Wikipedia talk:Consensus

“ Consensus is a partnership between interested parties working positively for a common goal. ” — Jimmy Wales

rv pending resolution of discussion

user:Nikkimaria, in my edit summary I said "Some wp:Verifiability editors may argue that wp:ONUS applies to more than verifiability. But there is no consensus that this is so." Can you point to a discussion that puts into question the content of the very limited text you have reverted? If so, please provide a link to a talk page post that argues content should be included when Verifiability is at issue. - Butwhatdoiknow (talk) 01:32, 2 December 2022 (UTC)

This ongoing discussion calls into question the limiting of ONUS to only verifiability-related disputes. Nikkimaria (talk) 01:35, 2 December 2022 (UTC)

What if we keep the text and link to wp:BURDEN instead? - Butwhatdoiknow (talk) 03:35, 2 December 2022 (UTC)

That's certainly closer, but the text there is a bit more nuanced than what was proposed here. Nikkimaria (talk) 00:38, 3 December 2022 (UTC)

Please help me wp:FIXTHEPROBLEM you see. What should it say to accurately reflect ONUS or BURDEN? - Butwhatdoiknow (talk) 01:37, 3 December 2022 (UTC)

Per BURDEN, material lacking a reliable source that directly supports it is excluded. Nikkimaria (talk) 01:49, 3 December 2022 (UTC)

So "If the dispute relates to verifiability and the disputed content lacks a reliable source then it is excluded"? - Butwhatdoiknow (talk) 02:26, 3 December 2022 (UTC)

If material lacks a reliable source supporting it, it is excluded. Nikkimaria (talk) 04:11, 3 December 2022 (UTC)

Of course WP:ONUS is not limited to Verifiability related disputes. The entire POINT of the ONUS section of WP:V is to remind editors that there is more to inclusion/exclusion than just Verifiability. Blueboar (talk) 02:10, 2 December 2022 (UTC)

Well, that's the first point. Then there's the second point, which is whatever the "include" sentence means. - Butwhatdoiknow (talk) 03:35, 2 December 2022 (UTC)

Revert number 2

user:Kolya Butternut, with regard to your revert, I respectfully suggest that the text does not conflict with the RFC because the text applies here only after (a) a dispute has arisen, (b) a good faith discussion has taken place and (c) the discussion has resulted in no consensus that the content is not verifiable. This circumstance would be unlikely to occur for verifiable content because an editor would supply the verification and the discussion would end with consensus that the problem is solved. - Butwhatdoiknow 21:24, 3 December 2022 (UTC)

I restored your edit. I did not carefully read it in context. I have no opinion on your addition at this time. (Your ping didn't work.) Kolya Butternut (talk) 22:38, 3 December 2022 (UTC)

Thank you for being willing to look at this issue dispassionately. Too few editors share that trait. - Butwhatdoiknow (talk) 23:50, 3 December 2022 (UTC)

@Kolya Butternut I have more evidence then you can change or convince anyone of granting it any further 2600:6C54:7800:2E20:3441:E9DF:BDC0:C52D (talk) 05:24, 14 December 2022 (UTC)

Revert number 3

Regarding this revert, I fail to see the need for this clarification. Typically if disputed content lacks support from a reliable source, it will fail WP:V and not be subject to a discussion. Though there could be exceptions, it is a WP:CREEP move to specify what to do in rare circumstances. --GoneIn60 (talk) 21:00, 14 December 2022 (UTC)

Note: Comment below added from a separate subsection

@GoneIn60, you raise three concerns in your revert:
(1) "Shorthand links are unnecessary changes." I must admit to ignorance, please tell me what a "shorthand link" is.
(2) Grammar. I would hope that, once we get past your other concerns, you will help me wp:FIXTHEPROBLEM rather than simply object.
(3) Not an improvement. As you may read in the edit summary for one of the edits you reverted, the purpose of the edits was to fix a problem posed above: editors applying NOCON in the midst of ongoing discussions (before there is "no consensus"). With that explanation, do you still think my edits were not an improvement? - Butwhatdoiknow (talk) 21:58, 14 December 2022 (UTC)

Discussion determining text subject to GoneIn60's third concern

Butwhatdoiknow, hope you don't mind but I moved your comments here. The comment about shorthand was a reference to the "more targeted" link (as you phrased it), as seen here. I think "external links" should continue to point to external links, and a different phrase like "disputed links" would make more sense to redirect to ELBURDEN, or simply state "See ELBURDEN" at the end in an additional sentence. While those are acceptable workarounds, I don't really see a need to make any changes, which also goes along with your concern in #2. As for #3, see my comments above regarding CREEP. I don't yet see a convincing case that shows why this is a helpful addition; as of now it seems like unnecessary bloat. Would you mind explaining why you think it's necessary? --GoneIn60 (talk) 22:21, 14 December 2022 (UTC) @GoneIn60, your comments at the beginning of this section appear to refer to your edit removing pre-existing text regarding BURDEN. My three points were in response to your earlier revert undoing my edit changing wording without adding any new content. I don't see how CREEP applies to my edit. - Butwhatdoiknow (talk) 23:53, 14 December 2022 (UTC)

Yes, in retrospect, I probably shouldn't have moved your comments to this section. My apologies. The comment about CREEP refers solely to your edit here. As for the other changes you made in an attempt to add clarification, I still found those to be unnecessary. Let's look at one example:

In discussions of proposals to add, modify, or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit.

PaulT implied that the wording here could be confusing to newer editors, that it might encourage the deletion of "contested content" prior to discussion rather than after. I think we need to drill into this further before concluding that a change is warranted. This is a policy page, and the text in this example has essentially remained unchanged for over a decade. --GoneIn60 (talk) 04:22, 15 December 2022 (UTC)

You do not say the proposed change would make NOCON worse, just that it won't improve NOCON. If that is the case - and maybe I've misunderstood - then I gather that the rationale for your objection is that we shouldn't change the text of policies unless the new text is an improvement. Do I have that right? - Butwhatdoiknow (talk) 06:05, 15 December 2022 (UTC)

Before we get into the weeds of whether or not a specific change is good, bad, or meh, I think we should first establish the need. Is there a need for change? I am of the mindset that if it ain't broke, don't fix it. --GoneIn60 (talk) 06:17, 15 December 2022 (UTC)

PaulT and I think it is broke because folks are citing NOCON during active discussions. I'm having trouble seeing why you don't take the position that "well, I don't see the problem but they do, and their change doesn't make things worse, so I'm not going to object just because it is a change." I would think that position would be particularly attractive to someone who recently said "it's worth pointing out that NOCON favors the longstanding version after discussion fails." - Butwhatdoiknow (talk) 07:16, 15 December 2022 (UTC)

@GoneIn60, I would appreciate you responding to the substance of my post above: what is so sacred about the existing text that it must not be edited to solve what you perceive as an insignificant problem? This is not the case of adding a new topic (in which case CREEP would come into play), it is simply a matter of changing the text of an existing topic. If it does no harm, why object? - Butwhatdoiknow (talk) 15:22, 15 December 2022 (UTC)

The "why" in my objection has been explained already, and there are additional comments below explaining this further that you may find helpful. There is a valid concern that the cosmetic changes may be unnecessarily repetitive warranting further discussion. I'm not opposed to changing existing text; I'm opposed to changing existing text unnecessarily. --GoneIn60 (talk) 16:48, 15 December 2022 (UTC)

Have we established that "folks are citing NOCON during active discussions" is indeed a common pattern, or are these simply anecdotal accounts by a handful of editors? And if that is indeed an established trend, then is it a problem with the wording in NOCON, or is it simply an educational issue involving inexperienced editors? I think it's important to establish what exactly the problem is before attempting to find a solution. BTW, you didn't mention you personally observed the citing of NOCON during active discussions until just now (unless you're referring to an earlier thread I didn't participate in). --GoneIn60 (talk) 07:30, 15 December 2022 (UTC)

It's both.

I've done a few inappropriate reverts because I thought NOCON applies to any longstanding changes, interpreting its wording that longstanding text shouldn't be changed or removed until there's a positive consensus to do so.

I've also seen a few major edit wars, one currently discussed at WP:AN, apparently stemming from a similar misunderstanding.

While you're correct that it's an "educational issue" to an extent, I don't think it excuses the policies being worded in an unnecessarily arcane way. PaulT2022 (talk) 13:07, 15 December 2022 (UTC)

Absolutely. If there's widespread confusion that can be linked to the wording, we should definitely do something about it. The concern was that we were jumping immediately to that conclusion without hashing it out first. The first line of NOCON states:

"No consensus" occurs when good faith discussion results in no consensus to take or not take an action. What happens next depends on the context:

To me, if there's confusion about each bullet point that follows, perhaps this initial statement is being glossed over; "results in no consensus" is essentially what Butwhatdoiknow was attempting to add to each bullet point, but is that really necessary repetition? "What happens next" is also a key indicator that the actions listed should follow discussion. --GoneIn60 (talk) 16:05, 15 December 2022 (UTC)

PaulT and I answer that question "yes." Now, here's a question for you: what is the downside to the repetition? It adds an insignificant 8 bytes to the article size (even less now that the BURDEN sentence is removed). As I asked above, if it does no harm, why object? - Butwhatdoiknow (talk) 16:35, 15 December 2022 (UTC)

I would supportprefer phrasing without repetition, but the one that's clear.

The current version says: In discussions of proposals to add, modify, or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit. I get the point about the first sentence, but "in" is massively misleading as one if its meanings is "during". It's less than obvious that the first sentence overrides it. PaulT2022 (talk) 16:43, 15 December 2022 (UTC)

Thanks, repetition is something I think is best avoided. What if we change "lack of consensus" so that it reads something like: "In discussions of proposals to add, modify, or remove material in articles, a conclusion of "no consensus" commonly results in retaining the version of the article as it was prior to the proposal or bold edit." Thoughts? --GoneIn60 (talk) 17:03, 15 December 2022 (UTC)

I like the direction, but I don't think it alleviates ambiguity of "in" and introduces one of it's own (whose conclusion? if I see a discussion and conclude there's no consensus, do I go ahead and restore per NOCON? - ultimately that's the problematic behaviour).

If the only problem with Butwhatdoiknow's edit is repetition, then perhaps we could rephrase somehow else? When discussions of proposals to add, modify, or remove material in articles end without consensus, the version of the article as it was prior to the proposal or bold edit is commonly retained.? PaulT2022 (talk) 17:21, 15 December 2022 (UTC)

Sure, sounds reasonable, however not a huge fan of "commonly retained". Perhaps we lead with your suggestion but end it with: "the common result is to retain the version of the article as it was prior to the proposal or bold edit." Keeps it closer to the existing text. Final thoughts/suggestions? --GoneIn60 (talk) 21:25, 15 December 2022 (UTC)

No objections, sounds good to me. PaulT2022 (talk) 21:42, 15 December 2022 (UTC)

I'm in as well. Unless someone beats me to it, I'll make the changes tomorrow. Butwhatdoiknow (talk) 22:31, 15 December 2022 (UTC)

This sounds like the ONUS dispute...I don't think we should be making changes which will affect that. Kolya Butternut (talk) 02:15, 16 December 2022 (UTC)

The proposal is to change this:

Example from Butwhatdoiknow
In discussions of proposals to delete articles, media, or other pages, a lack of consensus normally results in the content being kept. to this: When discussions of proposals to delete articles, media, or other pages end without consensus, the common result is to retain the version of the article as it was prior to the proposal or bold edit.

I suggest that the new version, if it affects the ONUS dispute at all, brings NOCON closer to ONUS by excluding new content when a discussion results in "no consensus." - Butwhatdoiknow (talk) 04:41, 16 December 2022 (UTC)

None of this belongs on this policy page. SmokeyJoe (talk) 11:37, 16 December 2022 (UTC)

That is a separate discussion. Please feel free to start a separate section to discuss removing the text entirely. - Butwhatdoiknow (talk) 17:19, 16 December 2022 (UTC)

It is bloat on bloat. Why do you think this text is important even to discuss? What do you think are the implications of this text to editing or anything? SmokeyJoe (talk) 00:29, 17 December 2022 (UTC)

Kolya Butternut and SmokeyJoe: would you mind elaborating a bit further? The most recent discussion is about a slight cosmetic change to existing text of NOCON. The substance of the current policy will remain intact.

