Wikipedia talk:Contentious topics/2013 review

Discretionary sanctions is a remedy used by the Arbitration Committee in final decisions regarding problematic topic areas. This text is currently used as the remedy. This is a consultation on replacing the text with this new remedy.

Draft v1

Comments made during the consultation of the first draft have been archived to /Archive 1. Feel free to cite that page in comments on the second draft. AGK [•] 23:31, 16 November 2013 (UTC)

Earlier comments made during the consultation of the second draft have been archived to /Archive 2. Issues with the second draft that are still being discussed remain on this page. 13:55, 18 January 2014 (UTC)

Draft v2

Behavioural expectations (comments)

Editors editing within the area of conflict are expected to:

1. adhere to the purposes of Wikipedia; and
2. comply with all applicable policies; and
3. follow editorial and behavioural best practice; and
4. comply with any page restrictions in force within the area of conflict.

Failure to meet any behavioural expectation may result in sanctions.

i) Why all the "and"s?

ii) What is the point of this section because as far as I can tell what it states (other than point 4) is true for all articles. Are the authors of this draft really saying that an administrator is going to treat an uncivil foul mouthed editor with a large support base of acolytes any differently than they are treated when making comments on other pages? If not then the decision by an administrator whether to sanction someone involves exactly the same ANI political considerations as those for ever other page, so why include points one to three in this draft?

How does an casual editor with a little knowledge of the topic decides whether a specific page is under arbcom sanctions when many arbcom sanctions state "Articles which relate to {topic}, broadly interpreted"? For example is Warsaw a central European city or an East European City? How about Berlin or Vienna? Is the Berlin Philharmonic and the Vienna Boys' Choir articles under [Wikipedia:Requests for arbitration/Eastern Europe#Standard discretionary sanctions|Eastern Europe discretionary sanctions]] or are such topics about Central European issue and outside the scope of the Eastern Europe discretionary sanctions?

iii) I do not think that an uniformed editor can possibly be expected to know the scope of the articles under wording such as "Articles which relate to {topic}, broadly interpreted". Because of this point four is so loosely defined as to be uninformative for any arbcom sanctions that include the sentence "Articles which relate to {topic}, broadly interpreted".

iv) "Failure to meet any behavioural expectation may result in sanctions." Why only may result in sanction? Why not "will result"? Or is sending the signal that if you have a bigger enough support within the community base you will not.

-- PBS (talk) 19:37, 18 January 2014 (UTC)

If the above warning is going to be used then the Arbcom needs to include a link to the Arbcom case that implemented the sanction in that article. At minimum it should include a link to a summary of the sanction so the editor doesn't have to wade through a massive arbitration. Kumioko (talk) 14:28, 19 January 2014 (UTC)

What "above warning"? -- PBS (talk) 12:18, 20 January 2014 (UTC)

The comment in the Quotation box. Alert, comment or whatever you want to call it. Unless I am completely missing the intent of the message the intention is to notify the editors of a sanction on an article or topic. My point was that its a big place and we have a lot of articles with a lot of sanctions in place. We need to include a link to the sanction in question. Kumioko (talk) 15:03, 20 January 2014 (UTC)

Issuing alerts (comments)

Alerts/notifications: to log or not to log

This has been a recurrent issue, and there's been some discussion among arbitrators. So where are things up to?

* Alerts/notifications: these will likely be vanilla notifications than a topic is covered by discretionary sanctions and that misconduct within the topic can leave someone open to sanction. Someone cannot be sanctioned purely because they have been notified: they have to engage in misconduct AFTER they have been notified. While they will not be appealable, they do not carry any stigma.

• To log: The arguments in favour of logging alerts are:
  1. It makes life much easier for enforcing admins to have a central reference point;
  2. central logs will highlight abusive/trolling notifications;
  3. if there isn't a central log, people will create their own unofficial ones.

• Not to log: The arguments against logging are:
  1. it upsets people being on a central list as it carries a stigma;
  2. it makes someone a target;
  3. as it cannot be appealed it's a permanent blot on their copybook (a sort of "block lite" I suppose).

Let's have arguments for and against please.  Roger Davies ^talk 19:27, 7 January 2014 (UTC)

• Copy-pasting comments I've made elsewhere:

I strongly favor not logging notifications. Logging is more trouble than it is worth, and scrapping it would solve several recent incidents I'm aware of (and similar ones I probably don't know about) where the notified party took offense at having the notice "officially logged". I guarantee that each "side" in any dispute knows whether the people on the other side have been notified or not, and will joyfully link to the notification if needed. The change to the wording (no longer "warnings") is a tremendous improvement, but scrapping the logging of notifications would be even better.

I suggest thinking of it like an edit warring notice; AN3 sees a lot more action than AE, and the reporting party is relied on to provide a diff of a notice, and it generally works.

The problem is, as clear as we try to be, and as carefully as it is worded, notices are going to continue to be used occasionally as weapons by opposing parties, much as an edit warring notice is often used as a weapon by the other edit warrior. But there's less "sting" involved in an EW notice, and I believe that's largely to do with the lack of official logging. No one asks for an EW notice to be "vacated". Let's take the sting away from DS notices too; the notification will still be there. If someone appears to be using notifications as a harassment tool, we can address that the same way we might address someone using EW notices as a harassment tool.--Floquenbeam (talk) 19:58, 7 January 2014 (UTC)

• There must be a distinction between the editor who is merely going about their normal day-to-day activity and happens to edit a contentious article, and an editor who focuses on a topic that happens to be contentious. Logged warnings against the former are highly inflammatory, while they generally should apply to the latter. If there is no objective way to distinguish the two cases, there should be no logs because claiming "don't worry, we tar everyone who edits this article with the same brush" is at best misguided. Johnuniq (talk) 20:02, 7 January 2014 (UTC)
• My view is that logging is far better than the alternative, which only gives more weight to people using notices as cudgels in debate. The Committee has been working on removing the stigma from the notifications, but there's only so much we do as long as the community keeps on calling them warnings or lending them such a stigma; I suppose you can't just say 'you're wrong', but a notice that you've been notified of sanctions in an area ≠ some misapplied block on your log, and we (the collective we) should stop acting that is the same. Der Wohltemperierte Fuchs^(talk) 20:09, 7 January 2014 (UTC)
• I believe there should be no logging until an uninvolved admin determines that a user has been behaving badly. At that point a warning could be logged. It can then be followed with a sanction if the behavior recurs. Jehochman ^Talk 20:11, 7 January 2014 (UTC)
• I tend to agree with Jehochman. If we are no longer treating alerts like warnings then they should not be logged. But if as a consequence of misbehaviour after an alert an editor is warned/admonished that should be logged.
  The process question then is what implication does this have for enforcing admins. Well if the edit summary was always the same or similar a post to a user page would be quickly found in the last 999 edits. So we could make a template that has a standard edit summary--Cailil ^talk 20:15, 7 January 2014 (UTC)
• No logging. I agree with Floquenbeam; his comparison with the way edit warring notices function (=better, and without any logging) is compelling. Bishonen | talk 21:32, 7 January 2014 (UTC).
• On balance, I think that the hassle associated with logging alerts/warnings/etc outweighs the benefit of doing so. It is too easy to (mis)use disputes and appeals, etc., about these notifications as a sideshow that distracts from the core issue about whether there is actionable misconduct in a topic area. We should instead adopt Cailil's idea of recommending a standard edit summary (e.g. "Discretionary sanctions alert"), which can be searched for in talk page histories. We should also consider expanding the fourth criterium for receiving alerts, "through their actions, clearly demonstrated that they are already aware that the area of conflict is under discretionary sanctions", with something like "through their actions, or by participating in discussions in which discretionary sanctions are mentioned ...". This would enable administrators to ascertain relatively easily whether editors know about discretionary sanctions even without referring to a log.  Sandstein  21:41, 7 January 2014 (UTC)
• The best way to handle this is with a bot. When someone edits an article under DS for the first time, a message is automatically posted to the user's talk page. The bot can take care of logging, if desired. This also makes the process impersonal, which is better in this case. vzaak 21:59, 7 January 2014 (UTC)
• I agree with Johnuniq's point that there must be a distinction between the editor who is merely going about their normal day-to-day activity and happens to edit a contentious article, and an editor who knowingly focuses on a topic that is contentious. Some editors have tens of thousands of edits. Myself, I have nearly 20,000 edits[1] over the course of almost 5 years. I've participated in dozens (hundreds?) of dispute resolution discussions, some of which were at AE, others where discretionary sanctions came up. I don't remember them all. According to the way that this is worded, such editors can be sanctioned without warning for a single innocent mistake or even a temporary lapse of judgment (perfection is not required). It doesn't seem realistic to expect editors to recall every single discussion that they've ever had on Wikipedia. A Quest For Knowledge (talk) 22:41, 7 January 2014 (UTC)
  • I would submit that all articles under DS should have a prominent editnotice regarding DS. In my proposal of a bot that handles all talk-page notices, a user re-editing a DS article after some time interval would receive the notice again. vzaak 23:12, 7 January 2014 (UTC)
• As I wrote somewhere above, I don't see any useful purpose in logging alerts. Some people are still writing as if an alert is an admonition or somehow unfriendly. It should instead be thought of as useful, friendly information and be worded as pleasantly as possible. Receiving an alert should carry no stigma whatever. Zero^talk 23:25, 7 January 2014 (UTC)

I find it chilling that a last-notice-before-block notice, that bypasses normal talk page discussions and the official warning policy, would be referred to as "friendly". —Neotarf (talk) 07:03, 9 January 2014 (UTC)

That's because you don't understand what is being proposed. There is no "last-notice-before-block" aspect to it whatever, nor is anyone proposing to bypass discussion. I suggest you read the proposal again. Zero^talk 08:41, 9 January 2014 (UTC)

Answered below; the new proposal has no such language. —

• Stigmas are a cultural construct. ArbCom cannot simply announce upon high that such alerts/warnings/notifications/whatever-we-call-them "do not carry any stigma". Only culture can make such a determination. To think otherwise is the height of hubris. A Quest For Knowledge (talk) 01:25, 8 January 2014 (UTC)

Actually it is exactly the sort of thing that should be made clear. What it means is that nobody can correctly say "you were warned about your behavior" merely on the basis that you received the standard alert. All they can correctly say (in the absence of more information) is "you were notified of the presence of discretionary sanctions". Zero^talk 05:18, 8 January 2014 (UTC)

• While I'm somewhat agnostic on how specifically to do it, we need a good way of determining whether or not an editor can reasonably be said to have been made aware of discretionary sanctions in an area. Notifications are not an indication of wrongdoing, they're just a heads up to say "Hey, this is a particularly sensitive and touchy area, and so there are some special restrictions in place." I think that's far superior to having an editor whacked out of the blue with discretionary sanctions when they weren't even aware such a thing was applicable. That being said, I think there are several ways we can reasonably presume an editor is aware of the discretionary sanctions, including (but not necessarily limited to):
  • The editor has explicitly referred to the applicability of DS in a remark or edit summary, or has replied to another editor's remark in which that editor did so.
  • The editor has joined conversations regarding clarifications of the DS or alleged violations of them.
  • The editor has been explicitly notified of the DS.
  • The editor has made reasonably significant edits or reverts to a page containing a prominent editnotice specifying the applicability.
• What I would be categorically opposed to is allowing the "warriors" on either side be in charge of keeping track of who has been warned on the "other side". Logging is infinitely preferable to that. But regardless of our ultimate solution, notices aren't warnings. Indeed, sometimes an AE request is closed with an official, logged warning that an editor did violate the sanction, and that while no action will be taken at that time, future violations will lead to harsher remedies. Those warnings may be applied only by an uninvolved admin and are logged essentially as sanctions (and somewhat are a type of suspended sanction, since they are there to indicate that future violations will be treated more harshly as a repeat breach). They are very distinct from the simple "Hey, be careful, there are some special considerations when editing here" notices. Seraphimblade ^{Talk to me} 06:33, 8 January 2014 (UTC)
• What's with all this templating? If you want people to know that an article is under special conditions, why not just tell them? And let them know what the conditions are so they can act accordingly. Here is an example of a consensus discussion where editors were starting to cycle a policy page without getting prior consensus. A note was posted on the discussion itself--as a reminder to *everyone*--and the discussion was able to go forward.—Neotarf (talk) 16:44, 8 January 2014 (UTC)

• People can (and do) edit articles without going anywhere near the talk page. And many of those that do simply don't read all the clutter at the top of the article's talk page. A talk page notice is therefore considerably less effective at drawing someone's attention to something than a message left on the editor's talk page.  Roger Davies ^talk 18:12, 8 January 2014 (UTC)
  • If someone edits a page, the edit is either a problem or not a problem. If an edit is not a problem, why chase after some good-faith editor with a "last notice before block" template slammed on their talk page. If the edit is a problem, what good is a template that links to a 300-page document of closely written insider jargon? The edit needs to be reverted, without the user who did a good-faith revert being dragged off to WP:AE. And the the user who did the problem edit needs to be brought into the talk page consensus discussion.—Neotarf (talk) 05:42, 9 January 2014 (UTC)

Where did you get the "last-notice-before-block" idea from? It is not present in the proposal. Zero^talk 08:41, 9 January 2014 (UTC)

Collapse ... no, it hasn't been removed. It's elsewhere,  Roger Davies talk 10:45, 9 January 2014 (UTC)

Zero, Compare with the language of the current procedure: Discretionary sanctions are a fast-track procedure to tackle misconduct within defined topic areas and/or to prevent disputes from within the defined topic area overflowing freely into other areas of the encyclopedia; Discretionary sanctions may be imposed by any uninvolved administrator after giving due warning; Best practice includes seeking additional input prior to applying a novel sanction or when a reasonable, uninvolved editor may question whether the sanction is within the scope of the relevant case; Warnings should be clear and unambiguous, link to the decision authorising the sanctions, identify misconduct and advise how the editor may mend their ways; Notices of imposed sanctions should specify the misconduct for which they have been imposed as well as the appeal process; Discretionary sanctions have an established and clearly defined appeal process, which must be adhered to; Overturning arbitration enforcement actions out of process is strictly prohibited per longstanding principle; Discretionary sanctions should be used with caution where the community is already dealing with the specific issue through dispute resolution processes. Now check the new draft. This section has been removed completely. Imposing sanctions is totally at the whim of one arbitration enforcement admin. The only requirement for flushing the toilet is that the user in question has been informed of the existence of an arbcom case on the topic. Regards, —Neotarf (talk) 09:52, 9 January 2014 (UTC)

So, Roger Davies, I'm confused. Are you saying that a vital part of the proposal is not available for the community to examine? How then can anyone be asked to make a meaningful evaluation? —Neotarf (talk) 11:43, 9 January 2014 (UTC)

Now I get your point, I think. Previously, in principle, miscreants had to be warned about their behavior first, then could be smacked if they misbehave again. Under the new proposal they can, in principle, be smacked the first time they misbehave provided they knew there were sanctions in place. Did I understand you? I'm not sure that the difference will be as great in practice as you seem to think (the words "repeated or serious" will cover most cases), but you could start a new section of this page to discuss it. This section is about whether alerts should be logged or not, which is a separate issue. Zero^talk 11:03, 9 January 2014 (UTC)

• I personally support continued logging. I agree in particular with the third argument for logging in particular: regardless of whether notices are logged centrally, I think people will continue to keep these logs one way or another. Centralizing them will ease the job of trying to track them down. – GorillaWarfare (talk) 05:08, 9 January 2014 (UTC)
• Unlike the other warnings, DS alerts (which are not really warnings in the new draft) are required for the imposition of a sanction. By contrast, the other warnings are not necessarily a requisite for the imposition of blocks: in certain cases, a restriction may be imposed even on an editor who was never warned that his behaviour was disruptive, whereas nobody can be sanctioned unless he was aware that DS had been authorised for the topic area he was editing in (usually as a result of a warning, although all other relevant circumstances may be taken into account to determine whether a person knew about them).

  As a result, logging these alerts is quite useful, for it makes it possible for a person starting an AE thread to link to the relevant diff without having to waste a lot of time searching (for example, the alert may have been delivered years before or the editor in question may have switched accounts some time after the alert and before the current AE thread or other cases which would make it difficult to locate the diff).

  Doing away with logging would, therefore, in my opinion be problematic, because it would make that much more difficult to find the edit containing the alert; and I predict that if we were to scrap logging, quite soon the various parties would move the log to their userspace (or, at least, those parts they consider particularly useful...); so, basically, we'd be moving possible disruption from one place to another (and one where, by the way, an editor is generally granted considerable leeway).

  The way I understand it, the main problem is the location of the log: right now, people who have been informed that DS have been authorised are lumped together with those who have actually been sanctioned as a result of those DS, which may give the impression even the former somehow misbehaved. In my opinion, the most efficient solution (the one which would allow us not to lose the advantages of logging the alerts, while trying to minimise the disadvantages) would be to move the log to its own page, with a prominent notice at the top reading something along the lines of "inclusion here does not imply any sort of misconduct". Salvio ^{Let's talk about it!} 10:32, 9 January 2014 (UTC)

• In practice, it's the expectation at WP:AN3 that warnings are made before reporting editors: the reporting template expects a diff, and I've observed reports declined due to lack of warning. NE Ent 10:54, 9 January 2014 (UTC)
  • The four of us (Noetica, SMcCandlish, OhConfucious, and myself), were not provided with any diffs, although we were plainly accused of misconduct. Apparently this is how WP:AE works. The only reason given for the warnings was that we had given evidence without being admins. —Neotarf (talk) 11:35, 9 January 2014 (UTC)

• You may be misremembering (after all, it was a year ago). The warnings did contain a diff to this and the reason was not for giving "evidence without being admins" but for "using AE as a soapbox". Which may also, perhaps inadvertently, be what you're doing here. If you have any further meta comments, could you please make them in a separate section?  Roger Davies ^talk 14:24, 9 January 2014 (UTC)

No, there is no "misremembering". I have provided the diff to the exact wording. This is the exact definition of "casting aspersions": to accuse someone of egregious misconduct without evidence. If anyone thinks the link provided here contains some smoking gun that I violated some policy, it doesn't. It doesn't even show that I have ever edited in that subject area. It's a link to the archive of the case Apteva filed against Noetica.

