Break 1 re ONUS: add text re both pro and con must support positions?

@Rambling Rambler, I suggest you spend less time trying to devine my intentions. If you proceed regardless of my suggestion, please consider Hanlon's razor. - Butwhatdoiknow (talk) 06:36, 9 February 2026 (UTC)

The reason people try to divine intentions is because sometimes, knowing the goal can help you think of a different, and possibly even better, way to achieve that goal. But if you can't figure out what the goal is, and the proposed solution doesn't seem like a good one, all you can do is say "No". WhatamIdoing (talk) 19:43, 9 February 2026 (UTC)

If helping is the goal then perhaps people should ask "what is your intention?" instead of using divination - particularly when that divination leads you to negative conclusions about another editor (such as "isn't great," "come across poorly," and "a years-long campaign . . . in the face of repeated dismissals"). - Butwhatdoiknow (talk) 21:35, 9 February 2026 (UTC)

Discussion is the second step to build consensus, the first is through editing as per CONSENSUS. Your change would make discussion based on policy a requirement for editing. -- LCU ActivelyDisinterested «@» °∆t° 21:09, 7 February 2026 (UTC)

Reasons can be given in edit summaries. - Butwhatdoiknow (talk) 23:18, 7 February 2026 (UTC)

A new editor may not even know what an edit summary is, knowledge of Wikipedia's WORDSALAD should not be a requirement to edit. -- LCU ActivelyDisinterested «@» °∆t° 23:27, 7 February 2026 (UTC)

Think you're missing a not after your should. Nil Einne (talk) 06:14, 8 February 2026 (UTC)

Thanks for spotting that. -- LCU ActivelyDisinterested «@» °∆t° 12:19, 8 February 2026 (UTC)

Butwhatdoiknow, your proposed change would, in all good faith, quite possibly encourage Wikilawyering. I suggest the effectiveness of WP:ONUS is in its very simplicity; it is clear and absolute. Yes, editors may try and argue, but as it stands they are arguing the unarguable. This addition would probably give them grounds to argue the more. —Fortuna, imperatrix 20:37, 7 February 2026 (UTC)

• Editors are free to edit in any way that they think improves an article … that means both adding and removing content. Now, sometimes editors disagree on whether an edit improves an article or not - and when that happens, we seek consensus. That consensus can be reached via further editing, via discussion, or via a mix of editing with discussion.

The ONUS to reach a consensus should be upon ALL of us. As it is, one “side” of the inclusion disagreement can simply point to ONUS, stonewall and cry “Not it! I don’t have to convince you… You have to convince me.” Blueboar (talk) 21:38, 7 February 2026 (UTC)

"As it is, one “side” of the inclusion disagreement can simply point to ONUS, stonewall and cry “Not it! I don’t have to convince you… You have to convince me."

And the only alternative is that we don't preference either side so all we get is:

"I don't have to convince you, you have to convince me!"

"Nuh-uh, you have to convince me!"

The point is if someone tries to stonewall by tapping the ONUS sign well it suggests there's only two people currently disagreeing and then it's a hop skip and jump to 3O or DRN, and if that lone person continues to try and stonewall because "I'm not convinced" then it's obvious stonewalling and they get a rather quick block for disruptive editing and failure to abide by consensus. Rambling Rambler (talk) 21:53, 7 February 2026 (UTC)

@Rambling Rambler, it turns out that we're on the same page. My goal in adding the text that got all of this started was to have the policy explicitly say that you can't use ONUS to stonewall - if you contest inclusion then you have to say more than "convince me!" First question for you: Do you think it is a good or bad idea to say this at ONUS? - Butwhatdoiknow (talk) 13:04, 18 February 2026 (UTC)

I think that the regulars at Wikipedia:Fringe theories/Noticeboard would find such a change to be anathema.

I think that "don't stonewall" is unenforceable (due to a distinct lack of mind-reading skills in the community) and would lead only to drama. "Convince me!" has a defined path for Wikipedia:Dispute resolution. "Don't stonewall" would take the focus off the content and put it on behavior. We want this:

• Alice: Convince me.
• Bob: Here are my reasons.
• Alice: I don't agree.
• Bob: Let's get a third opinion/go to a noticeboard/start an RFC.

We don't want this:

• Alice: Convince me.
• Bob: Here are my reasons.
• Alice: I don't agree.
• Bob: You're just stonewalling! I'm done focusing on content. I'm taking this to ANI because of your stonewalling behavior!

WhatamIdoing (talk) 18:07, 18 February 2026 (UTC)

Do we have to heed to what the editors at that noticeboard think? There are many editors who believe that noticeboard should be closed. Katzrockso (talk) 04:52, 19 February 2026 (UTC)

It's the same as any other noticeboard or other page where editors congregate: To the extent that the group's view matches the community's view, then that's the consensus, and consensus is king. WhatamIdoing (talk) 05:36, 19 February 2026 (UTC)

Option 3: Alice says "I don't agree because [reason]." Maybe Bob will see Alice's point and either concede, modify his proposal, or provide an additional reason for his proposed addition - no need to ask the other parent.

Isn't that what we want? - Butwhatdoiknow (talk) 05:59, 19 February 2026 (UTC)

It's nice when it works that way, but the key point in this scenario is that we need them talking about content (e.g., "reasons") and not about editors (e.g., "stonewalling"). WhatamIdoing (talk) 06:33, 19 February 2026 (UTC)

Exactly! My thought is that we should include text in ONUS that nudges editors in that direction. Do you think that is a good or bad idea? - Butwhatdoiknow (talk) 08:12, 19 February 2026 (UTC)

I don't think that WP:V is the right place to achieve that goal.

I think one way to make that problem worse would be changing ONUS to apply equally to all edits. When something is everyone's job, it's nobody's job, because everybody thinks somebody else will/should do it. WhatamIdoing (talk) 00:04, 20 February 2026 (UTC)

To be clear, I am not proposing a change to the existing text. It will remain and apply only to "inclusion." How would adding something saying that blankers should (not must) provide reasons change the scope of the current text? - Butwhatdoiknow (talk) 07:58, 20 February 2026 (UTC)

@WhatamIdoing, I respectfully request you educate me regarding how (a) adding a second sentence to ONUS saying blankers should provide reasons would (b) change the first sentence to apply equally to all edits? - Butwhatdoiknow (talk) 21:05, 27 February 2026 (UTC)

I didn't say that. I said that "one way to make that problem worse", not "your proposal in your comment at 05:59, 19 February would make that problem worse". That is, I'm reminding you of a related but distinct problem that needs to be taken into account. Changes (such as, but not limited to, encouraging Alice to provide reasons instead of a statement of disagreement) must be written in a way that cannot be twisted by editors to say that ONUS doesn't apply. For example, we don't want this:

• Alice: Convince me.
• Bob: Here are my reasons.
• Alice: I don't agree.
• Bob: You didn't provide a reason [alternatively: You didn't provide a reason that seems valid or relevant to me], so the WP:ONUS to achieve consensus for inclusion is no longer on me.

WhatamIdoing (talk) 19:41, 28 February 2026 (UTC)

We'll talk about Bob but, first, please let me know (1) whether you agree that we ought to encourage (not "require") Alice to say more than "convince me" or "I don't agree" and, if you do, (2) where we should do that. - Butwhatdoiknow (talk) 20:41, 28 February 2026 (UTC)

I think that "Convince me" is enough: it indicates that Alice is open to a conversation. "I don't agree" is sometimes enough: the reason may be patently obvious (e.g., Bob might already be fully aware that his explanation is unimpressive; see, e.g., anyone copy/pasting content from a conspiracy theorist website).

If I were going to encourage explanations, I'd look at essays that are focused on improving communication and encouraging dispute resolution, or perhaps Wikipedia:Talk page guidelines. If you felt like only a page with a {{policy}} tag at the top would do, then you might look at WP:EPTALK and WP:DISCUSSCONSENSUS. WhatamIdoing (talk) 00:04, 1 March 2026 (UTC)

Returning to my question, do you agree that we ought to (somewhere) encourage - not "require" - Alice to say more than "convince me" or "I don't agree"? - Butwhatdoiknow (talk) 03:25, 1 March 2026 (UTC)

We already do. It's the second paragraph of Wikipedia:Consensus#Through editing. WhatamIdoing (talk) 01:45, 2 March 2026 (UTC)

Okay, so it's a good thing to say. If we can phrase it "in a way that cannot be twisted by editors to say that ONUS doesn't apply" (and maybe we can't), is there any reason to not say it at ONUS? - Butwhatdoiknow (talk) 03:41, 2 March 2026 (UTC)

Yes, there is a reason not to put information or rules about the proper way to conduct a consensus-oriented discussion in WP:ONUS.

Those reasons generally fall into two categories:

• Whether they should be in this policy at all, for which see the comments from @Blueboar in Wikipedia talk:Verifiability#Break 3 on ONUS and elsewhere.
• Whether they have anything to do with ONUS, because they don't. "Under the narrow circumstance of Bob wanting to include cited information that Alice doesn't want to include, Bob has the onerous task of getting consensus from the community" is not relevant to "It's generally a good idea for editors to provide explanations about why they hold a particular view on anything".

WhatamIdoing (talk) 04:18, 2 March 2026 (UTC)

Blueboar says "this policy needs to . . . point editors to WP:CON." That sounds good to me. Are you okay with that goal? If so, we can talk about how that might best be accomplished. - Butwhatdoiknow (talk) 15:47, 2 March 2026 (UTC)

Blueboar actually said "all this policy needs to do is point editors to WP:CON". Due to the word "all", I interpret that as implying replacement of the current text with a pointer. Which I oppose. Peter Gulutzan (talk) 19:25, 2 March 2026 (UTC)

Peter, do you think THIS (WP:V) policy needs to state the onus sentence? If so, why? Blueboar (talk) 21:16, 2 March 2026 (UTC)

Because WP:V is core. Because WP:V archives are where the history of WP:ONUS arguments is kept. Peter Gulutzan (talk) 22:41, 2 March 2026 (UTC)

WP:V is a core content policy.

WP:CON is a core behavioral policy.

I think Blueboar's question is closer to "Why should a behavioral rule should be in a core content policy instead of in a core behavioral policy?" WhatamIdoing (talk) 23:52, 2 March 2026 (UTC)

I don't see what makes WP:CON "core" but if it's behaviour that's another argument against moving to there a sentence that says "content". Peter Gulutzan (talk) 01:39, 3 March 2026 (UTC)

@WhatamIdoing, based on Peter's interpretation (which Blueboar seems to endorse), the Blueboar argument seems to be that a pointer to Consensus should replace the V ONUS sentence. I agree but, let's face it, that's never going to happen.

Nevertheless, my mistake may offer a solution to the problem of Alice using ONUS as a way to opt out of a good faith discussion. So I'll renew my question: Are you okay with adding text to ONUS pointing to the second paragraph of Wikipedia:Consensus#Through editing? (With the exact text to be determined later.) - Butwhatdoiknow (talk) 21:34, 2 March 2026 (UTC)

Short answer: No.

Longer answer: Alice isn't necessarily opting out of a discussion at all, and the situation might not be one in which reasonable people can come to a reasonable agreement.

You might have wondered why Bob's response at the end didn't look like this, if Bob wanted to know her reasons:

• Alice: Convince me.
• Bob: Here are my reasons.
• Alice: I don't agree.
• Bob: Tell me more about that.

Alice hasn't given her answers yet, but she also hasn't refused to give them. We shouldn't tell her that she has to do something in advance, when nobody has expressed interest in reading what she writes.

Bob might not be interested in hearing Alice's answers. Perhaps Bob is silently rolling his eyes because of course it's Alice again because Alice is always disagreeing about this. Perhaps it doesn't matter what Alice thinks, and Bob was laying out his reasons because he expected the question to be settled by other participants, e.g.:

• Alice: Convince me.
• Bob: Here are my reasons.
• Alice: I don't agree.
• Chris, David, Eve, and Frank: We all support Bob's idea.
• Alice: Well, I still don't agree.
• Geo: I find consensus in favor of Bob's idea, even though Alice still doesn't agree.

In some cases, good, honest intentions (wikt:good faith, as opposed to Wikipedia:Good faith) might be missing, or one of the parties might not have the necessary skills or information to have a productive conversation. This doesn't require malice on anyone's part; for example, some people will adjust the size and location of images so that the page looks pretty on their computer set up, but they have no idea that they've broken it for other people.

But sometimes malice, or at least stupidity, is also present, and there isn't any need for Alice to explain to Bob why he can't spam in a paragraph about the great sale on blue-green widgets at his company this week, or why he can't use Wikipedia to explain why his religious beliefs are correct and everyone else's are wrong. When the explanation comes down to "anyone with half a brain and a tiny amount of experience with Wikipedia can see that it's a massive and blatant policy violation", one wonders why Bob needs an explanation at all.

This last doubtless frustrates some of our neurodivergent editors, but the fact remains that in most human cultures, it's rude to directly tell someone a thing that the specific individual in question ought to know already. One does not try to teach one's grandmother to suck eggs even if she does not appear to know how to do it. Therefore, some restraint on Alice's part may be more socially appropriate than explaining all the reasons why the pseudoscience du jour is nonsense and anyone who believes it is a moron. WhatamIdoing (talk) 23:48, 2 March 2026 (UTC)

"We shouldn't tell her that she has to do something in advance." I am not proposing that we tell anyone that they have to do anything.

I am proposing that we somehow send the message that the ONUS content policy is not an exception to the DISCUSSCONSENSUS behavioral policy by, for example, pointing to DISCUSSCONSENSUS in ONUS.

