Talk:Universal Code of Conduct/Enforcement draft guidelines review

From Meta, a Wikimedia project coordination wiki
This page is for discussions related to Universal Code of Conduct/Enforcement draft guidelines review.

  Please remember to:

  Discussion navigation:


Universal Code of Conduct/Enforcement draft guidelines review


'Overview'

The code enforcement

It states Severe systemic issues is this a spelling mistake? or what is the meaning of systemic?Yger (talk) 17:01, 17 August 2021 (UTC)[reply]

wikt:systemic. Jonesey95 (talk) 20:39, 17 August 2021 (UTC)[reply]
Thanks, but why use such an unfamiliar English word? We should strive to use simple English in our communications.Yger (talk) 19:33, 19 August 2021 (UTC)[reply]

Code enforcement definition – this definition would, at a minimum, exclude admins. Can I check if deliberate? Additionally, are we all not supposed to be doing so. If it is limited to parts of the Wikimedia Community, should it not be limited to parts of the WMF? Nosebagbear (talk) 22:32, 21 August 2021 (UTC)[reply]

"Code Enforcement Officer"

I'm aware that this is just a placeholder name, but it definitely needs to be replaced. CEO as an acronym has already been taken by the ED. And the word "Officer" implies some sort of police power, rather than a position of responsibility. --AntiCompositeNumber (talk) 02:20, 18 August 2021 (UTC)[reply]

Agree that evoking an image of formal law enforcement figures isn't ideal for several reasons, not the least of which is approachability. Something like "CoC Administrator" sounds good to me, especially because "possesses ... technical rights" sounds like it presumes having admin tools. — Rhododendrites talk \\ 20:28, 20 August 2021 (UTC)[reply]

XY "Officer" is perhaps a very good (or very bad) choice here, because other than the commonly understood policing role, it more broadly designates a person with corporate responsibility. If this Code gives them none, it should be avoided. Crony Pony 2b (talk) 10:58, 18 August 2021 (UTC)[reply]

What does this term mean? What are this group's powers? It's defined at the top but it seems it's never used in the draft...? (It has one usage, and that's just to say they need to 'affirm' their respect for the document). Fuller comment here ProcrastinatingReader (talk) 16:51, 20 August 2021 (UTC)[reply]

  • Code enforcement officer definition: training and technical rights. Who is defining the training? What if the communities consider the training unsuitable or insufficient. Who gets to vet the WMF staff member’s training? Nosebagbear (talk) 22:32, 21 August 2021 (UTC)[reply]

How many people have reviewed this?

For the overview to begin with a sentence that includes

"Code Enforcement is the ... enforcement of violations of the Universal Code of Conduct."

says to me I need read no further. Not a typo, edit summary:

(Text written by the Universal Code of Conduct Phase 2 Drafting Committee)

It is not 'unkind' of any of us to require precision in a statement regarding enforcement. It seems obvious you didn't approach this with the correct frame of mind or care. Shenme (talk) 03:19, 19 August 2021 (UTC)[reply]

Would you prefer "is defined as"? The sentence quoted is reasonable in English as a standin for "is defined as" but perhaps you have issues with that phrase instead. Since you didn't actually clarify what you think it should say. Izno (talk) 18:00, 21 August 2021 (UTC)[reply]
The sentence defines Code Enforcement, that's right. And it says, I'll be forced to violate the code ("enforcement of violations") --Der-Wir-Ing ("DWI") talk 20:05, 21 August 2021 (UTC)[reply]

U4C Committee Definition

U4C – okay, the most critical first:
  1. I believe it is extremely unwise that the U4C can decide categories permitted for its own cases. It may involve little growth over time, but it could also include significant scope creep. That makes it impossible for users to do a fair assessment of whether to ratify the UCOC. This will mean a few more categories need to be added.
  2. I also feel that the U4C should take any instance of WMF employee action on-wiki or connection that would otherwise fall under the UCOC. That is, were they not an employee their actions would be handled by another function here, but they should automatically be raised here. Yes this may cause simultaneous on-wiki and off-wiki action being taken. Also an interesting consideration of vice-versa!
  3. Other than that very large BUT in point a, I actually quite like the U4C – I think both the concept and the initial scope categories are good!
  4. The U4C also needs to commit to not devolving any sanctioning authority to alternate authorities, and confirm that any individual case/appeal is to have a minimum quorum of 7 drawn directly from the U4C. Any systemic abuse case must be held by a quorum of 15. Determining rules for U4C size/size change worth doing? Nosebagbear (talk) 22:32, 21 August 2021 (UTC)[reply]
    I agree strongly with no devolving powers. This restriction should be spelled out explicitly in the document. (See enwiki ArbCom, for example, which has devolved powers to unilateral admin action through DS.) This Committee should not be allowed to single-handedly set new enforcement rules (eg giving more power to 'Code Enforcement Officers') through proclamations and delegation. ProcrastinatingReader (talk) 22:50, 21 August 2021 (UTC)[reply]

'Preventive work (articles 1 and 2 UCOC)'

adhesion

"promote voluntary adhesion to the code", "Recommendations of UCoC Translation for voluntary adhesion:", Did you mean adherence? Vexations (talk) 14:05, 14 August 2021 (UTC)[reply]

Unless the WMF has budgeted a few thousand gallons of glue for UCOC enforcement... Vermont (talk) 23:37, 16 August 2021 (UTC)[reply]
We shall see how long it takes to fix two words with the iterative process, continuous evaluation and evolution with community feedback. Vexations (talk) 12:46, 17 August 2021 (UTC)[reply]
I fixed it. Best, Vermont (talk) 14:01, 17 August 2021 (UTC)[reply]
Also, to clarify, the evaluation/editing of the document is not continuous. Vermont (talk) 00:45, 18 August 2021 (UTC)[reply]

Training requirements

Introducing a complicated training bureaucracy, as is proposed here, is likely to be a barrier to successful implementation of the UCoC. Most of the burden of creating the proposed training material, as well as the problems introduced by not creating the material, would fall on smaller, non-English-language communities. --AntiCompositeNumber (talk) 02:28, 18 August 2021 (UTC)[reply]

"legally binding"

Unless we're suing people over the UCOC, drop the words "legally binding". --AntiCompositeNumber (talk) 02:29, 18 August 2021 (UTC)[reply]

Recommendations of UCoC Training/Education amongst Community

First, I agree with AntiCompositeNumber that this section is a little concerning. Not because any of it is necessarily a bad idea, but because you're asking sometimes very small projects or organizations to do an awful lot of work just in setting up such a system, nevermind following through with execution.