As for the rest, I will answer on the appropriate talk page. —Neotarf (talk) 11:18, 11 January 2014 (UTC)

Evidently you do misremember. To quote the diff you give, "[Neotarf is warned not to] cast aspersions against others or to otherwise continue personalising […] disagreements". Kindly avoid derailing this consultation by reference to specific warnings and sanctions given to you in the past. AGK [•] 23:48, 11 January 2014 (UTC)

• @NE Ent: I don't question that (mainly because I've not been active there in quite a bit); however, the difference is that if I were to block for edit warring an editor who hadn't been warned beforehand, the block would still comply with policy (granted, it might be considered ill-advised, but it would still be valid), whereas if I were to impose a discretionary sanction on an editor who had never been given an alert, the restriction would be invalid, because one of the requirements for its imposition has not been met. Salvio ^{Let's talk about it!} 21:44, 9 January 2014 (UTC)

• I fall with Salvio here. The separation of the log might help easy the stigma that exists (hopefully in conjunction with an explicit notice). That being said as a result of the requisite nature of notifications, I find it important that they are logged somwhere. NativeForeigner ^Talk 18:04, 9 January 2014 (UTC)

If you call a tail a leg, how many legs does a calf have? Four, because calling a tail a leg does not make it so.
— Abraham Lincoln (attributed)

If notifications are logged centrally, and such logging is a prerequisite to sanctions, they're going to be treated as stigma carring warnings, regardless of what the committee may say. Which means editors will appeal them to WP:AE and the committee itself, no matter what editors / admins / arbitrators may state. Waste of time, and aggravation, for all involved.

• The first primary argument for central logging seem to be its lack will result in editors keeping hitlists against their wiki-enemies. So explicitly forbid editors placing diffs / links / notes of warnings on-wiki. (Obviously the talk page histories will serve as logs).
• The second argument seems to be concern admins would not know an editor has been warned. The Ent answer is .... so what? If in doubt, warn 'em again. If an editor is notified for R & I in January for saying Obviously, Ents are feebleminded and no one remembers / cares in July when they say Elobonians are dumber than Irish Setters, it ain't gonna hurt anything if they "escape" sanctions with another notification. The true warriors demanding Wikipedia post TRUTH™ in a DS area will be persistent enough that someone will remember the notification and dig out the diff when reporting their misbehavior (editor) or imposing a sanction (editor with sysop). NE Ent 01:37, 10 January 2014 (UTC)
• No log, some admins would abuse the log and harass editors. It happens under the current system and it would further be encouraged if it were to be logged. The templates themselves act as logs since the number of transclusions can easily be determined. Kumioko (talk) 14:39, 19 January 2014 (UTC)

• Call for ArbCom and the community to back the reform.

  I know several editors who have suffered emotional distress from the current DS system; thus I'm keen for the proposed reform to proceed. Leaving aside centralised logging for a moment, it's surprising that the proposal has been the subject of so much bloat. The new system comprises all of the elements of a much fairer, more functional system that will do two things at once: (i) enable the project to manage hot-button topics with limited resources; and (ii) remove the personal, punitive smell that surrounds the DS system.

  I'm satisfied that converting warnings to alerts is not just playing with semantics, and that a key motivation is to protect both regular editors of a contentious topic and those who might wander unknowingly into it and edit without realising that they might be exposing themselves to accusations of bad behaviour in context. As has been said above, one box of many at the crowded top of a talkpage alerting editors to take care is not likely to be noticed by many editors.

  I'm not persuaded by arguments that the new system will stigmatise; rather, it's a practical measure that once introduced will probably be seen as an obvious improvement and one that took too long to come. Nor am I persuaded by the view that an alert will expose an editor to subsequent sanction by an admin without due process.

  So I encourage people to get behind the reform proposal. Test antagonistic views against the facts, by actually reading through the text of the proposal—I fear there's misinformation abroad.

  Quick dispatch of this matter will allow the new committee to knuckle down and concentrate on further reforms. I encourage the committee to set a deadline and get it over and done with. We elect you partly to reform the system, not just to keep the seat warm. Tony (talk) 10:50, 10 January 2014 (UTC)

• Oppose the routine adding of random editors to cases. This system is too easily trolled; someone who is trying to game the system need only find one clueless admin, or shop enough forums, and sooner or later, the law of averages will catch up, and something will stick. New names should be added to cases in the same way the original names were added: in a comprehensive RFCU-similar forum that gives users the chance to participate before being sanctioned, to have any evidence presented against them, and to answer to any implications of wrongdoing. —Neotarf (talk) 14:21, 11 January 2014 (UTC)

  Are you talking about the giving of alerts/notices, the giving of sanctions, or both? In any case, there are a few too many misunderstandings of the procedure in there for me to concisely rebut, or indeed for me to take you seriously. AGK [•] 22:45, 11 January 2014 (UTC)

AGK, I am talking about the steps required for an Arbcom disciplinary procedure, whether it is called a "warning" or a "helpful, friendly notice" or a "fluffy bunny". If admins are required issue a "fluffy bunny" before indeffing someone, then the criteria for adding names to the "fluffy bunny list" need to be spelled out. I remember seeing the name of an editor who just ran for ArbCom on one of these lists, and wondering what he had done to get on it, or what powerful person he had gotten on the wrong side of. At least a hundred people see this list every month.

And yes, there do seem to be a lot of misunderstandings of the procedure. I wonder if the members of the committee understand exactly how it is being used in actual practice. There has been no explanation for the reason behind this latest proposal, or what perceived problem it is trying to solve. And now we find out the whole proposal is not here, there is more somewhere else, that has not been made publicly available. A summary statement might go a long way towards clearing this up. Regards, —Neotarf (talk) 08:28, 14 January 2014 (UTC)

• I second AGK's remark. Even though I know Neotarf personally online, I look aghast at what has turned into an unrelenting campaign against DS, against the reform proposal, against just about everything. Now I ask Neotarf to stop seeing spiders under every nook and cranny and to contribute in good faith to this page; otherwise, not to contribute at all. This is getting beyond a joke. Tony (talk) 09:45, 12 January 2014 (UTC)

• Insufficient clarity to offer opinion At top level articles unambiguously controlled by DS, (ie, Climate change, Abortion, India...), does an editor's initial article tweak or talk page comment - even the most helpful - justify "alerting" that editor about DS? If the answer is "yes", then central logging reaps benefits with no cost. If the answer is "no", then central logging will continue to be poorly installed lightning rod - still perceived as needed but runs high risk of burning down house. The best way to allay that risk would be better messaging about the meaning of alerts and when they should be given. NewsAndEventsGuy (talk) 12:14, 13 January 2014 (UTC)

• An alert is not a warning/caution/punishment. It's an alert - a notification. A person already knowing that an article is under AC/DS is the only reason a user should *not* be alerted. An an alert is a point of information (an FYI) not a warning--Cailil ^talk 14:58, 13 January 2014 (UTC)

• There is a discussion at ANI over a user added to a case decision. The consensus so far says the notification is being used as a chilling effect tactic. —Neotarf (talk) 00:06, 15 January 2014 (UTC)

• Neotarf listen to Tony above. What you list here is irrelevant - AQFK incorrectly used a template that can only be used by admins for formal AE warnings. What is being suggested here, in this reform, is the use of a new template as an FYI rather than a warning. This is being done to (amongst other things) reduce legalistic attitudes, and remove any perceived stigma, and address the chilling effect of the current uw-sanctions template. Your post above provides no counterpoint or additional information to this discussion that hasn't already been thrashed out over the last few months. Indeed it is actually an example of why this is needed--Cailil ^talk 02:02, 15 January 2014 (UTC)

• I also agree with Neotarf's concerns and I am also seriously concerned about the ramifications of this "review". DS's are heavily abused by some admins now but no one does anything to stop them. I don't see anything in this review that would make me think that is going to change. Especially since it has widely been agreed by the admin community that an admin who has been involved in actions against an editor isn't involved..which pretty much makes the first point of the review criteria irrelevant and nothing more than a hollow statement. This entire review, when reading between the lines does nothing to help Wikipedia or the community but gives some admins who deal frequently with DS's and arbcom sanctions more power. A review is a good idea and I support that, but this one is going to continue the existing pattern of abuse. If a review is seriously being considered then there needs to be something added that would allow someone to take action against the admins if they are abusing it. We shouldn't need to do a multi month arbcom case just to remove the tools fro an abusive admin. Let the bureaucrats have a discussion and decide if the case is obvious and warrants action or if it needs to be kicked up to Arbcom based on the circumstances. Kumioko (talk) 14:39, 19 January 2014 (UTC)

Logging: Alternatives to logging at the case page

At the moment, the biggest issue is with the stigma of having these notifications lumped in with sanctions. There's a number of things that have been suggested to improve this, rewording the template, renaming them to "alerts", separating them from the case page - but whilst we require an alert before any sanction we will need a way to check that the alert has been provided.

Currently we log them, but that has the stigma. We could remove all logs, but that leaves the old-fashioned trawl through talk pages and that's inefficient. What about tagging the alert? All we'd need to do is set up an edit filter on talk pages to display a tag, eg Tag:DS alert, when someone adds the alert. This will mean that anyone can check very quickly whether the user has been previously alerted to the DS (using the Tag filter), but does not log it for people to go searching against. It won't show up in any other searches, thereby removing the stigma.

Would appreciate thoughts and any other suggestions to take this away from the case page. Worm^TT(talk) 12:57, 15 January 2014 (UTC)

What's in a name? that which we call a warning
By any other name would smell as rotten;
So warning would, were it not warning call'd,
Retain that foul stigma which he owes
Without that title

I stand by my previous comments. Trawling is good! It means that an editor who stays on the straight and narrow after notification will have the what little stigma there is from a user talk page notice fade into dust ....; the recurrent warriors will have recent enough warnings that the "other side" can easily dig out the diff when they rat them out. "Efficient" is not a Wikipedia value, nor part of what makes it great. It says here Earth has 4,500 authors making 11,000 edits. That's the epitome of inefficient. NE Ent 13:42, 15 January 2014 (UTC)

• Let's see if I understand this....

(A) Jane makes an edit at (for example) Abortion

(B) Any editor can post to Jane's page {{Template:Whatever We Call The DSAlert Template}}

(C) By programming, the alert delivered in B gets a "tag"

(D) Jane deletes the alert from her page

(E) Six months later, Jane hits a rough patch and exercises bad judgment on a series of edits. Other editors want to know if Jane had ever been "alerted", so they simply look up Jane on a yet-to-be-named Special:Log

If that is the concept, in my current state of relative know-nothingness it looks like a great idea!NewsAndEventsGuy (talk) 14:03, 15 January 2014 (UTC)

Yes. And AGK put the basic stuff ready for this into the template some time ago,  Roger Davies ^talk 14:55, 15 January 2014 (UTC)

Cool, thanks for speedy response and thanks AGK for setting that up. Query, what's been the roadblock or opposing arguments? NewsAndEventsGuy (talk) 15:11, 15 January 2014 (UTC)

Because the template is not for the old warning but for the proposed alert system, which isn't in place yet.  Roger Davies ^talk 15:18, 15 January 2014 (UTC)

• Using tags sounds like a very good idea to me to avoid the stigma problem. There's some risk of false negatives, where the filter won't tag manual alerts that don't use the prescribed template (maybe the filter can be set to recognize links to the discretionary sanctions remedy?), but that's a minor concern. A bot could alert all previous recipients of warnings, making the current warnings logs redundant. It is even conceivable that a bot could auto-alert all editors who edit articles in certain categories associated with DS topics, which would take humans out of the loop in most cases.  Sandstein  22:03, 15 January 2014 (UTC)
• I'm not sure I follow all the details, but this sounds like a good idea. A Quest For Knowledge (talk) 22:13, 15 January 2014 (UTC)
• This is going to cause more abuse. If someone is in a disagreement with another editor then all they need to do is add an alert to a page to make that editor tracked. Then others who feel the same way can pile on, leading to the editor to be blocked or banned over nothing. This has happened even by admins trying to get people blocked, so letting anyone add the alert is only going to be trouble. Additionally, its also possible of false positives where the filter says the user had an alert but didn't. All too often the admins are too lazy to actually check and perform due diligence and just act thereby making the problem worse. Kumioko (talk) 14:53, 19 January 2014 (UTC)

AbuseFilter 602

After some teething problems, I have written AbuseFilter 602. Any alert (using the new {{alert}} template) will now be tagged as "discretionary sanctions alert". I believe this eradicates the need to have any sort of manual log at all for alerts. To discover whether a user has been alerted, all one would now need to do is open the history of their talk page, type "discretionary sanctions alert" beside Tag filter, and hit Go. AGK [•] 00:07, 18 January 2014 (UTC)

Well done! Which means we'll have to make use of the template mandatory, I guess.  Roger Davies ^talk 00:09, 18 January 2014 (UTC)

We're not supposed to a bureaucracy (see also WP:DTTR) NE Ent 00:11, 18 January 2014 (UTC)

Yeah, I know. I think we're down to least bad options. Or we stick with the status quo. Though thinking about it more, nobody objects to template delivery for the Signpost or DYKs. Maybe we should turn it into a barnstar. "Here's a barnstar for editing in a tpic under discretionary sanctions. You probably need to be aware etc etc"  Roger Davies ^talk 00:19, 18 January 2014 (UTC)

NE Ent, Roger may have meant making use of {{Z33}} (on which the AbuseFilter relies) mandatory. It seems tidier to make the actual alert template mandatory, though, and in any event anybody who thinks NOTBURO applies to arbitration probably needs a reality check... AGK [•] 00:23, 18 January 2014 (UTC)

It may become a status symbol to have received the Z33 warning. It sounds mysterious, like Area 51. EdJohnston (talk) 00:55, 18 January 2014 (UTC)

A Z3 is a much better status symbol than a {{Z33}}. NE Ent 16:45, 18 January 2014 (UTC)

Pillars apply everywikiwhere; while the overall buro level of AC is not in scope in this discussion, whether the specific proposal makes something more buro or less buro should be a factor taken into consideration by the committee. NE Ent

Test 1, Hooray, I alerted myself;

Test 2, Maybe there is server delay but I could not make the tag filter "discretionary sanctions alert" pull anything up even after ~7 minutes +/-.

I love the concept. Thanks for working on it. NewsAndEventsGuy (talk) 01:23, 18 January 2014 (UTC)

@NewsAndEventsGuy: I think the filter was configured last night to only get tripped by admins (because it was still being tested). Try it again; it should work. AGK [•] 12:50, 18 January 2014 (UTC)

Nope, I re-alerted myself today and waited ~45 min. Still can't call it up via the tag field. NewsAndEventsGuy (talk) 14:13, 18 January 2014 (UTC)success! Suggestion, is it possible for the return hits to indicate which subject area each alert was for? It would be a pain to get 12 hits for someone and have to open up each to find out if they were notified for a specific subject area. Ideally, the edit summary would take care of that, but given the iffy diligence of eds completing the edit summary that's hit or miss. NewsAndEventsGuy (talk) 16:34, 18 January 2014 (UTC)

Fair question; I've thought of it myself, actually. It's possible, with a lot of fiddly extra code and a reliance on the editor giving the alert following even more fiddly instructions. However, most editors only edit one (or maybe a handful of) DS-enabled topic area(s). We're probably never going to have an editor with a record alerts for every topic area, in other words, so distinguishing in the tag log between topic areas seems unnecessary, and will make the technical system a lot more breakable. AGK [•] 17:19, 18 January 2014 (UTC)

I guess "wait and see" is a perfectly valid approach instead of guessing what needs will arise. Sounds good.NewsAndEventsGuy (talk) 17:47, 18 January 2014 (UTC)

• Alerts are a bad idea see my comments below in a section called #Issuing alerts (Comments #2) -- PBS (talk) 11:42, 17 January 2014 (UTC)
• Thanks, AGK, this looks like a good solution to the logging problem.  Sandstein  18:34, 18 January 2014 (UTC)
• I don't know if the alert is a good idea. On the front end it appears to be but it also means every edit ever made has to go through the filter which increases the load on the servers. It was said in the past that people didn't want to put sanctions on a filter to prevent editors from editing against their sanctions. If that has now changed then I think it would be better to create a sanction filter. That would basically eliminate 75% or more (it would never be able to catch 100%) of the problems that are being encountered and be a big step to putting AE out of business. Kumioko (talk) 14:53, 19 January 2014 (UTC)
• Well done AGK, let's hope it works :) Worm^TT(talk) 08:43, 20 January 2014 (UTC)
• Possible problem. Maybe I'm misunderstanding something, but the {{Z33}} template is triggered each time the page is edited. So after a user has been alerted on his/her talk page, subsequent edits to the talk page produce the "are you sure you want to alert?" message along with the alert tag. Was the plan to manually remove {{Z33}} after an alert is given? vzaak 03:17, 22 January 2014 (UTC)

Issuing alerts (Comments #2)

Any editor may alert any other editor that discretionary sanctions have been authorised for the area of conflict.

As one editor can alert another that discretionary sanctions have been authorised for the area of conflict, without "the other" having ever edited an article in the area in question--the wording no contains requirement for editing--then one may as well create a page (similar to Wikipedia:Arbitration/Active sanctions) that contain all the current "discretionary sanctions" (eg the links in Wikipedia:Arbitration Committee/Discretionary sanctions#Affected areas but with more details) and warn all users simultaneously with a bot job that links to that page on all user talk pages. Then whenever another "area of conflict" is agreed by Arbcom, or a current one is changed, then another message can be sent to all editors, ordering them to look at the updated list. Job done with no arbitrary list!Y and if the link and warning is included in the greeting messages that is often placed on a new users talk page, no one will ever be able to claim that they have not been alerted.

I would like to use sarcasm here, but I know it does not work well, so I will just say that I think the whole idea of alert lists is harmful to the project as it just reinforces the idea that Wikipedia is becoming more and more bureaucratic and a less and less friendly place to which to contribute, particularly if one usually only dabbles in a few pages that are far from the vitriol that surrounds some of the areas of conflict.