Here's an example taken directly from the discussion (I collapsed the example above to avoid confusion):

In discussions of proposals to add, modify, or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit.

would change to this:

When discussions of proposals to add, modify, or remove material in articles end without consensus, the common result is to retain the version of the article as it was prior to the proposal or bold edit.

--GoneIn60 (talk) 15:50, 16 December 2022 (UTC)

Cosmetic tangential matters. The intensity of discussions and reverts on cosmetic tangential matters, on a concept outside of the scope of consensus, suggests that editors think what they are doing matters. I ask: What do you think is the implication of these words? SmokeyJoe (talk) 00:32, 17 December 2022 (UTC)

• I look beyond the words and favor stability, hence my participation here to understand why a change was being pursued. Personally, I think a bigger overhaul to get the whole longstanding stance out of NOCON would be more productive, but unfortunately it's not on the table at the moment. I'll presume this small potato cosmetic change is of no interest to you, point taken. --GoneIn60 (talk) 01:04, 17 December 2022 (UTC)
• I'm not sure whether "these words" mean the current text or the proposed text, but I can tell you that the purpose of the proposed change is to drive home that NOCON is about what happens after a discussion fails, not what happens during a discussion. If you think that distinction doesn't matter one way or the other, why are you wasting your time to participate in this discussion? - Butwhatdoiknow (talk) 01:26, 17 December 2022 (UTC)

@Kolya Butternut, please respond to GoneIn60's post (above). - Butwhatdoiknow (talk) 01:02, 19 December 2022 (UTC)

[T]he purpose of the proposed change is to drive home that NOCON is about what happens after a discussion fails, not what happens during a discussion. Isn't the current text consistent with QUO? I don't think we should make changes to lose that. Kolya Butternut (talk) 02:47, 19 December 2022 (UTC)

QUO applies during a discussion, is designed to prevent edit wars and, as it says, "should not be used for any other purpose." NOCON applies after a discussion (when the result is no consensus). Accordingly, I suggest, they do not need to be consistent. That said, I think the practical effect is the same in both circumstances: under QUO the status quo remains temporarily in place and under NOCON (either the current or the proposed text) the status quo remains "permanently" in place until there is a consensus to change it. - Butwhatdoiknow (talk) 05:15, 19 December 2022 (UTC)

I think if there's a possibilitly it would weaken QUO then we shouldn't do that. And you suggested that this might bring NOCON closer to ONUS, but I think it should be the other way around. Kolya Butternut (talk) 05:29, 19 December 2022 (UTC)

Regarding your QUO concern (I'll speak to your ONUS concern later), let me ask you these questions:

(1) Do you agree that the intent of QUO is that it be applied only during discussion and the intent of NOCON is that it be applied only after a discussion fails?

(2) If you do, do you also agree that NOCON is not intended to support (or, if you will, strengthen) QUO?

(3) If you do, do you also agree (a) that some people mistakenly cite NOCON during discussion when they should be citing QUO, and (b) that causes confusion?

(4) If you do, do you also agree that - if we can do it without affecting any conflict between ONUS and either QUO or NOCON - we should tighten up the text at NOCON to make it clearer that it applies only after discussion? - Butwhatdoiknow (talk) 15:57, 19 December 2022 (UTC)

(Side note… during discussions, I much prefer to cite m:The Wrong Version rather than argue about whether QUO applies) Blueboar (talk) 16:12, 19 December 2022 (UTC)

(1) a, yes. b. not sure. Kolya Butternut (talk) 17:25, 19 December 2022 (UTC)

Kolya Butternut, It's not clear why there is any confusion about the intent of NOCON. The first line of NOCON already declares we are talking about next steps after "discussion results in no consensus". The proposal is to drive that point home in the bullet list that follows. While I don't think that's absolutely necessary, since the first line of NOCON is already clear to me, I can see how others might find the proposed cosmetic changes helpful. If you think some meaning is being changed, could you please be more specific instead of hypothetical? --GoneIn60 (talk) 19:26, 19 December 2022 (UTC)

And the issue that this change brings NOCON closer to ONUS? Kolya Butternut (talk) 20:47, 19 December 2022 (UTC)

• I'm not sure why that was mentioned. The proposed change doesn't move the needle in any direction; it's simply a clearer way to state what is already being stated. --GoneIn60 (talk) 23:12, 19 December 2022 (UTC)
• @User talk:Kolya Butternut, does this question mean that we've resolved the "weakens QUO" issue to your satisfaction? If not, with your indulgence, I'd like to continue with that issue before moving on to NOCON/ONUS. - Butwhatdoiknow (talk) 01:39, 20 December 2022 (UTC)

Maybe, but I'm not comfortable making changes to a contentious major guideline without opening up the discussion to the village pump perhaps. Kolya Butternut (talk) 03:38, 20 December 2022 (UTC)

Let's talk about that if and when we resolve your ONUS concern. - Butwhatdoiknow (talk) 06:00, 20 December 2022 (UTC)

Regarding your ONUS concern, I said the new text "brings NOCON closer to ONUS by excluding new content when a discussion results in 'no consensus.'" What I should have said was "makes clearer that, when a 'no consensus' discussion about new content results in 'no consensus,' NOCON and ONUS reach the same result." Do you think that is not the case? If so, how do you see the proposed text changing NOCON? - Butwhatdoiknow (talk) 06:00, 20 December 2022 (UTC)

"[M]akes clearer that, when a 'no consensus' discussion about new content results in 'no consensus,' NOCON and ONUS reach the same result." How does it do that? Kolya Butternut (talk) 09:50, 20 December 2022 (UTC)

To recap, the current text:

In discussions of proposals to add, modify, or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit.

The current proposed text:

When discussions of proposals to add, modify, or remove material in articles end without consensus, the common result is to retain the version of the article as it was prior to the proposal or bold edit.

This change is designed to make it clearer that, as is the general intent of NOCON, this text applies only after discussion fails. The purpose of this change is to reduce the number of editors who cite NOCON during discussion instead of QUO or The Wrong Version (which is cited in QUO).

Turning to your question, now that I look at it I think you and GoneIn60 are right: the change is so innocuous that it doesn't change anything vis-a-vis ONUS. (In my defense, I did preface my statement to the contrary with "if it affects the ONUS dispute at all." That said, now that I look even more closely at my post, I see that in the collapsed text above I quoted the wrong text. I sincerely apologize to all for the confusion that caused.)

Does that resolve your concern about a NOCON/ONUS conflict? - Butwhatdoiknow (talk) 17:34, 20 December 2022 (UTC)

I'm still curious what you had thought the effect might have been, so I can evaluate whether there would be one. Kolya Butternut (talk) 18:12, 20 December 2022 (UTC)

I think that both times (when I said "closer" and when I said "clearer") I was looking at the erroneous collapsed comparison. So whatever I thought when I said that was based on the wrong comparison. That is why I re-stated the correct comparison in my recap. - Butwhatdoiknow (talk) 23:11, 20 December 2022 (UTC)

@Kolya Butternut: Do you have any remaining concern that the proposed change affects the NOCON/ONUS issue? - Butwhatdoiknow (talk) 03:12, 3 January 2023 (UTC)

I'm always nervous about the effects of changes to NOCON/ONUS, but I'll neither object nor explicitly support the change? Kolya Butternut (talk) 03:16, 3 January 2023 (UTC)

That works for me, thanks. Perhaps someone else will see a problem and then we can always go back to the prior text. - Butwhatdoiknow (talk) 03:43, 3 January 2023 (UTC)

Re: NOCON

Currently WP:NOCON says:

• In discussions of proposals to add, modify, or remove material in articles, a lack of consensus commonly results in retaining the version of the article as it was prior to the proposal or bold edit.

This certainly was common practice when it was written, but given the recent discussions about ONUS, I have to ask whether this is still practice? Have we shifted to NOT retaining when there is no consensus? Blueboar (talk) 15:22, 2 December 2022 (UTC)

The discussion at ONUS suggests that a lack of consensus commonly results in any of a number of outcomes and neither NOCON or ONUS reflect "the" common practice. - Butwhatdoiknow (talk) 18:01, 2 December 2022 (UTC)

Fair enough… but if NOCON no longer reflects common practice we should probably amend it.

So… what IS common practice these days? My personal sense of actual practice is that we retain less often than we used to, but are not quite at “usually omit” (except in BLP situations). Can we clarify? When do we retain and when do we omit? Blueboar (talk) 20:36, 2 December 2022 (UTC)

Based on the discussion at ONUS, I doubt very much that we can clarify. - Butwhatdoiknow (talk) 23:44, 2 December 2022 (UTC)

Can we agree that it's common practice to retain longstanding material pending dispute resolution? WP:QUO I guess? Kolya Butternut (talk) 23:51, 2 December 2022 (UTC)

That’s part of what I am asking… that used to be common practice, but I am not sure whether that is still common practice. Practice seems to have shifted (or, at least, to be shifting) away from retention. Blueboar (talk) 00:02, 3 December 2022 (UTC)

We could agree to that, but I don't think it helps us with NOCON (which says what happens when discussion fails). - Butwhatdoiknow (talk) 01:40, 3 December 2022 (UTC)

Maybe practice is "shifting" (without stats that's hard to know), but policy is to not retain. ONUS is policy, QUO is merely an essay. Peter Gulutzan (talk) 16:44, 15 December 2022 (UTC)

“but if NOCON no longer reflects common practice we should probably amend it.”

I suggest removing it. Split to an essay. Summarise in WP:Editing policy.

WP:Consensus is about consensus decision-making and how, on Wikipedia, it is found. NOCON concerns itself with not-consensus and when editors (justifiably usually) decide to give up trying (on the specific small question) and what to do. It’s after a fork to another road. SmokeyJoe (talk) 00:37, 17 December 2022 (UTC)

Again, changing NOCON will affect the huge ONUS dispute. If anything is out of place it is the last sentence of ONUS, which is used as a conduct policy. Kolya Butternut (talk) 01:12, 17 December 2022 (UTC)

Did you mean to place this post in #Revert number 3? - Butwhatdoiknow (talk) 01:28, 17 December 2022 (UTC)

Again, I encourage you to start a new "Delete NOCON" section to raise these issues. - Butwhatdoiknow (talk) 01:29, 17 December 2022 (UTC)

There is one serious argument against NOCON. Consider this situation: Some user finds a serious problem with some text. That may be WP:REDFLAG violation, WP:NOR violation etc. This issue has never been discussed before, simply because noone noticed the problem. If the discussion of the problem doesn't lead to a consensus, this problematic text stays, even if some users merely object to its removal without providing serious arguments. They usually cite NOCON: if the text is long standing, and some users disagree with its removal, NOCON ostensibly takes effect.

This situation is pretty common, especially in low importance articles, which means low quality content continues to grow in Wikipedia. I think we must clearly stipulate that if no consensus is achieved about compliance of some text with our core content policies (at least WP:V and WP:NOR; WP:NPOV is a little bit more tricky), it should be deleted.

It other words, in all discussions about compliance of some text with our content policies an explicit consensus must be achieved about compliance, not about non-compliance. The burden of proof of compliance should be on those who wants to keep the text.

It seems the main problem of NOCON is that it does not clearly discriminate two cases:

• (i) deletion vs keeping of some text when both new and old versions comply with our policy; In that case, "no consensus" means "keep a long standing version".
• (ii) deletion vs keeping of the text that violates our policy; In that case, "no consensus that some version complies with the policy" means "it does not comply", and it should be removed, no matter if it is a long standing or a new version.

Paul Siebert (talk) 20:29, 19 December 2022 (UTC)

NOCON cited during discussions

I joined this year and my impression is contested content more often deleted prior to discussion than after. I think the current wording is highly confusing to new editors. (And acts to inflame edit wars in a form of 'no consensus' reverts, which I've occasionally done myself mistakenly seeing NOCON as an overriding principle in disagreements about longstanding content.)
I did note that removal before discussion mostly happens when sourcing or BLP claims are at play, so perhaps changing "When the material in question lacks a reliable source" to something like "When the verifiability of the removed material is contested" (closer to the referenced WP:BURDEN) would suffice?
"No consensus" clearly still is a guiding principle for most changes overall - see Talk:Recession#RfC:_Phrasing_of_the_infamous_sentence closure for example. PaulT2022 (talk) 16:14, 14 December 2022 (UTC)

I'm going to edit NOCON to try to add clarity. Thanks for pointing out the issue. FYI, the recommended practice during discussion is wp:QUO. - Butwhatdoiknow (talk) 17:58, 14 December 2022 (UTC)

Thanks! I think WP:QUO contradicts WP:BURDEN, especially the part buried in a footnote (why?):

Once an editor has provided any source they believe, in good faith, to be sufficient, then any editor who later removes the material must articulate specific problems that would justify its exclusion from Wikipedia (e.g. why the source is unreliable; the source does not support the claim; undue emphasis; unencyclopedic content; etc.). If necessary, all editors are then expected to help achieve consensus, and any problems with the text or sourcing should be fixed before the material is added back.