Second, the "A link to the UCoC should be present on:" part takes for granted something that doesn't jibe with my (perhaps long outdated) understanding. Specifically, that the UCoC isn't supposed to supersede all other conduct-related policies and procedures, but is instead a starting point with which various groups can build their own policies. From the beginning conversations about the UCoC (I'm thinking back to the Wikimania session in Sweden, for example), it was articulated as something intended primarily for those many projects which did not have well-developed policies for dealing with harassment and other conduct issues. The idea that it should be the UCoC itself rather than the local policies the UCoC is adapted into runs contrary to this assumption. So where do things stand? — Rhododendrites talk \\ 20:16, 20 August 2021 (UTC)[reply]

  • In some ways the relative dearth of detail is absolutely right – it’s rightly avoiding too much detail in an area where that would just be difficult. Could I clarify is that local communities/affiliates can each create their own (perhaps with a “model code” for those that don’t?) The phrasing appears to intentionally have a fairly flexible inclusion, which is good, but is the U4C responsible for assessing violations? If so, given my interpretation above, I’d add that as a category. Nosebagbear (talk) 22:36, 21 August 2021 (UTC)[reply]

Recommendations of UCoC Consent amongst Community and Foundation Staff: Definition and Respect

  • Contributes to online and offline Wikimedia projects and spaces
  1. I am aware of different individuals who have come to define this differently. IRC screams to attention, and members of the UCOC drafting committee know others. I am aware this gets some detail in the “transparency of process”, but that is not immediately relevant. This paragraph is therefore one that could use its own full overview subsection – preferably soon enough to ease future discussion.
  2. I also disagree that there is any right to demand more than “will comply with/adhere” the UCOC. We have it as a specific acceptance that much as no editor is forced to edit, no admin is forced “to admin”, except so far as to conclude an action (admin accountability etc).
  3. However, while we’d certainly hope that the bulk of editors will respect the documentation, I don’t believe a) that’s required b) is a legitimate demand – editors could disrespect the UCOC and still acceptably uphold it. The subtitle (if holding authority) seems to contradict its content, but then on re-reading I think it might agree. So I’m now confused – could it be clarified how it’s a recommendation of consent – it would appear to be demand? Or does it want a specific consent from Community and Foundation Staff which would seem needlessly controversial
  4. Other than those facets, it seemed to cover all the significant groups that came to mind Nosebagbear (talk) 22:36, 21 August 2021 (UTC)[reply]

Translations for voluntary adherence

This is normal phrasing for treaties, and this could be one of the more multi-lingual documents ever. However. I believe that it has some flaws. A translation has been provided to a number of communities, who will vote to ratify that copy, or not. If a difference in meaning big enough for us to actually care about the wording is found; that ratification would surely be in serious jeopardy, pending a new one? Nosebagbear (talk) 22:36, 21 August 2021 (UTC)[reply]

'Responsive work (article 3 UCOC)'

'Principles for processing and filing of reported cases'

  • In regards to Cases should be forwarded or escalated where appropriate and Allow reports to be forwarded to relevant bodies, please remember that there is some element of trust in writing a private report. Please codify that consent of the person is required to forward the case to another "relevant body" (an Arbitration Committee, a public venue, or whatever else) outside of the U4C Committee. Of course, lack of consent may mean the report cannot be dealt with, but that should be up to the reporter and/or affected users. ProcrastinatingReader (talk) 12:08, 19 August 2021 (UTC)[reply]
  • It seems like there should be some material here talking about the relationship between this process and other processes/venues that may exist within the relevant project or organization. This could just be a matter of expanding "Cases should be forwarded or escalated where appropriate". — Rhododendrites talk \\ 20:37, 20 August 2021 (UTC)[reply]
  • To me, U4C is just the ultimate location that UCOC might end up at. We aren't deriving our conduct-handling authority from one small body at the top, it's the other way around. In terms of formal wording Cases should be forwarded or escalated where appropriate is fine Nosebagbear (talk) 22:46, 21 August 2021 (UTC)[reply]

'Providing resources for processing cases'

  • Here, as in some other places, it would be good to have some additional context. i.e. what "an ArbCom" is and why it is desirable. — Rhododendrites talk \\ 20:45, 20 August 2021 (UTC)[reply]
  • If "ArbComs" are the "resources for processing cases", what is their relationship to the [tentatively named] Code of Conduct Enforcement Officers? — Rhododendrites talk \\ 21:03, 20 August 2021 (UTC)[reply]
  • Shared ArbCom across projects of same language - This sounds fantastic if it can work. Projects would either need a major amount of policy unification (in my en-wiki bubble I’ve no idea if that’s practical), or arbs with major cross-wiki experience. Arbs are chosen to be among the best, which requires practical knowledge of policy, not just reading it case by case. Nosebagbear (talk)
  • Everything else here either sounds good or, again, if this works, fantastic – I’m not remotely qualified to assess viability, so let the U4C and all try. Nosebagbear (talk) 22:46, 21 August 2021 (UTC)[reply]

'Types of violations and enforcement mechanism / groups'

