should not act as administrators in disputes in which they have been involved. This is because involved administrators may be, or appear to be, incapable of making objective decisions in disputes to which they have been a party or about which they have strong feelings. Involvement is construed broadly by the community to include current or past conflicts with an editor (or editors), and disputes on topics, regardless of the nature, age, or outcome of the dispute.
In straightforward cases (e.g., blatant vandalism), the community has historically endorsed the obvious action of any administrator – even if involved – on the basis that any reasonable administrator would have probably come to the same conclusion.From WP:INVOLVED (local policy)
Although there are exceptions to the prohibition on involved editors taking administrative action, it is still the best practice, in cases where an administrator may be seen to be involved, to pass the matter to another administrator via the relevant noticeboards.
This policy is well-defined, and the excerpts above explain the crux of the expectations placed upon me while performing administrative tasks — do not take admin action against someone you have previously been in dispute with, nor when the situation in which you are acting is one which triggers strong feelings.
I also note the “any other reasonable administrator” exemption, something I later come to rely on in defending my actions.
Wheel warring (link)
With very few exceptions, once an administrative action has been reverted, it should not be restored without consensus.From WP:WW (local policy)
There’s not much which can be summarised further than the above, so I’ll give a relevant example:
- Administrator A blocks User B from editing
- Administrator C unblocks User B
Administrator D is now unable to re-block User B without first gaining community consensus (or, the explicit permission of Administrator C).
Grounds for performing a check (link)
The tool is to be used to fight vandalism, spamming, to check for sockpuppet abuse, and to limit disruption of the project. It must be used only to prevent damage to any of the Wikimedia projects.
The tool should not be used for political control, to apply pressure on editors, or as a threat against another editor in a content dispute. There must be a valid reason to use the CheckUser tools to investigate a user.From Checkuser policy#Use of the tool (global policy)
The English Wikipedia’s checkuser policy expands on the above global one, but for this post only the grounds for performing a check is relevant. In summary, the primary purpose of checkuser is the prevention of sockpuppetry and the abuse sockpuppets often cause — no explicit list of grounds for checking are provided, as “checkusers asked to run a check must have clear evidence that a check is appropriate and necessary. The onus is on an individual checkuser to explain, if challenged, why a check was run“.
“Fishing” is to check an account where there is no credible evidence to suspect sockpuppetry. Checks are inappropriate unless there is evidence suggesting abusive sockpuppetry. For example, it is not fishing to check an account where the alleged sockmaster is unknown, but there is reasonable suspicion of sockpuppetry, and a suspected sockpuppet’s operator is sometimes unknown until a CheckUser investigation is concluded. Checks with a negative result do not mean the check was initially invalid.From Wikipedia:Checkuser#Fishing (local policy)
As mentioned above, checkusers must have clear evidence that a check is both appropriate and necessary. Checks performed without clear evidence are considered “fishing expeditions” — i.e. a check performed to “see what you catch“.
Requests for adminship (link)
Requests for adminship, often just referred to as “RfA”, is a process in which editors wishing to become administrators are subjected to a week of comment and voting — our candidate advice pages are appropriately pessimistic:
Some editors have left Wikipedia as a consequence of an RfA that has gone poorly.From Wikipedia:Guide to requests for adminship
It’s suffice to say that RfA is probably the most stressful week an English Wikipedia editor will experience on the site.
In recent years, it has become very rare to see self-nominations, and it is expected that a prospective candidate will secure one or two co-nominators who will write a nomination statement and support the editor during the week.
Request for arbitration (link)
When a dispute on the English Wikipedia requires the assistance of the arbitration committee to resolve, a request for arbitration is created, often by the aggrieved editor. This case request is then reviewed by ArbCom, accepted (or declined) and case proceedings commence.
On a personal note, the request for arbitration process is incredibly complex, and it’s easy to get overwhelmed even just as an onlooker.
Before I begin, I’m going to make explicitly clear that this post:
- is not an attempt at rehashing the dispute in question.
- is not designed to plead innocence or shift blame.
- contains my own opinions and recollections of the events leading up to the loss of my oversight and checkuser permissions.
- may, where absolutely necessary, contain direct quotes from emails I sent, but never quotes from other editor’s replies.