The list is open-ended, explicitly includes NPOV violations (undue), and explicitly forbids restoration of status quo even for non-verifiability problems (note the "any problems"). Have no opinion on the remove vs preserve debate, but I feel the policies should be written in a more straightforward way. PaulT2022 (talk) 19:02, 14 December 2022 (UTC)

I'm not sure that CONSENSUS is the place to discuss whether QUO and BURDEN conflict, but here goes:

First of all, I'll bet dollars to donuts that you are the first person to read BURDEN footnote 3 in more than 4 years and 7 months. Good for you for doing that, and I share your desire for clarity in the Wikipedia space. But I'm guessing the reality is that no one pays any attention to BURDEN footnote 3.

Secondly, footnote 3 only comes into play after the following sequence of events: (1) Editor A adds unsourced content. (2) Editor B deletes it because it is unsourced. (3) Editor A restores it with a source. (4) Editor B deletes it for some other reason. (5) Editor A disputes Editor B's new rationale. Yes, at that point footnote 3 and QUO diverge, but how often will that occur?

If you nevertheless wish to pursue this issue then I suggest you follow up on the wp:V talk page. - Butwhatdoiknow (talk) 19:46, 14 December 2022 (UTC)

After browsing archives, I don't see a lot of appetite for changes in WP:V so raised it on WT:RV instead. Please feel free to post a notice to WT:V if you think it's useful for that page's watchers.

In my experience the situation you describe occurs often because Editor B's "other reason" would more often than not say that either the source isn't good enough, or text misrepresents it or something to the effect - which is covered by WP:BURDEN even without going deep into footnotes. PaulT2022 (talk) 20:20, 14 December 2022 (UTC)

There is no conflict in situations where the status quo ante bellum version doesn't contain the disputed content, but when it does, then unfortunately we need to keep in mind that QUO is just an essay. It wouldn't hold any weight over BURDEN, although it's worth pointing out that NOCON favors the longstanding version after discussion fails. --GoneIn60 (talk) 20:23, 14 December 2022 (UTC)

it's worth pointing out that NOCON favors the longstanding version after discussion fails - but this is not WP:NOCON says:

When discussions of proposals to add, modify, or remove material in articles result in no consensus the common result is retaining the version of the article as it was prior to the proposal or bold edit. However:

• When the material in question lacks a reliable source, it is excluded.

It sounds to me like it prescribes to exclude material when there's a disagreement on source even when no consensus to remove it is reached in discussion (which makes sense in light of the WP:BURDEN wording). Am I reading too much into this? PaulT2022 (talk) 20:44, 14 December 2022 (UTC)

No, it's a valid concern. When material lacks support from a reliable source (i.e. it is unverifiable), then it's in violation of WP:V and cannot be included. Its inclusion at that point is not subject to discussion and consensus, so NOCON does not apply. If the source is still being debated, then we haven't yet reached the determination that "the material in question lacks a reliable source", and NOCON still doesn't apply. That line you quoted was recently added with very little discussion, and given the attention it is already receiving, it should probably be removed. Discussing it currently above. --GoneIn60 (talk) 22:44, 14 December 2022 (UTC)

If you turn "the common result" into "is required", you are reading too much into this. WhatamIdoing (talk) 00:06, 13 February 2023 (UTC)

A question related to our NOCON discussions

In the numerous discussions about NOCON above, we often give examples where two editors (A and B… or Alice and Bob… etc) are in a dispute. This raises a philosophical question…can just two editors EVER form a “consensus”? Sure, two editors can agree… or disagree… but is that agreement/disagreement enough to say there is “consensus’ or “no consensus”. Don’t we need more than just two editors to opine before we can say whether something has a “consensus”? Blueboar (talk) 12:47, 16 December 2022 (UTC)

• No, we don't need more than two. In fact, we don't even need two. One person can opine (make a bold edit) and the edit has consensus until and unless another editor raises an objection. wp:SILENCE.

That said, if one editor does raise a concern then - unless one or more additional editors join in - the original editor and the objecting editor can reach a consensus to resolve the objecting editor's concern. And that consensus prevails until and unless a third editor raises an objection. - Butwhatdoiknow (talk) 17:13, 16 December 2022 (UTC)

• It's an intriguing question. The number of participants is irrelevant in the ability to achieve "consensus", although a minimum of two editors is required to determine "no consensus". However, the more participation that results in consensus, generally the more resistant that consensus is to change (in the short term anyway). --GoneIn60 (talk) 02:10, 17 December 2022 (UTC)
• I think that depends on the article's level. About a couple weeks ago, I proposed some change on the WWII talk page, and two users supported it, with no opposition. However, I am still not sure if there is a consensus, because that is a high profile article that is being watchlisted by thousands of users. However, is some article is much less popular, an agreement between two editors may form a consensus. Wikipedia is highly heterogeneous, so it would be deeply incorrect to propose simple uniform criteria.

I would say, if we want to propose some concrete figure, it should be derived from the number of users who were active in this concrete article during the last year.

And, with regard to "no consensus", it is tricky. Consider two different situations:

• A user A says the text X does not comply with WP:REDFLAG, and it should be removed, a user B disagrees, and .
• A user A says the text X is not relevant to the article's topic, and it should be removed, a user B disagrees.

In both cases, there is no consensus, but in the first case the text should be removed per WP:V, in the second case it should stay per WP:NOCON. Paul Siebert (talk) 02:15, 19 December 2022 (UTC)

• But shouldn’t users A and B seek out a third opinion in both cases? Surely at least one more editor should be consulted before saying whether a consensus exists or not. Blueboar (talk) 21:51, 19 December 2022 (UTC)

  That is not a solution. If "A" and "B" are not individual users, but two groups of users, that doesn't change anything. For example, we frequently have a situation when, during some RfC, one group of users ("the group A") says that the text doesn't comply with REDFLAG, and another group of users ("the group B") maintains that everything is ok. In that case, I see two different interpretations of the RfC's outcome:

  1. "There is NO CONSENSUS that the edit X violates REDFLAG", and

  2. "There is NO CONSENSUS that the edit X complies with REDFLAG"

  Formally, both conclusions are equally legitimate, but their consequences are different. In the first case, the edit X stays if it is an old edit (per NOCON, if there is no consensus, an old version is restored), and it goes if it is a new edit (for the same reason).

  In the second case, the edit X is always removed, because the users failed to come to consensus that it complies with our policy.

  In my opinion, the second approach is much more reasonable, because I see no reason why the text that looks problematic from the point of view of our policy becomes less problematic merely because the violation has been unnoticed for a couple of years. Paul Siebert (talk) 00:32, 20 December 2022 (UTC)

Revert re presumed consensus

@Nikkimaria:, you reverted my edit with the explanation that it "overstates the case not only in the specific wording but also the prominent placement." Before we talk about placement, I'd like to learn how the text overstated the case. Please let me know where I went wrong. - Butwhatdoiknow (talk) 03:38, 28 December 2022 (UTC)

The proposed text stated that edits "have consensus" until reverted etc. The text of the body states that they have presumed consensus - a subtle but important distinction. If I make a (new, bold, undiscussed) edit and you revert it a few minutes later, my edit didn't have consensus; as SILENT states, you "find out whether your edit has consensus when it sticks". Nikkimaria (talk) 03:47, 28 December 2022 (UTC)

So the wording "All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted" doesn't overstate the case (depending upon where it is placed), do I have that right? - Butwhatdoiknow (talk) 05:22, 28 December 2022 (UTC)

The specific wording in the policy is "An edit has presumed consensus until it is disputed or reverted" - "changed" is addressed in the following sentence. Nikkimaria (talk) 00:15, 29 December 2022 (UTC)

Is there anything in the "following sentence" ("Should another editor revise that edit then the new edit will have presumed consensus until it meets with disagreement") that conflicts with "All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted"? - Butwhatdoiknow (talk) 00:39, 29 December 2022 (UTC)

Yes, depending what the change is. If something like this is to be included in the lead, it makes sense to use phrasing consistent with the existing policy. Nikkimaria (talk) 02:31, 29 December 2022 (UTC)

Regarding placement, let's work one on thing at a time.

Please help me see why you view "All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted" as inconsistent with the two sentences in the policy. Please give me an example of a change for which the outcome would be different under the "following sentence" and the latest version of my sentence. - Butwhatdoiknow (talk) 02:41, 29 December 2022 (UTC)

Your version assumes any change to an edit means the edit doesn't have consensus. A change can fundamentally alter the original edit, but it can also be entirely congruent with the original - eg if you add a sentence and I add a wikilink to it, I've changed your edit but am not disputing it. And this process is what is reflected in the extant IMPLICITCONSENSUS section. What's wrong with the existing policy wording? Nikkimaria (talk) 02:53, 29 December 2022 (UTC)

In your example the presumed consensus for the text without a wikilink is replaced by the presumed consensus for the text with a wikilink. This is what the two sentences in IMPLICIT say and what my proposed single sentence says. I'm still not seeing any inconsistency.

It looks to me like the two sentences in IMPLICIT describe a process and my single sentence states a "rule." The single sentence can be useful to, for example, drop into wp:SILENCE with a link back to wp:IMPLICIT for more information. - Butwhatdoiknow (talk) 17:43, 30 December 2022 (UTC)

@Nikkimaria: I'm hoping we can continue this conversation. - Butwhatdoiknow (talk) 09:51, 16 January 2023 (UTC)

I don't agree that we want to have a "process" and a "rule" that differ in wording in such a way that they can be interpreted differently - even if you personally would interpret them the same. Nikkimaria (talk) 00:46, 17 January 2023 (UTC)

Would you please give me an example of how the two texts may be interpreted differently? - Butwhatdoiknow (talk) 07:26, 17 January 2023 (UTC)

I have, and you've disagreed. Which is your right, but I'm not sure how repeating that would be helpful. Nikkimaria (talk) 03:28, 18 January 2023 (UTC)

@Nikkimaria: Apologies for not making my question clearer. Please give me an example of a change for which the outcome would be different under the two different interpretations that you see. - Butwhatdoiknow (talk) 18:38, 29 January 2023 (UTC)

I'm afraid that doesn't make the question clearer. Outcome of what? Presumed consensus only carries weight if nothing happens. Nikkimaria (talk) 01:37, 30 January 2023 (UTC)

@user:Nikkimaria: Okay, your example is an edit that builds on - is congruent with - a prior edit. And you seem to say that my phrasing means the original edit loses "presumed consensus" at that point. Let's say you are correct. So what? The original edit is now history, it has been replaced by the edited text.

The new text - including the original edit (as changed) - now has presumed consensus. I don't see any practical difference in the outcome under my phrasing: under both formulations the old text as edited now has presumed consensus. Are you seeing something different? Butwhatdoiknow (talk) 06:06, 10 February 2023 (UTC)

The original edit hasn't been replaced, it's still there. No one has removed or fundamentally altered it. So we presume it has consensus until someone does.

Maybe this will help. Suppose I run a bot to automatically link all instances of a particular term, and that bot adds a link to your edit. Obviously a bot can't assess the merit of your edit, so we can't say the bot agrees. A person who comes along could disagree with your edit or the bot's edit separately.

(And I realize you wanted to address the specific wording first, but I think this conversation is meaningless because we haven't agreed to place something in the lead at all, regardless of its specific wording.) Nikkimaria (talk) 02:57, 11 February 2023 (UTC)

Thank you, that does help.

In your example the original text ("some words here") has been replaced by the new text ("some words here"). The new text (which includes the old text) has presumed consensus until a later edit changes it. Again I ask, once the new edit takes place (and has presumed consensus) what practical difference does it make whether the retained original text had presumed consensus before the new edit?