Doing that would not eliminate any of the existing DISCUSSCONSENSUS exceptions (some of which you describe above). But it might reduce the chances that editors will give "ONUS" as a rationale for removing text. - Butwhatdoiknow (talk) 01:38, 3 March 2026 (UTC)

Is that actually still happening? My impression was that this behavior had declined significantly. WhatamIdoing (talk) 06:09, 3 March 2026 (UTC)

My impression is that the behavior continues. Shall I start a new section asking the community what they have found to be the case? - Butwhatdoiknow (talk) 17:13, 3 March 2026 (UTC)

I wonder whether it would be quicker and more reliable to search for recent edit summaries that mention WP:ONUS. (I think there's some tool that makes it possible to search edit summaries.) WhatamIdoing (talk) 19:33, 4 March 2026 (UTC)

How do we determine whether there is such a tool? (Note: There's also this Special:WhatLinksHere/Wikipedia:Onus, but then someone would have to manually go through the list to see how often ONUS is given as a reason (as opposed to cited as a process).) - Butwhatdoiknow (talk) 20:53, 4 March 2026 (UTC)

Ask at Wikipedia:Village pump (technical)? WhatamIdoing (talk) 22:01, 4 March 2026 (UTC)

Sounds good. Since you think there may be such a search tool I'm thinking you'd be the best person to ask. Would you please do that? - Butwhatdoiknow (talk) 23:25, 4 March 2026 (UTC)

@WhatamIdoing: Renewing my request. - Butwhatdoiknow (talk) 15:45, 9 March 2026 (UTC)

 Done Wikipedia:Village pump (technical)#Searching edit summaries

Thanks for the reminder. WhatamIdoing (talk) 03:15, 10 March 2026 (UTC)

@WhatamIdoing: Okay, we have an answer: 1169 lines. Seems like a big number to me, but keep in mind that I have no idea what it actually means. - Butwhatdoiknow (talk) 00:51, 14 March 2026 (UTC)

1169 edits out of about five million, so about one out of 4,000 edits.

I've asked for a copy of the results. WhatamIdoing (talk) 04:39, 14 March 2026 (UTC)

After removing false positives (mostly REVISIONUSER, but also words like bonus) and these discussions (where we are talking about it but not invoking it), there were ~353 uses left.

ONUS is mentioned in the section heading for a few discussions, so it appears in all the edit summaries, even if the comment is unrelated or it's just a typo fix. There are 70 of these in the section heading for a single discussion at ANI. Another 42 were a discussion at Talk:Jawaharlal Nehru#WP:ONUS and WP:DUE, and two others have about 10 edits/comments each. If we treat a section heading as "one" invocation, no matter how many replies the section had, then there were maybe 220 instances left.  Nearly all of these were in the mainspace, and most of those were WP:UNDO or other reverts.  (25 to 30% of them got reverted in turn.)

This dataset covers the most recent two months.  I think it would be fair to say that ONUS is invoked about three or four times a day. For comparison, I believe that a typical day has about a quarter million edits. WhatamIdoing (talk) 22:47, 15 March 2026 (UTC)

Thank you. Accepting your analysis (which seems to err on the side of underreporting), we have a rough estimate that editors cite "ONUS" as a rationale for removing text more than 1,000 times a year. I suggest this is is a less than trivial behavior that we should discourage, and that we can do so by, for example, pointing to DISCUSSCONSENSUS in ONUS. - Butwhatdoiknow (talk) 23:40, 15 March 2026 (UTC)

As I've always said, I think the word "onus" in the Onus sentence should be a wikilink to a sentence in Consensus which says the onus to change consensus is on those who want to change consensus. Kolya Butternut (talk) 00:24, 3 March 2026 (UTC)

We removed the word onus from WP:ONUS a couple of years ago. I believe that one of the main complaints was that Kids These Days™ were suspected of being unfamiliar with the dictionary definition of the word ("burden; a disagreeable necessity; obligation"). Here's the modern version of the sentence: The responsibility for achieving consensus for inclusion is on those seeking to include disputed content. WhatamIdoing (talk) 06:08, 3 March 2026 (UTC)

To redirect back from your non sequitur, as I said, I think if the word onus at V linked to a sentence at Consensus that said the onus to change consensus is on those who want to change consensus it would clear things up nicely It would just invalidate your interpretation of policy. Since I already know you disagree I would like to hear what others think. Kolya Butternut (talk) 17:25, 5 March 2026 (UTC)

• Playing devil’s advocate… That works when a consensus pre-exists, but doesn’t address situations where there isn’t (yet) any consensus that someone is trying to change.

At the moment ONUS can be interpreted as saying: “OK… then those wanting to include need to get a consensus.” Blueboar (talk) 17:53, 5 March 2026 (UTC)

I'm not sure I understand. When adding something new, an editor is changing consensus. So if we say at V: "The onus to achieve consensus for inclusion is on those seeking to include disputed content", and it links to a sentence at Consensus which is broader than just this sentence, that sentence at Consensus includes adding things. Kolya Butternut (talk) 01:02, 6 March 2026 (UTC)

IF we say that adding something changes consensus, then wouldn’t removing something also change consensus? Both are changes to the text after all.

That said, I disagree with your statement… merely adding (or removing) something does not form a new consensus. For a consensus to exist, there must be some indication that the edit has been accepted by other editors - either via further editing around it (a silent consensus) or via discussion (an overt consensus). Blueboar (talk) 02:03, 6 March 2026 (UTC)

I'm confused about what I said that you're disagreeing with. I'm not talking about creating a new consensus, I'm talking about changing past consensus. Surely adding something that never existed is a change to existing consensus. Removing something that was just added is not a change to existing consensus. Removing something that has been there for a long time is tbd? To work within your comments:

[T]hen those wanting to include need to get a consensus. Yes, those wanting to add something need to get consensus, and those wanting to remove something or change something which has consensus need to get consensus. And by linking from the word "onus" at V to the policy at CON, we're making clear that the Onus sentence is part of the VNOT policy and is just reiterating something at CON. It doesn't make sense to have a behavioral policy at V after all. Kolya Butternut (talk) 19:59, 6 March 2026 (UTC)

An edit does NOT change past consensus. Blueboar (talk) 20:25, 6 March 2026 (UTC)

An edit is an attempt to change past consensus? Kolya Butternut (talk) 00:34, 7 March 2026 (UTC)

You are assuming that a past consensus exists. Consider:

• A creates article with one sentence "A".
• B adds a second sentence "B" five minutes later.
• C removes the sentence "A" ten minutes later.

Which version is the "past consensus"? I think the answer is "none of them". And therefore none of the edits can be sensibly characterized as an attempt to change the past consensus (because you can't attempt to change something that doesn't exist). WhatamIdoing (talk) 05:57, 7 March 2026 (UTC)

If the article survived an attempt at deletion, then in your example the consensus version is just that it exists, and all of them are an attempt to change past consensus. Kolya Butternut (talk) 10:44, 7 March 2026 (UTC)

The article is 15 minutes old at the end of those three edits. There hasn't been time to attempt deletion.

Are you claiming that the creation of an article is an attempt to overturn an existing consensus against having an article? I don't think that's sensible at all. WhatamIdoing (talk) 15:53, 7 March 2026 (UTC)

Sounds like a non sequitur, but our policy says in article deletion discussions we default to keep. Kolya Butternut (talk) 17:39, 7 March 2026 (UTC)

That argues against your point: If An edit is an attempt to change past consensus (according to you), then the edit creating of the article would have to be an attempt to change past consensus. Since there was no article, the consensus that the edit is "attempting to change" would have to be against creating the article. Claiming we have a consensus against creating articles is nonsense. WhatamIdoing (talk) 17:54, 7 March 2026 (UTC)

I agree that your framing is nonsense. Kolya Butternut (talk) 20:46, 7 March 2026 (UTC)

• When you say "the word onus at V" do you mean (a) the ONUS shortcut, (b) the word "responsibility," or something else? - Butwhatdoiknow (talk) 18:09, 5 March 2026 (UTC)

  I meant the word "onus" in what I thought was the longstanding version of the sentence: The onus to achieve consensus for inclusion is on those seeking to include disputed content. Kolya Butternut (talk) 00:52, 6 March 2026 (UTC)

  Does your proposal change now that you know the longstanding version is no longer standing? Or maybe it's just tweaked to propose (a) changing "responsibility" back to "onus" and then (b) linking "onus" to Consensus? - Butwhatdoiknow (talk) 02:12, 6 March 2026 (UTC)

  The word doesn't change the meaning or function of my proposal, but I think it should be changed back to "onus" so that a wikilink makes sense. The word "onus" at V would link to WP:ONUS at Consensus. The onus sentence at V should be merged back to VNOT. Kolya Butternut (talk) 20:04, 6 March 2026 (UTC)

  I don’t have a strong opinion on returning the word “onus” to the text (or linking it to CON)… I do have a strong objection to merging the onus section back into VNOT. They address different things and were separated for a reason. Blueboar (talk) 20:24, 6 March 2026 (UTC)

  But if the onus section were actually at CON (in an expanded form) then couldn't the onus sentence at V be merged back into VNOT? Kolya Butternut (talk) 00:32, 7 March 2026 (UTC)

  Onus sentence == onus section. If the onus section is moved to CON, then there won't be anything left here to merge back into VNOT. WhatamIdoing (talk) 05:55, 7 March 2026 (UTC)

  You're missing the point, i.e., that's a non sequitur.

  The new Onus section where WP:ONUS would link to would be at Consensus. The old Onus section at V could be restored into its older form but remove the WP:ONUS wikilink, so it would just be VNOT. Kolya Butternut (talk) 10:38, 7 March 2026 (UTC)

  So... are you proposing to duplicate the ONUS sentence?

  That is, WP:V will say "The responsibility for achieving consensus for inclusion is on those seeking to include disputed content", with a link to WP:CON, which will get a new section (and the WP:ONUS shortcut), which also says "The responsibility for achieving consensus for inclusion is on those seeking to include disputed content"? WhatamIdoing (talk) 15:55, 7 March 2026 (UTC)

  Not a duplicate. The WP:ONUS link at Consensus would say The onus to achieve consensus for changes to consensus is on those seeking the change. Kolya Butternut (talk) 17:48, 7 March 2026 (UTC)

  So you'd like to set up a contradiction between WP:V and WP:CON?

  You seem to be proposing that WP:V says the onus is on would-be includers and that CON says the onus is on would-be changers. In that case, when the change in question is Alice removing something over Bob's objections, then WP:V says the onus is on Bob and CON says the onus is on Alice.

  We are supposed to be avoiding that kind of contradiction.

  Or did you mean that you want to move this sentence over to CON, in the hope that you'll have an easier time changing that bit about the onus being on editors seeking to include disputed content to say instead that the onus is on editors seeking to change content? WhatamIdoing (talk) 17:58, 7 March 2026 (UTC)

  There is no contradiction, but we already know we disagree on each other's interpretation of the policy, so maybe let others respond. Kolya Butternut (talk) 20:50, 7 March 2026 (UTC) Kolya Butternut (talk) 20:55, 7 March 2026 (UTC)

  I think we do need to articulate our understandings again on this point. How do you see this as a contradiction? If Alice is removing something to change consensus, the onus is on her, but if Alice is removing something to restore consensus, the onus is on Bob. Kolya Butternut (talk) 00:28, 10 March 2026 (UTC)

  Since nobody else seems particularly interested, I'll post on your talk page. WhatamIdoing (talk) 03:17, 10 March 2026 (UTC)

• @Kolya Butternut, I don't think this was a non sequitur at all, and since you do, I suspect that you did not understand what I said. Here's the conversation in simplified form:

• You: I want to put a link on the word onus in this policy.
• Me: You physically can't put a link on the word onus in this policy, because the word onus isn't actually in this policy any more.

Sensible responses include things like "Um, maybe put the link on responsibility, since that word is actually in the policy" or "I'd like to overturn that consensus and put the word onus back in the policy so I can put a link on it" or even "Oh, never mind", but probably not "you just disagree with me and the word actually still is there". WhatamIdoing (talk) 21:48, 5 March 2026 (UTC)

An editor who removes content, who then doesn't take part in a discussion started by someone else, and reverts the re-additional of that content is being disruptive. That communication is required is already part of policy. Consensus doesn't mean that everyone has to be convinced, if two editors disagree find a third opinion per WP: Dispute resolution. "Convince me!" only holds up until there is a third voice. Content being in or out of an article while consensus is found is not a problem, there is no dead line. -- LCU ActivelyDisinterested «@» °∆t° 22:00, 7 February 2026 (UTC)

ONUS is more like a default position if there's a disagreement over content and if it wasn't there, it would encourage further back and forth edit warring. This was a frustrating discussion: Talk:City_College_of_New_York#RFC_on_Phrase_on_Honors_College. I feel that one strong willed inclusion proponent forcibly rested to inclusion and it turned the situation into ONUS where I was put in place to establish consensus to remove. If WP:ONUS was being followed, it would have defaulted to exclusion and the burden was on the one who wanted to include it to start an RfC and reach a consensus. Not like how it went in that discussion.