  • Threats of physical violence should have more than just sending them to trust and safety. Yes, of course we should involve T&S, but there are other things that may be important to do in the short term: gathering information, making short-term blocks or protections, communicating with the person being threatened so they're not just in limbo until T&S can get back to them, etc. — Rhododendrites talk \\ 21:09, 20 August 2021 (UTC)[reply]
  • "Cases should be promptly sent to the Wikimedia Foundation Legal team, or, when appropriate, other professionals who can appropriately evaluate the merit of the threats" - what "other professionals" does this mean? If we formally include WMF Legal here, that implies sufficient funding for the legal team to be able to reliably and promptly respond to these inquiries -- is the WMF on board with that? A legal threat would be incredibly stressful for a volunteer, so I'd like to see this part of the flow chart better developed. — Rhododendrites talk \\ 21:09, 20 August 2021 (UTC)[reply]
  • "Off-wiki violations (examples such as, but not limited to: in person edit-a-thons or off-wiki instances such as on other platforms similar to: social media platforms, discussion lists)" - this conflates two very different phenomena. An edit-a-thon is typically (though not always) an event formally connected to a project or affiliate and thus should be governed by some form of the UCoC (or similar policy/system). Social media platforms, message boards, etc. aren't under the purview of the same rules, necessarily. Hence why people aren't typically sanctioned for what they say on Twitter (with exceptions). — Rhododendrites talk \\ 21:12, 20 August 2021 (UTC)[reply]
  • Interim acts – mentioned by Rhod. Now, in the event of a threat of harm I squash the violator while communicating the needful. Those aren’t really suited to further action. But a case of self-harm I dealt with was vastly more complicated. Without going into details, that was an instance of multiple individuals acting at the initial outset, while pouring the information into WMF Emergency, before they very rapidly could take everything over. Nosebagbear (talk)
  • Violations involving litigation or legal threats I have to ask why we’re sprinting off to Legal here. The large majority of legal threats are of the vein “block me and I’ll sue”, and the average response time to one being posted on ANI is minutes…and a block. Sure there’s lots of instances where a report (again, probably a block and report) is the right thing, but I can’t imagine a) we should start doing this for all and b) that Legal is going to get anywhere close at a 24hr response time for all languages. Nosebagbear (talk)
  • Handled by AffCom – this is obviously the usual course of affairs, but makes me think that the U4C permissible cases should specifically note systemic failure by AffCom. AffCom themselves need to clarify how they’re handling the very different cases of WMDE, regular chapters, all the way down to the smallest user groups. Nosebagbear (talk)
  • Systemic Failure – point “i” is good. Regarding point ii: Is this meaning that desysopping procedures for x-wiki abuse are falling into U4C jurisdiction? I’m not positive of what “administrative level” means here. If so, that could only be Stewards, global sysops and rare cases where someone with two local rights causes issues on both? Suspect I’m misunderstanding, cheers Nosebagbear (talk)
  • off-wiki instances such as on other platforms similar to: social media platforms, discussion lists needs clarity. Is this “any such case”, cases where someone has followed someone from off-wiki, a case where it might even have begun off-wiki, just cases where it’s local affiliates on those. These are extremely different things to have lumped together. Nosebagbear (talk) 22:46, 21 August 2021 (UTC)[reply]


On-wiki UCOC violations
  1. Single-wiki is mostly fine, except for one aspect I cover during ratification section below Nosebagbear (talk)
  2. But cross-wiki here is a concern. It formalises gaming issues – let’s say my problematic behaviour is all on one wiki, but I can make a few problematic edits to drop myself into U4C direct jurisdiction if I think is better. Alternatively, someone might get yanked even when this is inappropriate. In the event that there is clear misconduct on one wiki but questionable conduct on another, will the case half proceed, and then the U4C realise they have lost any moral right to proceed with the case because one wiki’s charges turned out to be non-issues and remand it to the single-wiki? Nosebagbear (talk) 22:46, 21 August 2021 (UTC)[reply]
Escalation and de-escalation mechanisms

(Following Wikimania discussion) It would be helpful to have well-defined processes for escalation and delegation. Some suggestions:

  • Reports that can be solved locally will be referred back to a local project community, either by pointing the complainant to relevant processes in the project, or by U4C members supporting editors to engage with local dispute resolution.
  • The U4C will investigate and mediate disputes where local dispute resolution has failed and escalation is necessary, or in cases of cross-wiki disputes that require external mediation.

--Deryck C. 12:19, 16 August 2021 (UTC)[reply]

'Recommendations for the reporting and processing tool'

Reporting tool

What stops you mentioning a email addresss (ending up in an OTRS queue) as tool for reporting complaints, incidents of harassments? 212.182.155.232 11:15, 16 August 2021 (UTC)[reply]

mediawiki extension

The requirement that the UCoC tool be implemented "as a MediaWiki extension" isn't very useful. Aside from requiring a "not invented here" approach, it doesn't convey whatever intentions the drafters may have had. If the intention was that a separate user account shouldn't be required, state that. If the intention was that the tool must be WMF operated and freely-licensed, state that. Previous requirements that something be implemented as a MediaWiki extensions have ended up with things like SecurePoll. --AntiCompositeNumber (talk) 02:43, 18 August 2021 (UTC)[reply]

I think it's smart to use tried-and-tested external software, even if it's commercial. The WMF is not the first organisation to need case management. The desire to use made-in-house software, and/or FOSS software only, constantly leads to us using bad products that waste volunteer time and productivity. Use the stuff that's been used at various orgs, refined to work, and is actively maintained, rather than creating yet another tool that will join the graveyard of unmaintained tools (mw:SecurePoll, mw:FlaggedRevs, etc). And on that note, WMF Legal literally uses Zendesk (proprietary commercial software), and Design uses Figma/Sketch AFAIK; why are volunteers confined to mailing lists, private wikis, and inadequate unmaintained extensions? ProcrastinatingReader (talk) 11:57, 19 August 2021 (UTC)[reply]
@AntiCompositeNumber and ProcrastinatingReader: I was the one who suggested this provision got added to the draft. My intention behind it is to ensure other MediaWiki-based (non-WMF) communities have easy access to this tool. With what you both have said here, I understand that steps will need to be taken to ensure its long-term viability if it is going to be a MediaWiki extension (assuming the committee doesn't go with an external commercial solution instead based on this feedback). –MJL ‐Talk‐☖ 21:41, 19 August 2021 (UTC)[reply]

Tool Functionality nature

  • I want to know how we are guaranteeing we won’t see any of the issues we have seen with other great tools we’ve heard along these lines. That is being implemented before they’re ready, failing to handle technical debt, insufficiently quick response times, non-stressed statement of technical admin access to extremely private data, etc etc Nosebagbear (talk)
  • I certainly back there being a tool, so long as it doesn’t become compulsory. While I am aware that the WMF would probably not like to make it readily possible for individuals (or themselves) to make it link into pre-existing processes where that is viable, I suspect that while it would slow initial take-up it would increase it in the long-run. Nosebagbear (talk) 22:48, 21 August 2021 (UTC)[reply]