(This particular conversation is an attempt to achieve consensus regarding whether it is important for this policy to say that retained retained original text retains the presumed consensus it had before a new edit.) - Butwhatdoiknow (talk) 19:30, 11 February 2023 (UTC)

As above, the "new text" is a link only. One could agree or disagree with the addition of the link, separately from whether one agrees or disagrees with the addition of the original text; the addition of the link shouldn't impact how the previous text is assessed. But really this particular conversation is a discussion of how to word something that we haven't agreed should be included in the first place. Nikkimaria (talk) 00:48, 12 February 2023 (UTC)

We agree that "the addition of the link shouldn't impact how the previous text is assessed." But that still doesn't answer my question. Do you believe my proposed text would impact how the unchanged/unlinked part of the previous text is assessed? If so, please explain how. If not, what practical difference would my text make? - Butwhatdoiknow (talk) 15:50, 12 February 2023 (UTC)

If you agree, then why continue? Under your proposal, any change, including the addition of a link in the scenario described, would impact how the previous text is assessed, as it is now the link only that is under the presumed-consensus proviso. Nikkimaria (talk) 16:58, 12 February 2023 (UTC)

@Butwhatdoiknow, I think the problem with your version is the word "until". It implies that if Alice adds a sentence, and Bob changes the sentence, then Alice's contribution no longer has any (presumed) consensus. That's not really true when editors are working in harmony. If Alice writes "Reese's Peanut Butter Cups are made with chocolate and peanut butter", and Bob adds the word filling to the end of her sentence, then it would be more accurate to say that both of their contributions (not to mention all of the contributions by all the editors who wrote the rest of the page) have presumed consensus. (talk) 00:14, 13 February 2023 (UTC)

@WhatamIdoing, thank you for joining the conversation. - Butwhatdoiknow (talk) 17:30, 13 February 2023 (UTC)

The practical effect of presumed consensus

@Nikkimaria: and @WhatamIdoing: If you'll permit, let's take a look back at what it means to have presumed consensus under the current CON text:

An edit has presumed consensus until it is disputed or reverted. Should another editor revise that edit then the new edit will have presumed consensus until it meets with disagreement.

My understanding, using the example Nikkimaria and I are working with above, "some words here" has presumed consensus until it is changed to "some words here." At that point, the presumed consensus for not linking "words" evaporates but (a) each word in the text has presumed consensus dating back to the edit that created it and (b) the link has presumed consensus dating from when it was added. In both cases, the presumed consensus for a word or the link evaporates if it is disputed, reverted, or changed in the future.
Do I have that right? - Butwhatdoiknow (talk) 17:32, 13 February 2023 (UTC)

Yes. WhatamIdoing (talk) 01:30, 14 February 2023 (UTC)

For clarity, the presumption of consensus for a word or the link evaporates if/when it is disputed, reverted, or changed in the future, but it might still have real consensus. Also, we put common-sense limits around this, so that blatant vandalism doesn't count as a (relevant) "change". WhatamIdoing (talk) 01:34, 14 February 2023 (UTC)

Right, "real" consensus trumps presumed consensus every time and, one would expect, the new edit would be quickly returned to the original text (which would then have presumed and real consensus). It doesn't matter whether the new edit was in good faith or vandalism, it's presumed consensus disappears when it is reverted. - Butwhatdoiknow (talk) 17:03, 15 February 2023 (UTC)

Okay, let's look at my proposed text:

All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted.

Applying this text to the example Nikkimaria and I are working with above, "some words here" has presumed consensus until it is changed to "some words here." At that point, the presumed consensus for not linking "words" evaporates and (a) each word in the text has presumed consensus and (b) the link has presumed consensus. In both cases, the presumed consensus for a word or the link evaporates if it is disputed, reverted, or changed in the future.
Would you agree that is how the proposed text would work out? - Butwhatdoiknow (talk) 16:28, 16 February 2023 (UTC)

@Nikkimaria: and @WhatamIdoing: I would appreciate your answers to this question. - Butwhatdoiknow (talk) 18:30, 20 February 2023 (UTC)

The edit (of adding text) has been changed. There's no basis in your proposal to split out the choice of not-linking and the choice of what words were added to say that the former lacks presumed consensus but the latter still has it. The edit has been changed, period; under your proposal the substance of that change is immaterial. Nikkimaria (talk) 02:29, 21 February 2023 (UTC)

Okay, let's say that my proposal means the unchanged text loses presumed consensus because a link was added. So what? What difference does that make? In other words, does "presumed consensus" give text any greater protection than any other text that has not been through the discussion process? - Butwhatdoiknow (talk) 03:33, 21 February 2023 (UTC)

@Nikkimaria: I look forward to your reply, - Butwhatdoiknow (talk) 22:20, 26 February 2023 (UTC)

At this point I'm going to defer to the question raised below rather than continuing this thread. Nikkimaria (talk) 02:17, 27 February 2023 (UTC)

I don't think it's true that "All Wikipedia edits have presumed consensus until...". Some are known in advance to not have consensus (consider an edit summary that says "Just a demonstration for the talk page – will revert"). Others are known in advance to have "real" consensus (e.g., edits made pursuant to agreement on the talk page). WhatamIdoing (talk) 06:58, 21 February 2023 (UTC)

Well then, the current text ("An edit has presumed consensus until it is disputed or reverted.") is wrong. Butwhatdoiknow (talk) 03:50, 22 February 2023 (UTC)

Do you agree or disagree regarding how the proposed text would work out? - Butwhatdoiknow (talk) 03:54, 22 February 2023 (UTC)

@WhatamIdoing: I look forward to your reply, - Butwhatdoiknow (talk) 22:21, 26 February 2023 (UTC)

I think your example about linking "some words here" is fair, but the choice to add the link might reasonably be construed as an expression of support for including the words (and therefore a form of demonstrated consensus). WhatamIdoing (talk) 18:00, 27 February 2023 (UTC)

Preservation of presumed consensus distinction

IMO there's another problem with your proposed change. "Presumed consensus" can be interpreted as a different weaker type of consensus (which I think is a good idea) rather than a mechanism for it to achieve regular consensus. Your proposed change eliminates that. Sincerely, North8000 (talk) 18:48, 20 February 2023 (UTC)

I wholeheartedly agree that presumed consensus is weak and should be distinguished from "real" consensus. But I'm having trouble seeing how All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted - which says that presumed consensus for text evaporates when the text changes - eliminates that distinction. - Butwhatdoiknow (talk) 19:38, 20 February 2023 (UTC)

@North8000: Please help me see the problem. - Butwhatdoiknow (talk) 22:16, 26 February 2023 (UTC)

Presumed consensus is different than consensus and much content has only presumed consensus. So your edit was a substantial change and one that conflicts with that. Sincerely, North8000 (talk) 00:20, 27 February 2023 (UTC)

What if the sentence read "All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted"? - Butwhatdoiknow (talk) 16:00, 27 February 2023 (UTC)

Placement of text

Nikkimaria, above you express concern regarding where the text we're discussing ("All Wikipedia edits have presumed consensus until they are questioned, changed, or reverted") would be placed in the event we can agree that the text is accurate. The answer includes (1) the into is one option (we can talk about whether that is a good idea after we have text that is not objectionable on its own), (2) other pages (as a summary of IMPLICIT), and (3) perhaps as a replacement for the two sentences in IMPLICIT (which, I believe, are redundant). - Butwhatdoiknow (talk) 15:50, 12 February 2023 (UTC)

What is the point of presumed consensus?

If we are saying that a “presumed consensus” only lasts until there is a subsequent edit (and then evaporates) what is the point of mentioning the concept in the first place? WHY do we even need to address it? Blueboar (talk) 01:13, 27 February 2023 (UTC)

• A choice of wording by someone with limited experience in the process clearly. Moxy- 01:32, 27 February 2023 (UTC)
• It’s pointless. It’s meaningless waffle likely to confuse people. SmokeyJoe (talk) 01:52, 27 February 2023 (UTC)

If it is pointless, shouldn’t we simply cut it? Do away with the concept of “presumed consensus” entirely? Blueboar (talk) 13:02, 27 February 2023 (UTC)

• I guess someone thought it was nice to be able to say that everything was in articles by consensus and then invented this fictional concept to make it true. I have never seen a useful purpose for it and thoroughly support it disappearing. Zero^talk 13:36, 27 February 2023 (UTC)

  That's a good way of putting it. DFlhb (talk) 02:59, 1 March 2023 (UTC)

"Presumed consensus" is a synonym for "implied consensus." Here's what WP:EDITCON says about that:

Wikipedia consensus usually occurs implicitly. An edit has presumed consensus until it is disputed or reverted. Should another editor revise that edit then the new edit will have presumed consensus until it meets with disagreement. In this way, the encyclopedia gradually improves over time.

Are you proposing to delete this text? - Butwhatdoiknow (talk) 16:14, 27 February 2023 (UTC)

• I am certainly considering proposing we do so. I think we should at least examine the possibility. What are the up sides? What are the down sides? Etc. Blueboar (talk) 18:54, 27 February 2023 (UTC)
• I think it is a useful concept. This has to be looked at from two different directions:

• "How dare he edit my article without getting written permission in advance to prove that there is consensus!" – Um, all edits have presumed consensus. It's okay for editors to make bold (=significant but undiscussed) edits. We know that people check RecentChanges and their watchlists to revert anything they disagree with, but there's no way for them to mark their agreement with an edit. So unless and until someone visibly disputes it, we assume that non-removal is some sort of acquiescence to the edit.
• "My edit must stay on the page forever, until you can get near-unanimous consent to remove it, because it has consensus!" – Um, your edit only had presumed consensus, and that presumption of consensus disappeared when the other editor challenged it. Now your edit doesn't appear to have consensus at all. Please try to work out an agreement that satisfies (almost) everyone.

I think it is a mistake to say (without qualification) that presumed consensus only lasts until there is a subsequent edit. It is probably more accurate to say that presumed consensus only lasts until that material is disputed. (A "subsequent" edit could be to a different part of the page, an AWB typo-fixing run, a bot edit, or some other edit that doesn't really confer any sort of expression of support or opposition.) WhatamIdoing (talk) 17:57, 27 February 2023 (UTC)

Something important I would add (which I detailed below) is that the nature of implicit consensus means that it is generally not acceptable to object to an edit solely on the basis that you don't feel it has consensus, at least not unless there's an existing dispute to point to or the edit is so manifestly controversial or drastic that the objection is obvious. A lack of consensus is supposed to be based in some specific objection. "I think it makes the article worse" is fine (if a bit weak and vague); but it's not acceptable to revert someone purely with a demand that they get consensus, unless some other objection is backing that up. This doesn't mean that an editor should immediately restore the disputed text, of course; WP:AGF means you should assume they have a valid objection and try to tease it out of them. But if someone just repeatedly brushes them off with "no, you need to get consensus, WP:ONUS is on you, I'm not going to say anymore" without ever articulating why they object to an edit beyond that, then that's inappropriate of them. This gets to part of my underlying objection to the "WP:ONUS maximalist" proposals in that I believe discussions work best when everyone is encouraged to come to the table, put all their cards on it at the start, and try to find a way to address the underlying content dispute; I'm concerned that some of the more radical interpretations of WP:ONUS amount to saying that people who want to remove material or to keep it out of an article have less of an obligation to engage at all and can simply point to ONUS without articulating a position that can be meaningfully engaged short of an WP:RFC. In the rare cases where I've encountered people who have actually tried to apply this belief during discussions, it has been completely unsatisfactory and lead to useless, circular discussions focused on the minutiae of policy rather than actually productive discussions of article content or meaningful efforts to resolve disputes. --Aquillion (talk) 21:03, 27 February 2023 (UTC)

I particularly dislike it when someone reverts an edit with the explanation that they support it, but they believe somebody else (usually an unnamed/hypothetical person) will object. Reverting needs to be done by the person who objects, so that the bold editor knows who to talk to. WhatamIdoing (talk) 05:48, 28 February 2023 (UTC)

