Wikipedia:Requests for comment/When there is no consensus either way

When editors are unable to form a consensus either way about including or excluding a particular fact in a Wikipedia article, should they default to including or excluding that fact? (This question is not about living persons.) WhatamIdoing (talk) 02:40, 6 January 2024 (UTC)

The story

Template:Box-rImagine that an editor recently created a new article. The subject is obviously notable, and all of the information in the article is correctly cited to an appropriate reliable source. All editors agree that all of the information in the article is accurate, verifiable, and neutrally phrased.

However, editors currently disagree about whether one particular fact in the new article should be in this article at all. There is no dispute over the wording of the sentence that presents this fact; the dispute is specifically and completely about inclusion or exclusion of the fact at all. (The exact nature of the potentially unwanted information is unimportant, except for the fact that the disputed material has nothing to do with any living person, so Wikipedia:Biographies of living persons does not apply.)

As normal editing and discussions on the talk pages were unable to resolve the question, an RFC was held on the article's talk page, which produced a large number of thoughtful and policy-based comments from a large number of editors. However, at the end, the two sides were equally balanced. At the end of this discussion:

• Both sides had equally strong arguments based on policies, guidelines, common sense, and Wikipedia's values.
• Both sides also had equal numbers of supporters.

After considering both the equal strength of the arguments and the equal numbers, the closing admin felt there was no option except to declare a true case of "no consensus" in the closing summary.

Obviously, if editors had formed a consensus to include this fact, then they would include it, and if they had formed a consensus to exclude it, then they would exclude it. But since they were unable to form a consensus, they are uncertain what to do. These editors are now asking you: When there is no consensus to include information, shouldn't it be removed, because policy A says to remove it? But when there is no consensus to remove information, shouldn't it be retained, because policy B says to retain it?

The Fine Print:

• The dispute centers on Wikipedia:Editorial judgement. Some editors believe the article is better off including it, and other editors believe the article is better off excluding it.
• There are no policies dictating The Right Answer™ for whether this particular fact needs to be in this article.
• The fact in question is already cited to a reliable source, so WP:CHALLENGE (which is for uncited material only) does not apply.
• The objections cannot be solved through copyediting. No amount of re-wording is necessary, wanted, or helpful. Everyone agrees that it's already a very nicely worded sentence. The problem is not the wording. The problem is that half of them have good reasons for removing that fact from this article, and the other half of them have good reasons for keeping that fact in the article. If the fact is to be kept in the article, then everyone agrees that this sentence would be a very good way to present the fact.
• The sentence in question was added in the first version of the article, and it has been disputed ever since. All editors agree that there is no "stable version" or "status quo" version. There are only disputed versions available, in which some editors remove it and the other editors restore it.
  • Also, WP:STATUSQUO (go read it!) only applies during a discussion or other dispute resolution process; it doesn't say what to do when dispute resolution processes have ended.
• You don't need to personally evaluate whether the reasons editors gave are any good (in your opinion) or if the editors made policy-based arguments. The admin already did that part and determined that the strength of the arguments are exactly equal.
• The article does not qualify for deletion (the subject is notable, the text is not a copyvio, it is not an attack page, etc.), so you can't avoid the problem by trying to get rid of the whole article.
• "Have another discussion!" will not result in a different outcome. Editors want your advice for what to do now, not in some magical future when their views have changed.

The problem

Usually, when article content is disputed, it's easy for editors to reach an agreement ("consensus") about what to do. In rare instances, even after extensive discussion, editors are unable to make a decision. Several policies and guidelines provide advice about the default actions in such cases. For example, contested ==External links== are removed, and an Wikipedia:Articles for deletion discussion ending in no consensus results in the article being kept.

We have attempted to create a central list of what to do in cases of no consensus. However, in the course of doing this, we have inadvertently created a conflict between two policies. One policy says content without consensus is usually kept, and the other says content without consensus should be removed. According to our policy on conflicting written rules, we need to figure out what the best practice is and make all of the relevant advice pages match the right thing to do. If we are able to reach a consensus about what the best practice is, then we will later be able to make specific proposals about which page(s) to change and what those changes should be.