In addition to the above, I have, where possible;
- attempted to remain neutral when explaining and summarising Wikimedia policy and procedure, and instead made it clear when making opinionated comments in respect to policy.
- refrained from linking to user pages of other editors.
For context, it may be useful to note:
- I have publicly disclosed a significant mental health condition (the same which I have posted about here previously) on my user page, in an effort to reduce the stigma editors face.
- In my disclosure, I note that “I consider myself as accountable as any other administrator/checkuser/oversighter, and explicitly waive any “potential leeway” an investigating body may wish to apply”.
- I work for the Wikimedia Foundation, though of course I don’t represent the Foundation. These opinions are my own.
Lastly, I need to stress that although some readers may feel strongly enough to contact those involved, I must insist that you do not.
Background to the RfA
At the start of October 2022, after a few weeks of private discussion via email, I co-nominated an editor named Isabelle Belato for adminship.
Part of my nomination statement, a feeling which I still hold today, read;
I’m thrilled to be co-nominating an editor who exemplifies what being here is all about — Isabelle has proven their competence both in policy and in the curation of content.
Normally the contents of a candidate’s user page is not germane to any discussion of them (except, of course, where that content is problematic) — a cursory review of Isabelle’s user page, a step I believe a lot of nominators will do, didn’t reveal anything remotely concerning nor indicative of likely controversy.
The RfA went live on the evening of the 7th, and immediately attracted stalwart support from the community. Isabelle appeared to remain a little nervous via email (a feeling that every RfA candidate can understand and share in!), but it was my hope that the overwhelming support at least mitigated that slightly. The days passed, and I reflected on my own RfA experience from seven years prior — a familiar pit in my stomach returned.
A single oppose
Often on Wikipedia, a series of events can be traced to a singular, definitive cause — in this case, the casting of a single oppose vote on Isabelle’s RfA by an administrator named Athaenara.
Oppose. I think the domination of Wikipedia’s woman niche, for lack of a better term, by males masquerading as females as opposed to welcoming actual, genuine, real women who were born and have always been female, is highly toxic. Go ahead, “cancel” me, I don’t care.From Special:Diff/1115345450
This vote was made at 01:44 AM on the 11th, so I was probably asleep.
The community reaction to such a vitriolic statement was thankfully swift — another admin blocked Athaenara at 02:21 AM, and a request for arbitration for the removal of Athaenara’s admin permissions was made not long after.
I saw the comment at around 09:00AM, emailed Isabelle, and then left Athaenara the following talk page message:
I recognise your username Athaenara, but I do not recognise your words. To hold such a belief is one thing, but to say it so plain when it did not need to be said… I’m disappointed. I’m hurt for my candidate to whom we’d already explained RfA can be a nasty experience, but we never expected comments like this. I’m hurt for those who read your remarks and have internalised them. I’m hurting for you, who must have such hatred in your own heart to feel and believe such things, where people are merely trying to live their lives. Simply — how dare you?From Special:Diff/1115401703
At this point, I apparently should have swallowed my pain and stepped away. This direct attack on someone already going through one of the most stressful weeks of their time on Wikipedia, for just being who they are…
I could not stand back. The LGBTQ+ community has had to fight, must continue to fight, and needs to send the message that this is unacceptable on every level in situations like this.
I was angry, and I remain angry that this happened.
The next day (12th October 2022), a very inactive administrator unblocked Athaenara. About an hour later I realised this, and contacted the unblocking administrator (both on their talk page, and privately via email), as well as opening a thread on the Administrators’ noticeboard for incidents:
Although the original block stood, both for the reason given in the block log and by consensus at the above thread, Lourdes has decided to unblock Athaenara without discussion, nor even an explanatory comment. Given how much of a controversial topic this block and the events leading to it are, this absolutely should have called for discussion with the blocking administrator (Floquenbeam) first, as clearly stated in Wikipedia:Administrators#Reversing another administrator’s action. As to not wheel war, I am seeking consensus to reinstate the block.From Special:Diff/1115606391
Use of checkuser
After the surprise unblock of Athaenara, I suspected collusion or compromise of the unblocking admin and Athaenara — multiple editors had already made clear their suspicions of compromise, and now with this unexpected behaviour I was beginning to worry that multiple admin account compromises had taken place for the purpose of causing disruption.