When it comes to WP:VNOT and WP:NOTEVERYTHING, it is a huge grey area with a significant community editorial discretion on article-by-article basis and from reliability of factual standpoint, there is nothing barring putting something like XYZ Middle School has strawberry milk on menu on Fridays; or that students who enroll at ABC Community College full time for an academic year receives a calendar year bus pass. If the school's site is cited, there's no reason to doubt the factual reliability, however whether such thing is due under WP:ABOUTSELF rests in consensus. If the policy defaulted to inclusion with the burden to reach consensus on those seeking to remove it, Wikipedia would drown in minutiae and advertisement added by marketing and public relations effort. Graywalls (talk) 02:17, 9 February 2026 (UTC)

• When was WP:ONUS moved out from WP:VNOT, and where was this discussed? It is a corollary to that; part of the confusion above seems to be because it was moved to its own section at some point. --Aquillion (talk) 19:36, 17 February 2026 (UTC)

  December 2024: Semi-bold edit, preceded by discussion at Wikipedia talk:Verifiability/Archive 84#New Shortcut re “ONUS” section?. WhatamIdoing (talk) 20:37, 17 February 2026 (UTC)

  I tried to undo it last December. Special:Diff/1329927175 Kolya Butternut (talk) 20:39, 17 February 2026 (UTC) Kolya Butternut (talk) 23:46, 17 February 2026 (UTC)

  The revert said "This was discussed and approved at the time", but I saw very little prior discussion -- certainly not enough to establish consensus. Call for opinions: given the present far larger discussion, is the consensus for the December 2024 version as restored by User:Kolya Butternut, or the version that User:Blueboar reverted to (which is the current version)? --Guy Macon (talk) 13:34, 18 February 2026 (UTC)

  Most of this discussion has nothing to do with the line breaks. It's about whether ONUS should say "include" vs "include, exclude, or change". But if you want to discuss the line breaks, we can do that. Here's the chart with the arguments as I've heard them:

What would you add or change to that list? WhatamIdoing (talk) 18:29, 18 February 2026 (UTC)

Some history… when I added the original paragraph (what is now VNOT) it did not include the sentence that is now ONUS. I mentioned “consensus” in my original paragraph simply to warn that verifiable information might be omitted if/when a consensus to omit exists (or is formed).

The addition of the ONUS sentence (and eventually the ONUS shortcut) flipped that original intent… requiring a consensus to include.

If we put ONUS back into VNOT the paragraph contradicts itself. Blueboar (talk) 20:13, 18 February 2026 (UTC)

Can you say more about how the paragraph would contradict itself? Kolya Butternut (talk) 22:51, 18 February 2026 (UTC)

The first part (VNOT) says a consensus can omit… the second part (ONUS) says you need consensus to include. So which is it? do we need a consensus to omit or a consensus to include? It’s contradictory. Blueboar (talk) 23:36, 18 February 2026 (UTC)

I don't think it is. Everything is ultimately supposed to have consensus; in situations where we have no consensus, we decide whether to omit things or retain the status quo depending on several interlocking policies. But no outcome that lacks consensus should ever be considered stable and "we still haven't settled this" is a reason to continue discussions (this is one of the things I consider most poisonous about how WP:ONUS is frequently misinterpreted, in that it is sometimes cited as if it has some sort of power to end discussions, and frequently derails them as a result.) If we don't make that clear, people will be like "well we have no consensus so we omit, discussion is over" or "WP:ONUS, it's up to you to demonstrate consensus; as someone who wants to omit, I can remove this citing ONUS and have no responsibility at all to engage beyond that", which is absolutely not how things work - guidelines like WP:ONUS and WP:QUO essentially just propose ways to keep things going until consensus is reached, because we do have to have an actual live article at some point even when no version has a clear consensus. If we could leave it in a liminal state we would, but we can't. But anyone who intentionally leans on them as the endpoint of a discussion is misusing them ala WP:STONEWALL; once there is a dispute, you are always supposed to seek consensus for your preferred version, whether you want to include or omit something. Therefore it is correct to say that both omission and inclusion require consensus. --Aquillion (talk) 02:46, 19 February 2026 (UTC)

No disagreement on what the policy should be… but that isn’t what ONUS says. It says inclusion requires consensus… period. That is why we need to either completely re-write ONUS … or cut it entirely (my preferred option). Blueboar (talk) 12:30, 19 February 2026 (UTC)

Those are terrible ideas. 1) The text of articles are presumed to have consensus. 2) Which means, if it does not have consensus, it should not be the text.

The issue has always been, how do we get to sufficient consensus, and it is always been 'it depends', although we know it's often arrived at by who holds out the longest with enough craft to get by. We do not need more ammunition for getting whatever into articles, just because the proponent can say 'I have a source'. Indeed your prescription would lead to a worse Wikipedia on balance. And consensus is not just nice, it does what we think our reliability and usefulness arises from, more eyes taking care on the text. Alanscottwalker (talk) 12:43, 19 February 2026 (UTC)

But VNOT already addresses “I have a source”… It warns that other factors (including consensus) can omit. Blueboar (talk) 12:56, 19 February 2026 (UTC)

That phrase has never been good enough, alone, which is why the follow up to that phrase has always discussed the onus. The second part clarifies the first, on "consensus". Alanscottwalker (talk) 13:01, 19 February 2026 (UTC)

I don’t think it clarifies… I think it confuses. Blueboar (talk) 13:11, 19 February 2026 (UTC)

I suggest you close read the first without the other and you will find the first suggests., include must occur, never to be ousted, without a donnybrook. Alanscottwalker (talk) 13:20, 19 February 2026 (UTC)

Given that I wrote the first part (now VNOT)… I know what it says quite well. I also know what the intent was. Blueboar (talk) 15:09, 19 February 2026 (UTC)

That would suggest you are not an independent, nor impartial observer on its clarity and lack thereof. Alanscottwalker (talk) 15:13, 19 February 2026 (UTC)

Whatever… I have a good enough reputation that I am not worried about that. Others can judge on their own. Blueboar (talk) 16:09, 19 February 2026 (UTC)

Alan, I don't tend to agree with you about the current wording, but I think all three of us have the same idea of what VNOT is meant to say and all want the VNOT wording to be clear. Since you can see a different way to read the current wording – and I know your views on getting nonsense out of articles are absolutely sound – then I have to conclude that a motivated wikilawyer would also be able to see that different understanding. So how about we work on re-write? What do you two think of this:

It must be possible for someone to find a source that supports any information in an article. However, being able to source the content is only the minimum threshold for inclusion, and citing a reliable source does not guarantee that the information belongs in that particular article, or in any Wikipedia article at all. Editors are allowed to remove both cited and uncited material from articles through the ordinary consensus-oriented processes of editing and discussion. They normally should remove information, even if it is fully cited, if they determine that its inclusion does not improve an article or if other policies and guidelines indicate that the material is inappropriate. Sometimes, the removed information can be presented instead in a different article.

Does that sound approximately like what VNOT is meant to say? Do you think that it would be more wikilawyer-proof than what we've got now?

[Tangent] While I'm thinking about material that may or may not belong anywhere in Wikipedia, let me invite you all to an overlooked CTOP RFC about Moroccanoil#Controversy. WhatamIdoing (talk) 00:27, 20 February 2026 (UTC)

Perhaps, although I doubt I and others have the strength or motivation for a wholesale rewrite. I am reminded by your comment perhaps inadvertently, that possibly part of the issue is the suggestion that 'one source' somewhere in the world is always the minimum. Perhaps, one source is not even that, depending. I get why this policy focuses on 'a source' but if we are summarizing accepted knowledge on a topic, just one source seems a slender reed, indeed. Alanscottwalker (talk) 20:04, 28 February 2026 (UTC)

Relying on "just one source" in the whole world can be a significant WP:DUE problem, but it can also be a WP:BALASP win. We wouldn't omit the ordinarily expected facts (e.g., someone's birthplace, a company's founding year) solely because "just one source" mentioned it.

Perhaps I'll go make part of this change. Then you can see what you think in situ. WhatamIdoing (talk) 00:17, 1 March 2026 (UTC)

Break 2 re ONUS: add text re both pro and con must support positions?

Onus says inclusion requires consensus because it's in the context of discussing adding things. Removing things and changing things also requires consensus. Kolya Butternut (talk) 23:03, 19 February 2026 (UTC)

True enough, but that's in Wikipedia:Consensus and not in ONUS. WhatamIdoing (talk) 00:06, 20 February 2026 (UTC)

Onus is included in Consensus, so we don't need Onus. Kolya Butternut (talk) 04:28, 20 February 2026 (UTC)

The idea that additions require consensus is included in WP:CON.

The idea that if an addition is contested, is it the responsibility of the would-be adder and not the would-be remover to form/demonstrate a cosnensus in favor of addition is AFAICT not present in WP:CON. WhatamIdoing (talk) 04:30, 20 February 2026 (UTC)

If content with consensus is removed do you think it is not the responsibility of the remover to form/demonstrate consensus in favor of the removal? It's the same as when an addition is made to change consensus. Kolya Butternut (talk) 04:37, 20 February 2026 (UTC)

How do I know that this newly removed content actually has consensus? WhatamIdoing (talk) 04:43, 20 February 2026 (UTC)

Because it's stipulated in my question. Kolya Butternut (talk) 05:02, 20 February 2026 (UTC)

I asked, because what I need to do depends on how I know that the newly removed content actually has consensus. WhatamIdoing (talk) 06:10, 20 February 2026 (UTC)

The question isn't what you need to do it's whether you have the onus. Kolya Butternut (talk) 15:19, 20 February 2026 (UTC)

Same difference? According to the dictionary, you just wrote that "The question isn't what you need to do it's whether you have the obligation". Having the onus on you = it's something you have to do. WhatamIdoing (talk) 05:33, 21 February 2026 (UTC)

You're not answering my question. Kolya Butternut (talk) 09:59, 21 February 2026 (UTC)

I have answered your question; did you not see my answer?

Let me spell it out in more detail:

Imagine that Alice adds cited content to an article. Bob dislikes it (for some plausible reason; Bob is not being unreasonable) and wants it removed. The discussion ends up with an RFC that produces a clear result, namely that the content will remain in this article.

A month later, Chris – who has no idea about the prior history – sees the cited content in the article and decides that this doesn't seem appropriate for this article (like Bob, Chris is not being unreasonable). So Chris removes the content.

I notice Chris's removal in my watchlist. I want to keep the content in this article.

"Now what?" you seem to be asking.

The answer is:

Under ONUS, it's my job and not Chris's to do whatever work is necessary to 'achieve consensus' for keeping the content in the article. In this case, due to the recent RFC, having the onus on me is not very onerous: I can revert Chris with an edit summary that says "Per Talk:Example#Huge RFC", and we're done.

If ONUS didn't exist, it would be Chris's job to get consensus for their desired change. But ONUS does exist, so that's not the rule that is in effect.

.

So what if there wasn't an RFC, or any evidence that a consensus actually exists? We can tell that story, too:

Imagine that Alice adds cited content to an article. Bob removes it.

Alice notices Bob's removal and wants to keep the content in this article.

Now what? The answer is:

Under ONUS, it's Alice's job and not Bob's to do whatever work is necessary to 'achieve consensus' for keeping the content in the article.

This means that Alice needs to start a discussion with Bob. (Bob is allowed to start a discussion, but the onus for starting the discussion isn't on him.) If they can't agree, then Alice needs to pursue ordinary dispute resolution processes. That means Alice needs to post a request at Wikipedia:Third opinion. (Again, Bob is allowed to do so, but the onus for doing so isn't on him.) If a third opinion isn't enough, then Alice needs to post a request for mediation at the Wikipedia:Dispute resolution noticeboard. Bob is still allowed to do so, but he's still not required to do so. If we need an RFC, then the onus is on Alice and not on Bob to get the RFC organized.

And so forth, through the whole process.

If ONUS didn't exist, then starting these discussions would instead be the responsibility of whoever is making a "change", and – critically – whenever editors both claimed that the other person's change was more of a change (because 'my' change is always the WP:STATUSQUO consensus version, so it's always 'you' that has to do the extra work), we would in practice have drama instead of a predictable process. ONUS stops that debate in all cases of addition vs removal. WhatamIdoing (talk) 23:15, 21 February 2026 (UTC)

What evidence is there for your interpretation? Kolya Butternut (talk) 02:18, 22 February 2026 (UTC)

The plain reading of the words on the page. Maybe also the fact that AFAICT nobody except you thinks I'm wrong? WhatamIdoing (talk) 04:09, 22 February 2026 (UTC)

I don't know where you've been, but Onus is very much disputed. I plainly read the words differently than you do, but you're not providing evidence of practices. Kolya Butternut (talk) 05:12, 22 February 2026 (UTC)

The practice is that many editors can't keep BURDEN and ONUS straight in their minds. See, e.g., Talk:John Mearsheimer#WP:SOAP, in which @Hipal is accused of violating ONUS by removing something. See this comment and this one and many, many others, in which ONUS is named but the relevant rule is BURDEN. See Talk:Great Translation Movement#Removal of details critical of the movement, in which an IP says that ONUS imposes a WP:STATUSQUO rule.

Of the ones that aren't obviously referring to something else – try this search – the practice is that ONUS is invoked when the would-be includer is being told to stop trying to put the content back in the article and/or to start a discussion about whether to put the disputed content back in the article. WhatamIdoing (talk) 20:51, 23 February 2026 (UTC)

So as I said, Onus just means you need consensus to add something. This is consistent with Consensus which requires consensus for any change to consensus. Yes, the hard part is figuring out what version has consensus. Kolya Butternut (talk) 00:17, 24 February 2026 (UTC)

• WP:ONUS: You have to achieve consensus to add something, but the other editor doesn't have to achieve consensus to remove it.
• WP:CON: Everyone ultimately has to work together to achieve consensus for everything.