• No, it's important that edits be presumed to have consensus until someone voices a specific objection, otherwise we run into the "rv, get consensus" issue where people who dislike an edit but who don't want to articulate why they dislike it can WP:STONEWALL it by demanding consensus without giving an objection that can be answered. This is a serious problem in controversial topic areas (where people will sometimes dislike an edit because they don't like what it says, even though they can't come up with a reason to oppose it; and where there are sometimes WP:OWN issues where people will try and demand consensus for edits by default), and would generally discourage useful discussion by allowing people to remove content from articles without providing any hint of how their objections to it could be addressed. An objection doesn't have to be detailed (at least not at first); even something as simple as "rv., undue" or "rv, tone issues" is fine, since that at least points to a place for discussions to start. But you're supposed to voice an objection, and the point of implicit consensus is "you need to get consensus before editing this article" is not a valid one. --Aquillion (talk) 21:03, 27 February 2023 (UTC)
• What it does in the Wikipedia system is give some weight to the status quo in discussions which consider many factors. And, in turn, the amount of weight is determined by other factors such as longevity and amount of review it is likely to have had. This is how most of our policies and guidelines work, even though they are not worded that way. North8000 (talk) 20:19, 27 February 2023 (UTC)

  If we removed “presumed consensus” do you think that the “some weight to status quo” would disappear? Blueboar (talk) 21:03, 27 February 2023 (UTC)

  Taking your question literally, I think that the extreme "disappear" would not happen. I haven't done the analysis to opine on whether it would be significantly weakened.....my guess is that full removal of the term would weaken the concept. North8000 (talk) 22:05, 27 February 2023 (UTC)

  Weight to the status quo is provided by WP:PRESERVE. The fact that PRESERVE does not even mention presumed consensus is evidence that the concept of presumed consensus is not necessary. Zero^talk 01:08, 28 February 2023 (UTC)

  @North8000, I think you might be right (What it does in the Wikipedia system is give some weight to the status quo), and that feels very weird to me right now. So there's presumed consensus for Edit #1, that presumed consensus disappears when someone disputes it, and yet... and yet... and yet many editors treat that no-longer-presumed no-longer-consensus as still being evidence of real-and-right-now-consensus. WhatamIdoing (talk) 05:53, 28 February 2023 (UTC)

  @WhatamIdoing: This and a lot of policy quandaries all make sense in the grand unification theory. Wikipedia:How editing decisions are made  :-) Seldom explicitly acknowledged, IMO it's how a big part of Wikipedia works. North8000 (talk) 15:26, 28 February 2023 (UTC)

Does it help if we relabel "implicit," "presumed" consensus as "silent" consensus, making it clear that it only lasts until it is disputed or reverted (or perhaps "changed," but that is a conversation for another day)? - Butwhatdoiknow (talk) 16:16, 28 February 2023 (UTC)
This is inelegant draft wording, but how about something to the effect of: Material that has been in article for some time is presumed to have an implicit consensus, the strength of which is based on the length of time it has been in and the amount of review it has presumably received. While this implicit consensus does not carry the same weight as an actively-arrived at consensus, it should be taken into consideration and given some weight in discussions about the material. North8000 (talk) 17:00, 28 February 2023 (UTC)

I think only the the amount of review it has presumably received is important. In other words, how much consensus did it have when it was added. Because often, additions (especially silent additions) may go unnoticed for some considerable time. Selfstudier (talk) 18:54, 28 February 2023 (UTC)

@North8000: What purpose would your concept serve that is not already served by WP:PRESERVE? Zero^talk 00:51, 1 March 2023 (UTC)

I wasn't trying to put forth a new concept, rather I was trying to acknowledge how Wikipedia operates. Infusing reality into our wordings of course has benefits. But the purpose in the specific debate here is just to try to bring a resolution to it. Sincerely, North8000 (talk) 01:44, 1 March 2023 (UTC)

Ok, but Blueboar started this section with the proposal that presumed consensus be ditched. It seems to me that so far the only argument put forward for keeping it is the weight it gives to existing content. But that's not a good argument as PRESERVE already does that more thoroughly without the need for imaginary constructs such as "consensus" for text never discussed. Furthermore, the confusion over what presumed consensus actually means is further support for removal. Can you (do you want to?) offer a better rationale for keeping it? Zero^talk 02:56, 1 March 2023 (UTC)

Another rationale: Saying a bold edit has consensus has the effect of discouraging reverts with no explanation beyond "no consensus."

A better solution to this problem might be to remove the presumed consensus text and replace it with the text from wp:PGBOLD: "you should not remove any change solely on the grounds that there was no formal discussion indicating consensus for the change before it was made." - Butwhatdoiknow (talk) 16:02, 5 March 2023 (UTC)

Sounds good, but I think a reference to WP:PRESERVE would be better, since it's more comprehensive and gives specific, practical advice on what to do before reverting (rather than giving a commandment). DFlhb (talk) 16:32, 5 March 2023 (UTC)

I don't see that the PGBOLD text (dealing with "no consensus" reverts) and saying something in addition about PRESERVE (one of four alternatives to a revert) are mutually exclusive choices. - Butwhatdoiknow (talk) 22:01, 5 March 2023 (UTC)

"Wikipedia:Presumed consensus" listed at Redirects for discussion

An editor has identified a potential problem with the redirect Wikipedia:Presumed consensus and has thus listed it for discussion. This discussion will occur at Wikipedia:Redirects for discussion/Log/2022 December 30 § Wikipedia:Presumed consensus until a consensus is reached, and readers of this page are welcome to contribute to the discussion. SmokeyJoe (talk) 01:27, 30 December 2022 (UTC)

How long does it take for an edit to acquire silent consensus?

“An edit has can be presumed to have consensus until it is disputed or reverted”.

This statement implies that implied silent consensus occurs immediately. I’d doesn’t. It takes time. More precisely, it takes evidence of acceptance. Pageviews? Subsequent edits? SmokeyJoe (talk) 03:19, 30 December 2022 (UTC)

• "presumed consensus exists only in the absence of a dispute" I take this as the status quo is best during a content dispute... the status quo being the version that has evidence of being longstanding..... as in years over weeks or months. Moxy- 03:33, 30 December 2022 (UTC)
• It does occur immediately. And it ends as soon as it's edited. Per PGBOLD, that's the case even for policy pages ("you should not remove any change solely on the grounds that there was no formal discussion indicating consensus for the change before it was made" (italics in original)).

That said, I think most of us accept that, particularly for policies, the fact that text has been in place for some time does add a feather or two to the scale when discussing a change. However, we need to be very careful to avoid status quo calcification. - Butwhatdoiknow (talk) 03:53, 30 December 2022 (UTC)

• Time is a piss poor metric for determining consensus… for the simple reason that the number of page views an article gets over time varies greatly from article to article. Some get hundreds of views a day… but others get one or two views a year.

A much better metric would be subsequent edits… if lots of edits (by different editors) have taken place after some bit of content was added, and none of those edits have reverted/changed/challenged that particular bit of content, then we can say the content has a degree of “silent” consensus. If it is reverted/changed/challenged by the next editor to view the page (no matter how much time has passed) we can not say it had “silent” consensus. Blueboar (talk) 14:23, 30 December 2022 (UTC)

• Whether the metric is time, views, or edits, does anything different happen when an edit achieves "silent consensus" (as opposed to what happens before that point)? Butwhatdoiknow (talk) 15:04, 30 December 2022 (UTC)

Yes… once an edit is perceived as having “silent consensus”, editors use this to argue that the text should not be changed or reverted without a clear and explicitly discussed consensus to change or revert. And any future BOLD editing is labeled as disruptive, and rejected out of hand. Perhaps those who make this sort of argument misunderstand the policy and shouldn’t do so… but far too many editors do make this argument nevertheless. It is quite common in content disputes. Blueboar (talk) 17:54, 30 December 2022 (UTC)

So, whether or not it should happen, often "presumed consensus" (which evaporates when edited or disputed) becomes "silent consensus" (which doesn't) after some undefined matric is met. Is that what you are saying? Butwhatdoiknow (talk) 18:09, 30 December 2022 (UTC)

Ah, but these editors don’t believe that a “silent consensus” actually does evaporate upon being edited. They argue that the text is “long standing” and thus must be preserved until you hit them over the head with a metaphorical two-by-four in the form of a full and complete discussion that unequivocally determines that a change has consensus. Blueboar (talk) 20:12, 30 December 2022 (UTC)

These hypothetical editors also believe, I hear you saying, that there is an interim period when edits do evaporate upon being edited. So for those editors there are two states for edits not previously changed or objected to and the dividing line is some undefined metric based on views, edits, or time. Do I have that right? - Butwhatdoiknow (talk) 21:10, 30 December 2022 (UTC)

Correct… the way many editors think, a recently made BOLD edit does not have any presumption of consensus… but one made a while ago can gain that presumption due to silence. Blueboar (talk) 08:42, 31 December 2022 (UTC)

• Time as a metric is bad, but isn't it also nonexistent? I do not know of any PAG that says longstandingness is definable or significant. WP:CRP mentions it but I'm glad that essay only applies for special page restrictions. Peter Gulutzan (talk) 15:14, 30 December 2022 (UTC)

• Time is a piss poor metric for determining consensus. A much better metric is subsequent edits. I agree with Blueboar. —SmokeyJoe (talk) 21:43, 30 December 2022 (UTC)

  Are you saying that before a sufficient number of subsequent edits the original edit has no consensus? If so, how would you describe its state during that period? What would you call it? - Butwhatdoiknow (talk) 23:12, 30 December 2022 (UTC)

  Yes. An edit that no one else ever reads cannot be said to represent consensus. Further edits in the same direction demonstrate approval.

  In the meantime, what would I call it? How about “dynamic”? Consensus implies stability. If in doubt, use the talk page. SmokeyJoe (talk) 23:18, 30 December 2022 (UTC)

  Wait, is the test you are proposing based on edit numbers or view numbers? Or maybe you are saying consensus can be established by either one? - Butwhatdoiknow (talk) 01:38, 31 December 2022 (UTC)

  I don’t wholly disagree with you, but I think you are exploring the wrong path. The determination of consensus, where there is disagreement, is not subject to tests, algorithms, or rules. We know it when we see it. It’s different every time. In the most beautiful cases, the question was amended before the consensus statement is agreed, and even those holding minority positions explicitly agree to the final statement representing consensus. Sometimes Wikipedia produces a “false consensus”, typically involving those disagreeing giving up out of exhaustion. Accordingly, in wide recognition that this is a bad thing, Wikipedia:Drop the stick is a powerful rebuttal to someone who tries too hard. The outcome is neither “consensus”, nor “no consensus”. Often, the answer is “not now”, this is not the current priority. Consensus is always the goal, but it may be a very distant goal that we take small steps towards. SmokeyJoe (talk) 07:37, 2 January 2023 (UTC)

  I'm sorry, I've evidently misunderstood the purpose of this section. It begins with you asking when a dynamic edit becomes a silent consensus edit and giving some possible criteria. I meant my "what is your test?" question to find out which criteria you believe are appropriate and how you would apply those criteria.

  Maybe I should have asked the underlying question: "what happens when a dynamic edit becomes a silent consensus edit?" Does it make any difference in how a BOLD edit is treated? - Butwhatdoiknow (talk) 15:29, 2 January 2023 (UTC)

Different Types of Consensus

We're having some interesting discussions regarding silent "consensus" (compare the sound of one hand clapping). I thought it might be helpful to compare the consensus that arises from other paths as well. - Butwhatdoiknow (talk) 18:06, 31 December 2022 (UTC)

Step One: Identifying edit pedigrees

To start, let's take a step back and consider the ways that text ends up on (and off) Wikipedia. So far I have identified three paths:

• text that is added (or removed) without discussion (before or after edit) or modified
• text that has been modified without discussion
• text that results from or has been approved by discussion (including edit summary “discussions”)

Are there more? - Butwhatdoiknow (talk) 18:06, 31 December 2022 (UTC)

I would love to see User:Andrewa, author of User:Andrewa/Consensus is consensus, make the first answer here. SmokeyJoe (talk) 07:40, 2 January 2023 (UTC)

Andrewa's essay seems limited to policies and guidelines, wp:EDITCON is not (although wp:PGBOLD is). Perhaps that is part of the problem as we try to figure out silent consensus. As you say, Andrewa may well help us move forward. - Butwhatdoiknow (talk) 15:34, 2 January 2023 (UTC)

It was supposed to be about anytime we work towards consensus. But consensus is a good servant not a good master. The term was a favourite of arguably Australia's most manipulative Prime Minister to date. Rumour has it he now doesn't even get Christmas cards from his former Cabinet. Andrewa (talk) 10:34, 3 January 2023 (UTC)

I am indeed flattered. But I think my best comment on consensus is actually at wp:creed#consensus.