I fail to see the advantage of issuing alerts as most editors most of the time are disinterested parties to the issues for which "discretionary sanctions" have been issued. The only reason for logging warnings is because many arbcom announcements contain the wording "Articles which relate to {topic}, broadly interpreted" and as an administrator may not see behaviour on a specific page as a repeat of behaviour over many pages, a list of warning of disruptive behaviour in an "area of conflict" is useful to administrators in reducing the very time consuming business of trawling through a large users talk page history to see if they have ever been warned in that area before. -- PBS (talk) 11:33, 17 January 2014 (UTC)

I recently closed an move request on a page called Galicia (Eastern Europe) as no consensus. One of the options was to move it to Galicia (Central Europe). If there is another request to move it to Galicia (Central Europe) for which there is a consensus to do so, does the page stop being covered by Wikipedia:Requests for arbitration/Eastern Europe as the consensus is that it is in Central Europe and not Eastern Europe (then is it still under "Articles which relate to Eastern Europe, broadly interpreted")?

"broadly interpreted" covers lots of ground and many articles can fall under more than one grouping. For example someone who grew up way down south in Invercargill may be interested in classical music (but have little interest of knowledge European geography or politics) and by chance edit an historical biography article about an East European composer. Should they automatically be warned about "Wikipedia:Requests for arbitration/Eastern Europe". If so who decides the boundaries the phrase "Articles which relate to Eastern Europe, broadly interpreted"? If they have never had the need to engage in the rough-house of the Wikipedia name space before, I think that placing them on a list about possible sanctions is not the way to encourage such an editor to feel welcome.

So to sum up "Any editor may alert any other editor" is too broad. Trying to define it as "Any editor may alert any other editor [who has edited an article an area of conflict]", causes too many problems with definition because of the phrase "broadly interpreted". So I say drop the formal "issuing of alerts" and keep the system of warnings but restrict it to a voluntary list which is added to by administrators which can be appealed to ANI in the usual way (but noting my worries over ANI made above) -- PBS (talk) 11:33, 17 January 2014 (UTC)

If there was per se a punitive implication to receiving an alert under the proposed system, that would be a reasonable post; but this will explicitly not be the case. Centralised logging is not a creature of the Stasi or the NSA: it's a practical measure, for example, to prevent people being alerted multiple times. "I fail to see the advantage of issuing alerts"—you don't have the horror of trying to manage battlefield wars on hot-button issues with limited human resources. I doubt there's a practical alternative to this proposal, which is why I support the basic thrust. Tony (talk) 11:45, 17 January 2014 (UTC)

"If there was per se a punitive implication to receiving an alert under the proposed system, that would be a reasonable post" So you do not think it a reasonable post! Tony you write "explicitly not be the case", Where is the explicit comment in the proposal? "you don't have the horror" do you mean "one does not have the horror" or do you mean that PBS does not have the horror? "I doubt there's a practical alternative to this proposal, which is why I support the basic thrust" what exactly do you think is the basic thrust of this logging because all I see is the devil in the detail making it an nasty suggestion.

I would be more interested if you would explain to me if my two practical objections about any editor can alert any other editor with out any restrictions on applicability and how does one decide the scope of "Articles which relate to {topic}, broadly interpreted"? -- PBS (talk) 20:04, 17 January 2014 (UTC)

It's a reasonable post. 3rr / edit warring works reasonably well, without much drama, and requires no centralized logging. NE Ent 23:28, 17 January 2014 (UTC)

Belief that "3RR/edit warring works reasonably well" reminds me of the fellow who fixes headaches by hitting his toe with a hammer. If there's an alternative, we should take it seriously. NewsAndEventsGuy (talk) 23:41, 17 January 2014 (UTC)

@NewsAndEventsGuy: I am sorry (maybe I missed a converstaion, or perhaps I am rather dense), but please expand on "If there's an alternative..." because I do not know what it is that you think it is that needs an alternative. -- PBS (talk) 12:10, 18 January 2014 (UTC)

Could be a miscommunication, and I don't feel like working it out will move the review process along, so please forget I said anything. NewsAndEventsGuy (talk) 12:30, 18 January 2014 (UTC)

Now I am even more confused. Moving the review process along to what? -- PBS (talk) 12:42, 18 January 2014 (UTC)

To its conclusion. We are holding this review so as to eventually adopt an updated, better version of the arbitration standard DS remedy. AGK [•] 12:52, 18 January 2014 (UTC)

Yeah a remedy that will give the Arbcom and supporting admins even more power that gives us regular editors even less rights. Its adds more flexibility to blocking admins and less AGF to editors who have "violated" a rule. This DS review doesn't level the playing field or make things more fair, its just another power grab. This will have the inevitable conclusion of expanding further the rift between admins and editors. DS's are abused now it will be even worse when this joke of a process is pushed through. This whole process improvement is a disgrace. Kumioko (talk) 13:41, 18 January 2014 (UTC)

The new system retains the same checks and balances as the old one, tries to emphasise softer remedies over hardline ones, and reduces the oppression exerted on templated editors. I therefore find little credibility in this analysis. AGK [•] 13:53, 18 January 2014 (UTC)

And I see no credibility in the argument that this is beneficial to anyone but the admins doing the blocking. Broader discretion on when, how and what they can do, and calling it reform. This "reform" is just window dressing for allowing a few abusive admins to have more power with less justification or avenue for recourse. This reform allows the same admins who have abused this system in the past to continue abusing it with less accountability. Anyone who thinks this reform will be a benefit to anyone besides those who are blocking are kidding themselves. This policy is going to end up as the policy version of the Visual Editor disaster. This system needs to be reformed, what you are doing here is taking spam and covering it with mushrooms (or other Fungi) and BBQ sauce and calling it a steak. As long as the ability exists for abusive admins to have unlimited discretion over when and what they can do to editors without any reasonable appeal process for those editors this project is a joke. The sheer volume of discussion here in this venue that few watch ensures that the vast majority of editors, outside those who have a vested interest in the status quo, do not get involved. This venue amounts to posting a community notice on the 5th page of the paper knowing no one will see it and be able to complain. You say my statements have no credibility but the problem with that statement is I am one of the few outside this circle of admins that understands the process and can see through this farse for what it is. Kumioko (talk) 16:55, 18 January 2014 (UTC)

Without seeing you articulate the other sides claims of benefits and then reason your way to your conclusion, this lowly editor equates multiple lines of protest as ignorable rant.NewsAndEventsGuy (talk) 17:10, 18 January 2014 (UTC)

Honestly it is a bit of a rant. I have articulated time and time again and it lead to nothing because the admins that support this have all the power. The editors have none because we, according to the Arbcom and the admins are clueless because we are not one of them. They don't really care what we editors say, they are going to do whatever they want, that is also what I have seen them do time and time again. So if you feel like dismissing my comments as rant feel free, just remember it a few months from now when they ban you from the project for making an edit that some admin feels is within the scope of a sanction because some alert told them so. I don't know everything about Arbcom or these processes but I know that a lot of admins who are attracted to this venue because it gives them a feeling of power. They abuse it and get away with that because there is no accountability for admins. So for every 100 admins that screw up, only one gets desyssopped after a month or more of conversation. Learn more about the process before you dismiss my statements. Its an abusive process and too open to the individual admins discretion. All it takes for an admin to ban an editor is to say they edited against some sanction, then the burdeon of proof is on the editor to prove otherwise...but they can't, because they can't edit. So it moves offline to Email that isn't transparent where they can be told no, without anyone seeing anything. Kumioko (talk) 17:28, 18 January 2014 (UTC)

• @PBS: An enormous list sent to every account on the project? I don't think that's a sensible way of doing things. What about people that join the project after the list is sent out? What about the vast majority of contributors, who don't care about or have anything to do with these topics (and with arbitration itself)? Can you rely on people remembering a topic was mentioned in a list of twenty or thirty topics a bot delivered two or three years previously? This proposal seems as though it would resolve non-issues, and perhaps also introduce significant new problems. AGK [•] 12:56, 18 January 2014 (UTC)
  • The list would not have to be sent to everyone only a link to a list. As I said above the link to the information can be contain within the greeting templates. BUT the point I was making was that singling out people for alerts are not desirable. Notice in this section not one person has addressed any of the specific issues I have raised against the use of alerts. If those specific reasons are not clear to you I can list as bullet points. -- PBS (talk) 19:28, 18 January 2014 (UTC)
    • Simply so we're both talking about the same thing (and everybody else knows what we're talking about too), it actually would be helpful if you would list your objections to alerting in bullet points. Sorry to make you jump through another hoop! AGK [•] 19:36, 19 January 2014 (UTC)

@AGK: you write "We are holding this review so as to eventually adopt an updated, better version of the arbitration standard DS remedy., but that sounds a little to me like the arguments in favour of manifest destiny or as Mrs. T government used to say TINA (there is no alternative). Your premise seem to be that there is going to be an adoption and that the adoption is for a "better version". However what is the current "standard DS remedy", and how is this better? There seems to be little discussion of the details of this proposal. It reminds me a lot of the whole WP:ATT fiasco. -- PBS (talk) 19:28, 18 January 2014 (UTC)

Are you participating in this consultation without having learned what the current remedy is? It is here, in the grey box. AGK [•] 19:30, 18 January 2014 (UTC)

See the links given by me at the start of this section which include the link you have provided, but it is not called the "standard DS remedy" it is called "Discretionary sanctions". However now that is answered do you see my point about TINA and the similarities with process that was used for adopting WP:ATT? An assumption over ATT was made by those that drafted it that NOR and V were broken and should be fixed and that they adopted a TINA attitude to all those who criticized their draft. -- PBS (talk) 19:49, 18 January 2014 (UTC)

user:AGK did you miss this post or do you not want to comment on TINA? -- PBS (talk) 14:40, 20 January 2014 (UTC)

As you did not elucidate what the "alternative" is, I see nothing substantive that needs my comment. AGK [•] 14:45, 20 January 2014 (UTC)

Concerns surrounding "any editor may alert any other editor" have some merit. No matter how many disclaimers there are about alerts being non-stigmatizing and advisory-only, an editor issuing an alert to another editor who perceives the alerter as being "on the opposite side" will, despite all the disclaimers, perceive some degree of intimidation and uninvitedness. Further, many DS topics involve (how shall I put this delicately?) conspiracy-minded participants, and there are actual cases of people calling DS notices "threats". A bot that automatically hands out alerts will solve this problem and save everyone's time in the process. (Nobody is intimidated by a bot.) vzaak 17:36, 19 January 2014 (UTC)

Having a bot issue the alerts was previously proposed. The consensus then held that because so many articles are peripherally related to each DS topic area, it would be impractical to have a bot issue the alerts. The bot would require an index of thousands articles affected by DS, and the index would have to be maintained by hand. For now, it seems like having alerts issued by hand, and instead de-stigmatising them, is the best solution. AGK [•] 19:36, 19 January 2014 (UTC)

I don't believe it's been adequately considered. Just going down the list of DS topics, in many cases there are categories already set up: Abortion -> Category:Abortion, Afghanistan -> Category:Afghanistan, Arab–Israeli conflict -> Category:Arab–Israeli conflict, etc. There's no reason to maintain a list of thousands of articles by hand. For an uncategorized article, a bot may record the first {{alert}} manually issued, and thereafter handle subsequent alerts. So even in the worst case -- an uncategorized article -- a bot is still an all-win and no-loss time- and intimidation-saver.

It's very important to avoid the intimidation/uninvitedness factor. Remember there was even a (dismissed) ArbCom case which included complaints about DS notices as "threats". vzaak 20:01, 19 January 2014 (UTC)

I pushed the bot hard, and am now biding my time. If this is adopted, as I expect it will be, somebody will eventually start making alert-related tools of various sorts and kinds for individual editors to run. Some amount of time later, bot automation for the top articles will be worthy of revisiting, and we'll have actual data about how its all working out to provide context. NewsAndEventsGuy (talk) 21:09, 19 January 2014 (UTC)

OK, but I must stress the intimidation problem again. As a compromise, how about a very simple bot that reads a username+category entry from a page, removes the entry, and places the DS message on the user's page. That sounds roundabout, but it's all the difference in the world. This is not only for the benefit of the person being notified, it is also for the notifier. As someone who has been harassed by conspiracy-minded (and frankly weird) individuals, I very much wish to avoid poking these people directly. vzaak 21:46, 19 January 2014 (UTC)

If you really want to go that way then disseminating in the way you propose is pointless as any editor at any time can edit a page that falls under a arbcom ruling particularly when they include the term "Articles which relate to {topic}, broadly interpreted". So you may as well publish a new item to all user talk pages similar to Wikipedia:Wikipedia Signpost but for official arbcom announcements, frocking and defrocking of admins etc. You could call it the "the Wikipedia Gazette" after The London Gazette, there is no need for it to be as complicated as the Wikipedia:Wikipedia Signpost it only needs to be a page of links with the latest news in one paragraph. Personally I have yet to see a strong augment in favour of alerts. It seems to me that ad-hock alerts on user talk pages are enough and do not need to be logged. Only warnings by admins logged ought to be logged and then only because it save administrative time. -- PBS (talk) 14:40, 20 January 2014 (UTC)

It doesn't make sense that some editors are issued with a warning for a particular subject area, and not others. It is unfair. Either the subject area applies to everyone, or, a particular editor is under sanctions, regardless of subject area. Since subjects areas are not in themselves bad editors, it makes sense that certain editors are the potential problem. Of course it might be that most editors are OK, but that certain Admins unfairly target certain editors, and the Admins should be sanctioned! --Iantresman (talk) 21:57, 19 January 2014 (UTC)

Do you mean Warnings or Alerts? -- PBS (talk) 14:40, 20 January 2014 (UTC)

@AGK: you wrote "Simply so we're both talking about the same thing" yes we are but I am opposed to logging alerts.

1. As many arbcom decisions contain the wording "Articles which relate to {topic}, broadly interpreted" who decides whether an article is or is not related to the decision?
2. As any editor at any time may edit an related article simply for example because they are fixing minor citation errors, putting in a link to a new article or whatever, if the list is is inclusive then over time it will have 100 or possibly 1,000 of name in it if it is exclusive then anyone one in the list will think that they have been singled out. Either way I think the list is bureaucratic and not helpful.

As I have said elsewhere administrative warnings are another matter because they are useful for other administrators when following up with administrative actions (and serves a similar function to the block log) -- PBS (talk) 14:40, 20 January 2014 (UTC)

A perfect example of the problem

In my very first use of the DS alert, the alerted editor came to my talk page to complain that the alert was "ominous", a "threat", a "warning", an "attempt to create a 'chilling effect'", "as if there might be consequences for disagreeing".[2] This is exactly the problem that I have been saying needs attention. It is not sufficient to say, "Oh let's just de-stigmatize alerts". This understanding cannot be uploaded to the minds of editors. However benign an alert looks to us (because it is actually benign), from the perspective of a newly-alerted editor it is entirely different.

Earlier I pointed to an ArbCom request in which the "alerts are threats" idea played a significant role. Indeed, the case begins, "This request for arbitration is to resolve recurring threats...".[3] See my response in #3 "Threats".[4]

Early use of the new DS alert-not-warning system portends disaster. Please don't blow off this problem. vzaak 03:05, 22 January 2014 (UTC)

Logging (comments)

Draft text

• Added a provision about logged modified or overturned sanctions.  Roger Davies ^talk 23:03, 16 November 2013 (UTC)

• Since the old warnings have been replaced by non-accusatory alerts (and a good thing too), what is the purpose of requiring alerts to be logged? Previously it was useful to know that an editor had been warned about bad behavior, but now that purpose has gone away. Of course someone filing a case will still need to provide a diff proving that an alert has been issued. Sanctions, on the other hand, obviously should be logged. Zero^talk 03:55, 17 November 2013 (UTC)

• I suggest: "All sanctions must, and alerts may, be logged ...". Logging alerts facilitates processing AE threads later, but depending on circumstances editors may not want to do it.  Sandstein  17:43, 17 November 2013 (UTC)

• It's probably best to keep it as a bright line as failure to alert is spelled out later. Maybe this can be reviewed again after it has been in operation for a while? The logging data can also be useful for deciding whether DS for a topic can be rescinded.  Roger Davies ^talk 08:42, 28 November 2013 (UTC)

• The point of alerts in the first place is to avoid editors saying "but I didn't know there was a DS for that topic" after being sanctioned. Without alerts (or the current warning system), there's no viable way to tell an editor who really didn't know about the contentious topic area vs. an editor that is being disingenous after being disruptive in an attempt to avoid sanctions. Logging the alert is thus positive proof that the editor was indeed aware of the DS. —Darkwind (talk) 08:03, 30 November 2013 (UTC)