One of These Things (Is Not Like the Others). WhatamIdoing (talk) 00:24, 25 February 2026 (UTC)

but the other editor doesn't have to achieve consensus to remove it

Your "it" is referring to the bold edit to add something, not an edit removing something with silent consensus. Kolya Butternut (talk) 02:20, 25 February 2026 (UTC)

No, my "it" refers to any material on the page at all. WhatamIdoing (talk) 07:20, 25 February 2026 (UTC)

Onus doesn't say that. Kolya Butternut (talk) 11:47, 25 February 2026 (UTC)

ONUS says "The responsibility for achieving consensus for inclusion is on those seeking to include disputed content."

Note the absence of qualifying or restrictive words, such as "if it was a bold edit, but not if it's been there a long time" or "doesn't apply if the inclusion has silent consensus".

This sentence applies to all content whose inclusion is disputed.

Also, have a look at WP:SILENT, and see how many times it basically says that when there's a noisy dispute, you don't have silent consensus. WhatamIdoing (talk) 21:44, 25 February 2026 (UTC)

And if consensus cannot be reached the result is the status quo version stays. If something has already been included and you want to remove it 20 years later, what's happening is you're trying to remove it. Kolya Butternut (talk) 22:12, 25 February 2026 (UTC)

Nope. The onus is on the would-be includer to prove that there is a consensus to keep it, full stop. There is no "and if you can't, then after a short while, the onus stops being on the would-be includer to achieve consensus to include it, and starts being on the would-be excluder to achieve consensus to exclude it".

See also WP:STATUSQUO, which says it only applies during active, consensus-oriented discussions and not after they've concluded. WhatamIdoing (talk) 22:19, 25 February 2026 (UTC)

See WP:NOCON. Your Onus argument only applies to biographies. Kolya Butternut (talk) 22:53, 25 February 2026 (UTC)

Yes, let's see WP:NOCON, especially the bit where it says that it's not imposing any rules and you should follow the policies and guidelines that it imperfectly summarizes. WhatamIdoing (talk) 00:51, 26 February 2026 (UTC)

NOCON is not a summary of Onus or less specific than Onus. They are not contradictory; NOCON is just more clear. And Onus is a verifiability policy not an editing or consensus policy. Kolya Butternut (talk) 03:35, 26 February 2026 (UTC)

Say there was an RfC that established that portion of text last month. Katzrockso (talk) 05:04, 20 February 2026 (UTC)

If there was an RFC that established that portion of text recently, and there's no reason to believe that the situation has changed, then I revert the removal with a link to the RFC in the edit summary.

ONUS requires me (not the blanker) to achieve consensus. In the case of a recent RFC, I have pre-achieved that consensus. WhatamIdoing (talk) 05:21, 20 February 2026 (UTC)

The person who wants to remove the content has the onus to show that consensus has changed. Kolya Butternut (talk) 05:30, 20 February 2026 (UTC)

WP:ONUS doesn't say that. WhatamIdoing (talk) 06:09, 20 February 2026 (UTC)

Irrelevant. This is how Consensus works. Kolya Butternut (talk) 06:19, 20 February 2026 (UTC)

Irrelevant. This is how WP:ONUS works, and since we're actually talking about the single sentence in WP:ONUS, the whole, complex, vague, and variable approach to consensus in general is mostly off topic. Yes: Consensus is king, and it can be arrived at through many routes. No: Nothing about consensus on wiki gets the would-be adder free of the onus ("duty, responsibility, obligation, requirement") to demonstrate a positive consensus in favor of inclusion. WhatamIdoing (talk) 05:37, 21 February 2026 (UTC)

If we removed the onus sentence nothing about our policies would change because Consensus results in the same standards (as I interpret Onus). Kolya Butternut (talk) 10:03, 21 February 2026 (UTC)

I don't think you ever "interpret Onus". I think you instead interpret an integrated idea of how consensus-based editing ought to work. WhatamIdoing (talk) 22:51, 21 February 2026 (UTC)

Onus just means you need consensus to add things. It's not complicated. Kolya Butternut (talk) 05:50, 22 February 2026 (UTC)

WP:ONUS doesn't say that. It says "achieve consensus" not "achieve or show." According to the ONUS text you still have to convince the blanker. - Butwhatdoiknow (talk) 07:52, 20 February 2026 (UTC)

Not quite: ONUS doesn't say who you have to convince. You have to achieve a consensus, but the blanker could be on the losing side of that consensus. WhatamIdoing (talk) 05:40, 21 February 2026 (UTC)

Well, in your example there is only the proponent and the blanker. And ONUS does not say that the proponent can pre-achieve consensus. So -

• If we can't look at anything other than what ONUS explicitly says, your statement that "In the case of a recent RFC, [the proponent has] pre-achieved that consensus." is incorrect, and
• If we can look elsewhere then we can take WP:Consensus into consideration. And the proponent can show a pre-achievement of consensus.

So, can we look elsewhere? Butwhatdoiknow (talk) 06:45, 21 February 2026 (UTC)

If consensus is already achieved, ONUS doesn't say that you have to (rather pointlessly) go through the motions of re-achieving what already exists. WhatamIdoing (talk) 22:52, 21 February 2026 (UTC)

And it doesn't say that you can "achieve" by showing a prior achievement. So saying "it doesn't say that" doesn't help us decide what ONUS means regarding a prior consensus.

I agree that it makes sense that you can point to a prior consensus, and if we look past the text of the sentence, that is the practice.

It seems to me the solution is to not make an "it doesn't say that" argument when editors point to other practices or policies that inform ONUS. Instead, explain why they don't apply. Butwhatdoiknow (talk) 07:43, 22 February 2026 (UTC)

(Outdent) Just a thought… perhaps one way to interpret ONUS is that it is focused on (and negates) the concept of “silent consensus”… Ie when some bit of material is removed, the editors wishing to keep the material can not claim a silent consensus… instead they need to go to the next level and achieve an actual (discussed) consensus to keep it. Blueboar (talk) 12:23, 25 February 2026 (UTC)

Whatever Onus means, it is subordinate to Consensus, which dictates these things. As I said, considering the policy to its logical conclusion, I don't think content in the lead of a highly trafficked article that has been there for 20 years would be removed if there were no consensus to keep it. Kolya Butternut (talk) 15:12, 25 February 2026 (UTC)

Sure, but not all articles are “high traffic”. Blueboar (talk) 16:11, 25 February 2026 (UTC)

• One thing that I think that is clear from these discussions (and the past... five years of discussions, I think it has been at this point) is that there is a lack of clear agreement on what ONUS means, how it's enforced, and so on; so I've added a note to it to that effect. I tried to word it as cautiously as I could; I'm open to alternative wordings but I do think that we need some sort of indicator so people understand that they could be stepping on a minefield by citing it or by applying what they personally see as an "obvious" implication of its text. --Aquillion (talk) 04:55, 26 February 2026 (UTC)

  I've removed that for the moment. Open to seeing if anyone can come up with alternative wording, but it's not at all clear to me what someone is meant to do with that note. In practice, what does it mean for someone to "be cautious in citing it"? Nikkimaria (talk) 00:12, 27 February 2026 (UTC)

  What about just leaving: The precise meaning and intent behind this sentence is heavily disputed. Kolya Butternut (talk) 02:53, 27 February 2026 (UTC)

  Again, what is someone meant to do with that? Nikkimaria (talk) 03:06, 27 February 2026 (UTC)

  I don't know what someone is meant to do with that, but I feel confident predicting that what some editors would do with that is claim that ONUS can be ignored and that the onus is definitely not on them to achieve consensus for disputed content that they want to put in an article. WhatamIdoing (talk) 03:20, 27 February 2026 (UTC)

  If you could re-write the ONUS sentence – "The responsibility for achieving consensus for inclusion is on those seeking to include disputed content" – to express the meaning of this one sentence more clearly, how would you rephrase it? NB that I'm specifically not looking for a whole explanation of all points of consensus. I'm looking for a rephrasing of this one/sole/isolated sentence. For example, would be be clearer to turn it around grammatically, from

  The responsibility for achieving consensus for inclusion is on those seeking to include disputed content

  to

  Any editor who wants to include disputed content has the responsibility to form a consensus to include the disputed material?

  Or would clarity require a bigger change, such as:

  Any editor who wants to include disputed content has the responsibility to form a consensus to include the disputed material before the disputed content can be re-added to the article?

  or

  Any editor who wants to include disputed content has the responsibility to form a consensus to include the disputed material; however, if no consensus can be found, BLP content should be removed and non-BLP content should be kept

  or something like that? WhatamIdoing (talk) 03:10, 27 February 2026 (UTC)

  This covers it: While information must be verifiable for inclusion in an article, not all verifiable information must be included. Kolya Butternut (talk) 03:18, 27 February 2026 (UTC)

  That isn't the ONUS sentence. "Let's just remove it" is not an answer to a question about how we could re-write this sentence to make it clearer. WhatamIdoing (talk) 03:19, 27 February 2026 (UTC)

Break 3 on ONUS

• Perhaps: “When there is a dispute as to whether to include or omit verifiable content, follow the guidance set out at WP:Consensus.”.

  Blueboar (talk) 13:37, 27 February 2026 (UTC)

  Which part of WP:CON do you think is the equivalent/relevant guidance?

  ONUS imposes a 'responsibility' on a particular person ('those seeking to include'). AFAIK there is nothing similar in WP:CON. WhatamIdoing (talk) 20:11, 27 February 2026 (UTC)

  That can and should be discussed at WP:CON… all this policy needs to do is point editors to WP:CON. Blueboar (talk) 15:51, 1 March 2026 (UTC)

  Except for the common sense inference that whoever wants to change consensus is responsible for achieving consensus. Kolya Butternut (talk) 21:26, 27 February 2026 (UTC)

  It might be common sense.

  It might be the community's actual practice.

  It might be the best possible practice.

  But it is not actually in the ONUS sentence, which is the point I've been trying to communicate to you for many years now, or even in the WP:CON policy.

  Working from the perspective of the actual words of the ONUS sentence, let me translate it to a different setting. Compare:

  • The responsibility for achieving consensus for inclusion is on those seeking to include disputed content.
  • The responsibility for washing the dishes is on those seeking to eat the cookies.

  Does that sound fair?

  And if you saw a note in the kitchen saying that "The responsibility for washing the dishes is on those seeking to eat the cookies", would you say this is a fair way to put this into practice?

  • If Alice wants to eat the cookies, Alice has to wash the dishes.
  • If Bob doesn't want to eat any cookies, Bob doesn't have to wash any dishes. (He's allowed to, but he doesn't have to.)

  WhatamIdoing (talk) 23:13, 27 February 2026 (UTC)

  Your question was Which part of WP:CON do you think is the equivalent/relevant guidance

  Onus is a verifiability policy it does not create new editing or consensus policies. Kolya Butternut (talk) 12:14, 28 February 2026 (UTC)

  That's (unfortunately) not true. Any policy can impose any type of rule. WhatamIdoing (talk) 00:18, 1 March 2026 (UTC)

  Here's why I was asking that:

  • If ONUS ("Alice has to do the dishes, because she ate the cookies") provides a 'unique' rule, then removing ONUS from this page means this rule ceases to exist.
  • If ONUS is duplicative, then we could remove it here and tell people to go read the same rule elsewhere ("WP:BBQ says that Alice has to do the dishes, because she ate the cookies").
  • So if there is a section in Wikipedia:Consensus that says the same thing as ONUS does here, then we could take the proposal that @Blueboar makes and point people at WP:Consensus#Cookie eaters must wash dishes, and nothing's changed in practice; we're just rearranging the deck chairs and people will usually get over it as long as the shortcut takes them to a place that says the rule is still the rule.
  • But if there is no equivalent rule there, then removing the ~18-year-old and famously fought-over ONUS rule and pointing people at a different rule – one that doesn't say that Alice has to wash the dishes – is the kind of significant WP:PGCHANGE that needs clear evidence of community consensus.

  WhatamIdoing (talk) 00:26, 1 March 2026 (UTC)

  So… my take… (to belabor your analogy) Imagine a policy called WP:Dieting… it mentions eating cookies as one of several ways to break a diet.

  Now… imagine we have a separate guideline called WP:Eating Cookies… it is appropriate for WP:Dieting to link to WP:Eating Cookies, when it mentions cookies…

  However, it is inappropriate for WP:Dieting to talk about who has to wash up after eating cookies. That should be discussed at WP:Eating Cookies. Blueboar (talk) 01:29, 1 March 2026 (UTC)

  I agree in principle. However:

  1. We haven't been able to get agreement to move the ONUS sentence to WP:CON (no matter how logical that might be).
  2. We haven't been able to get agreement to remove ONUS from Wikipedia's ruleset, and re-writing it as "please see WP:CON, which doesn't say anything like this" amounts to wholesale removal.