Tool Features

  • This process doesn’t state who is responsible for allocating to right body (where it’s not so clear the system can’t just do that itself) – how will this work? Nosebagbear (talk)
  • The general information that are set out are generally good for a non-exclusive list. I do think that specifically mentioning some wording of “diffs” would be beneficial. Making it so the system goes “click to insert URL of problematic edit 1” so new editors can provide evidence in a way that regulars are used to, then fantastic Nosebagbear (talk)
  • The process doesn’t seem to handle boomerangs inherently – these are a common need on en-wiki, and they can’t just be excluded Nosebagbear (talk)
  • Good principles, more on how that’s going to be ensured, not requested/urged please [happy to get it in a binding appendix!] Nosebagbear (talk) 22:48, 21 August 2021 (UTC)[reply]

Privacy and Anonymity

  • So...this section is looking pretty lightweight given the scale of discussion that I’ve both seen and participated on it over the last 30 months or so and it’s not a specific open question. Now, if this is because how to handle privacy options is going to be decided by local communities (and the tool will just build all the options, with some being enabled – like userrights) then I’m completely game for that and you can skip the below. Because it’s not covered I have to mention a couple of posed options. Nosebagbear (talk)
  • Privacy in cases is sometimes a negative for the accuser – both the example given, but others as well. But it is also a massive negative for both the accused and the whole community. And I’m a little surprised that neither factor gets mentioned. Nosebagbear (talk)
  1. Community-hidden, accused-gets data. This is the form that ArbComs are often formed to handle. On en-wiki at the moment, that’s usually when handling off-wiki data, but can be others and certainly could have an expanded scope. Negatives include things like not being able to audit those we have selected to hold these positions and not knowing “where the axe will fall”. Nosebagbear (talk)
  2. Some/all data hidden from accused, all data hidden from community. Here the accused basically gets the bum-rap. Many of us know instances where context has been key to explain why you (or someone else) did something; or where the other did something in a forum the reviewers may not know about. This is especially the case for anything with off-wiki evidence. And how are they ever to make an appeal? How can one show remorse for an act they don’t know – even if they get the broadest strokes how could that be utilised? CUBlocks are already difficult, but for a harassment charge, any appeal could be the appellant: lying, believing they did nothing wrong (incorrectly), believing they did nothing wrong (rightly), accepting they did something wrong. Nosebagbear (talk)
  3. Logically, the large majority of raised cases are desiring that person to be directly sanctioned – rather than, say, mediation. As such, opting for the most-private option allowed makes sense Nosebagbear (talk)
  4. How to handle scale – this depends on exact clarity given by UCOCDC, but the type allowed in 2a usually necessitates very broad trust in the assessors – for example I respect the current en-wiki arbcom, firmly. But I wouldn’t trust, say, 1-3 admins or CEOs selected to hear cases privately because the scale grew. Nosebagbear (talk) 22:54, 21 August 2021 (UTC)[reply]

Processing & Transparent Documentation

  1. How are we supposed to ensure that this early-stage processing, which could be important, is done correctly? Nosebagbear (talk)
  2. Yes to filtering out bad-faith reports, but doesn’t raise how to handle boomerang cases. Without that it encourages a wave of first-mover attempts. Nosebagbear (talk)
  3. Interesting concept, but any great vagueness at all and it would struggle to give out information about UCOC enforcement. When the system will process many thousands of cases, each growing % of vagueness represents many more chances for unseen error. Nosebagbear (talk) 22:54, 21 August 2021 (UTC)[reply]

'Recommendations for local enforcement structures'

Training and Support

Fairness

  • Obviously equivalents to en-wiki’s INVOLVED are good. But this is why any private case needs a full arbcom – it’s the only way to have both reliable enough people and enough ‘’opinions and eyeballs’’ to ensure it’s actually executed. Just having one amazingly fair person does not ensure a fair outcome. The issue here is not bad-faith errors Nosebagbear (talk) 22:56, 21 August 2021 (UTC)[reply]

Clear communication between local administrators

Can I check if this just means “every project with admins should have a page for them to discuss” or if there will be private spaces cut off from local editors being able to read, or comment and at least ensure those comments will be read (even if not posted into the same literal page)? Nosebagbear (talk) 22:56, 21 August 2021 (UTC)[reply]

Transparency of process

  • This is really interesting – would love to see how different communities view different things of different severity. But it should also contain a “why” – perhaps some communities find mediation fails more, or IBANs especially effective. Nosebagbear (talk) 22:56, 21 August 2021 (UTC)[reply]
Well, in German wikipedia almost everything is handled publicly. There are very few exeptions. Not in the policies, there were very few cases that were handled non-publicly. --Der-Wir-Ing ("DWI") talk 23:44, 21 August 2021 (UTC)[reply]

'Recommendations for how to process appeals'

Open questions for the Community

Escalation: Where do the complaints go, what instance/body/judge is supposed to process them.

  • Am I right in thinking this is “error of process”? But in non-appeal aspects? Sorry - I’m just a little lost here so am going to skip for now. The problem is that many immediate appeals are based on these grounds, so there’s major crossover with "appeals". Welcome any detail. Nosebagbear (talk) 23:07, 21 August 2021 (UTC)[reply]
  • I’ve seen this raised before, but can we get a clearcut distinct phrasing for “appeal because I think the block was unwarranted” and “appeal because I’ve learnt my lesson”? Is that what escalation/appeal is meaning? The processes for handling the two are very different indeed, so sharing of terms is not clear. I wonder if there's a word from another language we can borrow? Nosebagbear (talk) 23:07, 21 August 2021 (UTC)[reply]


Job descriptions

(Also following Wikimania UCoC roundtable)

There seems to be at least four job descriptions emerging for the UCoC. It will be helpful to have separate roles and descriptions for these, not just one "U4C membership". The different groups can have overlapping membership but the role differences need to be clear:

  • "Supreme Court Judges" who mediate or arbitrate cross-wiki disputes or disputes escalated from local projects that cannot resolve the dispute on its own;
  • "Clerks" who triage incoming reports to decide whether they should be resolved at UCoC level or delegated to local projects
  • "Mentors" or "Ambassadors" who help local projects improve their internal governance
  • "Cultural interpreters" who provide cross-cultural advice to the Judges and the Mentors, so as to understand local (both real-world and project) culture and avoid cultural pitfalls during dispute resolution.

Deryck C. 12:24, 16 August 2021 (UTC)[reply]

Regulations for appeal (after the previous question "Where do the complaints go" has been answered).