• The mention of edit notices here is unclear about what it means as it's the first mention in the document. Maybe it's an oversight from removing the other edit notices? In any case, I really don't see the point in logging the placing of an edit notice as it has no enforcement or notification role. Thryduulf (talk) 08:35, 17 November 2013 (UTC)
  • Logging alerts is good housekeeping. It has an enforcement role as people can't be sanctioned unless they're demonstrably aware of DS.  Roger Davies ^talk 10:47, 2 December 2013 (UTC)
    • Yes, logging alerts is good. My comment though relates to edit notices. "While failure to log an alert, an edit notice or a sanction, does not invalidate it" (emphasis added). That is the first mention of edit notices in this draft, and (unlike in the first draft) they are not evidence of awareness. I suggest just getting rid of the text I italicised (and the preceding comma). Thryduulf (talk) 10:14, 3 December 2013 (UTC)
      • @Roger Davies: it looks like you might have missed my above comment. Thryduulf (talk) 13:22, 20 January 2014 (UTC)
        • Don't take it personally, they generally ignore my comments completely, especially when its something they don't want to hear. That's why I stopped taking the time to add suggestions or tell them about the problems, they just ignore them so anything more than a comment to raise awareness of said problem is just wasted time. The only reason I even bothered comment in this is because this review is a good idea generated by good intentions that is poorly written, will be poorly implemented, poorly supported and will cause more problems than it solves. Unless the Arbcom does something active to deal with abusive admins that abuse it, this review and associated policy change will fail because it gives the admins more flexibility to abuse the system. Kumioko (talk) 15:15, 20 January 2014 (UTC)
• Consider omitting "... repeated failures to log may result in sanctions for the issuing editor or administrator." as rules creep. Seriously, the Arbitration Committee is going to take the time to vote on a motion for egregious misconduct of this sort? If not, who imposes the sanctions?  Sandstein  17:45, 17 November 2013 (UTC)
  • It's not rules creep. Sooner or later someone will issue alerts pointedly or disruptively; we'd may as well anticipate the day. I don't see why the AE admins can't handle this kind of stuff themselves. ArbCom doesn't need to be involved in anything.  Roger Davies ^talk 10:44, 2 December 2013 (UTC)
    • I agree with Sandstein it is rules creep and should be removed. -- PBS (talk) 16:34, 11 January 2014 (UTC)
• I don't find here a valid argument why alerts should be logged. Darkwind wrote "Logging the alert is thus positive proof that the editor was indeed aware" but it isn't anything of the sort; the editor is most unlikely to find a logged alert instead of the alert itself on its own talk page. Roger Davies also thinks that the logging makes the editor aware of the alert; it simply isn't true. The proof that an alert was issued is the diff of it; no other proof is necessary. It seems to me that logging alerts serves no useful purpose. Zero^talk 14:51, 22 December 2013 (UTC)

Take pity on the poor backlogged admins. The log is for their easy reference, so that they can readily verify procedures were followed prior to imposing sanctions. Also provides evidence that they acted properly if their exercising their admin authority is questioned. NewsAndEventsGuy (talk) 13:04, 9 January 2014 (UTC)

• Anyone can issue an alert, but only administrators should do the logging. I see it as a useful interim step to warn a participant that their behavior has become enough of a concern that discretionary sanctions are being considered. Logging someone's name to a case page is generally seen as a very big deal, so shouldn't be done lightly, and should only be done by administrators. --Elonka 15:19, 22 December 2013 (UTC)

If DS only becomes authorized after this logging, then the logging would a de facto warning, thus invalidating this converation's assumption that by fiddling with the word "alert" we are really changing anything. Instead, a "warning for cause" from an admin (via logging) would still be required, and an "alert" from a regular editor like me would be optional. Seems a lot like what we do now to me. Check my contribs, I give regular-editor head's ups about ARBCC all the time. But I don't log it. NewsAndEventsGuy (talk) 13:00, 9 January 2014 (UTC)

Yes, that is a sensible and rational view. However, if a notification is an actual warning, some would like them appealable. Historically though, admins have long had broad discretion is to give whatever warnings they consider necessary and, in this context, I'm not clear why DS-related warnings should be appealable when, say, 3RR or copyvio notifications aren't. I suspect that, as usual and despite ArbCom's best efforts, this will prove to be irreconciliable and ArbCom will get grief in perpetuity from whichever side feels it has lost the debate.  Roger Davies ^talk 14:48, 9 January 2014 (UTC)

I suggest bifurcating the logging section to treat alerts and sanctions differently. Also, I think "edit notices" have gone away so delete those references. Some possible text tweaks are

===Logging====

=====Alerts=====

Any editor may, at their option, log an alert in the motion authorising discretionary sanctions for the area of conflict. Such logging helps administrators who are considering imposing DS ascertain whether an editor has been given notice, but logging of alerts is not a prerequisite to discretionary sanctions.

=====Sanctions=====

All alerts and sanctions must be logged on the page specified for this purpose in the motion authorising discretionary sanctions for the area of conflict. Whenever sanctions are modified or overturned, the administrator amending the sanction must append a note recording the amendment to the original log entry. While failure to log an alert, an edit notice or a sanction, a sanction does not invalidate it, repeated failures to log may result in sanctions for the issuing editor or sanctioning administrator. NewsAndEventsGuy (talk) 17:06, 11 January 2014 (UTC)

Thank you. Though if it's entirely optional you can bet anyhthing you like that one side of a dispute will scrupulously log their opponents alerts but not their allies.  Roger Davies ^talk 17:14, 11 January 2014 (UTC)

That benevolent tool known as a hammer can always be used as a weapon.... until it is simply place out of reach. If the "alert" is perfunctory, then let bots handle the issuance and logging. That way neither "side" can try to tweak someone in the nose via logging. NewsAndEventsGuy (talk) 17:59, 11 January 2014 (UTC)

Bots can't handle all the notifications; it's too complex.  Roger Davies ^talk 19:00, 11 January 2014 (UTC)

I don't speak wiki, but I do speak relational database. All we need is a table of page names that are associated with each ruling, and a table of user names associated with each "notice" section (plus date if you like). If it takes too many resources to do that in real time, just process each page name in the list of all page names to find the daily edits on that one page, and just alert the new names that pop up. If something so bad happens in less than 24 hrs that waiting for the bot is not an option, then people can still do it manually. I don't see what the big bot deal is. Can you enlighten? NewsAndEventsGuy (talk) 19:10, 11 January 2014 (UTC)

Bots aren't really a practical option. Take Abortion for example. DS protect not only the main articles but also the spin-off ones, and intersectional ones.

So, firstly and obviously, all the articles in the category:Abortion are covered, as well as the articles on: abortion clinics, abortion practioners, prominent pro- and anti-Abortionists, researchers, scientists, writers, journalists, broadcasters, lawyers, litigants, popular culture items, relevant legislation by country, legislative history, and any of the sources for the preceding articles which themselves have articles (for example, books, learned journals, magazines, and the respective publishers and writers).

Secondly, there are then the articles with a significant intersection to the area of conflict and which are likely to include significant coverage of abortion. This broadly means articles in the following categories category:BioethicsBioethics, Methods of birth control, Medical ethics, Human pregnancy, Reproductive rights, Ethically disputed practices, Human reproduction, and Birth control.

Thirdly, then there are wiki-discussions about either the editors or the articles references on any number of pages (WP:COIN, WP:AFD, WP:AN, WP:ANI etc).

And this is all before we get to any unusual or novel intersections. Setting those up would be tricky, and as it would involve clear human intervention, is bound to attract criticism. ("You only include articles XYZ and XYY in order to make me and my mates sibject to DS.") Perhaps more worryingly, as a bot has no way of telling whether or not, especially in peripheral articles, the edit is covered by DS or not, the number of false positives will be very very high.

So again, does the solution you propose eclipse the perceived problem in terms of workload and disruption?  Roger Davies ^talk 21:24, 12 January 2014 (UTC)

Well sheesh, it doesn't have to be perfect to achieve the MAIN BOT GOAL: Cultivate a culture of "Your edits may be fine but did you know...." Just run the bot on the articles pre-flagged as being directly in the subject area. That will (a) save labor, (b) cover most of the eds in the subject area, (c) and give birth to that cultural attitude towards alerts/loggings. For all the other types of articles you mention, (A) a high percentage of eds will have also fiddled at one of the flagged articles so they will already be covered, and (B) all eds can still manually do the alert and logging to pick up the small amount of slack. Notice that this labor for the small amount of slack is spent doing the same things that we would have to do for everyone without the bot. So the bots save labor, get lots of speedy coverage without getting bored, bring about a cultural paradigm shift over the alert/loggings, and everyone can still manually pick up slack bots miss. There is so much UP side, and essentially zero down side (that I see anyway) the only reason that really makes sense to me for not doing this is if they alerts are not really perfunctory ("Your edits may be fine but did you know....) but instead imply some sort of wrongdoing. Since actions speak louder than words, I suggest the mechanism we use to carry out the alerg/logging will do most of the communicating about what it means, and the text we actually place in the alert or log won't really matter. Perfunctory alerts should be run mainly by bots, specifically to de-stigmatize the whole thing. NewsAndEventsGuy (talk) 22:00, 12 January 2014 (UTC)

A bot simply can't identify whether an edit falls within the scope of a discretionary sanctions remedy. Until computers can perform general intelligent action, your proposal is so impractical that it seems like a non-starter. AGK [•] 23:58, 12 January 2014 (UTC)

We're not communicating so let's use the Tea Party ruling as an example.

QUERY T or F, "Any edit, including very helpful edits, at Tea Party or Talk:Tea Party merit the issuance of an alert to an editor."

• If you said "false" then please specify the language in the draft text on which you rely.
• If you said "true", then is there any technical impediment to having the brainless bot issue alerts to every editor who appears on those pages, and whose name is not already on the log?

For this exercise, please confine your thinking only to edits at Tea Party or Talk:Tea Party. If we get past this question, I'll go on to illustrate the concept by discussing edits at Michele Bachmann. NewsAndEventsGuy (talk) 01:56, 13 January 2014 (UTC)

No thank you. AGK [•] 11:45, 13 January 2014 (UTC)

Now that is a nonstarter.NewsAndEventsGuy (talk) 11:58, 13 January 2014 (UTC)

• Could I comment on a few statements above? "The point of alerts in the first place is to avoid editors saying 'but I didn't know there was a DS for that topic' "—well yes, but a nicer and more practical way of putting it might that the purpose is to protect editors who edit a topic without knowing that it's subject to DS. Who wants casual visitors to an article caught up in sanctions through their not knowing?

  Roger Davies says: "Logging alerts is good housekeeping"—yes, but let's be specific: we certainly don't want to waste time alerting someone more than once for the same article (even the same cognate area?); and that might irritate gnoming editors mightily (25 alerts when they work through a category one afternoon?). So centralised logging seems to be a must, and I'm sure the talents here can work out a streamlined way for everyone to search it quickly.

  NewsAndEventsGuy: I understand your notion that a bot could further depersonalise the alert system, since no human would issue the alert. That would be a good thing; but I think it's not within our grasp at this stage to create a bot so sophisticated that it could exercise judgement about gnomes and other one-off visitors, and repeat visitors who edit substantively. Better to start the system with complete human oversight, including the question of which cognate articles should be included in a DS domain (there will be grey areas). Tony (talk) 08:52, 13 January 2014 (UTC)

Tony, thanks for your comments but we're not communicating. First please answer the T/F question in my comment immediately before your reply. NewsAndEventsGuy (talk) 10:57, 13 January 2014 (UTC)

@NewsAndEventsGuy: Having given this considerable thought, I've concluded that automatic alerting by bot is not a good idea. First, the premise is faulty because it's pointy and disruptive: let's issue everyone in sight an alert to prove to the handful who regard them as stigmatising that they're not stigmatising at all. Second, it will issue alerts on an industrial scale and the people who will receive most alerts will be wiki-gnomes and vandal fighters, who will get one for each topic they edit (currently 27). It's true we could fix the bot to exclude AWB edits, but then that defeats the purpose of universality. Third, it wouldn't catch the intersections, without producing huge numbers of false positives. Frankly, all in all, this simply isn't worth the considerable set up effort and the inevitable fall out, and is - as they say - the tail wagging the dog.  Roger Davies ^talk 15:10, 15 January 2014 (UTC)

Answering your remarks...

(A)Re sentence starting "First" - you seem to be expressing opinion that follows from your other remarks, which I think were based on incorrect undestanding of the idea.

(B)Re sentence starting "Second" - False, because for each ARB ruling the bot would give a single alert to any editor. That functionality obviously requires some form of logging.

(C)Re sentence starting "Third".... False, because the bot would be told to monitor only those pages that are unambiguously 100% under a single ARB ruling's jurisdiction, and the bot would be blind to the "intersections".

Projecting the future, if we go to mere "alerts" what will happen is that regulars in the controversial pages will just alert everyone as soon as they show up. Given the penchant for human error, people will get re-alerted multiple times. For pages unambiguously 100% under a single ARB ruling, it's hard for me to see how having a bot would be any different, except it would be easier, facilitate the attitude that the alerts are not stigmatizing, and would reduce repetition in alerting eds. HOWEVER, I can let go of the "bot" idea if it will help this review along. It can always be proposed again later, in context of seeing how things are going after the revision is deployed.NewsAndEventsGuy (talk) 15:33, 15 January 2014 (UTC)

Re: [C]. You misunderstand me, not vice versa. We have 27 topics under DS (ie 27 single Arb rulings). A busy vandal fighter or an active wiki-gnome could easily accummulate one alert for each topic over a few days. Anyhow, I'm grateful that you're happy to let go of this for now.  Roger Davies ^talk 15:47, 15 January 2014 (UTC)

Thereafter, they get none. A little bit of chaos is always the price of changing procedure. At most your vandal fighter gets a one-time receipt of up to 27 bot-generated alerts and then no more. Plus we could also provide an opt-in tool so any ed could register their user names in the log to prevent receipt of any bot-alerts. So I still do not understand why there is resistance, but that's fine - I can live with that for now.

Speaking of repeat alerts.... suppose five editors all decide to send an alert to me about ARB-xx after my first edit in that subject area. If there were a log, couldn't the template be told to abort delivery of all the alerts after the first alert? NewsAndEventsGuy (talk) 16:07, 15 January 2014 (UTC)

Role of administrators (comments)

Draft text

• Various tweaks:

  Verb concordance. "Considers" > "consider". (per Bbb23)

  Added "acts when involved" to the "Questionable sanctions" section and changed its subheader to "Questionable administrator conduct".

  Added vanchor template to subheads.

  Tweaked the text about logging page restrictions, and added a bit about edit notices.

   Roger Davies ^talk 23:13, 16 November 2013 (UTC)

  • In "Questionable administrator conduct", "Arbitration Committee" -> "Committee".--Bbb23 (talk) 23:56, 16 November 2013 (UTC)
• We now appear to have a situation where one editor can place a warning on another editor's page, immediately complain to a sympathetic admin, who can instantly issue a one year ban, with no discussion and no appeal. The "Community sanction noticeboard" worked in a similar fashion, except it featuring a discussion period, but it was abolished because it was unfair. There does not appear to be any due process, nor checks and balances. --Iantresman (talk) 00:30, 17 November 2013 (UTC)
  • That is incorrect on every level. AGK [•] 00:48, 17 November 2013 (UTC)
    • Except that it could still happen, just with a bit more time in between, in theory (or my reading) a user can be alerted then the next enforcement action is a site ban, leaving them banned while they try to appeal from behind the ban and block. I agree that it is very unlikely to happen, but it could. Why a site ban rather than a one year block? Callanecc (talk • contribs • logs) 01:00, 17 November 2013 (UTC)
      • Is there really any difference? I mean, the main difference between a block and a ban is that a ban requires consensus to be lifted, just like an AE-block (although the requirement may be slightly different, in that if a banned editor appeal to AE, then only the opinions of uninvolved administrators count towards the determination of consensus); also, an AE-ban may not exceed one year in duration, so it's actually impossible for an editor to be indefinitely banned after an AE thread. Although I agree with Thryduulf that it's probably a good idea to require that remedies be proportionate. Salvio ^{Let's talk about it!} 10:55, 17 November 2013 (UTC)
        • @Salvio giuliano: the main difference I see is use of the user talk page. If someone is banned they can't use their talk page to do anything other than appeal if they are blocked they still can use it to discuss other things with other people. Also the sigma attached to being banned for a year is much more so than being blocked for a year. For example, we have Wikipedia:List of banned users but not Wikipedia:List of blocked users. There is also the precedent of allowing one user to ban another which, at the moment, is impossible. Callanecc (talk • contribs • logs) 01:50, 18 November 2013 (UTC)
          • The part about the possibility of editing one's talk page after being blocked/banned is actually controversial, at the moment. There are editors who believe that both blocked and banned editors should only be allowed to use their talk page to appeal their block; others disagree. I am not aware of anybody maintaining that blocked users can use their talk page to discuss whatever it is they want to discuss whereas banned ones can't (but it's entirely possible this has escaped my notice). So, really, the only difference is that a ban has more stigma attached to it, which is something I may agree with, but, to be entirely honest, isn't something I find particularly troubling. The end result is practically the same: the editor can't edit unless there is a consensus that he should be unblocked. So, if there are editors who feel really strongly about it, I'll not stand in the way (and will vote to change the wording), but, otherwise, this is a bit of a "meh" issue for me... Salvio ^{Let's talk about it!} 00:51, 21 November 2013 (UTC)
    • Perhaps add some text that remedies are required to be proportionate. That doesn't define what is and is not proportionate, but it is a requirement against which sanctions can be explicitly judged. Thryduulf (talk) 08:42, 17 November 2013 (UTC)
  • @AGK, could you explain how it is incorrect at every level? Are you suggesting that editors don't warn people, and uninvolved administrators can't then just ban someone? --Iantresman (talk) 16:21, 17 November 2013 (UTC)
    • @Iantresman: It is incorrect that administrators can ban people for no reason (which you implied) and that sanctioned editors cannot appeal (which you stated). A number of other assertions you made were also not correct. Regards, AGK [•] 19:50, 17 November 2013 (UTC)

• @AGK Thank you for the clarification. It seems that an editor can alert another that an article is under DS. The alerted editor can then edit an article under DS to which the first editor takes umbrage, alerts an uninvolved admin who can immediately instigate a ban without further warning or discussion. The banned editor may even have missed the alert, and finds themselves banned without having put their side of the situation. ie. no due process, no checks and balances? --Iantresman (talk) 20:05, 17 November 2013 (UTC)

• There are checks on the enforcement request, in the sense that administrators cannot impose sanctions for any conduct that is not disruptive. If they do, the sanctioned editor can then appeal to three different venues. If an editor fails to notice an alert and edits disruptively enough to then be sanctioned, that is nobody's fault but their own. The DS system has been used for years; "no checks and balances" has never been an issue. "I did not deserve this alert/sanction" sometimes is an issue, but the draft provides plenty of opportunity for them to make their case. AGK [•] 20:53, 17 November 2013 (UTC)

• @AGK Presumably the way to completely avoid disruptive editing is to discuss proposed edits in the talk pages, and on various noticeboards first? --Iantresman (talk) 22:58, 17 November 2013 (UTC)