  So it might be appropriate, but it doesn't seem to be feasible at this moment. WhatamIdoing (talk) 02:46, 2 March 2026 (UTC)

This is a long discussion so I'm sure I've missed many details. Overall, I agree with WhatamIdoing's general take. ONUS, how ever it is worded, basically says that even if some claim has been in an article for a long time (silent/presumed consensus), if it's challenged then those who want to keep it need to either show prior consensus or establish a current consensus to keep the material in. Based on previous discussions I think it's worth adding that ONUS doesn't stand alone. Suppose I see some 10 year old claim in an article that I think is UNDUE thus I delete it. My justification for the removal can be "UNDUE" (hopefully with some additional explanation). My justification should not be "ONUS" as that implies that I don't think there was a consensus to add that text a decade back. Since the claim wasn't challenged back in the day we can assume an implied consensus. Where ONUS comes in is if someone reverts my removal saying something like "restoring long standing text" or "get consensus to remove". I can challenge that restoration by giving both my reason for removal and noting that per ONUS, this content is now in a NOCON state and thus should be deleted if we don't establish a consensus to include. If the policy text is to be edited I think it should make this more clear rather than water down the function of ONUS. IE, ONUS isn't the reason to initially remove the disputed text, rather it tells us how we should evaluate a NOCON claim regarding the content. Springee (talk) 14:41, 27 February 2026 (UTC)

I think a lot of the arguments over ONUS could be solved if it was made clear it's not a valid reason to challenge content, rather it backs up challenges (I'd even argue nothing in ONUS as it stands states you can remove content). Unfortunately I said the same thing three discussions and 100k words ago. -- LCU ActivelyDisinterested «@» °∆t° 14:55, 27 February 2026 (UTC)

• I'm late to the party and have only skimmed the (very large) discussion above, but believe that ONUS can, unfortunately, be abused to remove large amounts of text one doesn't like and argue that it's the other party's responsibility for achieving consensus for inclusion. If you want a recent example, just look at Crime in Minnesota and its revision history and talk page. Some1 (talk) 00:09, 12 March 2026 (UTC)

  On the Crime in Minnesota ONUS is not being used to remove large amounts of text, it's being used to stop the re-addition of text that is disputed on the talk page. The editor who initially removed the content even makes clear on the talk page why they removed the content, "I removed some content for bias and neutrality issues." No comment on if the content should or shouldn't be in the article, it whether consensus has or hasn't been reached to retain the content, but ONUS was not used to remove the content that was due to other concerns. ONUS was brought up once editors started to edit war over the contents inclusion. -- LCU ActivelyDisinterested «@» °∆t° 10:07, 12 March 2026 (UTC)

  Sigh… yet another reason why we should move ONUS out of WP:V… if the material is being challenged on Neutrality grounds… that is what people should focus on. ONUS is a distraction. Blueboar (talk) 13:24, 12 March 2026 (UTC)

  Onus in its current state is too narrow to be moved. What about: The Onus to achieve consensus for changes from consensus is on those seeking the change for text at Consensus? Kolya Butternut (talk) 15:47, 12 March 2026 (UTC)

  Because:

  • Nobody knows what you mean by "changes from consensus".
  • The main point of ONUS is that the content-adder and not the content-blanker has to get consensus, no matter which of them can claim that their version is the One True™ Status Quo Consensus Version.

  WhatamIdoing (talk) 19:01, 12 March 2026 (UTC)

  No that is not the point of onus. The point of onus is simply to say that verifiability is not enough to add something, and you have to get consensus. And the reality is that you have to get consensus for anything you want to do that doesn't already have consensus, it's a simple concept. Kolya Butternut (talk) 19:59, 12 March 2026 (UTC)

  The point of WP:VNOT is to say that verifiability is not enough. It's in the name: Verifiability does NOT guarantee inclusion.

  The point of WP:ONUS is to say which editor the onus is on. It's in the name: The ONUS is on someone. Whom is it on? WhatamIdoing (talk) 20:04, 12 March 2026 (UTC)

  Ok - let’s go step by step and see where we might agree, and not focus on where we disagree.

  Do we agree that when some bit of material HAS BEEN DISCUSSED, and that discussion results in a consensus to omit it from an article, it can (and should) be omitted? Blueboar (talk) 20:05, 12 March 2026 (UTC)

  I agree with that. WhatamIdoing (talk) 20:14, 12 March 2026 (UTC)

  I don't know, it depends on how pedantic we're getting. That is commonly true. Kolya Butternut (talk) 20:18, 12 March 2026 (UTC)

  Ok, wow… I thought I was starting with the basic “of course” stuff. If you are going to quibble that following a clear and unequivocal consensus is “pedantic”, then the conversation is over. We are never going to agree. I’m out. Blueboar (talk) 20:32, 12 March 2026 (UTC)

  This is a reaction to what I have been dealing with. Do you see the responses that I've gotten here and on my talk page when I try to make simple points? I tried to answer you in the affirmative while acknowledging that things like RFCs may not be considered "discussions". I have been trying to say that any kind of change from consensus needs consensus and that is being treated as a complicated concept so I feel like I can't trust my words. Kolya Butternut (talk) 21:00, 12 March 2026 (UTC)

  But I was not saying your words are pedantic; I was saying I could imagine pedantic reasons why my yes answer would not be accurate. Kolya Butternut (talk) 21:01, 12 March 2026 (UTC)

  Nobody knows what you mean by "changes from consensus".

  The same thing as "make an edit" from "previous consensus" in the chart in WP:EDITCON. Kolya Butternut (talk) 03:18, 13 March 2026 (UTC)

  How are editors to know whether a "previous consensus" existed?

  What are they to do when there is no previous consensus?

  What are we to do when the person making the blanking edit claims there was no such previous consensus, but the person wanting to re-add the blanked material claims there most definitely was? WhatamIdoing (talk) 04:30, 13 March 2026 (UTC)

  These concerns have no bearing on the interpretations of the policies I am discussing. Kolya Butternut (talk) 04:50, 13 March 2026 (UTC)

  The main point of ONUS is to solve the problem of editors arguing over who has to do the work of seeking consensus, by assigning it to the person who wants to keep the content. These concerns are the reason that sentence exists. If they're no concern of yours, then why are you here? WhatamIdoing (talk) 05:11, 13 March 2026 (UTC)

  I disagree. But I think it is clear what I mean by my suggestion. Kolya Butternut (talk) 08:17, 13 March 2026 (UTC)

• If we can not agree on what to do when there IS a clear consensus, we will never agree on what to do when there ISN’T a consensus. Blueboar (talk) 12:12, 13 March 2026 (UTC)

  I think we agree on what to do when there is a clear consensus. Kolya Butternut (talk) 14:50, 13 March 2026 (UTC)

  Ok, I’ll bite… what do you think we should do when there is a clear consensus? Blueboar (talk) 20:33, 13 March 2026 (UTC)

  When there is a clear consensus to omit something? It is removed. I am confused by the question because it is so obvious. The first exception that comes to mind is that consensus cannot override NPOV. (But of course we depend on consensus to determine NPOV.) But rather than step by step questions, can you tell me your full interpretation/opinion on Onus? Kolya Butternut (talk) 23:24, 13 March 2026 (UTC)

  Finally… a blunt answer to the obvious. Thank you.

  Now for a more nuanced discussion: WHEN a clear consensus to omit exists, but someone thinks that this consensus should change, what should happen? Blueboar (talk) 01:11, 14 March 2026 (UTC)

  If there is a pre-existing consensus to omit, that sounds like the existing version of the article is without this content, and someone wants to add it. In that case, they simply go through the normal consensus building process. This could start with a bold edit if they have a convincing justification such as a new source, or a new discussion at the talk page, etc. But can you tell me your interpretation of onus now? Kolya Butternut (talk) 02:01, 14 March 2026 (UTC)

  I don’t have an “interpretation”… I read the sentence literally: “The responsibility for achieving consensus for inclusion is on editors seeking to include disputed content.” Blueboar (talk) 15:54, 14 March 2026 (UTC)

  The dispute seems to be over the phrase "seeking to include". I interpret that to mean trying to include something that is not yet there. Do you interpret it that way or that the onus is also on people who want to maintain long-standing text with potential pre-existing consensus? Kolya Butternut (talk) 16:54, 14 March 2026 (UTC)

  Ah… ok, I would again take a literalist stance… ONUS currently says “include”… it does not distinguish between “adding new” vs “maintaining old”. Perhaps it should make that distinction, but it currently does not. Blueboar (talk) 18:09, 14 March 2026 (UTC)

  I feel like the word "seeking" suggests content that is not already there. Kolya Butternut (talk) 18:12, 14 March 2026 (UTC)

  "Seeking to include" encompasses someone who is seeking to:

  • put something new in an article,
  • restore text (long-standing or otherwise) after it's been removed, and
  • keep something from being removed from the article.

  All of these situations involve someone seeking to have material "included". "Included" is bigger than "added initially". WhatamIdoing (talk) 18:37, 14 March 2026 (UTC)

  I know what you think. I was asking what Blueboar thinks. (You're missing the point.) Kolya Butternut (talk) 19:00, 14 March 2026 (UTC)

  I would agree with WAID … “seeking to include” is not limited to new additions. The ONUS sentence is focused on “in vs out” not “new vs old”.

  Now, in any discussion about disputed material, I would certainly expect that those “seeking to include” would raise the issue of longevity (or “longstandingness”) to argue for inclusion… longevity is after all a valid (if weak) argument to keep. However, we have all seen times when that argument fails to convince. And when it does fail to convince, then a consensus to include has not been achieved - and according to ONUS the material is omitted. Blueboar (talk) 21:05, 14 March 2026 (UTC)

• Considering WP:QUO, would you say it is omitted before or after an RfC? Kolya Butternut (talk) 23:57, 14 March 2026 (UTC)

  ONUS does not care… it focuses on the end result, not what happens on the way there. Blueboar (talk) 00:29, 15 March 2026 (UTC)

  How things work in the context of Onus is something people care about. If longstanding content without explicit consensus remained in place throughout the dispute but was then removed after a finding of no consensus in an RfC, that might influence me to compromise on your interpretation of Onus. Kolya Butternut (talk) 01:04, 15 March 2026 (UTC)

  The purpose of WP:QUO is to stop edit warring during a discussion. It doesn't matter which m:Wrong version is on the page during the dispute (except for the multiple scenarios in which QUO specifically says to remove the disputed content). As long as there's no edit warring, then QUO doesn't care if long-standing non-BLP, non-COPYVIO, non-spam, etc. material is kept on the page during the discussion. WhatamIdoing (talk) 04:23, 15 March 2026 (UTC)

  Regardless of its purpose, which version remains pending an RFC, the version including the disputed long-standing content, or the version without? Kolya Butternut (talk) 12:14, 15 March 2026 (UTC)

  Do we need to care or specify which version remains? The vast majority of RfCs show consensus in days. Do we need policy that ensures some "right" version is in place for that tiny number of cases where consensus takes a while to form and the "wrong" version is in the article?

  If we decide we do need a policy, I guess I'd have to argue the version that removed disputed content -- even content that had been in the article for years without objection -- is the only way to favor the "possibly not complete" wrong version over the "possibly incorrect" wrong version. Valereee (talk) 12:47, 15 March 2026 (UTC)

  ONUS does not address which version should remain during discussion. Again, it is focused on the end result, and does not care what happens before that point.

  That said (playing devil’s advocate here), ONUS does imply that when there is “no consensus” we would default to omit (as the inclusionists have failed to achieve a consensus to include) and I suppose a deletionist could Wikilawyer that into: “during discussion, since there is not (yet) a consensus to include, we should default to temporary omit (we can always return it if the end result is consensus to keep).”

  Personally, I think that argument is a stretch… but, I could see it being argued. Blueboar (talk) 13:13, 15 March 2026 (UTC)

  Yes, I understand that Onus does not address this, but if other Wikipedia policy did, then it would affect what would happen around Onus. I think that if longstanding content is not removed until after an RfC it would make it easier to deal with civil POV pushers. Kolya Butternut (talk) 18:35, 15 March 2026 (UTC)

  Perhaps … but personally I have never understood why people care which m:Wrong version state the article is in while content is under discussion. It’s not like you win some sort of advantage in the discussion by having the material temporarily in or out.

  If we temporarily remove and the consensus is to keep, the material in question will be restored - and if we temporarily keep and the consensus is to omit, the material in question will be taken out.

  That temporary status is a pointless issue to argue about.

  So… whatever! Blueboar (talk) 19:45, 15 March 2026 (UTC)

  It depends if it makes it to the RfC stage. Kolya Butternut (talk) 19:57, 15 March 2026 (UTC)

  Why? Blueboar (talk) 23:42, 15 March 2026 (UTC)

  A POV pusher can more easily stonewall and make other editors give up if the content is removed before an RfC. It creates more shared responsibility. Kolya Butternut (talk) 00:20, 16 March 2026 (UTC)

  Are you assuming that the only POV pushers are on the remove side? Inclusionist POV pushers can stonewall just as much as deletionist POV pushers. The key is to ignore the POV pushing and get on with determining consensus. Blueboar (talk) 01:04, 16 March 2026 (UTC)

  No. I am simply discussing my concern about deletionist POV pushers. Kolya Butternut (talk) 01:07, 16 March 2026 (UTC)

  Not sure I see how it's easier to stonewall when arguing remove than when arguing keep. Valereee (talk) 12:24, 16 March 2026 (UTC)

Onus part 4

We have no rule saying which version should be visible during discussions. Anyone who believes that WP:QUO is such a rule is encouraged to go read it again. WhatamIdoing (talk) 22:40, 15 March 2026 (UTC)

Well, as I said above, I could see someone arguing that ONUS implies that the “omit” version should be visible during discussions… as there is not YET a consensus to include. But that really is a stretch. Blueboar (talk) 11:55, 16 March 2026 (UTC)

If it's a stretch, it's a concern we could work to compromise on. Kolya Butternut (talk) 13:55, 16 March 2026 (UTC)

What do you think a plausible compromise could look like? WhatamIdoing (talk) 17:12, 16 March 2026 (UTC)

I was thinking something making it clear that the status quo version should be kept pending discussion, but I realize I'd have to look at the previous discussions relating to Onus potentially contradicting other policies to come up with a compromise. Kolya Butternut (talk) 19:52, 16 March 2026 (UTC)

As long as it's not a BLP issue (or some other reason not to), I don't have an objection to favoring a last-current-stable version during discussion. If anyone is at all reasonably arguing BLP (etc.), remove during discussion. Valereee (talk) 20:07, 16 March 2026 (UTC)

The problem with the Wikipedia:Stable version concept is that it's always "my" version and never "his" version.