  • I think this is going to have to be something to leave to communities to determine, so long it meets the “yes there’s an appeal, heard by a third party”. We can’t just have a “moves one step up”. On en-wiki, if any basic vandalism single-admin block could force an appeal to the Community conduct boards, we’d be overwhelmed. So, an appeal is heard by a different admin.
  • In a similar vein, something heard by the Community can only really be heard by the Community for it to have any weight at appeal. But there are cases where conduct can “step up” appeal level: e.g. if the appeal-review admin disagrees that it’s clear-cut, it goes to Community. Or something like discretionary sanctions [which is its own fun ball of connection with UCOC].

In short: Ultra basic cases (e.g. AIV) -> decided by admin -> appeal by different admin -> in event of disagreement, heard by community “Regular” complex cases -> heard by Community -> appeals held later by Community Very complex cases and cases with rare acceptable grounds for being heard away from community -> (potential prior history at other conduct routes) -> ARBCOM -> appeal route as provided for (I can well imagine some ARBCOMs devolving to U4C)


Should the U4C committee also decide individual cases or process appeals?

  • So I believe the U4C should decide (or have the right to decide) individual cases “original jurisdiction” in the following cases:
  1. where projects don’t have sufficient conduct capability to hear them;
  2. cases where projects have devolved such to the U4C
  3. all cases involving the conduct of a WMF Staffer
  4. x-wiki cases that are sufficiently mixed to not allow ready prioritisation to a single wiki and too complex for standard resolution Nosebagbear (talk)
  • I believe the U4C should have the right to hear appeals in the following cases:
  1. All cases as above, except as specifically designated otherwise with regards to the U4C hearing its own appeals
  2. In the event of a systemically stable (in conduct terms) project losing that status, they may review appeals to a date agreed as being the start of major issues arising.
  3. WMF-imposed sanctions, except as specifically explained why a particular instance could not legally be subject to U4C review. General categorisation of a case is insufficient, the full U4C must be made aware of why a particular case is not subject to review, and may provide a public statement accepting or denying that judgement (but not revealing details). Nosebagbear (talk) 23:13, 21 August 2021 (UTC)[reply]


When should someone be able to initiate an appeal for a UCoC violation?

  • If someone sees that there is a person who violates with its behaviour from the own view the Universal Code of Conduct. In this case everybody who sees it should be able to report this. To avoid duplicate appeals the reporting should be public as far as it is possible.--Hogü-456 (talk) 19:12, 20 August 2021 (UTC)[reply]

What kinds of behavior or evidence would we want to see before granting an appeal?

  • Correct sanction felt to be given – this is probably the 2nd strongest reason to be against totally private cases. How on earth is someone supposed to meet whatever these criteria are without knowing the case?
That aside, general criteria: understanding of what got them blocked, remorse for what they did, explanation of what they’d do differently if unblocked. We like to see helpful activity on other projects but that’s hardly required. In socks, they must assist on that.
  • Sanction disputed as being correct - this needs to direct against either procedural or substantive grounds. The former is most seen in highly complex areas like Discretionary Sanctions. The latter needs to demonstrate either that there was missing information or that the judgement could not reasonably have been found from that. On that latter aspect, a lower bar is applied for a single admin blocking and a high one for Community/ArbCom.


Who should handle the appeals process?

How often should someone be allowed to appeal a UCoC violation decision?

To what extent should individual Wikimedia projects be allowed to decide how they enforce the UCoC?

  • Firstly, we were repeatedly assured that the UCOC would be the minimum. The inherent corollary of that is that Communities must retain the maximum local authority, interpretation, and implementation method control as possible. Either the UCOC shouldn’t be possible to interpret differently (don’t think we can manage that), or it’s an unstated facet and therefore the principle of subsidiarity and minimalism sends it right back to the local communities.
  • However, a sense of practicality does apply in two facets here: a) communities just stating that the UCOC doesn’t clarify something when it clearly does b) communities not having sufficient size and scope to be able to viably assess such things.
  • For a) could be appealed to the U4C who would use a “clear and convincing” breach. Mediation would absolutely be an option. In the event of an established breach not being resolvable by this method, it could be overturned. In the event of established communities a single breach wouldn’t be grounds for demonstrating “systemic abuse”, but multiple breaches within the medium-term would be a major cause for concern.

I should note that in the transition year, we should expect a very high rate of good-faith disagreements here, which should be bourne in mind.

  • For b) Here I think is a good time to mention something that’s been mentioned throughout this whole tome: model documents. Who they’re made by is up for discussion, but having very core, slimline, documents and interpretations that apply unless a Community conduct page is already there seems really beneficial. Otherwise it puts a huge amount of work onto 10 editors re-inventing the wheel. Nosebagbear (talk) 23:13, 21 August 2021 (UTC)[reply]

How will people be chosen for the U4C committee?