• "Sanctions must be logged" is redundant to the preceding section and can be omitted. As is "Administrators are expected to log page restrictions though failure to do so does not invalidate it", but restrictions could be mentioned above as a type of sanction to be logged. Besides, "is expected" is a "should" rule, which contradicts the "must" rule in the preceding section.  Sandstein  17:49, 17 November 2013 (UTC)
• "Any duly notified editor may be sanctioned for any repeated or serious failure to meet Wikipedia's behavioural expectations" should be omitted because it is redundant to the preceding section "Behavioural expectations", and has nothing to do with the role of administrators.  Sandstein  17:51, 17 November 2013 (UTC)
• The punctuation in "Page restrictions" is inconsistent. Either commas or colons and semi-colons should be used after "may impose on any page relating to the area of conflict". AGK [•] 21:40, 25 November 2013 (UTC)

• The whole Jclemens case request that is on WP:A/R/C right now brought something to mind: I would suggest that "regularly" be removed from the "Questionable administrator conduct" section. There's no reason to handicap ArbCom in this fashion; a single instance of extremely poor decision-making should not always be pushed aside simply because it was the first such occurrence. NW (Talk) 02:00, 15 December 2013 (UTC)

The "questionable Administrator conduct" section seems more likely to deter good conduct than bad. NuclearWarefare is correct that ArbCom needs to maintain flexibility. Overall, there are no positive standards: such as expecting administrators to be respectful, explain their decisions, communicate with other administrators, apply sober-minded judgement, and anything else. I mention these because DS is a place where adminship is a very big deal, especially given the degree to which an admins decisions are irreversible and have serious trust consequences in areas which already have a trust deficit. "Accountability" is an empty word without standards. --Tznkai (talk) 02:09, 21 December 2013 (UTC)

A situation just came up on AE and I think a clarification in the new draft is in order. The wording of "any uninvolved administrator may impose warnings, admonishments, editing restrictions, interaction bans, topic bans, site bans of up to one year in duration" suggests that, given the degree of power and discretion administrators have, the maximum length of any measure is one year. However the same provision also allows "and/or any other measures which the imposing administrator believes are reasonably necessary to ensure the smooth functioning of the project" which could include durations longer than a year, or running indefinitely. Same goes for the standard enforcement mechanisms. Is this a feature, or a bug? If the 1 year is meant to be a hard cap, it should be written so.--Tznkai (talk) 21:30, 29 December 2013 (UTC)

I now notice, too, that the wording of the provision "... warnings, admonishments, editing restrictions, interaction bans, topic bans, site bans of up to one year in duration ..." should be reconsidered. Admonishments aren't sanctions (and I find them patronizing and insulting; they should be omitted), but blocks as the most common type of sanction aren't mentioned. Is it really the intent to limit e.g. topic bans to a duration of one year? Indefinite topic bans are now relatively common at AE, and haven't to my knowledge caused any particular problems. In general, I prefer sanctions that are not limited in time to sanctions that are, because timed sanctions encourage a punitive approach rather than a preventative one - a sanction should end when it is no longer needed (e.g. after a convincing appeal), not when an arbitrary amount of days has elapsed. I therefore recommend to edit the provision to mention blocks, omit admonishments, and remove the one-year limit. That would also resolve the ambiguity perceived by Tznkai.  Sandstein  21:44, 29 December 2013 (UTC)

I would be fine with that with a stronger appeal process, and/r weaker overturn protection and/or higher admin conduct standards or other devices to counteract the first and harshest mover incentives. Otherwise, there are no real constraints on a single administrator being punitive, or for sanctions to slide into severity without anyone even trying to do so.--22:03, 29 December 2013 (UTC)

Sandstein wrote above 'I prefer sanctions that are not limited in time to sanctions that are'. I find that comment disturbing, as it runs counter to the entire Western system of justice. Indefinite sentences in the Western justice system are a rarity. Why should Wikipedia specialize in them? The same goes for sanctions 'broadly construed'. The entire Western justice system runs on sentences which are very strictly defined. It seems to me Wikipedia is going in the wrong direction on these two points, and that the premises should be re-examined. It's also very administratively time-consuming to deal with appeals on both these issues, and appeals arise much less frequently by definition when sanctions have time limits and when sanctions are not broadly construed. NinaGreen (talk) 15:52, 30 December 2013 (UTC)

Nina Green, we are not operating a criminal justice system that issues punishments to evildoers. We are, if we must use legal analogies, operating a system of administrative justice that should result in measures tailored to prevent conflict that hinders work on the encyclopedia. I am guided by the principle of our policy WP:BLOCK that "blocks should not be punitive". Neither should other sanctions. They should be aimed at preventing conduct that is deemed unhelpful. And they should last exactly as long as they are needed for that purpose. This means that even a block for the most egregious misconduct should be lifted after a day if we receive credible assurances that the misconduct will not reoccur, but it also means that a sanction for very minor misconduct should last indefinitely if we must assume that the same conduct will be repeated if the sanction is lifted. There is one practical advantage to timed sanctions – they do not need to be actively reviewed to expire. This makes them suitable for first sanctions or very routine situations such as edit-warring, but every subsequent sanction should, in my opinion, require affirmative action to lift following discussion with the user at issue. For the purpose of this policy, setting a upper limit of one year may have the detrimental effect of inducing admins to think in terms of timed sanctions only and therefore in terms of punishment rather than prevention.  Sandstein  16:17, 30 December 2013 (UTC)

Sandstein, no justice system in the Western world, whether criminal, civil, or administrative, uses indefinite sentences or sentences 'broadly construed'. The essence of the Western justice system is that nothing is left to the 'interpretation' of those administering the sentences. Wikipedia has lost sight of that principle. And it's also important to keep in mind the point I have brought up several times, and which has never been answered (see this page and the earlier archived discussion), namely that the purpose of an arbitration is to clear the decks of 'troublemakers', and once those decks are cleared by the arbitration, there should be no troublemakers left, so why are discretionary sanctions then imposed on a topic? The only answer can be, 'Well, the arbitrators are anticipating trouble in the future from some as-yet-unidentified trouble-makers'. Again, the entire Western justice system is based on punishment of offenders. Sentences are never imposed in the likelihood that someone will offend in future. Wikipedia again runs counter to the principles of our Western justice system, and the outside world would find that peculiar, if not downright offensive, if it were aware of it. NinaGreen (talk) 16:29, 30 December 2013 (UTC)

Nina, Sandstein is entirely correct that we're not in the justice business here, and I'm afraid you're describing a conception of "Western justice" that does not mach up with reality or theory. If I can wander into some theories of justice for a moment, the distinction that Wikipedia makes between punishment and prevention is a little bit nonsensical, since we accept deterrence as prevention. What we really seem to mean is that our use of blocks is not retributive, but utilitarian. For that reason, I don't see why more than "a day" is anything but entirely arbitrary amount of time to declare blocks punitive, when the measure is actually "about as long as we think it'll take." The advantages to time limits are prophylactic and pragmatic: no administrator has an assigned beat, no bureaucracy exists to transfer responsibility, no structured hierarchy exists to funnel action into unified policy or control authorized agents. There are many, many ways that a sanctioned user can get lost in the shuffle, or simply get an unintended message of rejection and denunciation, despite any of our protestations otherwise. For all of these reasons I think it is better to have the onus on administrators to renew sanctions than otherwise, but there are other ways of addressing that concern.--Tznkai (talk) 16:41, 30 December 2013 (UTC)

Tznkai wrote 'Nina, Sandstein is entirely correct that we're not in the justice business here'. That's absolutely wrong, both in terms of the reality of what happens on Wikipedia and in terms of how the outside world would see it. I know it's enormously difficult to change things once people have gone down a certain path and have come to see things in a certain way, and Wikipedia seems always at great pains to state that it's 'not in the justice business' and that arbitrations are not trials, and therefore no elementary rules of justice apply, but the reality is that an arbitration is a trial, whatever Wikipedians choose to call it, and the result of an arbitration is the passing of sentence, whatever Wikipedians choose to call it, and it all runs counter to the principles of the Western justice system. The 'fix' is a very simple one, and would greatly reduce the administrative workload on Wikipedia, so fixing the problem and bringing Wikipedia policies into line with the principles of the Western justice system would be a 'win-win' situation. NinaGreen (talk) 17:16, 30 December 2013 (UTC)

Nina, to a certain degree, how the outside world looks at us is also not our business, but let me focus on the rest of what you have to say. The fix is incredibly difficult. You've been talking about "elementary rules of justice" as if they were a culturaly neutral, reasonably universal, easy to implement, self evident set of assumptions and practices. They are not. Actual justice systems are incredibly complex, imperfect, reliant on submerged cultural assumptions and habits, and perhaps most importantly, paid professionals. To seriously integrate justice into what we do would be catastrophic, for we would have all of the disadvantages of a justice system and absolutely none of the advantages and resources those systems bring to bare. Wikipedia is not a sovereign authority. It does not govern people. It is a website. Insofar as possible, expecting just conduct is good, expecting production of justice is a nightmare. I am on my very best days here, a wise-enough, lucky-enough man. I have never achieved the philosopher-king zen required to promulgate justice.--Tznkai (talk) 17:23, 30 December 2013 (UTC)

Tznkai, if Wikipedia raises funds from the public, which it does, 'how the outside world looks at Wikipedia' is our business, both from the point of view of encouraging the public to donate to an institution which it considers to be properly run, and from the point of view of Wikipedia's integrity in requesting the public to donate. Re your second point: it would be entirely impractical (i.e. insane) to try to import the entire criminal justice procedure, or civil law procedure, or administrative law procedure into Wikipedia. But that's not at all what I'm suggesting. I'm suggesting a few easy 'fixes' which would go a long way toward altering public perception of the Wikipedia arbitration and sanctions system if the public ever happens to have a look at it, and which would drastically reduce administrative workload on Wikipedia. What's wrong with that? It sounds like a win-win situation to me. NinaGreen (talk) 17:43, 30 December 2013 (UTC)

I don't think the general public knows or cares to know much about our inner workings. I am concerned with new editor recruitment and retention, but that I think is more of a problem of newbie biting and our arcane just complex enough to alienate and not strong enough to do anything useful sour spot of a bureaucracy sucking all of the fun and satisfaction out of a hobby. Implementing the so called simple fixes - of which I am not exactly sure you refer, would make that problem worse with no appreciable gain. Encouraging people to further think of themselves as wronged parties will just encourage editors to treat administrator decisions as battlegrounds to find justice, again, bringing us the difficulties of a court without the advantages. For practical reasons aforementioned, I think broad construction and indefinite sanction can bring trouble as well, but thinking in terms of justice, or worse, "Western justice" a frame that will lead rather quickly to error. The tl;dr is this: editors care about getting fair shake and good experience, and we should do our best to give them one, but implementing the due process from justice systems would make that problem worse not better.--Tznkai (talk) 17:53, 30 December 2013 (UTC)

Since you admit you don't know what specific 'fixes' I'm speaking off, why go off on tangents about what you think I'm speaking of with statements such as 'Encouraging people to further think of themselves as wronged parties'? Who said any such thing? NinaGreen (talk) 18:03, 30 December 2013 (UTC)

I'll set out the specific 'fixes' in a new section below to clarify. NinaGreen (talk) 20:07, 30 December 2013 (UTC)

• My first problem here is that Iantresman is exactly right on what could and has happened in the past. I could list at least 3 admins off the top of my head who were desysopped or came close because of doing exactly that. I also have a problem with point one that refers to "Uninvolved admins". It has long been the case that admins acting on a case were not involved and even in cases where they clearly were, the general agreement was that they were admins and were trusted and thus were acting accordingly. I wholeheartedly disagree with that logic and have stated so repeatedly. Including this as a line is either pointless because it will never be enforced or it needs to be backed up someone such as the admin needs to refer to an uninvolved admin for agreement on the remedy. Which could also be abused...but would be better than the involved admin acting alone without any fear of reprisal. Kumioko (talk) 14:59, 19 January 2014 (UTC)

Reducing or overturning sanctions (comments)

Draft text

• The WP:AN section is now back to how it currently is. Much more to the point, it now occurs to me that the process for reducing or overturning sanctions is effectively the same as for an appeal. Easiest is probably to fold them into one. Thoughts?  Roger Davies ^talk 23:14, 16 November 2013 (UTC)
  • You're probably tired of me saying this: "Arbitration Committee" -> "Committee".--Bbb23 (talk) 23:58, 16 November 2013 (UTC)
  • Yes, the two sections should be combined.--Bbb23 (talk) 00:01, 17 November 2013 (UTC)
  • The two items should start lowercase for consistency and because the items are sentence fragments, although item #2 muddies things by having a fragment followed by a complete sentence. You could cheat a bit by putting the sentence in parentheses. The first item should have a semi-colon before "or" for consistency with the other longer lists of this kind.--Bbb23 (talk) 00:13, 17 November 2013 (UTC)
• I would prefer to see "the status quo prevails" (in Reducing, and below in Appeals) be phrased more clearly in the policy. In the past the community has had disagreements about whether, when consensus is unclear about something, "status quo" refers to "the state before the sanction (status quo ante)" or "the current state". Yes, you could argue that if you meant "status quo ante" you'd say that, but I rather think it would just be easier and cause less argument later on to say "if consensus is unclear, the sanction stands", which is what (I think?) you actually mean. A fluffernutter is a sandwich! (talk) 01:34, 17 November 2013 (UTC)
• As in the previous round, I recommend striking "(b) uninvolved editors at the Administrators' noticeboard" for practical reasons. In controversial cases attracting many commentators, figuring out who is uninvolved will be very difficult at best, and can provide fodder for endless wikilawyering. Besides, the provision doesn't tell us what "uninvolved" means in this context (unlike for administrators, we don't have a policy defining uninvolvedness for non-admins). Are people who edit in the same topic area with the same or the opposite point of view as the sanctioned editor uninvolved? Good luck figuring that out in the calm and collegial climate that we all know prevails in noticeboard discussions. This appeals provision is a recipe for paralysis and endless additional conflict.  Sandstein  17:57, 17 November 2013 (UTC)
• And yes, the sections "Reducing or overturning sanctions" and "Appeals" should be merged. Also, you should consider merging this part of the rules with Wikipedia:Arbitration Committee/Procedures#Reversal of enforcement actions, because it makes little sense to have separate sets of rules for appealing DS and non-DS AE actions.  Sandstein  18:06, 17 November 2013 (UTC)
• As I have commented elsewhere on this page, this provision combined with others creates a first and most severe mover problem. MBisanz had some choice words for the situation that I endorse again, 4 years later, "it is a bit like the sheriff of an old west town coming into the bar and throwing some loaded handguns on the table". Relying on the community or a clear consensus of AE administrators is dubious protection. Disputes for which discretionary sanctions are authorized are almost by definition areas with contentious edit warriors full of passion. Separately, even here, Wikipedia should err on the side of restricting less, not restricting more. The underlying conceit of nearly all of Wikipedia's norms and procedures, maybe the very structure itself, is that anyone can edit, because all mistakes and errors can be undone. This provision not only turns that assumption on its head, it flouts it by protecting a class of administrative action.--Tznkai (talk) 01:46, 21 December 2013 (UTC)
• Not sure if it should be in this section or another one, but there should be something somewhere to cover the situation of multiple administrators active in the same topic area. For example, if there are three administrators monitoring an article, and one administrator tells an editor, "Don't do that again or I'm imposing a sanction," but then a second administrator comes in and imposes a sanction without discussing it with the first admin (effectively undermining them). Another potential situation is where an administrator imposes an indefinite revert restriction on a certain section of the article, and then their attention turns to other areas of the project. If a second administrator then becomes more active in monitoring the article, how much authority do they have to deal with "old" sanctions from an administrator who hasn't been around for several months? A note to the previous admin's talkpage may be a good idea, but if there's no timely response, do they have to go through the step of a WP:AN appeal, or can they just proceed with their best judgment? Perhaps an addition to this section might be: "If the original enforcing administrator has been inactive on monitoring a particular dispute for 90 days or more, and is not responding in a timely manner on their talkpage, another active administrator may make reasonable adjustments to sanctions. If, however, the original administrator returns and disagrees, administrators are expected to engage in civil discussion (modeling the correct way to handle disputes) as to how to proceed." --Elonka 15:34, 22 December 2013 (UTC)

Restricting admin actions

Just to highlight a small but important proposed change, which needs discussion. Earlier versions have enabled admins to increase sanctions of their own volition but not to decrease them or overturn them. It strikes me that ALL admin intervention should be subject to the same rules. This would not prevent an admin increasing a sanction if fresh misconduct takes place but would prevent increasing a block unilaterally from say a week to a month without going through the process. Thoughts?  Roger Davies ^talk 09:30, 29 November 2013 (UTC)

Shouldn't all sanctions (block and topic bans) be based on a set escalating scale. ie. The first sanction is always 1 week, then the next infringement would be 2 weeks, then 1, 3, 6, months, and finally a year. There currently seems to be no rationale for the duration set by admins.

Shouldn't the same duration be applied to both blocks and topic bans? It seem inconsistent that a block of up to one year can be set, but topic bans can be indefinite. This would probably reduce the number of appeals. --Iantresman (talk) 11:01, 29 November 2013 (UTC)

@Iantresman, short answer 'no'. Long answer: the reason AC/DS is discretionary is because the action required may need to use the sysop's best judgement. If the system only used the normal 1 week 2 week, 1 month, 3 months, escalation cycle then it wouldn't be discretionary.

@Roger: I see where you're coming from but for clarity this applies to sysop escalating sanctions imposed by other admins - not their own. Or would you see the need for that with sysops increasing their own sanctions. Either way there is a layer of added bureaucracy here (but I see your point).