Due to these concerns, I performed a checkuser on both Athaenara and the unblocking admin (Lourdes) — it was my opinion that these checks were within policy1, 2, as the check was made to “[investigate, prevent, or respond to] disruption (or potential disruption) of any Wikimedia project“.
At the time, I did not consider that this check would be considered involved.
Use of admin tools
After letting the “Athaenara unblocked by Lourdes” ANI thread sit for around half an hour (during which editors noted their support for the reinstatement of the block), I reblocked Athaenara.
At the time, I considered the ramifications of being potentially involved (a point I personally reconciled2 as a reblock in case where any reasonable administrator3 would do the same) and wheel-warring (I depended on emergent consensus at the ANI thread — that is to say, consensus which was evident4 from a few experienced editors giving their opinions).
An Arbitration case opens
On the evening of the 12th, I received a talk page message from an arbitrator requesting that I “contact the arbitration committee at my earliest convenience” — this message is universally recognised on Wikipedia as something serious (in fact, Athaenara had received the exact same sort of message).
I emailed the committee less than a minute later, and awaited a response. Roughly a quarter-hour later, I received an explanation that the arbitration committee considered the checks I ran on Athaenara and Lourdes to be inappropriate, both per checkuser policy and involvement policy, and that they would like me to explain my actions.
My response, in which I attempted to neutrally explain the thought processes which led to making the checks, and where I explicitly made the statement that I “agree[d] in this case I should not have checked Lourdes, and by extension Athaenara.” was sent to ArbCom about an hour later — it took a significant amount of time to find supporting documentation/diffs and ensure my reply was not emotionally compromised.
I then heard nothing for six absolutely nerve-racking hours, during which everything from the public humiliation of having permissions revoked for cause, to my job being impacted, ran repeatedly through my head. I followed up to confirm the email had been received — it had, though no one had thought to let me know.
It is disheartening to know that the events which followed from this point on are the more likely contributors to the loss of my permissions — I was, at this point, unable to regulate my emotions. I rapidly changed my position5 between resigning my permissions, wanting a private case and wanting a public case.
I finally decided on a public case5, one of the first for misuse of checkuser permissions, and thus Wikipedia:Arbitration/Requests/Case/Reversal and reinstatement of Athaenara’s block began.
The entire case, I felt nothing but anger, betrayal, hurt, and disappointment towards a system which was meant to protect editors. Why was I likely to “come off worse” than the editor who posted those vitriolic comments? What ulterior motives were at play from those who were baying for my blood? Why was the standard of involved, and its explicit exception, not fairly applied to me? On reflection, I was psychotic and did not remove myself from the situation as I normally would — admittedly, this was a situation where it was made clear that I could not remove myself easily.
The ArbCom case resolved in the removal of my checkuser and oversight permissions, and my admonishment. I resigned my admin permissions on the English Wikipedia, logged out, and didn’t look back for quite a while6.
I received significant (and valid) votes for removal at this year’s reconfirmations, and as such decided to resign as a steward.
As it stands, I’m no longer able to do any of the things I’ve enjoyed doing (protecting editors from abuse and harassment, investigating large-scale disruption etc.) on Wikimedia projects for the ~7 years I’ve held advanced permissions.
It seems unlikely I’ll stick around as a volunteer in the long run, and honestly it’s taken a toll on how much I enjoy my job working for the Foundation.
I’m afraid this post has no “grand realisation of an opportunity for growth”, nor any sort of positive lesson or moral — this whole situation is one I regret, and worst of all, it was pointless. The same outcome (transphobic editor gets banned) would have occurred without me taking any actions.
All I’ve accomplished is ensuring that the next time a transgender editor feels unsafe on the English Wikipedia, I’m not going to be in a position to help.
1. — A later Ombuds Commission investigation found this action to be against policy.
2. — The ArbCom case into my actions found this action to be against policy.
3. — The Involved Admins section of the English Wikipedia administrator policy contains the exception “the community has historically endorsed the obvious action of any administrator – even if involved – on the basis that any reasonable administrator would have probably come to the same conclusion“.
4. — This was later disputed.
5. — I was probably manic/delusional at this point.
6. — I have since returned to the English Wikipedia as an admin, though admittedly at a very low activity level.