Our current recommended list of users include, but are not limited to: CheckUsers, oversighters, bureaucrats, administrators of local projects, arbitration committee members, Wikimedia Foundation employees, Affiliates, etc.
  • The Wikimedia Foundation elections committee is mostly filled with Stewards and other esteemed editors. But you needn't look further than here or here to see what a fiasco that turned out to be. The same issue is reflected in the Ombuds Commission. These are not accountable bodies, and (respectfully) other discussions leave it unclear whether they're even competent bodies (sorry, I can't figure out a lighter way to put it). Compare all that to the English Wikipedia's Elections Committee, composed of editors who run for, and are elected specifically for, that position (here), or to the ArbComs themselves.
    Members of the U4C Committee should not be 'chosen' at all, as history shows that 'WMF appointment of esteemed editors to Committees' doesn't work. Candidates should be elected into the role by their peers, the editors who they will be helping and/or judging, in annual elections. In that election, they can set out their platform and why they would be effective at the job, and the community will have the option of picking those they trust for this job (conducting CU checks, determining consensus to appoint admins, preventing LTA abuse, etc, is not at all the same as deciding on conduct problems with established editors). Since this body will have even more power than stewards, they should be subject to reconfirmations (like stewards) to ensure there is still community confidence in their continued tenure. Any other method of appointment is not legitimate IMO. ProcrastinatingReader (talk) 18:28, 17 August 2021 (UTC)[reply]
    I strongly agree. While the Foundation technically doesn't need the buy-in of local communities in order to implement this new Code, things will go a lot more smoothly if local communities are given the ability to ensure that these committees don't stray from their intended purpose. If, on the other hand, these appointments are made by the Foundation, I strongly believe that the local communities will come to resent the committees and their members as an outside governing force that is not answerable to them. Lepricavark (talk) 04:30, 18 August 2021 (UTC)[reply]
    @ProcrastinatingReader: In your opinion, who should organize said election? –MJL ‐Talk‐☖ 01:25, 21 August 2021 (UTC)[reply]
    @MJL: Interesting question, though I don't think I have a qualified answer to give. I'd probably start by looking at history: what has worked and what hasn't. As I understand, enwiki's ArbCom electoral processes system seems to work well (an RfC en:WP:ACERFC2020, followed by empanelling a commission en:WP:ELECTCOM2020). However, that's developed over two decades and probably requires a critical mass of interested editors to keep functioning, and I dunno if the U4CC will have that level of interest. In comparison, the WMF Board Elections Commission seems like a farce to me, so that clearly doesn't work. I suppose other projects probably have functioning election organisation systems, too. And, of course, there are the SE which seem to be organised by stewards and those seem to work out well too, although I don't think self-oversight is the way to go for this body. Steward oversight could also be a good idea (as they're competent, trusted, and experienced with global communities), but on closer examination this may not work really (can expand on why I think that), and of course depends on whether they'd be interested in the first place.
    The preceding paragraph assumes one big global election should be held on meta. Perhaps it shouldn't, and a series of elections should be held locally or in project groups (the current draft guidelines stipulate communities grouping together by language, for example, in shared ArbComs). An alternative approach might be making the Committee composed of current ArbCom members, which fits nicely in line with the ideas in "Providing resources for processing cases" (expansion of ArbComs), but may not necessarily be representative of the editorbase, and of course some ArbCom members may not be interested. This method would be a more indirect election, and would save time/energy spent on another set of annual elections. Lots of different ways to look at this question, I suppose. ProcrastinatingReader (talk) 01:31, 21 August 2021 (UTC) e: 01:52, 21 August 2021 (UTC)[reply]
  • I agree, and I hope that the U4C committee has regular elections by the communities, and not appointments. Jackattack1597 (talk) 00:07, 21 August 2021 (UTC)[reply]
  • I would firmly note that ElectCom has demonstrate that its process does not work, whereas en-wiki's electcom does work, though may not necessarily here. I have serious issues with the way Steward reconfirmations are run - unlike the above I'm not saying they can't run good elections, but I'm saying that using their current mechanism as evidence for it is not necessarily warranted. The WMF also definitely can't have any input into the process. They have demonstrated on several occasions an inability to do crucial mathematics, which makes me reticent to trust them in something they have such a vested interest in. Nosebagbear (talk) 23:07, 21 August 2021 (UTC)[reply]

Removal from office

  • If a completely new body with staggering powers is going to be formed, I assume it's reasonable that while they can (and should!) have an internal impeachment process as well it clearly would be seriously odd to trust in it. There's a question about having different needs: individual issues and group issues. For the moment, I'm going to say it's the latter that needs a focus. Perhaps RfCs by at least 10 projects representing 10% of the active editors of the project should in effect force an immediate recall election, and, if appropriate, move the timing of the next UCOC review period sooner. Nosebagbear (talk) 23:07, 21 August 2021 (UTC)[reply]

Should an interim committee be formed while the "U4C" committee is being created?

Should global conduct committees, such as the Technical Code of Conduct committee, be merged into the proposed U4C?

  • Could we get a list of these, to aid discussion? One specific exemption is that of the Ombuds Committee. Though that needs a rewrite. It’s needed for U4C overview (which may require them to sign the same nondisclosure agreements to be able to get absolute access for review) and would provide a route for tool users to get a route of appeal to specialists.

General comments

Appeals

I think the rules must clearly establish "Yes, the accused is allowed to file an appeal based on 'I did not perform the action in question and here is evidence that I did not" OR "No, the accused is not allowed to file an appeal based on the idea that they did not perform the action in question; doing so is a sanctionable offense" with a possible "No, the accused is not allowed to file an appeal based on the idea that they did perform the action in question but said action should not be considered bad; doing so is a sanctionable offense." The second scenario is very counterintuitive and it may not be obvious that it's not allowed. It is best to tell people ahead of time that they are not allowed to do this (or that they are). The third scenario is somewhat counterintuitive and many people have filed appeals like that and been surprised when the result was negative. This will save everyone time and aggravation. Darkfrog24 (talk) 14:46, 17 August 2021 (UTC)[reply]

"...many people have filed appeals like that..." On what project, against what decisions? --Der-Wir-Ing ("DWI") talk 16:15, 17 August 2021 (UTC)[reply]

Timelines

I see more than one reference to "We must handle complaints in a timely manner." This is good. However, I feel it would be best to add some acknowledgement that we have to give the accused person time to read and compose a response to the accusations, ideally before they're presented to the people who are going to make the decision. Perhaps a day and a half for every 500 words of accusing text. So if someone files a 3000 word complaint, the accused would be given a copy and nine days to write a reply, and then the complaint and reply would be presented to the community or committee at the same time. At the very least, the accused should have a chance to read and respond to the complaint before the decision is made or sanctions imposed. Otherwise, we leave open a loophole for abusive complaints. Darkfrog24 (talk) 14:55, 17 August 2021 (UTC)[reply]

Reason for sanctions

Until 2011, the English Wikipedia had a rule--I think it was for AE--saying "Every sanctioned person has the right to be told exactly what they were sanctioned for." I do not know why that rule was taken down; it strikes me as a very good idea, but there may have been a good reason for the change. Perhaps someone who knows why it was removed could comment on whether "Always tell the sanctioned person exactly what they did" backfired on the English Wikipedia or created any problems there. Publicly stating what action merited what sanction doesn't only tell the sanctioned person what they need to work on; it proactively establishes to anyone watching that a given action counts as violation of the UCoC, and then they'll know in advance not to do it. Darkfrog24 (talk) 14:55, 17 August 2021 (UTC)[reply]

I would imagine it was for the reason that if you said to an editor, in your capacity as a volunteer, you have been banned from the Wikipedia projects for [insert criminal act of your choice here], you were leaving yourself open to being sued as an individual if that person vehemently disagreed, and considered it a matter of defamation. Likewise if said by an Officer of the WMF, leaving the WMF open to being sued. We now know, at least in the latter case, and perhaps in the former, there is no such risk, not since the WMF successfully argued before a Californian judge that you can be banned from Wikipedia for any reason, including no reason at all (so who would care what the reason given was?). Crony Pony 2b (talk) 11:05, 18 August 2021 (UTC)[reply]

Right to be heard

I also tried to argue on numerous occasions that we should include the principle of audi alteram partem in the UCoC enforcement guidelines, along the following lines:

Allow all individuals involved in a case tell their side of the matter before making an enforcement decision. In urgent cases, a preliminary decision may be made before that, but only if an appeal is possible.

