Wikipedia talk:Emptying categories out of process

Request for feedback

@Marcocapelle: as you are the editor I work with most closely at WP:CfD, I would like to ask if you've got any feedback for this essay.

It's the result of me being a bit frustrated when at Wikipedia:Categories for discussion/Log/2023 July 15#Category:Wikipedia essays about controversial topics (which you and I both participated in), someone emptied the category out of process.

I asked a question about it at Wikipedia talk:Categories for discussion#Emptying nominated categories.

There, I was made aware of Wikipedia talk:Categories for discussion/Archive 19#Where does it actually say you should not just empty a category you don't like?, an informal 2020 discussion leading to a failed informal proposal and vote to establish a formal guideline on it as part of WP:CAT.

I concluded I didn't want to start a whole RfC about it, at least not yet, and write this essay as an informal start to establish some basic rules and agreements.

I anonymised the case studies, but did link to relevant pages so that everyone can read them for themselves, draw their own conclusions, and see if I summarised things fairly.

I asked closer Qwerfjkl's opinion, who said: I suggest you ask the CfD regulars, because I mostly close discussions. That said, it looks good to me.

So that's why I'm asking you now. What do you think of it in general? And can you see places where corrections, additions or omissions may be needed or helpful? I may ask other people's feedback later, but first I'd like to hear your opinion. Cheers, Nederlandse Leeuw (talk) 12:03, 18 July 2023 (UTC)

I wouldn't advocate calling every form of emptying disruptive, not the least because I have occasionally removed articles during discussion, haha. If this would convert to changing the text of the guideline I would propose something like: "If an editor plans to nominate a category for CfD or if a category is already discussed on CfD, this editor may remove articles from that category if and only if (1) removal is expected to be uncontroversial and (2) if they disclose at CfD which articles they removed and why. Not disclosing this is considered disruptive. The removals may be reverted during the CfD discussion if there is no clear consensus about it." Marcocapelle (talk) 17:29, 18 July 2023 (UTC)

@Marcocapelle Thanks! Certainly not every form of emptying is disruptive, but some forms are. I've adopted some of your wordings for the lead section, as well as that emptying may just be a mistake. I've also hatted the precedents and case studies to not make the essay needlessly difficult to navigate, and for a better protection of the identities of the people involved. It's not about those people, it's about examining what has happened in the past, and what we can learn from it for the future. Do you see any other improvements? Cheers, Nederlandse Leeuw (talk) 17:52, 18 July 2023 (UTC)

@Johnbod: Amicable reminder that you've been invited to give feedback on this essay: compliments or constructive criticism, suggestions for corrections, additions or omissions. As the editor who proposed a new guideline in July 2020, you've already made signficant efforts in this area which are appreciated. Cheers, Nederlandse Leeuw (talk) 06:53, 20 July 2023 (UTC)

I said at Cfd talk: "I've been complaining about this for years, and while the essay is fine as far as it goes, really we need a clear determination that this is wrong, so that the language in it can be made less tentative, and it can be given a more official status. I'd completely forgotten that I started Wikipedia talk:Categories for discussion/Archive 19#Where does it actually say you should not just empty a category you don't like? back in the day - it's a great pity the draft there wasn't adopted. Johnbod (talk) 16:28, 18 July 2023 (UTC)" - I'm a bity busy now but will take a closer look soon, I hope. Johnbod (talk) 18:58, 20 July 2023 (UTC)

Thanks for copying your comment. It is indeed a pity that your proposal was not adopted, although I understand some of the objections raised against certain wordings in it. For example, as several colleagues pointed out, the phrase clearly meeting the definition is not very helpful, because what may be clear to user A may not be clear to user B. The Roman Walls case study shows that what was "clear" and "obvious" to one user disagreed with community consensus.

This is why I wrote: You should explain your thinking, preferably in a nominator's rationale at CFD for why you think the category is inappropriate and should be deleted, or in exceptional cases perhaps in a disclosure for why you removed (almost) all items out of process. A lot of conflicts stem from the fact that people don't explain what they're doing and why. They just assume a priori that their point of view is a fact, and that it is (or should be) "clear" or "obvious" to everyone, and that anyone who does not understand or disagrees is a bit of a fool. However, in many cases, they themselves are actually wrong, but they can't or refuse to see it, nor acknowledge it.