The problem here from my POV is that a second sysop's actions could legitimately be described as a new and individual sanction under AC/DS as opposed to a tweak/alteration of another admin's action, and requiring sysops to go to AE for approval (when in most cases it'd merely be a rubber stamping excercise) this would have the unintended consequence of wasting time--Cailil ^talk 17:18, 29 November 2013 (UTC)

@Cailil. I would argue that the giving of a sanction is the discretionary bit, and doesn't have to apply to the duration. There is also a good argument that one admin shouldn't be judge, jury and executioner for hopefully obvious reasons. Since the option is available, I can't see why we wouldn't want to take it, in order to maximise transparency and accountability. --Iantresman (talk) 17:54, 29 November 2013 (UTC)

@Roger Davies: In which situation could this be a problem? I can imagine situations where a user is given a brief block, then continues ranting on their talk page or offwiki in the vein of "just you wait until the block expires, then I'll show these dirty Syldavians", which makes the blocking admin (or another admin) decide that a longer duration is preventatively needed. This wouldn't be a problem with a normal block, so why would it be in this context? And what does "going through the process" mean? There isn't much of a process for discretionary sanctions; they may be imposed without prior discussion.  Sandstein  18:23, 29 November 2013 (UTC)

I think that situation would be covered under Roger's comment that it "would not prevent an admin increasing a sanction if fresh misconduct takes place". Only if one admin were to block for say 2 days then another comes along (and the blocked user hasn't done anything anywhere) and unilaterally makes it a one year ban under the protection of DS. So likewise I assume this means that if an admin semi-protects a page under DS then autoconfirmed accounts continue to disrupt that that admin is free to upgrade protection to full but would be unable to extend the semi or lower it to PC? Callanecc (talk • contribs • logs) 00:30, 30 November 2013 (UTC)

@Sandstein: Generally speaking, in a situation where an administrator thought another administrator's enforcement decision too lenient, and unilaterally increased its severity. AGK [•] 11:44, 2 December 2013 (UTC)

OK, I see that possibility, but that hasn't ever happened that I'm aware of. I don't think it's worth the instruction creep.  Sandstein  16:43, 2 December 2013 (UTC)

I think I'm more or less with Cailil on this one. An administrator cannot overturn a sanction imposed by another admin acting under AE, but he can impose a new sanction if he sees fit - providing he does it based on misconduct by the sanctioned user since the original sanction and not on older evidence, because if he did the latter that could probably be considered a change of the original sanction which is not permitted under the existing rules. Gatoclass (talk) 13:34, 3 December 2013 (UTC)

Yes. That's exactly what I had in mind.  Roger Davies ^talk 09:36, 4 December 2013 (UTC)

Okay, I can't see much of an objection to that. Gatoclass (talk) 17:36, 6 December 2013 (UTC)

If consensus is unclear

• The draft says "If consensus is unclear, the status quo prevails".

this is contrary to Wikipedia:Consensus which says in the "No consensus section"

• "When actions by administrators are contested and the discussion results in no consensus either for the action or for reverting the action, the action is normally reverted."

If there is an uncivil editor who has sufficient acolytes to overturn administrative actions in areas where there are no arbcom sanctions, then if an arbcom that contains the wording "Articles which relate to {topic}, broadly interpreted" they can be sanctioned. Has the unforeseen consequences of two different procedures been discussed. For example under which process would the determination be made if a particular article falls within the scope of "broadly interpreted"? Who decides which process is the appropriate one to use? It also begs the question why are administrators given the benefit of the doubt when imposing sanctions but not when administrative actions taken over behaviour on non-sanctions pages? -- PBS (talk) 13:17, 19 January 2014 (UTC)

Appeals (comments)

Draft text

• I've brought the language for this more or less in line with the language for modifying or overturning,preparatory to merging the two sections,  Roger Davies ^talk 23:15, 16 November 2013 (UTC)
  • There should be a colon after " three possible stages for appeal". I don't see the need for "and/or" in item #1; "or" would be fine. Each instance of "Arbitration Committee" should be changed to "Committee" and no wikilink.--Bbb23 (talk) 00:06, 17 November 2013 (UTC)
    • Overall - Can we safely assume that this only applies to formal appeals? IOW, continued discussions are allowed on the admin's talk page? For example, if an admin sanctions an editor, it's permissible for an editor to say, "Hey, I think you make a mistake. How about X?" A Quest For Knowledge (talk) 00:42, 17 November 2013 (UTC)
    • Regarding provision 3, why is a consensus of ArbCom members required when it doesn't take a consensus to impose the sanction? This is a whole sale departure from current practice, correct? A Quest For Knowledge (talk) 00:42, 17 November 2013 (UTC)
      • Regarding provision 3, this was already fixed by me and NuclearWarfare in the previous draft.[5][6] How did it get back into the article? Were any other changes lost? A Quest For Knowledge (talk) 18:01, 17 November 2013 (UTC)
        • @Bbb23: This was already discussed[7] and fixed.[8][9] with draft 1 and as you can see, I started a second discussion above. Can you please self-revert? A Quest For Knowledge (talk) 20:05, 17 November 2013 (UTC)
          • Fixed? There was nothing broken. I don't agree with that change, so I was bold and took it out. It adds a lot of work and bureaucracy. Determining consensus is easy, whereas "a majority etc" requires a formal vote. Formal votes are usually unnecessary, and add delay and inconvenience. There is no point in a formal vote if the appeal doesn't have any support.  Roger Davies ^talk 20:24, 17 November 2013 (UTC)
            • @Roger Davies: Wait a second, let me see if I have this straight. You disagree with the current policy and practice (AKA consensus). When this issue came up for discussion, you declined to participate in the discussion and allowed existing consensus to stand.[10] You decided to keep silent and quietly changed policy without bothering to mention it to anyone? Honestly, I don't know how you edit Wikipedia, Roger, but when I attempt to make contentious changes, I don't try to quietly make them and hope nobody notices. Instead, I try to be as open and transparent as possible. I don't try to hide my edits. In fact, I do the exact opposite: I start discussions on the talk page to bring attention to the change to make sure that everyone is on board. I'm really taken aback that you would attempt to make a major change to policy apparently without even bothering to mention it. In any case, you boldly changed existing policy. If you want to change consensus, the burden is on you to change that consensus. A Quest For Knowledge (talk) 23:10, 17 November 2013 (UTC)
              • What consensus, AQFK? NW reverted you a couple of hours after you unilaterally made the original change. When I walked through the changes of that page one by one to arrive at the current draft, I reached the same conclusion as him.  Roger Davies ^talk 05:50, 18 November 2013 (UTC)
• Either as part 3 or subpart 2(c): "Approval of the Committee". (I.e. arbcom may authorize an admin to modify a sanction)

I still don't see the need for three separate stages of appeal - the appellant should have the choice of appealing either to AE, AN or ARBCOM, not both AE/AN and if that fails, to ARBCOM. It's highly unlikely that there are any facts worthy of further consideration after a user has already been sanctioned either by an admin or from an AE request, had his appeal turned down by the sanctioning admin, and then had a further appeal rejected at AE or AN. Moreoever, the provision for appeal first to AN and then ARBCOM raises the possibility of drama between the community and ARBCOM. Gatoclass (talk) 06:18, 17 November 2013 (UTC)

• • The way this is supposed to be, in my opinion, is that a sanctioned editor has the following three separate possibilities to appeal a sanction: in this order, a. to the imposing admin, b. either to AE or to AN and, finally, c. to ArbCom. It's not possible for an editor to first appeal a sanction before ArbCom and then to AE/AN and neither should it be possible to appeal to AN and then to AE or vice versa. Also, it's probably a good idea to say that an appeal to the community is not required to appeal to ArbCom, i.e. that a person can appeal directly to the committee, if he so wishes (which is technically known as an appeal per saltum), although, if we turn him down, he then can't go ask the community. Salvio ^{Let's talk about it!} 10:47, 17 November 2013 (UTC)

Yes, that is how I understand the appeals process, though I probably failed to make that clear in my previous post. My concern is that allowing three separate appeals is excessive, and also that allowing someone to appeal first to AN and then if that fails to ARBCOM raises the possibility of wikidrama between the community and ARBCOM in any situation where ARBCOM decides to overturn the result of the AN appeal. Gatoclass (talk) 11:09, 18 November 2013 (UTC)

Well, an appeal to the admin who imposed the restriction in the first place is not mandatory, but it should be included: every admin has the authority to undo his own actions and if he realises he made an error, to allow him to self-revert is IMHO the best option (it's a quick and unbureaucratic process and it also gives the opportunity to the imposing admin to demonstrate he's ready to change his mind and recognise he may have been wrong, which is a good thing).

Also, the appeal to ArbCom is necessary as well, in that admins imposing discretionary sanctions are exercising a power we delegated them and, so, we need to retain the authority to ascertain if said power is being exercised in a reasonable manner.

This leaves the two alternative appeals to the community and, as far as those are concerned, I guess it's a matter of personal preference: in my opinion, the community is already marginalised when it comes to DS and I'd rather we did not entirely exclude them from DS. If they somehow mess up, ArbCom may, as an extrema ratio, overturn their decision (just as we, theoretically, may lift a sanction imposed by them), which, it's true, may cause drama. Then again, most of what ArbCom does lately seems to cause drama... And I don't think that the risk of creating drama is so high as to really justify disfranchising the community when DS are concerned... Salvio ^{Let's talk about it!} 14:40, 19 November 2013 (UTC)

I agree that appellants should have the option of appealing to Arbcom, I just don't see why they should have a right to three appeals, firstly to the sanctioning admin, secondly to AN/AE, and thirdly to Arbcom - IMO, they should choose between AN, AE or Arbcom for their second appeal with no right to a third appeal if the second fails. All a third appeal is likely to achieve is to allow disruptive editors to create more wikidrama. Gatoclass (talk) 11:40, 28 November 2013 (UTC)

@Gatoclass: the first reason why we should grant three different venues is that appeals are not always merely in the interest of the sanctioned party; they can also be in the interest of the entire project. Assume that an editor was rightly restricted for disruption; he appeals to AN and the only people who show up are those who agree with him or have long worked with him and they all support overturning the sanction. In this case, it would be appropriate for the imposing admin or for the editor who started the original AE thread to appeal the community's decision and ask that ArbCom set it aside. But, also, we have to bear in mind that, sometimes, an AN discussion can degenerate into a mob and in that situation as well ArbCom needs to intervene to make sure that the appellant is treated fairly.

I agree that ArbCom need to take into consideration the fact that the community has already expressed an opinion and should be careful to only revert it when fairness demands it, but it still necessary IMHO to provide for such an eventuality. After all, ArbCom has the power to speedily decline to hear a case... Salvio ^{Let's talk about it!} 11:55, 3 December 2013 (UTC)

Well, I don't really see why users should be permitted to appeal to AN at all - IMO, they should be given a choice between AE and Arbcom and that's it. But, I've had my say on this and see no point in repeating myself, so I'll drop the matter now. Thankyou for taking the time to respond. Gatoclass (talk) 12:40, 3 December 2013 (UTC)

• Appeals do not seem proportionate, in that one person can ban an editor for a year, but you have to jump through hoops to appeal. --Iantresman (talk) 16:24, 17 November 2013 (UTC)

• And the sanctioning admin is expected to participate in this hoop-jumping by explaining their actions in public. That's a deterrent against frivolous or thoughtless sanctions, apart from the possibility of being sanctioned for such.  Sandstein  18:02, 17 November 2013 (UTC)

• A process which the banned editor is not part of. What form of justice excludes the condemned from taking part in their own defence? The principal of Habeas corpus is an important safeguard. --Iantresman (talk) 20:13, 17 November 2013 (UTC)

Ian, while much of what you say is technically true, it does not happen in practice. Part of that is because of the faith and trust given to admins to not be... Well, dicks. I don't think you will find an AE report where the admins didn't at least give the accused a right of responce. As for the solo rogue admin handing out bans, it is rare for an admin to not wait for a second admin opinion before acting. Usually even then it is because so few admins volunteer for the AE hassle. About the only one I can think of that comes close to the cowboy law you describe is Marshal20 being blocked for a month by Sandstien. Even that was overturned, by Sandstien, after a civil discussion on his talk page... Not some Kafkaesque appeal maze. I believe that your concerns, that I admit are technically true, are more alarmist than prudent. 204.101.237.139 (talk) 23:22, 17 November 2013 (UTC)

I can only go by my own DS case, where I feel this was not the case, which was applied to an article that was not under DS, and an article I was not even editing, and following to the letter, recommendations to "adopt Wikipedia’s communal approaches". This is why I'm trying to get clarification, and checks and balances.--Iantresman (talk) 00:02, 18 November 2013 (UTC)

I agree that the wording of discretionary sanctions could use some improvement with regard to scope; many DS cases refer to articles covered by the topic area rather than edits. IMO it should be made clearer that any edit pertaining to the topic area, whether on a related page or not, comes under the scope of DS. The existing wording has caused confusion in the past and will probably continue to do so until the problem is addressed. Gatoclass (talk) 11:01, 18 November 2013 (UTC)

That does indeed speak volumes about your motivation. Perhaps you should look into how other cases were handled from a less intimate view before tarring them all with the same brush. 204.101.237.139 (talk) 20:54, 20 November 2013 (UTC)

My motivation is to make sure that no-one else goes through the same thing. It has no bearing on my case which has been and gone. --Iantresman (talk) 22:04, 20 November 2013 (UTC)

• "is no longer an administrator and/or has relinquished the tools" means the same thing, does it not?  Sandstein  18:02, 17 November 2013 (UTC)
  • The wording does seem unnecessary—omit "and/or has relinquished the tools". Johnuniq (talk) 06:50, 18 November 2013 (UTC)
    • I think this is intended to address cases where the admin has voluntarily given up their tools not under a cloud. In the past, people in the circumstance of "not holding any bits, but could have them back with no trouble if they asked" have been seen to have a confused sort of "still basically an admin, just no buttons" status. So I would guess Arbcom is trying to make sure that people in that situation are covered under this wording even according to people who view them as still-admins. Possibly a better way to do this would be to say "does not currently have administrator tools" or something, which would encompass both people who have had their bits pulled and people who have given them up, temporarily or otherwise? A fluffernutter is a sandwich! (talk) 15:28, 19 November 2013 (UTC)
• The introduction "There are three possible stages for appeal" may cause confusion. Either mandate that an editor must use option 1 before 2, and 2 before 3, or change "stages" to "options". Clearly options 1 and 2 should be used before 3—perhaps say something like "There are three options for an appeal which should occur in the following order". Johnuniq (talk) 06:50, 18 November 2013 (UTC)
• AQFK has changed "where a consensus of arbitrators is required to overturn the decision" to "where a majority of active, non-recused of arbitrators is required to overturn the decision". Why? Is there a serious suggestion that three rogue arbs may arrange to hold a quick discussion and declare that User:VeryBadEditor is unbanned while the good arbs are asleep? There is no formal vote in options 1 or 2, so why does option 3 need one? Johnuniq (talk) 06:50, 18 November 2013 (UTC)

• This a bit of a technical point but given that the processes for appeal at AN are radically different from AE has closure of such AN appeals been thought through? For instance if at AN an appeal is granted by uninvolved editors and a sysop enacts that, but on review ArbCom determines that that action/appeal breached policy, well then who is accountable - the sysop who enacts consensus at AN or the !voters at AN?
  And there again if appeals are "granted" improperly at AN but no sysop will enact them - the the system implodes. For clarity we know that if an improper decision is made at AE by a sysop then that sysop carries the can.
  I mentioned this at the last draft, but it bears repetition, unless there is some level of structure & accountability at AN for editors granting appeals, comparable to that of sysops at AE, then the AN appeal system are in a limbo. As Sandstein points out above the accountability of sysops under AC/DS keeps things in order. If we eliminate accountability we open up a can of worms. And ultimately the person who suffers from that limbic state is the appellant. We also run the risk of facilitating "admin shopping".
  To be clear I'm not arguing against AN appeals I'm just suggesting we formalize it--Cailil ^talk 13:30, 19 November 2013 (UTC)
  • In my opinion, non-admins should not close anything requiring the use of tools, which means that IMHO a non-admin may decline an appeal (although he should only do so when it's entirely uncontroversial), but should never grant one. I also add that, as far as I'm concerned, an admin acting in good faith should not be sanctioned even if ArbCom later determines that he was wrong in his determination of consensus and, so, decides to overturn his decision... Salvio ^{Let's talk about it!} 01:06, 21 November 2013 (UTC)
• DS, under these guidelines, could result in an editor either being blocked or site banned for up to a year. Should some guidance be added for this, i.e., a blocked user may post an appeal on their talk page and request that someone copy it to the appropriate venue, and a site banned user should email BASC? Seraphimblade ^{Talk to me} 11:40, 29 November 2013 (UTC)
  • Yes. Though technically arbitration/AE appeals go to the full committee, not BASC (which only handles appeals of community sanctions).  Roger Davies ^talk 10:53, 2 December 2013 (UTC)

Appeals (grounds)

Reviewing this a bit more ... Given the volume of appeals, their sometimes directionless nature, and the desirability of reducing the committee's footprint, it would probably be a good idea to try to focus minds. Rather than just rehearing the whole thing again (which is what tends to happen), it might be better to have specific grounds. I suggest:

1. There was no actionable misconduct, or;
2. The sanction was not a reasonable exercise of administrative discretion.

The first covers situations where there was either (i) no misconduct at all or (ii) the misconduct was of such a minor or inconsequential nature that any sanction is excessive. The second covers administrative over-reaction or heavy-handedness. Thoughts?  Roger Davies ^talk 15:46, 2 December 2013 (UTC)

I agree that these are the possible grounds of appeal, and that it would be beneficial to request that appellants discuss them. However, that in and of itself is not likely to reduce the volume of appeals. To reduce the strain on the whole Committee, you could direct that appeals are heard by a subcommittee only.  Sandstein  16:40, 2 December 2013 (UTC)

I'm against limiting the cases where an editor may appeal. It's always impossible to anticipate all cases where we may find it necessary to intervene and binding our own hands in advance is always a bad idea. Not to mention that, in my opinion, the two grounds indicated here are so vague that they will not serve the purpose of limiting the number of appeals. Anyway, if we need to explicitly list the grounds of appeals we will accept, I think we should include:

• a. lack of competence on the part of the person who imposed the sanction (i.e. a sanction imposed by a non-admin or by a person who has relinquished the tools and hasn't requested them back),
• b. unreasonable exercise of administrative discretion (which includes the absence of actionable conduct, considering the presence of disruption is the logical prius which makes it possible for an admin to impose a sanction), and
• c. presence of formal defects (for instance, the sanctioned editor was not aware of the existence of discretionary sanctions).