It is the people that make Wikipedia work, not the rules. Andrewa (talk) 10:34, 3 January 2023 (UTC)

NOCON - only after vs when

Regarding this edit, I think "only after" is better because it makes clearer that this section does not apply during discussions. It's a small difference, but important because too many editors cite wp:NOCON (what happens when discussion fails) when they should be citing wp:QUO (recommended practice when discussion is taking place). - Butwhatdoiknow (talk) 16:10, 4 January 2023 (UTC)

It is true that the "No consensus" section focuses on discussion and what happens after one ends with no consensus. However, discussion isn't the only path to no consensus. An editor can determine through editing alone that they don't have consensus for an edit they've made (e.g. Alice makes a bold edit, Bob reverts bold edit, Alice sees the revert and moves on instead of pursuing consensus through discussion). Inserting "only after" in front of "good faith discussion" can be misleading in that regard. --GoneIn60 (talk) 16:37, 4 January 2023 (UTC)

I would interpret Alice's decision to leave the playing field as a form of consensus - she concedes, agreeing to let Bob's revert stand. There is in contrast to a "state of 'no consensus'" situation where Alice enters into a discussion with Bob and there is no compromise or concession.

As user:Thinker78 has pointed out, the problem with the sentence is that "no consensus" appears twice with different meanings. That's a problem I'm hoping to solve when I have more time. - Butwhatdoiknow (talk) 17:49, 4 January 2023 (UTC)

Guess it's a matter of perspective; at some point someone is conceding. If discussion stalls and neither party escalates further, you could argue by the same token "a form of consensus" was reached. To me, it makes no difference at what stage of the interaction Alice decides to walk. It's the same outcome, regardless if you call it "no consensus" or "a form of consensus". --GoneIn60 (talk) 21:19, 4 January 2023 (UTC)

If a discussion stalls? How do you tell the difference between one party exhausting another, and one party silently conceding? I don’t think that a discussion stall should ever be considered even silent consensus. SmokeyJoe (talk) 01:00, 5 January 2023 (UTC)

Exactly my point. We can't read between the lines. "Stall" in this sense is just like Alice walking away before discussion; neither situation necessarily means one party is conceding. But from a certain perspective, some may choose to call it that. I choose to call it "no consensus" until there is one. --GoneIn60 (talk) 16:15, 5 January 2023 (UTC)

Lack of consensus should not result in inclusion

I favor this revision, with changes by me and a refinement by User:Blueboar. It states:

When discussions of proposals to add, modify, or remove material in articles end without consensus, the common result is to remove the contested material.
* In biographies of living people, contested material that is removed should not be restored until a discussion results in affirmative consensus to include.

This enshrines both accepted practice, and long-standing consensus from policy discussions.

The current wording of this policy "locks in" material when there is no consensus for its inclusion. That contradicts the numerous (and extensive) recent discussions at WT:V, including the latest one, in which most editors agreed that material needs consensus to be included. That's also the plain wording of WP:ONUS, and repeated attempt to water it down have failed. Per WP:POLCON, this page needs to change to reflect the version above, which far better reflects wider Wikipedia consensus. DFlhb (talk) 15:37, 9 February 2023 (UTC)

Courtesy ping: @Butwhatdoiknow DFlhb (talk) 15:38, 9 February 2023 (UTC)

• The first part of the above version does not match my understanding of consensus re “NOCON-after discussion”. Except for BLPs, an end result of “No Consensus” does not always mean we “remove the contested material”… it’s more case specific than that: we return the article to whatever state it was in prior to the contested edit. The retention/omission of the material depends on that prior state. If the contested edit was an addition to the prior state… yes, we omit the addition. BUT, if the contested edit was a removal from the prior state… we return the removed content.

In other words, “no consensus” (in a non-BLP) rejects whatever the “change” was. But that rejected “change” can be either an addition or a removal. Blueboar (talk) 17:49, 9 February 2023 (UTC)

• That said… one issue that keeps coming up in debates is the question of what constitutes the “change”… this is why we have endless arguments about which version is “long standing” and has “implied consensus”. The NEW wording (above) does cut the gordeon knot of such debates. Thus I am not necessarily opposed to it. Blueboar (talk) 18:23, 9 February 2023 (UTC)

  If I may advocate for it a bit more: it may motivate the pro-inclusion side to try harder to come up with better sources or phrasing. The encyclopedia wins out. And it clarifies that we don't need affirmative consensus to remove content added by an IP ten years ago, on a page that may be obscure enough that such affirmative consensus is hard to form. It incentivizes editors to think about whether the content is actually good, rather than treating any content as if it were immanent. What is, stops being valued more than what could be. DFlhb (talk) 18:41, 9 February 2023 (UTC)

It is unclear exactly what change this section is about. Including whether or not it applies only to BLP's. The link is to a version of the entire page. If this section is to be meaningful, that should be clarified. North8000 (talk) 19:56, 9 February 2023 (UTC)

True. Here's a proper diff. DFlhb (talk) 20:05, 9 February 2023 (UTC)

I would oppose that bundle of many changes for several reasons. A blanket / categorical statement that anything needs a supermajority (which is basically what a consensus is, and yes, I know it's not a vote) to be added, kept or removed would be handing a giant hammer to wililawyers /and POV warriors. And also wading into an area where there are too many variables to be making such a simple categorical statement. This mess, including the conflict with ONUS is going to need more than just a tweak to fix. next, it removes any weight given to the "last stable version" . Also it changes "contentious" to "contested". For example a POV wililawyer warrior that wants an article on someone to be a hit piece would want to exclude that they won an academy award. So "won an academy award" is not contentious material but it is a contested inclusion thus requiring a full debate plus winning a supermajority in order to include. North8000 (talk) 20:15, 9 February 2023 (UTC)

I agree with changing back "contested" to "contentious"; that was poor wording on my part. But is the rest of this actually a problem in practice? It's purely equivalent to ONUS, which has been policy for years, has been widely invoked and applied, and hasn't led to this doom-and-gloom. I'll note that BLPs already function the way you describe: no consensus on contentious material already leads to its removal. DFlhb (talk) 20:39, 9 February 2023 (UTC)

Yes, @DFlhb, you're right about BLPs. But imagine that we're talking about, I dunno, a film from decades ago. Someone wants to the article to say that the director died the day after the film was finished. It's well sourced and the facts are undisputed. Someone else, however, thinks this is just unimportant trivia and belongs on some fansite (or only in the biography for the director). There's an RFC, and the outcome is split straight down the middle, with 50% of participants demanding its exclusion and 50% of them demanding the inclusion.

NOCON says we usually go with whatever the article "always" said, which could be including or excluding. You're saying that it should be excluded per ONUS, even if it's been in the article without any hint of concern for 15 years now. I'm asking you: Is having less well-sourced, accurate information always best for Wikipedia? That is the natural result of applying ONUS to such a case. WhatamIdoing (talk) 00:26, 13 February 2023 (UTC)

Fair points; but boldness is preferable to stability. Technology was untouched since 2012, but awful, and I needed to remove half of it, even sourced stuff, to improve dueness and remove stuff a professional encyclopedia wouldn't cover about the subject. Many articles contain sourced statements that come from a 10 min Google search, rather than books or studies. So our articles on operating systems can include utterly routine coverage of 0-days that's sourced, but completely undue. And sometimes, things may be cited but may not reflect the majority of sources (see my comment at Talk:Joe Biden#Taiwan).

Stability would be preferable if half our articles were FA-class. But most of them suck! Only 11% are rated above Start-class. Again, see Technology, before and after. Enshrining stability stunts article quality. If material was never scrutinized, why require affirmative consensus to remove? The stability mindset is the wrong one altogether: it leads editors to think "why should anything change", rather than "how do we make the best article?". The random walk framing in Paul Siebert's first comment at WT:V#Answers, is simply brilliant, and completely true. Many editors' takeaway from WP:NOCON is that the status quo has inherent validity, so they're not incentivized to scrutinize that status quo or consider whether it's any good. That benefits no one. DFlhb (talk) 07:36, 13 February 2023 (UTC)

Thank you for your work on Technology. Articles on fundamental subjects are incredibly difficult to write well.

There are two axes available to us:

• Boldness – Stability
• Include – Exclude

Your examples tend towards "boldly excluding". But if we write the rule as "boldness", that encompasses "boldly including", too. Do you want people to boldly dump new content into articles? WhatamIdoing (talk) 01:40, 14 February 2023 (UTC)

Do you want people to boldly dump new content into articles? That's fine, as long as that material can be challenged just the same if it's 1 minute old or 10 years old.

My fundamental issue is the fact that CON says that implicit consensus evaporates instantly when material is challenged, and yet I still hear people say: "it's been there for years, that's the consensus version". The only time that ever actually applies is when an RFC decided on the exact wording. Otherwise, the "consensus" is just stonewalling or wikilawyering. People should have to defend the status quo on its merits, not on technicalities. DFlhb (talk) 02:47, 14 February 2023 (UTC)

RFCs don't produce binding decisions. They help is identify the consensus (if any) at one point in time, but Wikipedia:Consensus can change even if you had an RFC about it in the past. Also, you can have a solid demonstration of consensus (at that point in time) without an RFC.

I think the way we've been talking about implicit consensus for the last several years has been misleading. Part of this is just because Wikipedia:Nobody reads the directions, so when someone says something like "WP:ALLCAPS says that I'm right and you're wrong", then people tend to believe that the rules are being accurately and fairly represented. After all, why would any editor ever lie, given that all I have to do to discover their duplicity is to click the link they gave me? But the editor talking to you probably hasn't read that page for years, if ever, either, because they had exactly the same thought. They aren't being duplicitous; they're just repeating the rumor they heard.

It might be helpful to suggest/model some more appropriate language in such disputes. Right now, some people are saying "This has been included in (or excluded from) the article for so long that it has implicit consensus". Imagine if people saw others saying "You've just disputed this, and WP:IMPLICIT says as soon as anyone's disputed something, we can no longer presume consensus for it. That said, I think the old version of the article was better..." WhatamIdoing (talk) 16:51, 14 February 2023 (UTC)

DFlhb, the current wording is necessarily non-prescriptive and relates to situations with an immense range of variables and considerations. Your proposed change is much more prescriptive; I think that in your proposal and discussions you have a particular type of scenario in mind and feel that the policy should be modified to provide a more desirable outcome for that particular type of scenario. I don't support that change, but perhaps the current systems is better at what you seek than you perceive from a simple reading. I think that implicit consensus is simply a way to put into words that, in consideration of multiple factors, some weight is given to the status quo based on it having been there, with amount of time it's been there and how many eyes it has had on it being additional factors that determine that amount of weight. Sincerely, North8000 (talk) 17:26, 14 February 2023 (UTC)

NOCON is among our most abused policies. People treat it as prescriptive when it benefits them. Some even use it to defend statements in BLPs, despite the BLP exception to NOCON. BTW, the assumption that longstanding content has been properly scrutinized is often false. Almost no one does spot checks outside of GANs. I've even seen articles pass GAN with zero spot checks, and numerous errors! DFlhb (talk) 18:44, 14 February 2023 (UTC)

Related RFC two months ago: Wikipedia talk:Verifiability/Archive 78#Straw poll about verifiability and consensus.

I was a little surprised to see few editors willing to support inclusion of well-sourced verifiable material. I know we've been trending that way for years, but this was more than I expected. Fifteen years ago, I think editors would have said that since an AFD closing with no consensus results in keeping an article, if a discussion about removing a paragraph ends with no consensus, we should keep the paragraph. I don't remember hearing that analogy for several years now. WhatamIdoing (talk) 05:01, 15 February 2023 (UTC)

I never understood that analogy. I think there is a huge difference between: “should we have an article on this topic?” (Default being “yes” if no consensus at AFD) and “should our article contain this specific bit information?” (Default being “no” if there is no consensus in discussion). Blueboar (talk) 13:57, 15 February 2023 (UTC)

The way the question is posed could make all the difference. Are we (1) removing a recently-added paragraph, (2) removing a longstanding paragraph, or simply (3) adding a new paragraph? Your straw poll addressed scenario #3, which seems to me like the easiest "exclude" conclusion out of the three. But is there really a difference between scenarios #1 and #3? Should #2 have the potential of reaching a different outcome? Some argue the framing of the question and what is considered the bold edit matters. I can see them changing their answer accordingly based on those factors.