I would like to understand why this was not included. --Gnom (talk) 17:33, 17 August 2021 (UTC)[reply]

Avoid unnecessary abbreviations

Please replace "ArbCom" with "arbitration committee or similar conflict resolution body", and "U4C Committee" with a more telling name. We need clear language for this to work, and to be inclusive. Thanks, --Gnom (talk) 17:39, 17 August 2021 (UTC)[reply]

I agree. I've looked all over and can't find what the other 2 C's are (presuming Code of Conduct are 2 of the 4). Abbreviations and acronyms are easier to remember when one knows (or once knew) what they stand for. — JohnFromPinckney (talk / edits) 23:11, 17 August 2021 (UTC)[reply]
"Parking Enforcement Officers are on patrol in this area! Have you Paid and Displayed?"
I think it's meant to be Universal Code of Conduct Coordinating Committee. But then the Committee is repeated. It's probably easier to just call it the "UCOC Committee" or something.
Additionally, I suppose local sysops/etc are all "Code Enforcement Officer"s. I suppose the term is meant to make you think of en:Law enforcement officers, but personally it reminds me of parking tickets (not that I'd know what those are). Both aren't great connotations IMO ('Wiki Police Force' has a better ring to it, though). ProcrastinatingReader (talk) 00:10, 18 August 2021 (UTC)[reply]

Criteria for recusal

I wonder if "anyone named in a dispute should recuse themselves from the case" would be better as "anyone with any conflict of interest in the case should recuse themselves."

Let's say two people are dating and they want to keep it private or business partners or cousins or friends. (It's a community and people meet each other.) One of them files a complaint, and the other one is on a committee. Maybe we should add an "all judges are expected to abstain in any case in which they have a conflict of interest" and "any judge may recuse themselves for any reason" so no one has to reveal private information to bow out. Darkfrog24 (talk) 17:58, 17 August 2021 (UTC)[reply]

That makes sense, rather than may recuse for any reason, it is reasonable to say that you don't have to explain who your conflict of interest is provided you have recused. WereSpielChequers (talk) 20:35, 17 August 2021 (UTC)[reply]
I think this will lead to a lot of argument about what a conflict of interest is. Jo-Jo Eumerus (talk, contributions) 09:00, 18 August 2021 (UTC)[reply]

Unable to handle

  • "...Where local structures are unable..."

This is rather broad latitude. Does this even need to be included? If they are "unable to handle", they should be able to request assistance. So this seems unnecessary. - Jc37 (talk) 20:14, 17 August 2021 (UTC)[reply]

Mm. I'm a bit confused here because Once formed, the permanent committee will decide on how often it should convene, and the nature of the cases that can be escalated to it. seems weird in combination with the "Types of violations and enforcement mechanism / groups" section. Does this mean the Committee can change its scope, or that of the cases allocated to it under article 3, it can choose which it will take and which it won't? Both interpretations seem in need of refinement, so perhaps there's a third. ProcrastinatingReader (talk) 21:08, 17 August 2021 (UTC)[reply]

enforcement of violations

" enforcement of violations " might be neater as enforcement of the UCOC or enforcement against violations. Maybe this is an English v American English thing, but do we really want to enforce the violations? WereSpielChequers (talk) 20:40, 17 August 2021 (UTC)[reply]

I also rise to speak against defining enforcement as "enforcement". The tautology is too tautological. Perhaps instead of "...and enforcement of violations...", we might say "...and sanctions (or 'punishment', maybe) for infractions of...". JohnFromPinckney (talk) 22:55, 17 August 2021 (UTC)[reply]
Also, what does "enforcement" mean? Blocking? Jo-Jo Eumerus (talk, contributions) 09:39, 19 August 2021 (UTC)[reply]
Presumably they would have power for bans, removal of user rights, etc., just like Trust and Safety's Office actions. But I agree there should be a definition on methods of "enforcement". ProcrastinatingReader (talk) 11:55, 19 August 2021 (UTC)[reply]

Addressing certain legal realities

I find this all very fascinating and it certainly does look like a lot of effort has gone into it. Can I just ask though, how much awareness is there among the people involved, of the basic legal reality, as confirmed by a recent Californian judgement where the WMF pressed this very argument, and succeeded: you can be banned from Wikimedia for any reason, including no reason at all. There doesn't need to be a reason, and any and all reasons for an enforcement given in the pursuance of this Code, can be nullified post-hoc, should the aggrieved party not be satisfied and wish to pursue their case externally. I suppose people might still want to pursue this Code in good faith regardless, but to me, it does rather still show that ultimately, it is all rather pointless, and if you get the sense your face just doesn't fit, or your complaint is valid but unpopular, you're probably wasting your time. Which was of course the problem with the previous way of doing things which has apparently led to this new Code. A cold hard legal reality that should be made clear to all parties (including Enforcement Officers) at the outset of any nominal proceeding. Although it has always been in the Terms, I am guessing nobody in the movement really thought it was true, until they saw the WMF Legal department hire very expensive lawyers to turn it into a precedent. Unless I am mistaken course, in which case, by all means, enlighten a sceptical soul. Crony Pony 2b (talk) 10:51, 18 August 2021 (UTC)[reply]

@Crony Pony 2b: ...confirmed by a recent Californian judgement .... Which one? I assume you can offer some links. --Der-Wir-Ing ("DWI") talk 16:48, 18 August 2021 (UTC)[reply]
I am curious about this as well. Darkfrog24 (talk) 01:52, 22 August 2021 (UTC)[reply]

On training

I see a lot of mentions of training, but no details on how trained persons are to be assessed? If there are no mechanisms to ensure the trained person has taken on board their training, and indeed no mechanisms to review whether their training is still effective down the line, perhaps this word should be downgraded to the more usual "guidance". As in, Johnny was provided with all the guidance he would have needed to effectively discharge this duty, so it's not really our fault if he didn't follow it, or never even understood it. Crony Pony 2b (talk) 11:13, 18 August 2021 (UTC)[reply]

Do we need a new 'enforcement process' at all?