(NB: We don't need you to tell us what the policies currently say, because we already know what they say, and – since they conflict with each other – at least one of them is currently wrong. Just tell us what you believe the best practice is.)

How to respond

Please add your viewpoint in the #Responses section.

• In this discussion, it's more important to think about common sense, practicalities, and Wikipedia's values than about which line of policy or guideline you'd quote in such an RFC. The goal for this discussion is to figure out what editors believe the best practice is, in principle, when there is no consensus about whether to include or exclude non-BLP-related material in a particular article.
• The policy What Wikipedia is not says "the written rules themselves do not set accepted practice. Rather, they document already-existing community consensus regarding what should be accepted and what should be rejected...Furthermore, policies and guidelines themselves may be changed to reflect evolving consensus." This RFC is trying to figure out what the current community consensus is. Once we know what the current community consensus is, we will be prepared to accurately document the community's best practice in the written rules.
• If you decide to cite a policy, guideline, essay, or other page that matches your view, please read it first, to make sure your memory of the page matches its current contents. (Previous discussions have seen editors linking pages that don't actually say what they thought it said.)
• The key and immutable point is that there is no consensus. Responses that reject this premise (e.g., "I need to know what the arguments were, because one side always has better arguments, so the admin was wrong to say the reasons each side gave were exactly equal in strength" or "The side with the stronger argument should get what they want") will not help us answer the real question.

Discussion

Do you have questions? Do you have ideas? Do you want to share some thoughts with other editors while you contemplate possible #Responses? Post them here. RFCs operate under the normal rules for discussions. If you have meta-comments about the RFC (e.g., Why is it happening on this page? Can I notify page X? Why didn't you give me specific examples?), then please post those on the talk page.

---

I'm having trouble making up my mind. I think that it's best for Wikipedia to include more information. If people are looking for encyclopedic information (broadly defined), I'd like them to be able to find it here. However, I worry about certain kinds of claims, such as health claims ("One study showed that Wonderpam™ kills cancer cells in a petri dish") or geopolitical disputes ("Blue Country says that Orange Country smells bad"). Usually, I think it would be okay. If the material is obviously bad, we'll form a consensus against it. However, every now and again, we make really odd decisions. Of course, those decisions tend not to last very long, so perhaps it doesn't matter that much if disputed content is included until a consensus is formed against it.

OTOH, the approach we take to WP:BLPs is to remove contentious material until there is a consensus to include it, and why shouldn't we take a similar level of care for other subjects?

Colin's comment about not preserving the old is a good one. WP:QUO is a fine approach during a dispute, to the extent that it discourages edit warring, but it's bad to extend it beyond that point. I've been catching up on my watchlist after years of neglect, and it has been discouraging to find articles editing 25 times over 25 months, without even a single sentence being added or removed. WhatamIdoing (talk) 19:48, 9 January 2024 (UTC)

It's been my general impression that this is what a maintenance phase looks like. We get oodles of smol gnomings as editors flit around fixing whatever is in their maintenance tasklist, and every once in awhile an article will undergo a complete rewrite over the course of weeks, without much action taking place in the space between.

My sample may be biased, and my memory might be faulty, but on my watchlist I very rarely see an edit that adds one or a few sentences of sourced information to an established article. Creation and repair seem to be the major spheres of mainspace editorial activity, with expansion and updating barely visible in the distance. Folly Mox (talk) 02:01, 10 January 2024 (UTC)

It's been like that for years, and has been getting worse. Many if not most of our current editors hardly ever add text. Some can't be bothered to find references, some are unconfident of their abilities to write sentences in English, and so on. Plus most are now on mobiles, where I certainly wouldn't want to do long edits. We should stop encouraging people to write new articles, & instead promote improving existing ones. Not sure what this has to do with your issue here, though. Johnbod (talk) 03:49, 10 January 2024 (UTC)

A lot of articles aren't changing anyway, so why would we ever want a bias towards the old version? The old version is happening too often anyway. If someone is attempting to (in their opinion) improve an article, maybe we should encourage that. Instead of excluding by default or including by default, maybe we should try the newest approach by default, or the least experienced editor by default. Really: If I can't, despite all of my wiki-friends, despite knowing the core policies so well, despite having written a non-trivial fraction of their contents, despite (as all my wiki-friends will agree ;-)) almost always being right, manage to scare up a consensus for my view, maybe the newbie deserves to "win". WhatamIdoing (talk) 05:25, 10 January 2024 (UTC)