Also, sometimes there genuinely isn't a stable version. WhatamIdoing (talk) 00:01, 17 March 2026 (UTC)

If content has been stable for over a year that would clearly be the stable version. Kolya Butternut (talk) 01:23, 17 March 2026 (UTC)

If content has been ignored for over a year that would clearly be an ignored version. Sometimes it takes years to notice a blunder, and when someone notices it other editors presume it's more stable than it ever was. Happened twice very recently. Selbstporträt (talk) 02:42, 17 March 2026 (UTC)

Consider:

(A) Good sir, there's a bug in that sentence.

(B) Hm...yes?

(A) *Shows the bug*

(B) Oh, I see.

(A) Now what?

(B) I believe you need to reach consensus?

(A) Why would I do that?

(B) You have ONUS.

(A) I just passed by and reported a bug.

(B) You still have ONUS.

(A) Pardon me? I must admit I feel no itch...

(B) No, that's lupus. I mean ONUS: if you want to change...

(A) That's what I mean: I feel no itch! Why would I scratch it?

(B) Then I'm afraid the bug will stay.

(A) So you spend your days telling bug reporters they have have ONUS?

(B) That's right. Now, if you wish to establish CON, that's the third door to your left, then you need to take the stairs until you meet a door that is not locked, and then....

A is long gone.

Perhaps bug trackers should do what they're supposed to do. Selbstporträt (talk) 03:19, 17 March 2026 (UTC)

Totally aware. It's often a judgement call. "Stable" is subjective. But two hypothetical cases:

1. Content A added a year ago
2. Significant subsequent non-minor edits by multiple editors
3. Content A removed
4. Removal reverted
5. Discussion started

I'd call version 2 the stable version, which means the content remains in during discussion.

1. Content A added a year ago
2. Few subsequent edits except minor, or non-minor editing only by editor who added A
3. Content A removed
4. Removal reverted
5. Discussion started

I'd call version 0 the stable version (though in practice I'd go with 3 to keep from removing any non-contentious subsequent edits) which means the removal is maintained during discussion.

Obviously there's a lot of different scenarios along this spectrum (and in the case of BLP/copyvio/etc., the whole thing's moot), but just to explain my general thinking. Valereee (talk) 12:33, 17 March 2026 (UTC)

I'd call version 0 the stable version. We need another word other than "stable", because this version is no less stable than version 2 in the first scenario. Kolya Butternut (talk) 12:47, 17 March 2026 (UTC)

No objection to any other term people can agree on. Valereee (talk) 13:06, 17 March 2026 (UTC)

It sounds like you're talking about what is the consensus version, which is different than the stable version. Kolya Butternut (talk) 17:04, 17 March 2026 (UTC)

Well...it depends on how you're defining consensus version. The fact a version of a page hasn't had any objections so far doesn't necessarily mean it's a "consensus" version. It's just a version that hasn't had any objections so far. If there wasn't discussion with consensus after something was added, an objection means the presumed consensus from no previous objections is called into question. Valereee (talk) 17:28, 17 March 2026 (UTC)

I would describe a sentence that has been in an article untouched and undiscussed for a year as stable. Kolya Butternut (talk) 18:30, 17 March 2026 (UTC)

IMO again it depends on number of active editors, length of article, length of content, types of edits since the addition, etc. A paragraph to the lede in an article that gets daily editing by multiple editors, sure. A few words in a section of an article that isn't edited heavily, probably not. Valereee (talk) 19:00, 17 March 2026 (UTC)

A few words in a section of an article that isn't edited heavily, probably not. I'm not sure how this has any bearing on the concept of stability. It sounds like you're talking about the strength of implicit consensus. Kolya Butternut (talk) 19:16, 17 March 2026 (UTC)

Because you can measure stability in terms of time or in terms of the number of edits. "One year" could be "two edits". "One year" sounds big, but "two edits" is small. WhatamIdoing (talk) 20:24, 17 March 2026 (UTC)

A distinction between stability and implicit consensus is fair in the abstract, but in practice the question here is what makes a version a useful baseline while a dispute is ongoing?

That’s why factors like the level of activity on the page and the number of editors matter. Content that has persisted through active editing by multiple editors -- whether that's over one day or 30 or 300 -- has effectively been exposed to scrutiny, so a lack of objection carries more weight. Content that has remained unchanged in a low-traffic area hasn’t had that same exposure, so the lack of objection is less meaningful.

In that sense, "strength of implicit consensus" is what determines whether being unchanged is meaningful here. A definition of "stable" that looks only at whether something has remained unchanged, without regard to that context, doesn’t provide a useful baseline for deciding what should remain in the article during discussion. Valereee (talk) 11:53, 18 March 2026 (UTC)

No, strength of consensus, implicit or otherwise, determines what happens after the dispute is over. During the dispute the version that is literally stable should remain, with a couple of exceptions. Otherwise we'd have to have a discussion over what remains during the discussion, and then discuss what content remains after the discussion. Kolya Butternut (talk) 12:43, 18 March 2026 (UTC)

This thread is about how to identify which version "should" be visible during discussions.

Above, Valereee gives five numbered versions and asks which is "stable". You replied that "It sounds like you're talking about what is the consensus version, which is different than the stable version."

Valereee asked "how you're defining consensus version". (Note: "consensus", not "stable")

You replied with an apparent non sequitur about how you would determine whether a sentence was "stable".

Valereee disagrees that an sentence could be "stable" if nobody even knows that it's there. You say "It sounds like you're talking about the strength of implicit consensus".

She builds on the idea of a "distinction between stability and implicit consensus", and you tell her that implicit consensus is irrelevant.

So I'm asking you a direct question: How do you, @Kolya Butternut, define or identify 'the consensus version'? WhatamIdoing (talk) 19:57, 18 March 2026 (UTC)

That's not relevant to what I've been talking about. I don't know that I'm interested in having a discussion about it if I don't see the connection. Kolya Butternut (talk) 20:27, 18 March 2026 (UTC)

You are the person who brought up the concept of a 'consensus version'. Why did you bring it up if you "don't see the connection"? WhatamIdoing (talk) 21:00, 18 March 2026 (UTC)

I brought it up to differentiate it from what I was talking about.... Kolya Butternut (talk) 21:44, 18 March 2026 (UTC)

And if the rest of us knew what you meant by "consensus version", then we might be able to figure out how it's different from what you were talking about. WhatamIdoing (talk) 22:06, 18 March 2026 (UTC)

What I meant? I was discussing my interpretation of what Valereee was talking about. Kolya Butternut (talk) 22:10, 18 March 2026 (UTC)

"If content has been stable for over a year that would clearly be the stable version" doesn't sound like someone else's interpretation. Selbstporträt (talk) 15:14, 19 March 2026 (UTC)

You're quoting my interpretation of "stable version", not consensus. Kolya Butternut (talk) 16:49, 19 March 2026 (UTC)

I'm rather quoting the claim that started this other round of hand waving. It shows that your "that's not what I was talking about" fails to capture the scope of the exchange, which wasn't about someone else's interpretation. It also fails (or perhaps succeeds) in not making your interpretation of consensus version any more explicit.

If your own interpretation of what's a stable version doesn't imply it's some kind of consensus, why bring it that concept in the first place? If you will go where everyone will go on the matter, say so. If you have an interpretation you prefer, say which one. If you wish to keep waving your hands in front of people trained to chased down referents on a page, do continue. Selbstporträt (talk) 17:51, 19 March 2026 (UTC)

I have been very clear. I am discussing which version may be present during a dispute, which is typically referred to as the stable version. I have been clear I am not particularly interested in discussing consensus at this time. Kolya Butternut (talk) 18:07, 19 March 2026 (UTC)

Artists are driven by the tension between the desire to communicate and the desire to hide:

(D) We have no rule saying which version should be visible during discussions. Anyone who believes that WP:QUO is such a rule is encouraged to go read it again.

(E) I could see someone arguing that ONUS implies that the "omit" version should be visible during discussions… as there is not YET a consensus to include.

(F) If it's a stretch, it's a concern we could work to compromise on.

(D) What do you think a plausible compromise could look like?

(F) I was thinking something making it clear that the status quo version should be kept pending discussion...

(G) As long as it's not a BLP issue (or some other reason not to), I don't have an objection to favoring a last-current-stable version during discussion.

(D) The problem with the Wikipedia:Stable version concept is that it's always "my" version and never "his" version. Also, sometimes there genuinely isn't a stable version.

(F) If content has been stable for over a year that would clearly be the stable version.

(H) If content has been ignored for over a year that would clearly be an ignored version.

(G) "Stable" is subjective. But two hypothetical cases: *presents two scenarios*.

(F) We need another word other than "stable", because this version is no less stable than version 2 in the first scenario.

What if I told you that this other word could be "consensus"? Selbstporträt (talk) 18:30, 19 March 2026 (UTC)

I'd ask what your point is? Kolya Butternut (talk) 19:28, 19 March 2026 (UTC)

To establish the relevance of the question you are dodging, mayhaps? Selbstporträt (talk) 20:00, 19 March 2026 (UTC)

Then I would say it would be more effective communication to speak in paragraph form, first establishing context, your opinion, and then your proposal? Kolya Butternut (talk) 20:53, 19 March 2026 (UTC)

How is this other round of Questions exemplify more effective communication than retracing the whole dialogue sequence, again? Selbstporträt (talk) 21:15, 19 March 2026 (UTC)

The dialogue sequence seems to show that the conversation resulted in no progress and ended with two parallel conversations of people talking past each other about consensus and stable versions. I don't want another round of questions I want you to tell me what you think. Kolya Butternut (talk) 21:26, 19 March 2026 (UTC)

Or, Tell me what the context of your question is? Or elaborate on your question? Because it feels like you're trying to back me into answering something and I don't know why because I don't know what I have not already answered in the past. Kolya Butternut (talk) 21:34, 19 March 2026 (UTC)

I think you brought up the concept of consensus version yourself. You said you wanted to distinguish it from the concept of stable version, but won't spell out why, or even what you had in mind when you made the distinction.

I think "And if the rest of us knew what you meant by "consensus version", then we might be able to figure out how it's different from what you were talking about" is direct enough, clear, and relevant. The concept of stable version isn't quite clear, if only because "stable version" refers to an admin call. Other reasons have been put forward. I suspect you suggested the concept of stability to establish some kind of criterion for consensus, but you won't tell us that you do.

I think it is mysterious why I have to demonstrate I can follow an exchange when hours have been spent to advance at snail speed. We're not trying to decipher Chauncer here. Everything being said and done is quite clear.

Is that good enough for you, or will we need to switch to playing Columbo? Selbstporträt (talk) 22:03, 19 March 2026 (UTC)

I suspect you suggested the concept of stability to establish some kind of criterion for consensus, but you won't tell us that you do. No, I don't. Maybe I can't follow because you are reading things into what I'm saying that aren't there. Kolya Butternut (talk) 23:10, 19 March 2026 (UTC)

Saying no, I don't isn't enough here, Kolya. Perhaps you would need to follow up with something like "what I intended was...", followed by some justification. At least if clarity is what you're after.

What was the point of responding "if content has been stable for over a year that would clearly be the stable version" to "sometimes there genuinely isn't a stable version", again? Selbstporträt (talk) 23:26, 19 March 2026 (UTC)

To say sometimes there clearly is a stable version. Kolya Butternut (talk) 01:20, 20 March 2026 (UTC)

Again, I think you're reading into things. There's nothing more than what I literally said. Kolya Butternut (talk) 01:28, 20 March 2026 (UTC)

"Sometimes there is a stable version" is already implied by "sometimes there genuinely isn't a stable version".

So you wanted to say...nothing at all? Selbstporträt (talk) 01:43, 20 March 2026 (UTC)

Onus part 5

Kolya, if you would describe a sentence that has been in an article untouched and undiscussed for a year as stable, that doesn't really answer the question of which version is the consensus version, which is different than the stable version. Which of those in the scenario above is the consensus version? And if the consensus version and the stable version are different versions, then which is it that you think should be privileged? WhatamIdoing (talk) 20:26, 17 March 2026 (UTC)

That's correct; it does not answer the question of which version is the consensus version, because that is not what we were discussing. There are two different privileges: the version which remains during the dispute without consensus and the version which results after the dispute is over without consensus. Kolya Butternut (talk) 21:04, 17 March 2026 (UTC)

Perhaps there is a need to distinguish what is meant in this discussion by "stable version" and WP:STABLE, which I now read was an admin power. To answer the question that has been asked: since this version has been identified by the WP:Wrong Version, I'd be tempted to favor the consensus version. Selbstporträt (talk) 13:55, 18 March 2026 (UTC)

The important step in all of the scenarios is step 5 - the discussion. To my mind this is where ONUS kicks in. It is step 5 that determines consensus on whether the material stays in, or is omitted. Blueboar (talk) 12:53, 17 March 2026 (UTC)

Yes, this is just about which version of the article is in main space during discussion on the talk. Valereee (talk) 13:04, 17 March 2026 (UTC)

Discussions about QUO should take place there.

P.S. - QUO already says:

Edit warring to maintain a "status quo version" is still edit warring, and you can be blocked for doing this. If a dispute arises regarding which version is the status quo ante bellum, be the adult in the room and don't revert. Tag instead. There is no rule on Wikipedia that requires anyone to revert, but if the page has already been reverted to an older, pre-dispute version, then it's especially helpful if you^[a] avoid reverting to a different version.