It doesn't seem to me that there is anything in the UCOC that isn't covered by WP:Civility already, and any 'violations' would have otherwise been reportable and actionable via the existing structures in Wikipedia for handling and resolving complaints within the community. Why do we need to invent a whole new layer of processes and 'enforcers' over what we already have? JeffUK (talk) 12:46, 18 August 2021 (UTC)[reply]

@JeffUK: I understand your confusion as to why this process is needed, but I do have to point out that the Universal Code of Conduct (and its subsequent enforcement) is meant to be applied to communities outside of English Wikipedia. While WP:Civility exists, you'll notice that it doesn't exist on every WMF-hosted site: Wikipedia:Civility (Q4654593). Beyond that, not every community has a particularly great track record with enforcing their own conduct policy. –MJL ‐Talk‐☖ 22:07, 19 August 2021 (UTC)[reply]

Weird phrasing and ambiguities

  • legally binding
  • Article 1/Article 2/Article 3 UCOC
  • Code Enforcement Officer
  • individuals charged with enforcing the Universal Code of Conduct must be fully acquainted with the regulations they enforce

etc...

The whole document reads like the drafters were writing a constitution for a state or establishing a police force. And more generally speaking, the structure and prose overall is not very polished and not precise.

Various parts are also ambiguous, inconsistent, or redundant. eg:

  • Systematic failure to follow the UCoC
    • Handled by "U4C Committee";
    • Cross-wiki violations of the UCoC at the administrative level will be handled by "U4C Committee"

Surely the second sub-bullet (cross-wiki systematic failures at the 'administrative level' [unclear meaning of term?]) is already included in the broader first sub-bullet (systematic failures to follow the UCOC)

  • Off-wiki violations (examples such as, but not limited to: in person edit-a-thons or off-wiki instances such as on other platforms similar to: social media platforms, discussion lists)
    • Handled by "U4C Committee", if the case is referred to them by event organizers or local affiliate groups

This sounds like it would include Discord groups as well (as 'social media platforms'). I don't know who the 'event organiser' would be in that case, but is this trying to say that the incident can only be investigated if the admins of the Discord server refer the incident? Either way, it's confusing when combined with For Wikimedia-specific conversations occurring off-project in unofficial or semi-official spaces (e.g. Discord, Telegram, etc.), Wikimedia’s Terms of Use may not apply. ... Nevertheless, the behavior of Wikimedians on these networks and platforms can be accepted as additional evidence in reports of UCoC violations. ProcrastinatingReader (talk) 17:58, 18 August 2021 (UTC)[reply]

Support after violation of the UCOC

From my point of view it is important to understand reasons for the behaviour of people and try to help them to change their behaviour if that was not compliant with the UCOC in the past. This is something where I wish support. From my point of view it can help to talk to the people in a video conference system like Jitsi. It can help to know the person better.--Hogü-456 (talk) 19:29, 20 August 2021 (UTC)[reply]

In the German arbcom, I and my succesors made good experiences with this idea. --Der-Wir-Ing ("DWI") talk 23:39, 20 August 2021 (UTC)[reply]
I think this ties in with the idea of "everyone should be told what they did." Darkfrog24 (talk) 01:50, 22 August 2021 (UTC)[reply]

Ratification

As a “meta”-topic, and an absolutely vital one at that, I’ve split ratification into its own section.

Ratification splits into three main questions: who gets a say; what are we getting a say on; and methodology

I’ve created a couple of proposals in the past which I know have actually been looked at, but now we can see it, here’s my summary idea to get the discussion started.


1) “Who gets a say?” – voting groups (VGs) I’m going to use the term “voting groups” for which a sufficient majority of each is needed in order to ratify the UCOC:

  1. VG1 – local projects (e.g. en-wikipedia, fr-wikinews)
  2. VG2 – active editors across all projects
  3. VG3 – recognised affiliates
  4. VG4 – WMF Board of Trustees

2) What are we getting a say on? Talks with members of the drafting committee have indicated that by-section ratification comes with significant issues that may outweigh its positives. As such, I believe dividing just once, with a ratification vote on each, is the best route:

  1. Content of the UCOC “Phase 1”
  2. Enforcement of the UCOC “Phase 2”

3) Methodology This is a set of critical details and a bunch of small ones. Some are like “should every local project count, or just those with 5+ active editors?”, is a sufficient majority of a VG 50%+1, or more? What definition of an active editor are we using?

They’re important, but can be settled once the first two questions have been.

Reasoning

1) These are the core groups who get affected by the UCOC. All will go through significant change. But the logical question is why I’ve separated VG1 and VG2 – after all, projects would go like how their editors supported. That’s true, and for a single local project is enough. But en-wiki has a huge % of the total editor base, but is just one project. Projects with fewer than 100 editors make up a huge % of all local projects. If just editors counted, small and medium projects would be disenfranchised. If just local-projects, huge numbers of editors at a couple of large projects would be. Neither is acceptable
2) The reasoning for not doing by-article is included, but I’ll gladly discuss with anyone with an interest (pro or anti). However, that “Phase 1” is key. Phase 1 never received community-wide consensus of any type. As such, multiple editors have raised a valid concern that it has no mandate. Should its supporters think it is good as it is written then why should a separate ratification vote be an issue?
Do you support its inclusion? Yes? Then Defend It.
Phase 1 gave most communities less time than they should have, failed to answer questions raised, and continued to include strange points (for example, “introducing bias or incorrect information” has somehow become a UCOC violation every one of us is guilty of).
3) Methodology includes things that absolutely need to be settled before we get the ball rolling on any actual process or even process discussion, but would be good for a month’s time if we’ve settled down on ratification ideas and preferences.

Nosebagbear (talk) 23:15, 21 August 2021 (UTC)[reply]

Discussion

Retrieved from "https://meta.wikimedia.org/w/index.php?title=Talk:Universal_Code_of_Conduct/Enforcement_draft_guidelines_review&oldid=21918724"