Honestly, I kind of like that idea: bias towards change in order to counter natural ossification. Feels unimplementable in practice though. Assuming we're still in the hypothetical here where there's a true no consensus on strength of argument or numbers, do we count up the total active editing days of both sides and award the lower total the victory? Does the newest editor present in the discussion grant an automatic win like a golden snitch? Folly Mox (talk) 13:46, 10 January 2024 (UTC)

I was thinking just then about a dispute as being a 1:1 thing, with the other voices not exactly being participants in it, but you're right that some big disputes have a group-vs-group dynamic. The newest editor rule would tempt people to engage in socking, but there are probably ways around that (e.g., the newest editor as of the day before the dispute started, the most recently extended-confirmed editor, the side with the most voters who have between one and five years' experience, etc).

We could also have a rotating golden snitch tiebreaker: A dispute can apply for a tiebreaker decision, and a randomly selected admin is stuck with making the decision. WhatamIdoing (talk) 19:29, 10 January 2024 (UTC)

If the old text has been there a long time, there is a MOS:RETAIN-style argument that if it was all that wrong someone would have changed it. Rfc's on most pages only attract a tiny fraction of our editors, often those most invested with a particular POV. The long-term wisdom of the crowds will often produce a better answer. Johnbod (talk) 17:25, 10 January 2024 (UTC)

There's equally a Wikipedia:List of hoaxes on Wikipedia#Extant for 10+ years argument that we can't rely on people to have already fixed it. WhatamIdoing (talk) 19:32, 10 January 2024 (UTC)

I'm reading this and thinking, when did I propose that we should completely reverse one of the defaults from retaining the status quo to be to adopt the new. I then read my text below and see that when I summarised our Editing policy it looked like I was proposing that. I wasn't really. I was saying that policy says we should try to work with and improve the new, but it also says that if you can't then you can excise it. It doesn't say what we should do if we disagree on whether the text can be fixed enough to be retained. But it does say that we do encourage change and for editors to try to incorporate change. What I was really saying is I think we should drop the "default to the old", because that doesn't encourage change, and I wasn't proposing replacing that with "default to the new". I can see problems with that default too. It is swinging too much the other way. I thought we should be more neutral about whether changes were better than what we had, on a project that does indeed encourage change.

MOS:RETAIN is absolutely nothing to do with "if it was all that wrong, someone would have changed it". It is about preventing regional English variety disputes. Because we always get random folk coming along and noticing that an article has spelled something "wrong" and "fixing" it. It is entirely a dispute-prevention and dispute-settling policy. Per WP:UPPERCASE, the words MOS:RETAIN sound like we have a MOS guideline that values keeping longstanding text. We don't.

I disagree with the idea that generally speaking text on Wikipedia demonstrates its goodness through age. Also the logic that says we should default to keeping the old because if the old was wrong, someone would have changed it, is, em, fundamentally flawed. You've got a rule that makes it hard to change, and thus a tendency to keep the crap. Most of our readers don't have the knowledge or inclination to fix the faults on Wikipedia. I can tell that because when I look at some random article, I can see absolute nonsense text added by some student class three years ago, on the assumption that "the wisdom of the crowd" would fix all their mistakes. Or I can read a barely literate lead that hasn't changed much for years despite being so painful to read you want to have a lie down after. Wikipedia is very much a work in progress and we absolutely shouldn't have a policy that discourages progress.