Butwhatdoiknow (talk) 16:35, 18 March 2026 (UTC)

We need to remember that there are other options… difs can be linked… the paragraph/section containing the disputed text can be copied onto the talk page so people understand what is being discussed. Obsessing over what version is “live” while the discussion is happening is Wikilawyering at its worst. It really doesn’t matter. What matters is the end result of the discussion. Focus on that. Blueboar (talk) 14:09, 17 March 2026 (UTC)

Totally. The pending state could the last state before the bug report, but a fallback solution, if negotiation fails, could be the last state where CON level has been the highest: the last important discussion, in terms of participation and amount of justifications. The need to preserve justifications from editing history (a "corporate memory") is felt all the time. Hence the new list on the ONUS discussions over the years.

It's hard to know what's going on with the development of a text without easy access to the reasons why it has become what it currently is. As long as the bug is solved and the focus is on justifications and not on voting, all should be well among editors who keep trying to solve a problem instead of getting their way. CON then wins as it should, and NPOV is preserved.

The problem here is that NPOV and CON are symmetric while ONUS is not. Symmetry is more stable. It's only through CON that ONUS makes any sense. Which means the problem is to find CON states. CON levels should help with that, at least conceptually. Selbstporträt (talk) 14:33, 17 March 2026 (UTC)

That is what I'm focussed on, too. Just trying to explore ways to find something that works for folks who have a different focus. Valereee (talk) 15:16, 17 March 2026 (UTC)

Here could be one way: put the ONUS section in BURDEN, and tell editors not to confuse the two. ONUS is related to BURDEN, but is truly about CON. Then add ONUS in CON too, and bring back the explanation of CON in terms of states, levels, or whatnot.

Where's the difficulty? Selbstporträt (talk) 15:45, 17 March 2026 (UTC)

Onus is a conduct policy in a content policy Kolya Butternut (talk) 18:33, 17 March 2026 (UTC)

That's what the solution solves:

Step 1: Add something like this to the end of BURDEN section:

The responsibility to demonstrate verifiability [i.e. BURDEN] should not be confused with the responsibility for achieving consensus for inclusion of disputed content [i.e. ONUS]. For more on WP:ONUS, see CON.

Step 2: Invert the redirect, and say

WP:BURDEN redirects here. For the responsibility to reach inclusion for disputed content, see WP:ONUS.

Step 3. Go put WP:ONUS in CON, where everybody here agrees it should belong.

So what I said obtains: those who want ONUS to be touched here should be happy. Those who want ONUS in CON should be too. And we have no duplication of content.

What were the arguments against the move? Selbstporträt (talk) 19:00, 17 March 2026 (UTC)

The argument against the move is probably not wanting to cement a sentence people don't agree on, so it's not about to move itself. I wanted this change. Even if you don't agree on my proposed section for consensus, it seems like a good place to link from at V. Kolya Butternut (talk) 19:21, 17 March 2026 (UTC)

Perhaps a nimble proposal sells the move too short. The whole "other issues" section isn't great. A Consensus subsection would work best in Verifiability and other principles, with the two points you mention.

"Principle" is too strong. "Other issues" is too weak. Redflag isn't just "another issue". This is not a great page. Selbstporträt (talk) 19:38, 17 March 2026 (UTC)

Onus is a conduct policy in a content policy. So is WP:BURDEN. WhatamIdoing (talk) 20:16, 17 March 2026 (UTC)

Content means nothing without conduct. It's a living website. It rests on editors. It would only maintained by LLMs that it'd still be true. If we froze the website and launched it into space so that it reaches the limits of our galaxy, then OK: content stands alone. But only once we can't reach it with Wi-Fi, and only if we can make sense of the final product without any editing trace.

In the beginning was a bold edit, BOLD was with an editor, the editor was BOLD. Selbstporträt (talk) 15:04, 18 March 2026 (UTC)

BURDEN directly relates to Verifiability… ONUS does not. Blueboar (talk) 20:02, 19 March 2026 (UTC)

It really doesn’t matter. The real life consequence is that editors may give up before it gets to an RFC, so the live version has power. Kolya Butternut (talk) 18:35, 17 March 2026 (UTC)

That's equally true for everyone involved, though, and it's a self-resolving problem: whoever didn't give up can change the page to whatever they want later, especially if there is no major discussion that rejects what they want.

I think the question here is why we should we bias the rules against change? The point of a wiki is that it's supposed to be quick and easy to change the contents of a page.

We really only have four options:

1. Whoever wants to include it has to get consensus to include.
2. Whoever wants to exclude it has to get consensus to exclude.
3. Whoever wants to change it has to get consensus to change.
4. Whoever wants to keep it the same has to get consensus to keep it the same.

You seem to have settled on #3, with #2 probably being your second choice, but why? Why do you think that an anti-change and/or pro-keeping-possible-errors is better than the conservative approach of "when in doubt, throw it out" or encouraging editors to edit freely? WhatamIdoing (talk) 20:22, 17 March 2026 (UTC)

Number 3, as in whoever wants to change it from consensus has to get consensus? That's obviously true. I understand determining consensus is the hard part.

Kolya Butternut (talk) 20:51, 17 March 2026 (UTC)

The easy part is to accept that ONUS works for any modification: addition, deletion, reordering, etc.

The less easy part is to accept that ONUS only needs to beat the previous CON, which should not require an AfC if that's just two-three editors having worked on a page after a year or two of waiting.

But the most important part is to make sure ONUS preserves NPOV: this is not a political exercise, we are simply trying to be faithful to the task, be it representing the world or Wiki principles. Selbstporträt (talk) 21:51, 17 March 2026 (UTC)

ONUS does not work for any modification: addition, deletion, reordering, etc. ONUS only speaks to addition/removal questions. It has nothing to say about other kinds of edits, such as rearranging or ordinary copyediting. WhatamIdoing (talk) 22:08, 17 March 2026 (UTC)

Wait. I could restructure the whole V page? That is, I could put ONUS under BURDEN and I have no ONUS to beat?

"Verifiability in" fails MOS: we're not supposed to reuse titles. The repetition is probably due to the miscellaneous section that indicates an inconsistency in the naming scheme. Selbstporträt (talk) 22:24, 17 March 2026 (UTC)

You can restructure the whole WP:V policy per WP:PGBOLD. And then someone can revert you per WP:PGBOLD and WP:CON.

ONUS only applies to articles, so it was never going to be relevant for restructuring anything outside the mainspace. Similarly, policies and guidelines are exempt from the MOS per the WP:NOTPART policy. WhatamIdoing (talk) 22:49, 17 March 2026 (UTC)

Taking public facing page as an article would be so much simpler. No need to duplicate BOLD for policies. So...NPOV doesn't apply here?

If CON applies on policy pages and there's no ONUS, that seems to change CON for policy.

I could revert the reverter and that reverter would have to stop first. Selbstporträt (talk) 23:07, 17 March 2026 (UTC)

NPOV only applies to the mainspace.

There are rules that apply everywhere (e.g., WP:COPYVIO) and rules that technically don't apply but will mostly be enforced anyway (e.g., WP:ACCESS), but mostly policies and guidelines do not apply to policies and guidelines. WhatamIdoing (talk) 00:18, 18 March 2026 (UTC)

Debatable whether Consensus only applies to content. Where else is the policy which describes the policy for changing policy? These pages do, however, need to comply with Wikipedia's legal and behavioral policies, as well as policies applicable to non-content pages. For example, editors may neither violate copyrights nor edit-war anywhere on Wikipedia. Kolya Butternut (talk) 01:12, 18 March 2026 (UTC)

That page is more or less what Larry wrote in March 2001. It shows he distinguished IGNORE from BOLD, and "no objections" from "widespread consensus". One could interpret an ONUS only for articles as an implicit assumption from CON that only needed to be spelled out after too many wars. Or as an appendage that serves no real function except prolong stonewalling. Either way, I would click on a link to the Consensus page to know more, and it isn't there. Selbstporträt (talk) 13:43, 18 March 2026 (UTC)

Not sure how much I agree that the live version has power, but for sake of argument: to the extent that's true, it reinforces the importance of what counts as our stable baseline. If being live carries weight, it seems more appropriate to give that weight to a version that has remained stable under scrutiny rather than to one that has remained static due to lack of scrutiny. Valereee (talk) 12:26, 18 March 2026 (UTC)

Wikipedia:Consensus applies to almost everything.

Wikipedia:Policies and guidelines, which is what you quoted from, applies to policies and guidelines as well as, in part, to some related pages (e.g., info pages, process pages, essays).

The three Wikipedia:Core content policies, which are Wikipedia:Verifiability, Wikipedia:No original research, and Wikipedia:Neutral point of view, apply only to the mainspace. WhatamIdoing (talk) 03:58, 18 March 2026 (UTC)

That depends on who gives up first…If we “include during discussion” and the inclusionists give up, then a consensus to omit will quickly form … which means the live version that included had no “power”.

Alternatively, if we “omit during discussion” and the deletionists give up, a consensus will quickly form to return the material… which means the live version that omitted had no “power”. Blueboar (talk) 20:38, 17 March 2026 (UTC)

Another kind of "onus" will actually be on the people who want to change the live version. The moving party has more work to do. Kolya Butternut (talk) 20:55, 17 March 2026 (UTC)

Meh… Not sure if you know ice hockey, but it doesn’t matter how many times your team wins the face offs if you can’t put the puck in the net.

Having “your” version be the live version is like winning face offs. Perhaps you have a momentary advantage, but to win the game, you have to score goals. Having “your” version gain consensus through discussion is scoring goals. Blueboar (talk) 21:40, 17 March 2026 (UTC)

The analogy doesn't apply well because teams rarely forfeit hockey games. Kolya Butternut (talk) 22:06, 17 March 2026 (UTC)

A team can stop skating and just tank for the season. Selbstporträt (talk) 23:20, 17 March 2026 (UTC)

I think the main "powers" of the during-discussion version are:

• If the discussions are lengthy, then The Right™ Version is seen by readers during the discussions. Discussions can be dragged out for months.
• If the discussions don't form a consensus for a specific change, then editors may leave it in place afterwards. Leaving it in place (doing nothing) can't re-trigger the dispute, but removing it will appear on people's watchlists.

WhatamIdoing (talk) 21:27, 19 March 2026 (UTC)

Onus is a behavioral policy. Kolya Butternut (talk) 10:37, 18 March 2026 (UTC)

Perhaps ONUS should be a behavior policy… but right now it is a provision of a content policy. That’s one reason why I keep saying it doesn’t belong in WP:V, and we should move it. Blueboar (talk) 11:56, 18 March 2026 (UTC)

And it is exactly as much of a behavioral policy as WP:BURDEN is. Both of them identify who has to do something, and therefore who does not. WhatamIdoing (talk) 19:34, 18 March 2026 (UTC)

I would, and have on multiple occasions, agree that it shouldn't be in V, unfortunately any discussion about moving it gets bogged down immediately as some editors want to change it. There's no agreement on changing it, so there can't be any consensus on moving it, and instead a new novel length discussion is created every few months or so. All the while the majority of the community seem to use and accept ONUS on a general basis. -- LCU ActivelyDisinterested «@» °∆t° 00:07, 13 March 2026 (UTC)

Remove "Verifiability and other principles" section?

I wonder whether the entire Wikipedia:Verifiability#Verifiability and other principles section could be removed. It's supposed to be a convenient summary of related policies and guidelines, but I think it ends up having problems with WP:REDUNDANTPOLICY. WhatamIdoing (talk) 04:00, 28 April 2026 (UTC)

We might be able to shorten and summarize it better, but I do think the section is important. WP:V is often the first policy page that new editors are pointed to, and it helps to highlight that there are other core policies that might impact their editing. Blueboar (talk) 12:49, 28 April 2026 (UTC)

What do you think about cutting it down to this?

==Verifiability and other principles==

This policy interacts with other policies. For example:

• Editors must write in their own words to avoid violating Wikipedia's copyright policy and avoid plagiarism. Never link to any web page that violates other people's copyrights.
• All articles must comply with Wikipedia's neutral point of view policy. Tiny-minority views need not be included, even if they are verifiable. Sources themselves do not need to maintain a neutral point of view.
• Wikipedia has separate, stand-alone articles only on notable subjects. Notability is based on the existence of suitable sources, not on the state of sourcing in an article.
• The no original research policy is focused on whether a reliable published source could be cited for a given piece of content, rather than on whether or not it is already cited in the article. Sources must support the material clearly and directly. Do not draw inferences from sources to advance a novel position.

(This is the same set of policies and guidelines that we already have on the page.) WhatamIdoing (talk) 17:25, 28 April 2026 (UTC)

Still not seeing any consensus for this edit

I still don't see any consensus for adding "independant" [sic] to the bullet point on secondary sources. Supposedly this edit was discussed, but I have not found any discussion in the archives about this. Sławomir Biały (talk) 12:38, 28 April 2026 (UTC)

It may not have a dedicated thread of its own (which would make it hard to find in the archives) but the need for independent sources certainly has been discussed many times. It tends to be brought up in just about any discussion relating to the need for secondary sources.