North8000 once described ONUS as putting "a finger on the scale towards exclusion of material". If you agree with tilting the scales in favor of exclusion, then you are probably more likely to frame the question in such a way that your opposition must defend inclusion (as opposed to them forcing you to defend removal). --GoneIn60 (talk) 07:17, 15 February 2023 (UTC)

Thanks. As I recall the context of that was to say how the fuzzy Wikipedia system successfully deals with the the prima facie conflict between onus and wp:consensus. IMO onus is a mis-worded attempt to to say something that needed saying in that spot (note that it is in wp:ver) and needs to be fixed. There was a multi-month discussion on that which ran out of gas. I was going to formulate an RFC but never got it done. North8000 (talk) 16:00, 15 February 2023 (UTC)

Thanks for the clarification. Sorry if I digressed into ONUS territory...also wasn't my intention to steer the focal point away from DFlhb. Just wanted to point out that it does seem that the framing of the question asked during debates can have an impact on the responses and how a "no consensus" result is interpreted. --GoneIn60 (talk) 16:31, 15 February 2023 (UTC)

This is true… “No consensus” can be interpreted as (a) “we don’t agree to do X” or (b) “we don’t agree to NOT do X” depending on how the original question was framed. There is also a third, more neutral interpretation: (c) “We could not agree either way… neither to do X nor not to do X.” The first two interpretations (a & b) imply rejection, the third (c) does not. Blueboar (talk) 17:01, 15 February 2023 (UTC)

I have wondered occasionally whether we ought to be clearer about that in NOCON. Sometimes "no consensus" means "this is a strong candidate for Wikipedia:Lamest edit wars, but I'm trying to de-escalate this situation by being vague." NOCON is about the (c) situation, not diplomatically phrased rejections. WhatamIdoing (talk) 17:18, 15 February 2023 (UTC)

My scenario was this:

Alice and Bob decided to create an article together. They were editing collaboratively until they ran into a serious disagreement. One of them wants to briefly mention a related subject, and the other doesn't think that related subject should be mentioned at all (in this particular article).

In this scenario, all the content is presumably "recently added" and "new", because the article itself is new. There cannot be any "long-standing" content when the article itself was just created. You don't know whether someone has added or removed it, but that doesn't matter, because it would obviously be silly to talk about a "long-standing" or "status quo" version in a brand-new article.

You also don't know whether it's "a paragraph". This scenario encompasses not just a short paragraph but also a sentence, a short phrase, or even just a single word ("briefly mention").

@GoneIn60, I think your #1 and #3 are the same thing. They're both about disputes involving new content in old articles. Let's imagine that Alice adds a new sentence to an old article, and Bob promptly removes it. If there had been a long-standing version (which, again, there is no long-standing version in the RFC's scenario, so you are talking about a different scenario), then QUO maximalists would argue for removal in both cases. ONUS maximalists would also argue for removal in both instances. What difference does it make to the outcome if the dispute is said to be "Bob removed Alice's new sentence" instead of "Alice added a new sentence that Bob rejects"? WhatamIdoing (talk) 17:12, 15 February 2023 (UTC)

Question: what do Charlie, Dave and Ellen think? Blueboar (talk) 17:21, 15 February 2023 (UTC)

In the original RFC scenario, the hypothetical respondents in Alice and Bob's dispute were exactly balanced, and both groups had equally strong arguments.

One (real) respondent to the (real) RFC said that in the scenario I gave them, the question ultimately came down to editorial judgement. I thought this a very clear way of phrasing it, exactly in line with what I intended. Either response could be considered reasonable, appropriate, and fully compliant with Wikipedia's rules. In that specific scenario, in that discussion, more respondents recommended exclusion. WhatamIdoing (talk) 18:22, 15 February 2023 (UTC)

In your example, letting "editorial judgement" tip the balance was the perfect outcome rather than trying to imagine a binary flowchart rule from policies. Well-intentioned Editors might consider the degree of WP:Relevance/germaneness, importance in relation to coverage of / understanding the topic, whether or not the addition or exclusion is POV'ish, how encyclopedic the addition is, the strength of sources for it, principles behind wp:weight (even though wp:weight is dysfunctional) etc. Even if it fails to produce the "supermajority" result of a wp:consensus, wp:consensus is also about that discussion process. North8000 (talk) 19:19, 15 February 2023 (UTC)

I agree, but when you have a discussion that takes into account all the relevant facts and circumstances, is unable to resolve the question of whether editorial judgement is "include" or "exclude" in the specific instance, then what? Tell the closer to flip a coin? WhatamIdoing (talk) 07:01, 21 February 2023 (UTC)

Agreed. From my perspective, #1 and #3 are essentially identical in your scenario, so the framing of the question makes no difference. We don't really see #1 and #3 diverging from one another until the disputed material in question becomes a "long-standing" concern. So unsurprisingly, the framing seems to always matter for #2.

Back when I weighed in favor of exclusion in the hypothetical scenario, I suppose the line of thinking was that the bar for inclusion should naturally be just a little higher, especially in regard to new content. If it ultimately belongs in the article, we'll eventually get there. And honestly, I'm not sure the idea of applying AfD "no consensus" outcomes to content disputes ever crossed my mind. AfD seems like a nuclear option, whereas content disputes are more like small battles over tiny steps forward or backward. --GoneIn60 (talk) 03:31, 16 February 2023 (UTC)

Contested material during vs after discussion

• We seem to be mildly edit warring about two related but distinct issues in the NOCON section. A) what to do during discussion, and B) what to do after discussion if there is no consensus. It may help to divide our instructions to deal with each situation separately. To aid in this, here is my summary of each situation:
  • A) My understanding is that during discussion, we usually keep contested material in place (so that editors can easily see what the discussion is about)… however, in a BLP we do the opposite (favoring removal during discussion, due to the extra care that BLPs require).
  • B) My understanding is that if there is still no consensus after discussion, we usually return the article to “status quo” (ie whatever state it was in, prior to the dispute - which might mean either keeping or removing the contested material, depending on what the previous status quo was). However, in BLPs we always favor non-inclusion if there is no consensus.

Please discuss if need be. Blueboar (talk) 15:51, 9 February 2023 (UTC)

I agree on A, and agree that the distinction between conduct during discussions, and after discussions, should be made clearer. But I disagree on B; there was "overwhelming consensus" that ONUS should not apply only to recent insertions, and therefore that the status quo does not take precedence over ONUS. That still seems to be the consensus. I just don't understand why we would ever want to include material when there is neither affirmative consensus, nor implicit consensus (since the material was contested). DFlhb (talk) 18:03, 9 February 2023 (UTC)

That is not what that consensus says. It was about whether to move ONUS to CONSENSUS and to rewrite it in one specific way; many of the opposes are premised on that aspect and clearly did not support the sweeping interpretation of ONUS that some people expressed in recent discussions. The closing statement summarized it purely as a rejection of that one change, not as a sweeping endorsement of one view of ONUS. Clearly there is not a more general consensus about the applicability of WP:ONUS given that a long-running discussion then and there has failed to reach a conclusion after months of discussions. --Aquillion (talk) 20:23, 9 February 2023 (UTC)

But rewrite it in what specific way? The proposal was to change ONUS to require affirmative consensus to remove, not just to keep. In other words, making ONUS substantively identical to NOCON. As Blueboar summarized it in the WT:V archives, most of the opposition focused on the language change, not on the move.

User:Buidhe's close found overwhelming consensus against the change, primarily due to concerns that it would undermine verifiability and efforts to remove poorly sourced information. That's a direct repudiation of the idea that ONUS should force us to keep things unless there is affirmative consensus. So why should NOCON? DFlhb (talk) 20:55, 9 February 2023 (UTC)

I agree with DFlhb, that is, not Blueboar's B. My reasoning is saying "status quo" is unhelpful, and it's not about being prior to a dispute, the wording is "prior to the proposal or bold edit", which means prior to the insertion in this sort of case. Peter Gulutzan (talk) 21:12, 9 February 2023 (UTC)

A blanket / categorical statement that anything needs a supermajority (which is basically what a consensus is, and yes, I know it's not a vote) to be added, kept or removed would be handing a giant hammer to wililawyers /and POV warriors. And also wading into an area where there are too many variables to be making such a simple categorical statement. This mess, including the conflict with ONUS is going to need more than just a tweak to fix. North8000 (talk) 20:06, 9 February 2023 (UTC)

I generally agree with this. I feel that a lot of people are trying to solve complicated issues with sweeping statements (a problem we have with WP:WTW, among other things), but that it rarely works out in practice. The reality is that most editing relies on implicit consensus and the fact that all edits are presumed to have consensus until they are challenged. I think that our policies beyond that - for when a challenge occurs - should be focused on trying to bring people to the table, get them to clearly state their objections or what they want to change / add / remove, and push them towards a consensus. Sweeping policies that people can point to to say "I win, I don't have to give you a detailed rationale, start an RFC if you disagree" with minimal engagement are bad because they encourage WP:STONEWALLing and generally make consensus-building more difficult; policies work better when they give everyone a reason to come to the table. At the very least, one thing I would like to see is policies that discourage "rv., get consensus" sorts of edits (unless there's an established objection, dispute, or contrary consensus that the editor is clearly referring to) - allowing people to revert or remove material without stating an objection tends to lead to more WP:TEND editing because people making tendentious edits often can't articulate their real reason for removing something. And even when someone is editing in entirely good faith, it's good to force them to think about and articulate their reasons - sometimes "this just feels wrong, somehow" has sound grounding, and sometimes forcing them to think through and articulate why an edit they object to feels wrong will lead to a productive compromise.--Aquillion (talk) 20:23, 9 February 2023 (UTC)

You have it mostly right… however, the same argument can be made about people who want to add disputed information… they (too) need to articulate why they think that the addition is an improvement. Why it should be retained.

Second, I think you are too focused on defending additions. “No consensus” isn’t always about additions. It can also be about removals. A lot depends on how the question is asked (are we seeking “consensus to add” or “consensus to remove”? Often that isn’t clear).

This is why I feel that (when there is a disagreement) the responsibility to build a consensus should be on those who want to change the article - regardless of whether that change is to add or subtract information. Those seeking change need to positively outline why the change is beneficial. And when it turns out that there is “no consensus” to change, we default to “change is not accepted - return to prior version”. Blueboar (talk) 21:05, 9 February 2023 (UTC)

I think this "changes" framing is imprecise. I made a table for what each policy recommends under different scenarios.

After a no-consensus discussion:	ONUS	NOCON
To add material:	Ng Not added	Ng Not added
To keep material:	N Remove	Y Keep
To change existing material:	Ng Keep as is	Ng Keep as is

The only question is whether we should have a bias toward inclusion or toward exclusion in matters where there is no consensus for inclusion, since that is the only substantive difference between ONUS and NOCON.

I can think of numerous reasons to have a bias in favor of exclusion (including verifiability and keeping out fringe POVs). If an editor cannot convince others that a piece of content is worth including, why should it be given a platform on a site that gets hundreds of millions of daily visitors?