On the other axis, of inclusion vs exclusion, there are just too many areas I think where there are clear arguments in favour of defaulting to exclude (certainly BLP but also biomedical and science). There may be some areas where editors think it really isn't an important enough topic to care, but I think mostly when editors get passionate about Wikipedia not misleading or misinforming our readers, then they tend towards exclusion. Exclusion is also a good default because long term, if something does become important or accepted, then we can always include it later. Whereas, as we have seen, once material is in Wikipedia, it can be hard to get it out, with people giving entirely non-policy based reasons like that it has been there a long time so must be ok. -- Colin°^Talk 10:51, 11 January 2024 (UTC)

You can certainly find a lot of "absolute nonsense text added by some student class three years ago", though it rarely sustains being completely wrong over any length of text, but then no-one is likely to mind in the least if you remove it. That is not what we are talking about here. Johnbod (talk) 18:47, 15 January 2024 (UTC)

Johnbod, I know we are only discussing contested text here. But the claim was that somehow our existing text is obviously better (and thus should be defaulted to) than any text (or absence) that some new contributor had offered simply because it has been here a long time. I think there are no end of examples of why that is not only wrong but also wrongheaded on a project that is based on "What we have here is not good enough, please help us improve it" model. If editors were more likely to make the text worse than make it better, we'd just lock all the pages and sit back and enjoy reading what people wrote in the past. -- Colin°^Talk 08:55, 16 January 2024 (UTC)

I don't see anybody claiming anything like that. Johnbod (talk) 17:18, 16 January 2024 (UTC)

Maybe not here, but we do see that belief in practice. Some editors revert copyediting because the new version is "only" equally good as the old version. Many editors incorrectly believe that WP:STATUSQUO is a policy requirement and that it requires them to prefer the old version over the new version if there's a dispute. WhatamIdoing (talk) 17:42, 16 January 2024 (UTC)

Ok, but "not an improvement" is a valid reason to revert. Where there is an actual change some editors are just mistrustful of new/different material, especially if not referenced (or worse, not changing the old ref immediately after). Johnbod (talk) 18:52, 16 January 2024 (UTC)

Is it really a valid reason to revert? If that's being used as a polite euphemism for "makes the article worse", then it's a valid reason, but if you believe that the change literally makes no difference to the overall article quality, I don't think you should be reverting it. Why would you waste your time making no difference to the article's quality? WhatamIdoing (talk) 19:45, 16 January 2024 (UTC)

Johnbod, you wrote "If the old text has been there a long time, there is a MOS:RETAIN-style argument that if it was all that wrong someone would have changed it." Surely that is explicitly favouring old text merely for being old, on the assumption that many eyes have reviewed it and found it acceptable. But it assumes that otherwise, if someone found it unacceptable, that they would overcome the inertia required to actually edit it to something better (for example, they may need to go look up sources, which they may find they don't have access to, learn how to cite them, and so on), or remove it, and then manage to get their edit retained rather than reverted by someone assuming the old text was fine because it is old.

Many editors may be considering situations where they supported existing text because they wrote it or are imagining situations where they have come into conflict with someone removing or changing text they wrote. Maybe they aren't so strongly remembering the text they improved and didn't get any conflict over because whoever wrote that is not here or because it was obviously better. I wonder if there is a bias in our recollection that assumes other people are making bad edits, rather than realising that to everyone else, we are the other people.

Don't you see a conflict between editors favouring existing text and our fundamentals which are that we want editors to change what's here because we hope you can make it better. Anyone coming to Wikipedia, in good faith, to make an edit, is doing so because they believe they are making it better. So surely that should be the default? I mean for articles that aren't protected in some way and editors who are established, our edits are automatically "live" and "retained". It isn't like we need to prove they are better to someone before they show up, or that they are removed after a few days unless anyone confirms it. And reverting is officially discouraged by editing policy in favour of trying to accommodate the change. -- Colin°^Talk 08:28, 17 January 2024 (UTC)

All these are factors, certainly, and there is such a conflict, but we know that quite a large proportion of (the minority of) edits that change texts aren't actually improvements. I don't see any workable "default" approach, either way. Each edit has to be looked at. We are still off-topic here. Johnbod (talk) 19:43, 17 January 2024 (UTC)

Size of English Wikipedia (1000 vol)

Speaking of subject-by-subject, this image splits Wikipedia's subject areas into these categories:

• Biology/health/medicine,
• Business,
• Science,
• Geography,
• History,
• Biography (living + dead),
• Society, and
• Culture/arts

I wonder whether editors would choose an include/exclude default differently for the different categories. Presumably for overlapping categories (e.g., BLP + pop culture), the more restrictive rule would apply. WhatamIdoing (talk) 16:46, 11 January 2024 (UTC)