What we need more of is people explaining their thinking, and fewer instances of people a priori asserting that they're right because it is "clear" and "obvious" (to them), and anyone who thinks otherwise is silly or stupid. Because that's just not true. And it breeds pointless interpersonal tensions and disruptive editing, but doesn't really get us anywhere in solving the categorisation issue. Nederlandse Leeuw (talk) 21:53, 20 July 2023 (UTC)

Maybe, but I still think a change at Cfd is needed to have real effect; an essay is too easily dismissed. You give examples of sanctions that may be applied - but are they ever in reality? I'm sure many cases of emptying and leaving the category for someone else to delete go completely un-noticed. Do you say that if a category where a deletion/merge CFD proposal has failed, this is especially reprehensible. I'm sure this happens too. Johnbod (talk) 03:14, 3 August 2023 (UTC)

@Johnbod Yes, sanctions are sometimes applied, see the "Sanction precedents" section. Unfortunately I haven't found many yet (4 so far), but that doesn't mean there haven't been any others (if you know any, please post them here). Plus, the case studies provide many examples which are sanctionable if the same editor repeatedly emptied categories out of process (without disclosure), and ignored warnings not to do so. Admins are also liable for deleting and especially undeleting categories out of process.

That ECOOPing often goes unnoticed is unfortunately true. But there is a script with which we can detect it: User:Nardog/CatChangesViewer.js. I haven't done that yet (I'm not sure how it works haha), but for CfD regulars this should be a handy tool to detect ECOOPing. Cheers, Nederlandse Leeuw (talk) 14:52, 3 August 2023 (UTC)

PS: Okay I've installed the CatChangesViewer script. It appears to work! Nederlandse Leeuw (talk) 15:05, 3 August 2023 (UTC)

@Nederlandse Leeuw, I might write some code to automatically check C1ed categories for recent emptying. — Qwerfjkltalk 20:52, 5 August 2023 (UTC)

@Liz: Would you also like to give feedback? You are the admin who deals with emptying out of process the most (like a few minutes ago). I would really appreciate any suggestions for improvement you could offer me. You know better than anyone else how persistent this issue is, and I hope that this essay may help reduce your workload and that of fellow admins and editors in the future. Cheers, Nederlandse Leeuw (talk) 01:23, 3 August 2023 (UTC)

SPEEDY policy 2005–2007

I've just written the new section SPEEDY policy 2005–2007. This was the result of documenting the 2006 case study on Forza Italia politicians. I'd like to thank @Qwerfjkl and especially @Redrose64 for helping me find this out.

So it turns out that between 7 October 2005 – 9 August 2007, there was a very explicit provision in the WP:SPEEDY policy, saying:

• (early text; 7–23 October 2005) Empty categories (no articles or subcategories) whose only content has consisted of links to parent categories. Be sure to check the history of the category. If it has a non-trivial history or isn't relatively new, it is likely that the category once did contain articles, and deeper investigation is needed before deleting the category to make sure that it wasn't emptied just to bypass WP:CFD.
• (middle text; 23 October 2005 – 25 November 2006) Empty categories (no articles or subcategories for at least 72 hours) whose only content has consisted of links to parent categories. Be sure to check the history of the category. If it has a non-trivial history or isn't relatively new, it is likely that the category once did contain articles, and deeper investigation is needed before deleting the category to make sure that it wasn't emptied just to bypass WP:CFD. This does NOT apply to categories listed on WP:CFD, as the discussion may merit renaming or prompt population.
• (late text; 25 November 2006 – 9 August 2007) Empty categories (no articles or subcategories for at least four days) whose only content has consisted of links to parent categories. This does NOT apply to categories being discussed on WP:CFD. If the category isn't relatively new, it possibly contained articles earlier, and deeper investigation is needed.

I think the 9 August 2007 removal of the "deeper investigation" sentence was an error that the editor in question was not aware of. They certainly had no consensus to do so, and ignored the consensus established between 4 other Wikipedians in October 2005 to include that sentence (originally 2 sentences) into the WP:SPEEDY policy. I think some version of the early, middle or late text (or perhaps a customised mix?) should be restored to WP:SPEEDY.

Likewise, I am in favour of the restoration of some form of the CFD:HOWTO policy September 2004 – July 2006, which stated:

• Unless the category to be deleted is non-controversial – vandalism or a duplicate, for example – please do not depopulate the category (remove the tags from articles) before the community has made a decision.

That, too, appears to have been accidentally removed from the WP:CFD#HOWTO policy, on 24 July 2006. Without consensus, overriding previous implicit consensus (customary law) since September 2004.

Restoring these two provisions to official policy, where they appear to have been accidentally removed from, could do a great deal of good in preventing and addressing the still-widespread problem of emptying categories out of process. Cheers, Nederlandse Leeuw (talk) 16:54, 5 August 2023 (UTC)