While we're discussing this, I'd also add among the formal requirements that the sanctioned editor should generally have been given a reasonable opportunity to make a statement before being restricted. Salvio ^{Let's talk about it!} 12:16, 3 December 2013 (UTC)

Salvio, some passing thoughts: Your (b) seems to be the same as Roger's (2) ("not a reasonable exercise of discretion"). Your (c) seems to be the same as Roger's (1) ("no actionable misconduct"; the draft elsewhere defines misconduct as actionable only after the subject has been duly alerted). Your (a) is a good point. AGK [•] 12:52, 3 December 2013 (UTC)

They are similar, but, in my opinion, there is a small logical difference: I basically lump Roger's two grounds together, saying that, by definition, imposing a sanction in the absence of actionable misconduct is a form of abuse of discretion. A different case is the one where the editor in question has actually misbehaved, but was never warned that his actions could lead to discretionary sanctions. I see this as a different case, because, in this case, a sanction might be reasonable, but it cannot be imposed due to formal requirements.

The difference, the way I see it, is that, in cases under b., if another editor had behaved in the very same way after receiving an alert, a hypothetical sanction would still have been inappropriate; in cases under c., on the other hand, the sanction would have been ok.

Then again, if we have defined as "actionable" only that misconduct which has occurred after an alert, this case ends up included under b., but, in my opinion, c. should still be included for those cases where particular formal requirements are present and in the event we were to include the need to give the editor in question the chance of making a statement before being sanctioned. Salvio ^{Let's talk about it!} 13:14, 3 December 2013 (UTC)

IMO adding a valid grounds for appeal section would not be practical and would just add to instruction creep. Also, I'm concerned that the grounds suggested above necessitate fault finding with the sanctioning admin, which will be likely not only to discourage admin participation at AE, but also to discourage admin participation in appeals. The existing appeals process gives adjudicators plenty of flexibility, either to reconsider the facts altogether, to dismiss the appeal outright, or do anything in between, and to do so entirely on a "no fault" basis if they so choose, which doesn't seem like such a bad approach to me. Gatoclass (talk) 13:04, 3 December 2013 (UTC)

While I agree that it would be better not to have specific ground for appeals, I fear that all appeals entail our second-guessing an admin. Even if we have never formalised it, an appeal is generally accepted only when the imposing admin abused his discretion, so I'm not really sure spelling it out clearly would make much difference. This is different, of course, from the cases where we find that the sanction as originally imposed was a reasonable exercise of admin discretion, but now is no longer necessary (which is, technically, not an appeal, though the end result is basically identical)... Salvio ^{Let's talk about it!} 13:14, 3 December 2013 (UTC)

Well I don't know about that. What I do know is that it's currently possible to change or undo a sanction at appeal with no suggestion of "abuse" by the original administrator, and I'd like to keep it that way. Gatoclass (talk) 13:21, 3 December 2013 (UTC)

I agree with Gatoclass here, to a point. AE should not be about micromanaging or tweaking sanctions (I think it was Tim Canens or Tarc that made this point at AE a number of times in 2011). Either the sanction is justified or not. The quibbles over "I think it should be 3 months" vs "it should be 3 weeks" are not fruitful. However, if an active consensus at AE decides through an appeal that a sanction should be modified, for what ever reason (demonstrated reform on the part of the appellate, a decision to bring the sanction on User A into line with User B and C, or simply if AGF is being applied), bureaucracy should not get in the way. This reform process has gone a long way to remove legalistic thinking around alerts/warnings it would be a shame if this was forgotten here when such progress has been made. Basing appeals solely on sysop misconduct doesn't help anyone. It'll create a cadre of ppl crying admin abuse even more than they do already (because they'll have to to get their AE sanction overturned) and fundamentally it fails to AGF.

Fundamentally I'm asking why are these grounds necessary? Since we're dealing with discretion perhaps a paragraph on "AE is not for micromanaging other admins actions." There needs to be a balance between allowing for a new consensus to form vis-a-vis a sanction and not tweaking actions unless there is a real need (due to mistake or misconduct)--Cailil ^talk 13:50, 3 December 2013 (UTC)

The test applied to appeals was not always "Is this sanction a fair exercise of administrator discretion?", though that does seem the best one to apply. I think it was Sandstein that introduced this line of thinking, but I'm not sure. AGK [•] 11:24, 20 December 2013 (UTC)

A concrete example might be helpful here, and reading this discussion calls one to mind. At this request, which was granted, there was no contention that the sanction was inappropriate or unjustifiable (indeed, the sanctioned editor acknowledged that it was). Rather, they were just asking for the scope of the restriction to be reduced, since it included some portions that were irrelevant to their case and the areas they had behavior problems in. The appeal was granted, but that was not any finding of misconduct or wrongdoing on the sanctioning admin's part, it was just a reasonable request that there was no harm in granting. This restriction on the scope of appeals to a finding of wrongdoing in the original sanction would also essentially remove the latitude to impose indefinite bans that are not meant to be permanent, but are intended to last until the editor reforms his or her behavior and can successfully demonstrate that in an appeal, rather than just allowing them to wait it out. The appeals section already contains provisions that a frivolous or groundless appeal can result in a restriction from filing another appeal for up to six months, and I think that's sufficient to deter filing an appeal just to throw things at the wall. I don't see any reason to restrict editors to come back and say "I've abided by my restrictions for a year, I realize I was behaving in a way I shouldn't have, and haven't had any problems while editing elsewhere", and I think those are valid grounds on which to appeal. This proposal is well-intentioned, but I think it could have significant unintended consequences. Seraphimblade ^{Talk to me} 15:50, 4 December 2013 (UTC)

Well, technically, that's not an appeal: by its very nature, an appeal requires a sanctioned editor to question the validity of the restriction he is under. If an editor says "I accept that the sanction was reasonable then, but in my opinion it's no longer needed now", that's not an appeal and should be treated differently, meaning that we should not just examine whether the original sanction was a fair exercise of admin discretion, but rather whether or not it is still necessary, but, as I said, these are two different concepts. Salvio ^{Let's talk about it!} 11:39, 20 December 2013 (UTC)

Yes, to a point. Having seen the way this discussion has gone, I am not convinced that we need to specify grounds for appeal (or rather it seems that the disadvantages of doing so outweigh the advantages). As people will always talk about appeals when they mean reconsideration or clemency, easiest is probably to define appeal as including any request for reconsideration of the decision. Which has the advantage of simplicity.  Roger Davies ^talk 12:01, 20 December 2013 (UTC)

Based on the language there, I'm not sure that's the definition of "appeal" that's intended. Why would we limit the future restrictions to six months in that case? If "appeal" only means an assertion that a sanction is groundless, and the appeals process determines that in fact there were legitimate grounds for it, future "appeals" should be barred in all cases and indefinitely. A sanction which was applied for valid reasons won't become invalid six months from now. Currently, however, the term "appeal" means a request to have one's restrictions lifted for any reason, whether on the grounds that they were invalid to begin with or that they were valid but are no longer needed. As this is already the wide understanding of "appeal", I think defining it more narrowly would be unnecessarily confusing. Seraphimblade ^{Talk to me} 03:17, 22 December 2013 (UTC)

I've added a definition of Appeal that is really broad, to make it clear that for our purposes the DS meaning is much wider than the legal one.  Roger Davies ^talk 16:08, 9 January 2014 (UTC)

Involvedness of admins who have previously expressed an opinion about the sanction

In a recent appeal at WP:AE, two editors have expressed the opinion that an administrator who supported (or, presumably, opposed?) a sanction in the discussion about the enforcement request that resulted in the sanction should not be considered uninvolved in the discussion about a later appeal. I think that's pretty clearly incorrect: Per WP:INVOLVED, interactions with others in a purely administrative capacity do not constitute involvement. That applies to discussions about whether or not sanctions should be imposed, because such discussions are about the exercise of administrative functions. Nonetheless, in view of the opinions to the contrary, this should perhaps be clarified in the rules (as should the rather more vexing question of what an "involved editor" is). If administrators may discuss only either sanction proposals or later appeals, then both enforcement requests and appeals will tend to attract even fewer comments from administrators than they do now. In the interest of a well-functioning sanctions and appeals system, that is not desirable.  Sandstein  19:37, 12 January 2014 (UTC)

• I tend to agree. AGK [•] 21:03, 12 January 2014 (UTC)
• As it happens, I agree with those two editors (in my opinion, an appeal presupposes a new set of eyes, i.e. people who have not examined the case yet and who have not yet made up their minds, so allowing administrators who have already voiced their opinion during the original discussion to review the appeal for me basically contradicts the very idea of appeal); anyway, the last time we discussed this with my colleagues, I was the only one holding this opinion. Salvio ^{Let's talk about it!} 21:55, 12 January 2014 (UTC)

• In principle, I agree that this would indeed the better way to go about appeals, and I'd probably support it if there were consistently more administrators working at WP:AE. But it does not really reflect policy or practice (we don't, for example, exclude from WP:DRV editors who have already commented in the prior AfD). And as things are, it is not likely to produce uncontested, predictable results. If admins who comment about requests are excluded from appeals we run the chance of low-quality sanctions (nobody comments on requests and enforcing admins make errors that could have been avoided through prior discussion), or low-quality appeals (nobody comments on an appeal and it fails by default, or it is granted or rejected by a "consensus" of one or two admins). The problem of an "echo chamber" of a few admins all confirming on appeal what they already agreed on the first time around can be avoided by the appellant addressing their appeal to the Committee or to WP:AN, as provided for by this proposal.  Sandstein  22:17, 12 January 2014 (UTC)

• Another possibility is to have community-based appeals forum. This could be separate from AN/I and AE. A Quest For Knowledge (talk) 22:21, 12 January 2014 (UTC)

• Please, no more drama forums. No flawless system of governance has existed anywhere at anytime, and Wikipedia will never have a perfect administrative system where everyone is happy. If three or four respected editors raise a concern about a particular case at WP:AN, they are likely to get a reasonable hearing, and that is all that is required for exceptional cases. I have disagreed with the outcome of some AE cases, but in the end its purpose is to reduce drama, not to guarantee justice with unanimous agreement. Re "involved": As Sandstein says, it would be nice if completely fresh admins were available for appeals, but they are not. Further, many AE cases require some familiarity with the background in order to understand the issue, and that rules out most admins. Johnuniq (talk) 23:00, 12 January 2014 (UTC)

• How about creating a new role for editors willing to hear AE appeals? To gain such a role, editors would have to go through a similar process that admins do. A Quest For Knowledge (talk) 23:42, 13 January 2014 (UTC)

• @Salvio giuliano:, was last time with AC 2013, or AC 2014? NE Ent 00:43, 13 January 2014 (UTC)

• The 2013 committee. AGK [•] 11:43, 13 January 2014 (UTC)

The problem is that you then have a situation where the same group of admins who favored a sanction can repeatedly reject it without any fresh faces even getting a chance to review it or can simply block consensus. Standard practice at AE has been for admins involved in imposing a sanction, whether the admin of record or merely one of those who supported the sanctions, to be considered involved regarding an appeal. At any time you have at least a dozen admins regularly contributing at AE. If a case is decided by four admins and three are free to repeatedly reject any appeal or block any consensus then the appeal process might as well go through ArbCom or the community at large. Discretionary sanctions are a lower tier of sanctions than arbitration and community sanctions. The appeal process should be less difficult.--The Devil's Advocate _{tlk. cntrb.} 00:55, 13 January 2014 (UTC)

I'm not aware of that being the standard practice at AE. To me, it's clear that the admin who decided to impose the sanction should not help decide the appeal, but all other admins who may have commented about it do not "own" the sanction. Because DS are imposed by decision of an indiviual admin, rather than by consensus, the other admins are not themselves responsible for the sanction and, as such, are not too involved to review it on appeal. Deciding otherwise would also open up immense new wikilawyering opportunities: How about an admin who, in a discussion, expresses weak support for a one-week block, but isn't convinced that a topic ban is needed – and another admin then imposes a two-week block and a topic ban? Does that make the first admin involved? Discussing these and similar issues in each and every case would add yet another layer of complexity and delay to appeals.  Sandstein  12:48, 13 January 2014 (UTC)

Would an admin who voiced support for a sanction imposed during in an AN/ANI discussion be acceptable as someone to close any appeal of said sanction? Come on, this is basic. I would say that goes for opposition as well. An admin who argued against a sanction should not comment as an uninvolved admin either. Editors formally appealing sanctions that were not imposed by ArbCom or the community are expecting an independent review and that is what they should get. To have the whole matter decided by the people who decided on the original case denies them the opportunity of an independent review.--The Devil's Advocate _{tlk. cntrb.} 23:32, 13 January 2014 (UTC)

I'm not getting the context here -- is this discussion about a sanction being imposed following discussion at WP:AE being appealed at WP:AE? NE Ent 12:58, 13 January 2014 (UTC)

Currently there is an appeal at AE where Sandstein weighed in as "uninvolved" and it seems now another admin who had taken part in the original discussion has taken his lead.--The Devil's Advocate _{tlk. cntrb.} 23:32, 13 January 2014 (UTC)

• Let's take a step back for a minute. Presumably, the whole point of the discretionary sanctions/AE approach is to enable admins to more efficiently and effectively handle disruption in problem areas. By creating a bunch of novel procedural hoops and wickets—which don't exist elsewhere on Wikipedia—we've actually made the AE process more cumbersome than standard administrative processes. In our quest to create a fully equipped virtual legal system (without the necessary personnel or support), we're undercutting the whole point of discretionary sanctions and Arb Enforcement.

  Again: discretionary sanctions are intended to streamline administrative handling of problem areas, but instead they immensely complicate things. Making matters worse, these excessively litigious technicalities provide positive feedback to disruptive editors at AE, many of whom have been sanctioned precisely because of their efforts to exploit technicalities and game the system. Personally, if this is where things are headed, I intend to cease participating in AE/discretionary sanctions. I'll simply apply whatever steps are appropriate as standard administrative actions, and they can stand or fall as such, with significantly fewer avenues for vexatious wikilawyering. MastCell ^Talk 17:08, 14 January 2014 (UTC)

• As a person who is normally quite at ease with complicated legalistic ways to go about things... I agree with the above. One should not forget that discretionary sanctions were originally intended as a fast-track procedure, and yet every year seems to bring more complications and venues for delay and appeal. (Of course, the general trend towards instruction creep is by far not limited to this area.)  Sandstein  17:17, 14 January 2014 (UTC)
• I would hope the purpose of discretionary sanctions would be to improve the encyclopedia. The easiest "streamlining" would be to allow any admin to ban any editor at will. We'd have no disruption in no time. Actually I've queried AE "talk" page for background info on history / purpose, in case any knowledgeable readers would like to share the wealth. NE Ent 02:14, 15 January 2014 (UTC)

• Yes, the purpose of discretionary sanctions is to improve the encyclopedia. Specifically, they improve the encyclopedia by making it easier to deal with areas that are chronically plagued by poor editorial behavior. At least, that's the intent. MastCell ^Talk 05:05, 15 January 2014 (UTC)

The last two comments illustrate the point I've made before. Nothing this complicated, legalistic and time-consuming can possibly be a good idea for Wikipedia. There is a better way. Get rid of discretionary sanctions entirely, and institute page blocks in areas where editing becomes unduly contentious. It's fast, efficient, effective, not legalistic, saves administrative time, and generally makes everyone happier because there is no feeling of individual 'injustice' when a page is blocked -- the feeling of 'injustice' in that case hits all editors equally, and essentially doesn't hurt them at all. A few days or weeks in which a particular page can't be edited doesn't ultimately hurt anyone. Most of the pages in question already say all that needs to be said, and what's going on is just POV-pushing. A few days or weeks of interruption of POV-pushing isn't a bad thing. NinaGreen (talk) 17:46, 14 January 2014 (UTC)

I should also add that the foregoing suggestion would free up administrator time, which could profitably be used in mediating the contentious issues which arise on a daily basis and result in the loss of good Wikipedia editors and which have nothing to do with DS issues. NinaGreen (talk) 17:59, 14 January 2014 (UTC)

The problem with the "page blocks" solution is it encourages edit warring to attempt to make sure that the "right version" is the one on which the block is imposed. Rather than helping to solve disputes it is like pouring petrol onto a fire. One then has the situation were both sides resolve nothing and just wait until the unblock to impose their preferred version. Removing those who are the most intransigent over content tends to allow the rest to work towards a compromise. Some of these issues are very complicated but a simple one for disinterested editor to understand easily is the Liancourt Rocks issue, where at the last requested move we had Korean newspapers encouraging their readers to sign up to Wikipedia and express their opinions. If it likely that a page block would work for an issue such as this? -- PBS (talk) 13:47, 19 January 2014 (UTC)

• @MastCell. My direct experience with WP:AC/DS, suggests that "streamline" means "taking shortcuts", "not having to present the evidence", "not having to follow the procedures". As a result, there is little due diligence and little due process. This is further compounded by some of my basic questions regarding WP:AC/DS, still remaining unanswered. I am all for streamlining and much appreciate all the hard work put in by Roger Davies, but in my opinion, all the extra work demonstrates that the process if flawed. I have many suggestions to improve the current system, making it quicker, fairer, and requiring less attention from Arbitrators, and requires no additional rules, regulations or special sanctions. --Iantresman (talk) 19:41, 14 January 2014 (UTC)
• @MastCell(2). I agree with you, "standard administrative actions" should be sufficient. The reason they don't work is because some admins don't under the process, don't understand evidence, and don't know how to investigate primary sources (of evidence). A lot of them are also biased, and apply double standards to different editors depending on what side of the fence they are editing on. I could provide hard evidence (diffs) supporting every single allegation here, but as is usual, they will be ignored (another problem with the system). If dispute resolution is not working properly, I vote we fix it, rather than bring in military rule. --Iantresman (talk) 19:53, 14 January 2014 (UTC)
  • • @MastCell I entirely agree with Iantresman. Some admins have little or no grasp of what evidence is, and some seem to me to have an understanding of the word "disruption" which amounts almost to "disagreeing too often with users I respect". While I agree with Iantresman that standard administrative actions should be enough, no registered user here should have to suffer the imposition of severe penalties (I am thinking of an indefinite block, or a long or indefinite topic ban; the meaning of "severe" would need to be defined) by an admin who has little grasp of how to assess the evidence of whether there has been disruption. In the English criminal courts, magistrates give summary justice, but they can only impose short sentences, while only a higher court, sitting with a jury, can decide cases which call for longer sentences. In the same way, I would give all admins here the power to administer short blocks and lesser topic bans, but have a cadre of more judicious admins (or they might have another title and include non-admins) who would take decisions on serious allegations. In any event, whatever system we have here, there needs to be some kind of appeal available against heavy penalties. I find it self-evidently flawed for an appeal to be decided by one or more people who were involved in the case at an earlier stage. Goodness only knows how that can be seen as fair. Moonraker (talk) 03:30, 16 January 2014 (UTC)
      • @Moonraker: I understand what you're saying. Really, I do, and I think your concerns are legitimate. Do you understand my point? We don't have the resources to create the sort of complex administrative and judicial structure proposed here. We don't even have the resources to provide consistent basic oversight of our most problematic and high-profile topics. We have to find a way to work with what we have, and use it to build a better encyclopedia.