While this specific RfC question has merit, I think a related question that actually comes up more frequently might be a more important thing to discuss: what to do when there is consensus for change, but no consensus on which. This happened at RFA Reform 2021, Article Creation at Scale, maybe Emoji Redirects?, and probably other mega-RFCs that I've forgotten. People agree the current approach isn't working, but every specific proposal for change fails, and we're left with the broken status quo. It would be nice if we had some sort of escape valve from those, like a mandate to implement the change that saw the least opposition, taking into account somehow when each proposal was made during the lifespan of the overarching RfC. Folly Mox (talk) 13:54, 10 January 2024 (UTC)

Oh, I thought we were talking about article content. Behind the scenes proposals are different, but ideas can change - compare the current Wikipedia:Village_pump_(proposals)#Increase_default_thumbnail_size_from_220px_to_250px with the last attempt, a mere decade ago. Johnbod (talk) 17:25, 10 January 2024 (UTC)

Responses

It's impossible to answer this because important information is missing. I need to know...

This is the place to say "I can't answer this, because you didn't say __________" (e.g., whether there were more admins on one side, whether the material in this new article is considered 'long-standing', what the article's subject area is, whether the material could realistically be moved to a different article, or whatever else is on your mind).

Feel free to ping WhatamIdoing (who wrote the hypothetical story above) in your comment, if you think she could supply the information you need.

• Your comment here...

We have enough information. In principle, when there is truly no consensus, I believe we should err on the side of...

Most responses to this will probably be either exclusion or inclusion, though you should not feel constrained by those options. Please consider explaining your view with:

• an example from articles you edit (e.g., "In articles about geography, editors might reasonably disagree about whether to mention _____, and in such cases I believe..."),
• a common-sense or practical explanation (e.g., "In my experience with Wikipedia:Contentious topics, what helps settle difficult disputes is...") , or
• a values-based explanation about what you think is best for Wikipedia's editing community and/or Wikipedia's readers ("In line with my favorite of our shared Wikipedia:Principles, namely...").

Remember, we're looking for editors' own opinions about what, in principle, is the normally right thing to do, when editors really can't form a consensus.

• I think most such cases are caused by a combination of a) not many people responding to the discussion/Rfc and b) the issue reflecting some contentious issue, especially where nationalism(s) are involved (usually the worst cases of pov editing in my experience). If there is a disagreement over facts, and lots of RS of equally appropriate weight (but perhaps from different countries) supporting each position, then both sides should probably be put in the article. If there is no disagreement over facts, but simply an issue of WP:UNDUE, different placing in the article, and examination of how RS treat the matter, will usually sort the matter. If a) is a big factor, a rerun with more publicity is probably best - vast numbers of Cfd discussions are these days relisted, often several times, because there are far fewer participants than there used to be. Johnbod (talk) 17:41, 8 January 2024 (UTC)
• To answer the question as posed (true no consensus, equal strength of arguments on both sides, etc), I personally believe we should err on the side of inclusion. We're an educational resource and more information ceteris paribus improves the encyclopaedia. I guess if I have to bluelink some WP:SHIBBOLETHs I'd pick WP:5P1 or WP:5P5? It's a personal value, as requested. In many actual cases it might be a good idea to have a second person take a look by opening a close review discussion at wherever we do those these days (still WP:AN I think?). This seems likely to occur anyway assuming the criteria in the problem statement are all met. Folly Mox (talk) 18:57, 8 January 2024 (UTC)
• I don't think the scenario here is common and is more of an intellectual exercise to make people not give unhelpful answers like "they should just keep arguing till they get consensus". If editors have been following the rules and have a decent and representative opinion, then their arguments have been about the encyclopaedic content on the page If then they truly cannot agree on whether it should stay or go or which version it should be, then there is no correct answer that is better than another, just do something arbitrary so we can all move on and do something else. What matters far more is the informal dispute between two editors or just a few editors. Are we to insist all such disputes must keep arguing and growing till they achieve the story on this page that after four RFCs and 50 editors of outstanding ability they still are absolutely at a stalemate? No, we shouldn't. We should figure out what to do when these two editors cannot agree and don't really want to argue more but want a solution that works and they can move on. Currently some of the time people say we preserve the status quo that existed before dispute (should such exist). Or some of the time people say explicit consensus is required to include text, otherwise we default to removal.