I also know that the need for independent sourcing has been discussed a lot at both WT:RS and at WT:NOTE (so you might have better luck finding a dedicated thread about it by looking through the archives of those pages). Also try the archives at WP:Village pump (policy) Blueboar (talk) 13:12, 28 April 2026 (UTC)

Independence appears to arise primarily in the context of establishing notability for a subject. I think outside of that context, independence is somewhat ill-defined. For example, often the best secondary sources are those written by established subject-matter experts. Requiring sources to be "independent secondary sources" could be seen as ruling out some of the best sources. I would happily look at any specific consensus that all secondary sources have to be independent. But I suspect that no such consensus exists. As a practical matter, the particular bullet point directs to WP:PSTS, which is about primary/secondary sourcing, not independence (which is only discussed in notability guidelines). Finally, the only place in the policy pantheon that independence is addressed in a sufficiently general way to pretend to apply more generally is the (pretty useless) WP:INDEPENDENCE essay, which appears to accept as given that sources that are not "independent" are in some way "promotional". So, if independence is to be worked into all of our policies, I think it should be done through a clear consensus of what the scope and meaning of that new requirement is. Sławomir Biały (talk) 13:23, 28 April 2026 (UTC)

For example, often the best secondary sources are those written by established subject-matter experts. Requiring sources to be "independent secondary sources" could be seen as ruling out some of the best sources. If someone is WP:WIKILAWYERING to somehow say being a subject-matter expert alone makes you not independent, that would pretty much be a non-starter anyways. KoA (talk) 13:51, 28 April 2026 (UTC)

For example, some of the best sources on Bott periodicity are by Raoul Bott and his collaborators, especially Michael Atiyah. Are those sources "independent" for purposes of this proposed wording? If so, what definition of independence is being used? If not, then the proposed wording seems too broad. Sławomir Biały (talk) 13:59, 28 April 2026 (UTC)

They're not really secondary either, so I don't think this wording change has much effect on your specific example. MrOllie (talk) 14:35, 28 April 2026 (UTC)

Bott (1970), "The periodicity theorem for the classical groups and some of its applications" is absolutely a secondary source, for example. Sławomir Biały (talk) 14:55, 28 April 2026 (UTC)

Yup. I see 'independent' in a guideline is widely endorsed. The new policy addition, not so. It IS frequently misrepresented as being actual policy, such as by {{AFC submission|nn}}. Something I'm trying to fix, but the discussion is being suppressed. RememberOrwell (talk) 17:25, 28 April 2026 (UTC)

If only other people could figure out what you have been harping on at WT:N. For example, {{AFC submission|nn}} does not contain the word policy, so how could it be misrepresenting anything "as being actual policy"? WhatamIdoing (talk) 17:47, 28 April 2026 (UTC)

Already answered. See Notability#Topic. RememberOrwell (talk) 18:17, 28 April 2026 (UTC)

Mind the gap between "I answered" and "Any human on the planet understood what I posted". I have read every word you posted at WT:N, some of it more than once, and I still have no idea what your actual problem is. I'm seeing no evidence that anyone else does, either. WhatamIdoing (talk) 19:10, 28 April 2026 (UTC)

Bott would be considered independent for the purpose of Bott periodicity. Any researcher working on a mainstream/generally accepted fact about a scholarly subject is treated as an independent source unless and until proven otherwise. Sure, it's named after him, and some inexperienced editor might claim that he was just doing all that math out of a desire for self-promotion, but the rest of us will just laugh at such claims. Bott would not be independent for directly autobiographical statements ("I developed this theorem in the summer of 1962") but is independent for the math itself. To put it more simply, in a form we've used in the past: No source has a conflict of interest with Algebra; all reliable sources are independent of the math itself.

To give you an idea of the metes and bounds, a pharma company's website is non-independent for a newly approved drug, but research done at a university hospital that the same pharma company paid for is usually considered (barely) independent, and editors have even claimed that the medication package insert that the pharma company directly wrote should be considered independent (on the grounds that the regulatory agencies have some say in the contents). We don't usually require complete and total disconnection. WhatamIdoing (talk) 17:38, 28 April 2026 (UTC)

My concern is that "independent" is being treated as self-explanatory when it is not. It may be clear for a company writing about its own product, or a lodge historian writing about his own lodge. It is much less clear for secondary scholarly sources by people closely involved in developing a subject. Those sources may be reliable and secondary, and may even be "independent" in the relevant sense, but that conclusion depends on an unstated definition. If by "independent", we simply mean that there is no direct conflict of interest (such as financial or institutional affiliation), that seems pretty uncontroversial, but it should be said and not assumed that this is what we mean. Sławomir Biały (talk) 18:02, 28 April 2026 (UTC)

I mean, nothing is self-evident; none of our policies are self-enforcing and they all involve judgment calls. If we used the same logic you're applying here to independence, we wouldn't be able to require that sources are reliable, either, because surely reliability is not always self-evident. We can trust the community to hash out disagreements about independence when they come up; but the key point is that we all agree that independence matters, and that writing an article using non-independent sources is not appropriate, in the same way that we might disagree about what qualifies as a WP:RS but are clearly in agreement that reliability is required. Including it here means that an editor can say "this article needs to be revised, because its sources are not independent" and someone who objects has to defend the sources as independent; they can't rebut that argument by saying "no, independence isn't required" and refuse to even engage with the question of whether the sources are actually independent. --Aquillion (talk) 18:38, 28 April 2026 (UTC)

Yes, but slightly a straw-man. We have an entire policy devoted to explaining what "reliable sources" means. This is not just treated as if it were self-evident. "Independent" just invites over-interpretation in my view. It's like saying we should just free-wheel with what "notability" means (even though Wikipedia's meaning is famously a term-of-art). Just as, in this case, "independent" means "lacking conflicts of interests like financial interest or institutional affiliation". That's not the WP:COMMONNAME sense of "independent", but is not actually explained in the policy! Sławomir Biały (talk) 19:01, 28 April 2026 (UTC)

I have to admit, the removal of the independent language comes across as fighting something the community has already widely endorsed throughout WP:PAG, so I think Blueboar's description of community stance on this is accurate. That minor addition is not controversial at all in that context. KoA (talk) 13:41, 28 April 2026 (UTC)

That bullet currently points readers to WP:PSTS, which concerns the primary/secondary/tertiary distinction, not source independence. Sławomir Biały (talk) 13:56, 28 April 2026 (UTC)

To give a quick example of why “Independance” is needed:

1) say a correspondent reports on an event where he was present (say a political protest that got out of hand)… his account of what he witnessed is primary… his reporting on what others say they witnessed is secondary. What makes the report usable is that the reporter is not directly connected to the protest and is thus considered independent. Blueboar (talk) 14:12, 28 April 2026 (UTC)

His reporting on what others say they witnessed is probably an eyewitness report, and therefore WP:PRIMARYNEWS. If, however, he combined eyewitness reports to provide some analysis of the even, then that analytical content is probably secondary source material. WhatamIdoing (talk) 17:49, 28 April 2026 (UTC)

There doesn't seem to be a problem there, and of course independent secondary sources are the gold standard described in the first sentence. No one should be reaching for non-independent secondary sources, so I really don't see this line of argument getting any traction. KoA (talk) 14:13, 28 April 2026 (UTC)

I guess I don't know what "non-independent secondary sources" means. It appears to invite misunderstanding (see my example above). The key distinction is reliable sources, and secondary sources. "Independent" is ill-defined for general sources (outside of, say, maybe things like, news items?? The example of a protest does not seem particularly illuminating for articles about, say, Bott periodicity. And if we are saying something that only applies to articles about protests, it doesn't seem helpful to gloss it in this way.) Sławomir Biały (talk) 14:21, 28 April 2026 (UTC)

Another example are press releases issued by corporations… the fact that they may be repeated (verbatim) in secondary media sources does not mean we can build an article based primarily on them. Blueboar (talk) 15:35, 28 April 2026 (UTC)

One last example… as you can see from my user page, I am employed as the curator of a Historical Society attached to a very old and somewhat prestigious Masonic Lodge. Part of my job involves giving public lectures on the history of Freemasonry and that specific lodge. So you could call me a “professional” historian (although not an academic one - I only have a BA).

Part of my job also included writing a history of that lodge for the historical society’s webpage. It is a secondary source (built on reviewing primary source material such as lodge meeting minutes and other records) and I did my best to make it accurate and reliable.

However, as I am employed by the historical society (which is directly connected to the lodge), my work is not independent. If someone were to try to write a WP article about this lodge (I would call it “marginally notable”, so that could be a possibility), that article should not primarily be based on my work. We might be able to cite my work for a few specific historical facts, but we would need independent sources for the bulk of it. Blueboar (talk) 16:16, 28 April 2026 (UTC)

Ok, so independent means not having a direct conflict-of-interests, like citing a publisher's own opinions of itself. But a researcher writing about their own area of specialization is independent? Sławomir Biały (talk) 16:54, 28 April 2026 (UTC)

Often but there will always be some exceptions. For example if it is someone trying to take personal credit for an disputed invention (like if one were to cite Newton or Leibniz directly for the inventor of Calculus). MrOllie (talk) 17:50, 28 April 2026 (UTC)

• Alice Expert writing about expertise is an independent source.
• Alice Expert writing about her own prior publications is a non-independent source.
• Alice Expert creating a meta-analysis of publications in her field is an independent secondary source.
• Alice Expert creating a meta-analysis of her own prior publications is a non-independent secondary source.

See also Wikipedia:Secondary does not mean independent. WhatamIdoing (talk) 17:53, 28 April 2026 (UTC)

Yeah, I don't support an interpretation of independence that would rule out citing an expert who writes a survey in which her own work features. A personal research statement? Probably ruled out on other grounds. Sławomir Biały (talk) 19:15, 28 April 2026 (UTC)

No policy has never said that non-independent sources can't be cited, and nobody in this discussion has said that, either. "Base articles largely on independent and secondary sources" does not mean "Cite exclusively independent and secondary sources". WhatamIdoing (talk) 19:19, 28 April 2026 (UTC)

Blueboar, that's a great example. WhatamIdoing (talk) 17:54, 28 April 2026 (UTC)

"Secondary media sources" aren't really a 'thing'. There are publications (e.g., Annual Reviews (publisher), university-level textbook publishers) that happen to focus on secondary sources, but all sources are primary for something. WhatamIdoing (talk) 17:51, 28 April 2026 (UTC)

• I think it should definitely be included; even the discussions here all acknowledge that independence is required. People above are disagreeing on what qualifies as independence, but that's fine, that sort of disagreement is true for every part of our sources - we disagree on what's a WP:RS sometimes, too, or on what qualifies as WP:BLP-sensitive, or all sorts of other things. And sometimes, inevitably, it's a sliding scale and not a red line. But it is clear that everyone agrees that "the sources for this article are not independent" is a valid objection, and that, facing this objection, the correct answer (if you think the sourcing is fine) is to say "no, they are independent" and not to say "no, independence doesn't matter." All that including it here is saying is that you have to base your argument on those terms - you can defend whatever you like as independent, and the community will decide as needed on a case-by-case basis, but the community seems to universally agree that you can't assert that independence is irrelevant. --Aquillion (talk) 18:38, 28 April 2026 (UTC)

What is "independent"? What should be included? We have an entire policy on reliable sources. Even just a simple working definition. For example, "independent" doesn't mean that the sources of an article are independent of each other, independent of the subject matter, or even independent of sources that have an interest in the subject matter (e.g., the originator of a concept often has written secondary sources that are the best sources). It means something very specific: independence requires that the sources have no conflict of interest, as narrowly defined by a financial or institutional affiliation with the subject. In fact, none of the common definitions of the word fit this very narrow usage. Sławomir Biały (talk) 19:07, 28 April 2026 (UTC)

@Sławomir Biały, you've been editing for 17 years now. In all that time, have you never encountered the Wikipedia:Independent sources page? It's linked four times just in this policy. WhatamIdoing (talk) 19:14, 28 April 2026 (UTC)

So "independent" means what that essay says? That is not clear from the policy text. If the intended meaning is lack of obvious conflict of interest, self-promotion, or direct affiliation with the subject, then I think the policy should say that, or at least link directly to a definition. Sławomir Biały (talk) 19:23, 28 April 2026 (UTC)

Or at least link directly to a definition, which it already does four times. Do you mean that every single instance of the word independent (and related variations) should be linked? That would be something like eight times on this page. I've thought about this in the past, because people tend to read just a few isolated lines in a policy, but I have hesitated because WP:OVERLINK is so ingrained in editors' minds. WhatamIdoing (talk) 19:27, 28 April 2026 (UTC)

The existing uses of "independent" in WP:V do not all mean the same thing. In "reliable, independent publications" under WP:SPS, independence means independent of the author/self-publisher, i.e. some external publication or review structure. In "reliable, independent, published sources", the meaning is less clear and seems to be a broad best-sources ideal. In the notability paragraph, "independent" is the familiar notability term of art: independent of the subject.

The proposed edit would add "independent" to a sentence on WP:PSTS. So I don't think the existing scattered uses of the word settle the question. If "independent" is meant in the notability sense, or in some COI/self-promotion sense, the policy should say so explicitly rather than relying on an already overloaded word. Sławomir Biały (talk) 19:32, 28 April 2026 (UTC)

Every use of the word independent (also third party) in this policy could be correctly linked to WP:INDY. It's not an overloaded word so much as it's one with different facets of a single whole, and in a different context, this side or that side may be more salient. WhatamIdoing (talk) 20:03, 28 April 2026 (UTC)

I think that it is clear that WP:INDEPENDENT enjoys enough support to be the rough working definition - which, to be clear, it already is, regardless of how you feel about it; this change doesn't alter policy, it just makes it more clearly reflect practice. You keep calling INDEPENDENT an essay, as if to dismiss it, but essays enjoy different degrees of acceptance and INDEPENDENT is near the top of the range. It is simple fact that WP:INDEPENDENT is used as one of the major guides for writing articles, especially ones where sources may lack independence from the subject. And I think it's fine to have something like that defined in an essay; it feels like you're trying to argue for a more clear-cut red-line "this is independence", but our policies and guidelines are best when they lay out broad principles while leaving enough room debate fuzzy areas on a case by case basis - otherwise you end up with policies that are used as rhetorical bludgeons rather than for consensus-building. --Aquillion (talk) 20:27, 28 April 2026 (UTC)