A great example is the debate over the inclusion of a flag and seal on Jerusalem. They were added with zero discussion and zilch verifiability (so bad that the Flag of Jerusalem article was itself later deleted). A discussion on removing them yielded no consensus, so a 151-comment RFC was needed to get rid of them. That's a huge waste of editor time, when ONUS would have avoided the need for the RFC. Can anyone provide examples of NOCON situations where we wouldn't want to delete, at least until proper sourcing/dueness is shown? DFlhb (talk) 22:18, 9 February 2023 (UTC)

Sourcing falls under the rest of WP:V - ONUS (verifiability does not guarantee inclusion) only applies once WP:V is satisfied, and should therefore never be invoked for sourcing concerns - in fact, by invoking ONUS, you are implicitly acknowledging that the contested material satisfies WP:V and is properly sourced. I've noticed a lot of defenders of an expansive interpretation of ONUS and a minimalist interpretation of NOCON tend to make this mistake; possibly the confusion is part of the reason for that position in the first place. Anyway, when talking about whether to include or exclude things that everyone agrees is verifiable and properly-sourced, the fact is that the vast majority of text on the wiki does not have explicit consensus. This means that if our resolution to no-consensus situations worked the way you're saying (ie. defaulting to removal), anyone in a contentious article could delete huge swaths of its text and demand a full RFC with a clear consensus to restore it. For an example where I'm involved, if ONUS / NOCON worked the way you describe, I would immediately delete most of the content on eg. Mermaids (charity), which I feel is undue and lacks an unambiguous consensus (and I don't think an unambiguous consensus could be demonstrated in favor of keeping the current text overall.) I think that this would improve that article, but as tempting as it is, if we applied that as the general rule, no article would be stable because massive removals of anything that hadn't previously achieved unambiguous consensus would become the norm; and any significant text on a controversial article would end up requiring reaching formal consensus, which is a time-consuming process that was never intended to be applied at that scale. This isn't workable - once the basic requirements of WP:BLP and WP:V are satisfied, article stability becomes a significant concern. Major changes to articles need to require consensus, including massive removals of text, or it becomes too easy for articles to destabilize; requiring a process-heavy demonstration of consensus just to retain the longstanding structure of a high-traffic article isn't feasible and would make editing in controversial areas even more time-consuming. --Aquillion (talk) 22:31, 9 February 2023 (UTC)

I agree with the overall picture given in Aquillion's posts. Another way that I would say it more mechanically to say that the decision making process is influenced by wp:Consensus, WP:Onus, strength of arguments given and other other factors per Wikipedia:How_editing_decisions_are_made which includes "we're here to build an encyclopedia". In turn, the implementation of wp:consensus is itself influenced by other variables such as the details of history of the questioned material in the article (such of time span and amount of scrutiny under an implicit consensus). Any categorical statement of which should be include or excluded based on just 1 or 2 factors is usually a recipe for trouble. North8000 (talk) 22:57, 9 February 2023 (UTC)

• I've noticed a lot of defenders […] tend to make this mistake I notice it too, but note that it is not my mistake. My reading of ONUS is the same as yours: it applies when WP:V is already met. I was imprecise: the flag was poorly sourced, but the coat of arms was fully verifiable. But the Jerusalem RFC was needed to prevent editors from edit-warring to enforce NOCON. That proves my point that it is NOCON that is harmful and oft-misinterpreted, not ONUS.

The Mermaids article is a highly contentious one, and I don't think it's too much to ask that its contents have affirmative consensus. That's exactly what we want with these kinds of articles, to prevent fringe POVs, WP:FALSEBALANCE, and especially stonewalling. But the amount of contentious articles is overall tiny, and WP:IMPLICIT consensus works just fine for everything else.

Perhaps an RFC would be good? It should ask whether ONUS should apply only to BLPs or to all articles? The former is utterly uncontroversial (per Blueboar), but I agree the latter is somewhat contested, though I argue it matches the consensus of previous discussions.

I also argue that if ONUS works well on our BLPs, including our most controversial ones, then there's no reason to think that making NOCON match ONUS would cause a torrent of instability. There's no wording in ONUS that states it only applies to BLPs, so this apocalypse would already have happened long ago. DFlhb (talk) 23:39, 9 February 2023 (UTC)

For good order's sake, the sourcing problem came up during the course of the RFC, it was not the basis for the initial challenge, which was based on conlevel (none and silence thereafter) and contradiction with an earlier RFC (so not NPOV) with a much greater level of consensus) for the included material. Selfstudier (talk) 15:44, 2 March 2023 (UTC)

I don't think that table is correct. I think it looks more like this:

After a no-consensus discussion:	ONUS	NOCON
To add material:	N Not added	Ng Keep as is
To keep material:	N Removed	Ng Keep as is
To change existing material:	(Silent)	Ng Keep as is

ONUS doesn't say anything about changing existing material (e.g., simple copyedits). It's all about inclusion and exclusion. WhatamIdoing (talk) 00:33, 13 February 2023 (UTC)

• One thing we have to make clear… “usually” does not mean “always”. This is an area where we have made lots of exceptions. We can find examples where we removed due to “no consensus”, and examples where we kept due to “no consensus”. The question is, what do we usually do? Answering that requires some statistical analysis, not a few specific examples. Blueboar (talk) 00:37, 10 February 2023 (UTC)

I'm on this same wavelength. We are not and should not be prescribing what has to happen. Whatever is decided, the phrasing should remain clear that it's simply a common result; there are always exceptions. Perhaps some who hesitate to change NOCON forecasting disaster are neglecting to take that into account and forgetting to apply a little common sense. Editors aren't all of a sudden going to start removing large swaths of verifiable article text just because they can. They'll still have to make a reasonable argument if they expect the ensuing discussion to reach the point of no consensus; without one, their position is vulnerable to a landslide of opposition. --GoneIn60 (talk) 02:48, 10 February 2023 (UTC)

As I've said below, our current practice is that longstanding text is retained with a no-consensus outcome (unless someone can affirmatively demonstrate that it lacks consensus); I think that this is so obvious that it hardly needs demonstrating. The reason why I'm so vocal about the need for WP:NOCON as it stands is because I absolutely believe that this is one of our "load-bearing" policies, one that we've frequently relied on for the outcome of key disputes on highly-controversial articles. This isn't a minor accident of policy that was added without people noticing and which only survived through the passage of time (the way, to be blunt, the problematic sentence on WP:ONUS is), this is a key pillar of how we've historically resolved disputes. If people are going to argue that that somehow isn't the case we can go through and demonstrate it, especially if demonstrating that fact would finally put this to rest, but I think the fact that no-consensus RFCs normally result in the status quo is obvious. Since that has been our policy for a long time, the burden is on anyone who wants to change it to demonstrate that a change to it would be beneficial. Beyond that, our policies on no-consensus outcomes are an extremely delicate balance - I agree that the phrasing only proscribes the default outcome, but I feel that the proponents of "everything must have overt consensus somewhere in order to stay" are the ones who fail to understand this. If consensus breaks down over longstanding text, and no one can convincingly demonstrate that it never had consensus for one reason or another, then that shows that, overall, it probably has sufficient support that it is not a problem, which means that we should prioritize article stability over WP:BOLD changes to an article's longstanding text made without consensus; if someone wants to make a significant change to an article, they should be the ones who have to make the argument for it and convince people, or at least they should bear the burden of showing that the text that they want to remove lacks even implicit consensus. And I feel that shifting too much weight towards a default outcome of removal in nearly all circumstances would make writing articles in highly-fraught topic areas - especially ones that have discussions break down along the faultlines of real-world disputes - much more difficult. Our default outcomes work better when they are cautiously-worded and function in a way that encourages everyone to come to the table; in my view the current practice (mostly) does that, while the proposed changes do not. --Aquillion (talk) 16:40, 11 February 2023 (UTC)

@Aquillion, I'd be more willing to believe your nice speech about how the NOCON section is a key pillar, except that you'd already been editing for about a decade when I started that section. We managed for more than a decade without it; if we got rid of it, I think we'd manage to survive.

I do agree that editors in certain areas (e.g., geopolitical disputes) are more likely to prize article stability over removing possibly bad content. Editors in other areas (e.g., pseudoscience) have the opposite view. Both groups seem reasonable to me. WhatamIdoing (talk) 00:40, 13 February 2023 (UTC)

Surely you recognize that WP:NOCON merely describes long-established practice. As early as 2007 (when it first started to really be fleshed out), this page stated that longstanding text was presumed to have consensus. To me, the real issue here is that I think that some people have mistakenly interpreted WP:ONUS as not accepting that sort of consensus as valid, and therefore quietly reversing longstanding practice in that regard. But clearly that isn't what it says; if we accept that "longstanding" text (for a certain idiosyncratic definition of longstanding that isn't just about time) has consensus and that a new consensus is therefore required to remove it, then the contradiction disappears. --Aquillion (talk) 09:13, 13 February 2023 (UTC)

@Aquillion, I created NOCON. I intended for it to reflect the community's view. But I didn't add that particular sentence because I was unable to establish that it truly represented the community's view. The disputed sentence was added later, by another editor, who has since been blocked. Since then, we've tried to make it work, but that's happened mostly by adding exceptions. The BLP exception alone means that the "QUO" notion of NOCON doesn't apply to more than a million articles.

The contradiction doesn't disappear. ONUS appears to require a real/demonstrable consensus, not just a presumption. The contradiction only exists in rare circumstances, though; it is not especially common for discussions to truly end in no consensus either way. It's much more common to find that there is no consensus in favor of X, and therefore we will do not-X. WhatamIdoing (talk) 02:00, 14 February 2023 (UTC)

Does anyone else see the irony with "that particular sentence" getting stuck as the longstanding version? Infinity mirror effect? --GoneIn60 (talk) 16:05, 14 February 2023 (UTC)

This is due to people asking: “Should we do X?” instead of the more open ended: “What should we do?” Blueboar (talk) 16:32, 14 February 2023 (UTC)

Having asked "What should we do?", in multiple discussions on several pages, is why I refused to add a statement about what we should do. I got such a variety of answers that I became convinced that whatever we wrote would be wrong in some significant fraction of situations, unless we wrote a really long (e.g., Wikipedia:Articles for deletion/Common outcomes-long) description.

@GoneIn60, I agree with you that it's ironic that this sentence has gotten stuck, and also unfortunate. Usually, if a policy statement doesn't reflect reality, we can get it removed without too much fuss. (For example, a mostly-wrong statement in this section about reverting admin actions was removed, and nobody was fussed about it.) In this case, though, I think that it's sticking because some people find it a very useful weapon: They can't get actual/demonstrable consensus, so they need a written rule that they can use to overrule the discussion outcome. Even if you only want this occasionally, when you want it, you really want it, and that can make you a staunch defender of its existence, even if you think it's wrong or unimportant 90% of the time. WhatamIdoing (talk) 17:00, 14 February 2023 (UTC)

Let me rephrase… before we ask: Should we do X?” we need to ask: “Should we do something?”… I find it is often helpful to first lay out what I think is problematic with the current state of the article, and establish a consensus that “something needs to change.” Then we can discuss what that change should look like. It may take time to reach consensus on the second part, but at least we know that the previous version doesn’t have consensus. Blueboar (talk) 17:37, 14 February 2023 (UTC)

• I mean, I don't think there is a reasonable dispute over what we do currently: As a general rule, "longstanding" text (a complicated and dicey term itself, but overall, any one where the removal can reasonably be called WP:BOLD) is retained following a no-consensus outcome, unless someone goes out of their way to proactively demonstrate that it has never had consensus. eg. if they can show it was disputed from the moment it was added, or the article is so low-traffic that they can reasonably argue that nobody has reviewed it before, then it can be removed - but they must affirmatively demonstrate this and convince whoever is closing the discussion of that fact, so it is clear that the "default" is otherwise for longstanding text to be presumed to have consensus and therefore to be retained when a discussion fails to reach consensus, with the burden of demonstrating that it lacks consensus being on whoever wants to remove it. That is current practice and has been our practice for well over a decade. We can survey no-consensus RFC outcomes if you want, but surely you aren't seriously arguing that, historically, any significant number of no-consensus outcomes over longstanding text or WP:BOLD removals have defaulted to removal? --Aquillion (talk) 16:13, 11 February 2023 (UTC)

  surely you aren't seriously arguing that, historically, any significant number of no-consensus outcomes over longstanding text or WP:BOLD removals have defaulted to removal - I've seen this happening more often than not in low-profile contentious articles where there isn't enough participation to generate a discussion that can be meaningfully closed.

  Something gets removed and if you have a disagreement between 2-4 editors, there's no affirmative consensus to restore and ONUS applies. I do agree with North8000 about the dangers if this were to be changed, but it does come at a price. PaulT2022 (talk) 06:57, 12 February 2023 (UTC)

I concur with and reinforce what Blueboar, GoneIn60 and Aquillion said. North8000 (talk) 16:36, 11 February 2023 (UTC)

• Are we having a mild edit war about what NOCON should say about contested material during discussion? If so, we shouldn't. NOCON is only about what happens when discussion fails. wp:QUO is about what happens during discussion. - Butwhatdoiknow (talk) 04:35, 10 February 2023 (UTC)

A way of resolving "No consensus"

There is a discussion about a way of resolving "No consensus" at Wikipedia:Village pump (idea lab)#A way of resolving "No consensus". Your input would be appreciated. Thank you, SchreiberBike | ⌨  03:58, 7 March 2023 (UTC)

Consensus dispute with editor

Hi all. I have a dispute with an editor at User talk:Bon courage#Please join active discussion instead of unilaterally reverting. It is about not following proper consensus procedures. If you can join the discussion to offer your insights that would be great. Regards, Thinker78 (talk) 01:11, 26 May 2023 (UTC)