I think the former, the default to what we had before you came along, is bad for the project. It is based on an "old hands" mindset that has crept in that thinks all us old hands have created something wonderful and all these new editors are just making it worse. I think that is fundamentally against the principles upon which Wikipedia was founded and preserved at WP:5P3: any contributions can and may be mercilessly edited and our default should be to try to incorporate and improve those edits. Our Wikipedia:Editing policy says "Even the best articles should not be considered complete, as each new editor can offer new insights on how to enhance and improve the content in it at any time."

Wrt the latter, that content has to justify itself and get consensus to be added or kept if challenged, I think on balance that is better than the opposite, that we make it really really hard to remove or keep out the shit. Not all additions are useful, educational or encyclopaedic information. Some of it is bad data. Some of it misleading. Some of it just doesn't belong on this project. Nearly of our policies are about keeping the bad out. There's a reason for that. Our heart wants to keep adding but our heads know we have to keep a check on it. So, when there's a dispute, and editors can't agree, we should go with keeping it out. -- Colin°^Talk 22:17, 8 January 2024 (UTC)

• (Saw this on Iri's talkpage, hope no one minds me chipping in). Mostly as per Colin. Everything that someone reads on Wikipedia should be proportionate, useful and relevant, as well as being simply verifiable. This is what makes us the type of encyclopedia we have agreed to be, and not all of the things that we are WP:NOT. In this scenario the content is immediately challenged for what appear to be legitimate reasons, which means that at least some editors think it unsuitable and those who disagree are unable to get enough support to demonstrate otherwise. In my opinion, if editors can't agree if the content is good or not, I believe the harm of presenting a fact that may not be encyclopedic is greater than the potential benefit. So, exclude in such cases until a consensus is found to include. Scribolt (talk) 08:21, 9 January 2024 (UTC)
• I personally believe that it is more harmful to remove otherwise relevant and well-cited, policy-abiding material for essentially, DONTLIKEIT. Excepting BLPs where there are risks of legal or privacy issues, and that sort of thing where I think erring on the side of exclusion may make more sense, given PUBLICFIGURES or other encyclopedic events and things, and if sufficient WEIGHT in sources, I think ultimately, in the case that no side clearly has the strength of arguments or a reason to exclude that has consensus, well-sourced material has a basis for generally being included per QUO, PRESERVE, and Wikipedia goal as a cohesive corpus of all encyclopedic information. Andre🚐 19:45, 10 January 2024 (UTC)

  I don't think it is fair to dismiss editors in good faith disagreeing that proposed (or current) material harms the article by its inclusion as an DONTLIKEIT argument. For a start, see WP:UPPERCASE. One big example left to editor consensus is NPOV, where we expect editors to get themselves familiar with the body of literature on a topic and form a consensus about what to include or exclude, and what to emphasise or minimise. It is exactly the sort of area where editors might disagree. Or they might disagree about where to include the material and some argue "not this article". -- Colin°^Talk 10:16, 11 January 2024 (UTC)

  Incel was a case of "not this article" years ago. An AFD concluded that it should be merged to Celibacy, and the editors there refused to have any of the content. (The difference between incel and 'nice guy who can't get a date' was not as obvious back then.) WhatamIdoing (talk) 17:59, 11 January 2024 (UTC)

  well, it's rather hypothetical, isn't it? but I'm thinking more about situations where the material is educational, and not necessarily harmful to the article, but harmful to perhaps the subject's reputation or their political campaigns. Andre🚐 18:18, 11 January 2024 (UTC)

  I don't think it is hypothetical at all, speaking from experience. Yes there are activist reasons for including or excluding material but also plain old editor disagreement. Several medical editors had a bust up a few years ago over whether to include prices in drug articles. Much of the argument against including the prices was that they were in fact wrong and highly misleading (what kind of "price" was it and what was that amount of drug used for was very unclear) so it isn't like editors who wanted them out were just worried about clutter, say.