        As a separate point, if we create a system which rewards endless disputation and litigation, then we'll attract and retain people who get off on endless disputation (and we'll lose people who just want to write an encyclopedia and who are put off by endless rounds of technicality-driven litigation). In fact, this is exactly what's happened over time. MastCell ^Talk 17:14, 16 January 2014 (UTC)

        • @MastCell I think WP:AC/DS is an example of "the sort of complex administrative and judicial structure proposed here". Regular dispute resolution and admin powers are sufficient, if they are used correctly, but in my opinion, are often used improperly. I don't think there is a need for more resources, I think many editors need to a better understanding and clarification of policy.

I still find personal attacks used frequently, and as an argument against an editor. It is often easy to identify, but editors are rarely criticised for it. I am currently witnessing an editor being banned for being a "single purpose account", despite this not being a sanction-able offence, and no evidence of poor behavior being offered. I myself was banned from an article using WP:AC/DS, that wasn't even under discretionary sanctions, and which I wasn't even editing, having chosen to discuss the matter first. The system is broken, and I don't think WP:AC/DS is the way to solve it.--Iantresman (talk) 22:19, 16 January 2014 (UTC)

• In principle, an AE thread isn't even required for discretionary sanctions to be imposed. An admin who saw an editor violating a discretionary sanctions remedy could unilaterally and immediately sanction that editor, and provided that there was indeed such misconduct, would have done nothing wrong. AE is a way to bring the matter to the attention of admins and allow for discussion of areas that may be more grey, but ultimately, the admin who applies the sanction "owns" it and is solely responsible for doing so. In any case, though, we've always held that interaction with an editor or area in a purely administrative capacity is not impermissible admin involvement for purposes of future issues involving the same thing. Discussing a possible sanction or remedy is pure administrative involvement, and so it is not a bar to the admin participating in future discussions as well. And all that aside, if the editor still feels wrongdoing has occurred, they may make a final appeal directly to ArbCom. Discretionary sanctions are meant to be a way that editors who are misbehaving in sensitive areas can quickly be stopped from doing so. There is still a multi-layer appeals process in case someone feels a genuinely wrong decision has been taken. Seraphimblade ^{Talk to me} 06:47, 16 January 2014 (UTC)

• What you are doing is wikilawyering. All that about being involved purely in an administrative capacity applies to actions regarding a topic area or editor generally, not specifically the issue of reviewing appeals. It is fairly clear that WP:INVOLVED is meant to apply when someone is contesting those administrative actions otherwise a blocking admin could reject unblock requests and the sanctioning admin could comment as uninvolved. You are suggesting a sort of loophole for admins where you just need to get a few admins to back your play and they can then shut down any appeal without further outside input. To say "well they can always take it to ArbCom" is pig-headed and corrupt reasoning. Editors should not feel like they have no choice but to go to ArbCom because a small group of admins who sanctioned them refuse to let them have their appeal heard by other admins.--The Devil's Advocate _{tlk. cntrb.} 18:24, 16 January 2014 (UTC)

• (in reply to a little above) @Iantresman: Why don't you list your ideas for improving the DS remedy, in a few short sentences? Under this bullet point would be best. I fear your proposals have gotten lost in all the talk (it's a huge page, after all). Thanks, AGK [•] 13:00, 18 January 2014 (UTC)

Suggestions for improvements

I am more concerned with improving standard policy, so that we don't need discretionary sanctions. But here goes: WP:CIVIL to be enforced. Today it is mostly ignored. For example, having my competency called into question is (a) fallacious (b) non-constructive (c) says nothing about why edits are problematic (d) is incivil. WP:TALK#FACTS Editors should present sources. If editors dealt more with verifiable sources per WP:RS and WP:V, there would be less inadvertent original research and synthesis per WP:OR and WP:SYNTH. This would reduce editors expressing their opinions. WP:ASPERSIONS. Editors MUST support criticisms of behavior with specific diffs. Diffs pointing to further aspersions are not admissible. Allegations MUST be struck through, if a request for diffs is not met Inadmissible criticism. Many behavioral criticisms are not actually sanctionable offences, and should not be admissible. A Single-purpose account in itself is not an issue. Likewise WP:PUSH: you're supposed to be civil, and there other policies that deal with inappropriate weight and sourcing. Editors who still think that NPOV is an actual point of view, rather than an editing style for describing POVs. Automatic escalating scale of bans. There is currently no consistency, and we shouldn't have to go through the appeal process. Bans should escalate 1, 3 days, 1, 2 weeks, 1, 3, 6, 12 months (reoccurring). No indefinite bans. Scale resets to zero after 12 months trouble-free editing. Editors must distinguish "no consensus" (no problem) with "against consensus" (potentially problematic) An allegation of a conflict of interest must not conflicts with WP:PRIV and WP:HARASS. eg. if you discover that someone is a vegetarian, you can not claim their orientation is a potential COI in relevant articles. If an editor is reported for violating a policy, investigation and sanctions must be undertaken. There must be no exceptions for "valued editors", nor claims of provocation by another editor. To improve accountability Admins/Arbitrators MUST respond to reasonable requests, especially requests for evidence (diffs) Arbitrators (a) to be split into teams of three, to enable faster processing of cases (b) One Arbitrator team to be responsible for appeals only, to ensure no conflicts of interest.--Iantresman (talk) 13:01, 19 January 2014 (UTC)

• I have said it before and I'll say it again here. The culture where the admin is always right and the editor has no rights is represented in the attitude that an involved admin isn't involved when enforcing policy against an editor who doesn't share the same view is flat out wrong. Until that changes, and the Arbcom takes the problem of admins abusing the tools seriously, this review is going to amount to a waste of time. Some admins make better decisions than others just like editors. Some admins start off good and let it get to their heads...they should have the tools removed. Some admins should have never been given the tools and after repeatedly abusing them still have them. Some admins never use them and use them as nothing more than a status symbol. For all the talk here in this review people are saying they want to make it more fair. Yet, in this entire discussion of review, after seeing repeatedly how to deal with editors who violate sanctions I have seen a word of talk about what would happens to an abusive admin, what happens when an admin violates the rules. If you want to make reform hold admins to the same or a higher standard than editors Stop letting them keep the tools when they violate policy. Stop making it so difficult to remove the tools from an abusive admin that editors simply give up and leave in frustration. Give someone (setup a committee or appoint group like the Bureaucrats) the job of policing the admins like the admins police the editors. Obvious abuse shouldn't have to be a multi month arbcom case. Just reforming the system to make it easier for admins to ban editors is not the answer and gives the impression that this is nothing more than a power grab to give more power to a few admins and arbs. Whether that's true or not doesn't matter. That is a perception that exists. Kumioko (talk) 15:15, 19 January 2014 (UTC)

• When admin is involved in dispute, sanctioning admins always take side of the involved admin. One involved admin behaved like editor, then threatned me in capacity of admin, then went to ArbCom to sanction me and I was sanctioned mercilessly by admins.[11] Wikipedia has become police state where there is no controlling authority to monitor admins. Although sanction period is expired 2 and half months ago, I do not edit that disputed article because that involved admin is still there and no matter what edit I do, he is going to again complain against me and I will be sanctioned. This DS is hopeless and disgusting procedure because of involved admins. Abhi (talk) 15:46, 19 January 2014 (UTC)

Sadly, this is my experience too. Accountability seems to be lacking. --Iantresman (talk) 16:06, 19 January 2014 (UTC)

Mine as well which is why I believe if the Arbcom is serious about reform they need to include some type of liability on the part of the admins. They need to be held accountable for making bad decisions and that is current lacking. Kumioko (talk) 16:33, 19 January 2014 (UTC)

Continuity (comments)

Draft text

• Apart from "full force" > "force", not a lot,  Roger Davies ^talk 23:21, 16 November 2013 (UTC)

Proposals and brainstorming

General comments from Harry Mitchell

I've been involved at AE, on and off, for several years. I seem to be a little late to this party, but I wanted to make some comments and suggestions (on arbitration enforcement/discretionary sanctions in general, rather than the draft text), both for the consideration of the committee, and for members of the wider community who do not closely follow AE.

Comments/observations on AE and discretionary sanctins as they are currently:

• The first thing I feel I should state, in the light of some of the comments I've read here and elsewhere, is that the vast majority of actions taken under the provisions of discretionary sanctions are entirely uncontroversial and usually irrefutable. The sanctions are usually for conduct that any reasonable person would think fell well below the level of decorum expected of encyclopaedia editors, much less editors working in toxic and controversial topic areas. In >90% of cases, no objective person could look at the facts and conclude that the admin acted unreasonably.
• Although any single admin has the power to enforce arbitration remedies (including imposing discretionary sanctions in the areas where they're authorised) unilaterally, my experience is that most violations are reported to WP:AE, because it's the best place to find admins familiar with the procedures. Thus, most enforcement requests get the attention of at least two or three admins, even if the result is uncontroversial, and several more will likely see it but not comment. AE is of course a public noticeboard. While it's not a venue for threaded discussion, any editor can comment on requests. I for one would welcome concise, clueful comments from uninvolved editors, especially those with no strong feelings about the topic area in question.
• Contrary to comments about admins having sweeping powers, we only enforce arbitration remedies. Although in the case of discretionary sanctions, we have close to carte blanche to decide on the exact method of enforcement (if enforcement action is necessary—"discretionary" also means that we have the option of taking no action or of a warning or something else intangible), we don't just make up the rules as we go along, and we can't sanction people willy nilly. In fact, our hands are tied by ArbCom much more tightly than is commonly believed. If you're unlucky enough to be sanctioned at AE, you would have to have done something that falls well below the standard of conduct expected of editors in controversial topic areas, after being advised of the expected level of conduct (or, if this draft becomes policy, clearly demonstrating that you are aware of it), and most likely after seeing edit-notices and talk page banners advising you of the discretionary sanctions, and probably after being advised that your conduct did not meet that standard. I'm an empathetic person, but I'd struggle to believe that that was just bad luck.
• Others have (correctly) observed that AE is dominated by a small number of admins. This is probably a result of a combination factors, including the toxicity that surrounds arbitration, the perniciousness of the topic areas subject to the sanctions, and the bureaucratic procedures involved with taking arbitration enforcement actions. This is not ideal, but it's not a crisis—some admins are more active than others, some comment on issues to do with particular topic areas, much like any other areas; of course, more objective opinions (from admins and non-admins alike) would be very welcome.
• There is a lack of critical thinking at AE. Many of the editors we deal with at AE are not bad people; they're not trolls or vandals. For the most part, although their actions are disruptive, it is not their intention to damage the encyclopaedia, and AE admins should, wherever possible, seek ways of limiting their disruption while allowing them to continue making constructive edits. AE admins too often (in my opinion) fall back on blunt instruments which either kick the can down the road (such as short-term blocks) or give the editor no chance to redeem themselves or make constructive edits in less controversial areas (such as broad topic bans or lengthy blocks).

Suggestions/proposals for improvements of AE/discretionary sanctions:

• Scrutiny of AE should be welcomed. It tends to be a mostly forgotten corner of the project space, but we (the admins who work there) should welcome comments from a broader range of editors, including non-admins. Extra objective opinions might not change anything, but people can have more confidence in the system if they feel they have a stake in it, and we may get some useful suggestions or outside observations which make the process work better.
• Add a clause to the standard motions/remedies authorising discretionary sanctions and to the log to encourage admins to use the AE noticeboard (and encourage parties to use it rather than approaching an admin directly) rather than acting unilaterally, especially when considering long-term sanctions or actions that might be controversial.
• Simplify the appeals process, and require admins invoking discretionary sanctions to inform the sanctioned editor of their right to appeal (to AE and/or ArbCom). Perhaps appeals of blocks could be conducted on the blocked user's talk page and transcluded to WP:AE? Regardless, it should be quick and simple to appeal arbitration enforcement actions, even for relatively inexperienced editors, and even while blocked.
• Arbitrators, especially the longer-serving arbs (and possibly some of the longer-serving AE admins), should approach experienced admins with a good track record of level-headedness and ask them to consider weighing in at AE from time to time. Hopefully this would deepen the pool of admins who handle AE requests and possibly bring in some fresh thinking and different perspectives.
• Arbitrator participation a AE in an advisory role. Where something contentious or novel is being proposed, especially when the editor in question or their supporters are claiming that the scenario was no one foreseen by ArbCom during the relevant case, it may be useful to hear from an arbitrator. This could be either an individual arb's perception of the case or one arb speaking for the committee.
• Periodical review of discretionary sanctions. Every few years, ArbCom (or perhaps an ad hoc body appointed by them) should review discretionary sanctions in a given topic area to evaluate whether i) they are having the intended effect, ii) they are still necessary in that topic area, and iii) whether admins' enforcement of them is in line with achieving the aim the committee had in mind when authorising them.

—HJ Mitchell | Penny for your thoughts? 16:35, 31 December 2013 (UTC)

Amended. HJ Mitchell | Penny for your thoughts? 16:42, 31 December 2013 (UTC)

HJ: Thank you very much for your submission. This will usefully inform the decisions we need to make at the end of this review. AGK [•] 12:15, 6 January 2014 (UTC)

Discussion of use of talk page

Refactoring this talk page

• Revision as of 19:51, 17 January 2014 user:PBS (revert reactoring of the section I initiated. user:Roger Davies If I had wanted to place it higher up the page I would have done so; Per refactoring guideline please do not do it again.)
• Revision as of 15:53, 18 January 2014 user:AGK (→‎Issuing alerts (Comments #2): move section back up to 'Issuing alerts'; do not move it back again -- it is not you who has to wade through all this at the end)]

Anthony no-one has to "wade through all this at the end" unless they want to. However there is a rule against re-factoring text on talk pages when an editor has specifically requested that it is not done. Please put the section back where I placed it. My placing was deliberated, because I know very well that people go to the last section on a talk page for current views, and I would appreciate it if you would revert your move. If I was not assuming good will I would be tempted to assume that you move for nefarious reasons, so to show that is not the case please revert your move. -- PBS (talk) 19:14, 18 January 2014 (UTC)

@PBS: In fact we (the arbitrators) will have to wade through this consultation at the end. The arbitrator charged with the subject of alerts will be looking at the alerts section above. They shouldn't have to jump between the archived and current alerts discussion section and your special section at the bottom of the page; discussion should be centralised. The section therefore needs to stay with all the other sections discussing alerts. AGK [•] 19:21, 18 January 2014 (UTC)

On the 'rule against re-factoring', the arbitration policy would take precedence: we are responsible for the "administration of arbitration cases and management of all the Committee's pages and subpages". AGK [•] 19:23, 18 January 2014 (UTC)

This is a discussion about the policy, not the policy. If it is the specific section title that worries you then as a compromise rename the section something like "PBS's comments on issuing alerts" -- PBS (talk) 19:31, 18 January 2014 (UTC)

(Is your first sentence a typo? I don't understand it, sorry.) Nobody wants to watch and read several places at once so that, as you say, there is a larger audience for your views. Changing the section title still does not centralise the discussion, and therefore is not a solution. Unless another arbitrator disagrees, the section can stay next to the other alert threads. Thanks, AGK [•] 19:39, 18 January 2014 (UTC)

AGK and PBS, please stop contributing like children here. NE Ent 19:50, 18 January 2014 (UTC)

No it was no typo you wrote "On the 'rule against re-factoring', the arbitration policy would take precedence" I replied this is a discussion about the policy, not the policy (so there is no "arbitration policy" to take precedence). the reason for the introduction of the re-factoring rule was precisely for circumstances like this so that the primary discussion did not get sidetracked. It seems odd to me to be discussing a draft that includes "follow editorial and behavioural best practice" and then have someone says that best practice should be ignored because they know best. But as you have stated that I can take it or lump it, I will say no more. -- PBS (talk) 20:11, 18 January 2014 (UTC)

Clear and substantial consensus

The phrase "clear and substantial consensus" is used twice in the draft in two sections (Reducing or overturning sanctions and Appeals). "What precisely does it mean? As the WP:CONSENSUS policy contains no such wording, I think that if it is defined elsewhere in the policies of guidelines then a link should be provided or if not and it is retained then a precise definition ought to be placed in the definitions section. If neither of these options are desirable then it should be replaced with a term that is defined in the policies and guidelines. -- PBS (talk) 12:46, 19 January 2014 (UTC)

See I know it when I see it NewsAndEventsGuy (talk) 13:22, 19 January 2014 (UTC)

But see WP:The duck test, where it becomes of concern if you are The Goose That Laid the Golden Eggs. --Iantresman (talk) 14:50, 19 January 2014 (UTC)

@NewsAndEventsGuy: do you think that the definitions section is not needed? For example "Committee" is defined and that is far more self-explanatory than "clear and substantial consensus". This is a multi-ethnic multi-cultural site, so why would you assume that "I know it when I see it" is an appropriate retort to "what precisely does it mean?"? -- PBS (talk) 12:13, 20 January 2014 (UTC)