  Material that smears a subject's reputation isn't necessarily "educational". You'll only regard it as educational if you think it is actually a fair and true account and you are unlikely to want to lead with it or have paragraphs of body on it unless you think this is clearly the most urgent and important thing readers have to know about the subject. If you don't think that, then actually your view is that including it is uneducational... that we would then be misleading our readers and leaving them thinking completely incorrect things about the subject. -- Colin°^Talk 13:38, 12 January 2024 (UTC)

  I'm really baffled by your implication that everything educational about a public figure that some people might want to remove would be a smear. Andre🚐 15:55, 12 January 2024 (UTC)

  "Public figures" are generally BLPs, which are already covered by the BLP exclusion rules.

  Also, I believe that this was a response to your statement above about situations where the material is educational, and not necessarily harmful to the article, but harmful to perhaps the subject's reputation.

  I'm sure you've encountered disputes involving businesses or countries, in which the subject would prefer a specific unpleasant fact not to be in the article (or didn't want it to be in the lead). Perhaps "Texas, a state that prides itself on a healthy business climate" would be preferred to "Texas, the state with the highest rate of fatal car wrecks". WhatamIdoing (talk) 16:47, 12 January 2024 (UTC)

  WP:PUBLICFIGURE: If an allegation or incident is noteworthy, relevant, and well documented, it belongs in the article—even if it is negative and the subject dislikes all mention of it. If you cannot find multiple reliable third-party sources documenting the allegation or incident, leave it out. Andre🚐 17:28, 12 January 2024 (UTC)

  Yes, but:

  • PUBLICFIGURE is part of BLP, which doesn't apply to the scenario at the top of this page, and
  • not every entity with a reputation that could be smeared is a BLP, or even a public figure.

  I think the question for you is: What example of article content can you give us that is both educational (your word) and also harmful to perhaps the subject's reputation or their political campaigns (your words) but that somehow could not be considered Material that smears a subject's reputation (Colin's words) even by people who want the material that is harmful to...the subject's reputation (your words again) out of the article? Can you think of something that "harms" a subject's reputation but doesn't "smear" that same reputation? WhatamIdoing (talk) 18:27, 12 January 2024 (UTC)

  Yes, all sorts of things. How about the The Daily Caller. There's a lengthy discussion in the history of that talk page (particularly Archive 7 I think, but good examples even in the current talk) wherein several good faith attempts were made to remove well-sourced information that I objected to along almost exactly these lines. To be clear, it is not a smear or a hit piece to talk about the Daily Caller's ties to white supremacists and climate denial, but some people in general want to remove this, effectively whitewashing their reputation. The charitable characterization of the opposing view is that the article as it stands is a COATRACK due to its extensive detailing of the Daily Caller's right wing controversial stances. Andre🚐 18:35, 12 January 2024 (UTC)

  Do you think that saying they have ties to white supremacists "smears" their reputation? WhatamIdoing (talk) 19:38, 12 January 2024 (UTC)

  I think you two might be working from different definitions. There's an understanding of smear that is restricted to false claims, and another that includes making statements that are true but unwanted. Back in the 1970s, you could "smear" Ronald Reagan's reputation by reminding people that he was divorced, even though he really did get divorced. These days, you can "smear" a fitness influencer by posting an absolutely true photo of them eating fast food, or by pointing out that this so-called healthy person has been diagnosed with a severe case of relative energy deficiency in sport. It all depends on whether you believe that truthful mud doesn't smear. WhatamIdoing (talk) 19:49, 12 January 2024 (UTC)

  Yeah, in my definition, smear means, "damage the reputation of (someone) by false accusations; slander." Andre🚐 20:31, 12 January 2024 (UTC)

  Two nations divided by a common language, and all that. WhatamIdoing (talk) 22:54, 12 January 2024 (UTC)

  I think we are getting a bit bogged down in specifics. I was reacting to your assumption that there aren't in fact encyclopaedic reasons to exclude something, only IDONTLIKEIT reasons. There are loads of encyclopaedic reasons to exclude stuff, otherwise nobody would find themselves in a no-consensus stalemate over it.

  But, fwiw, one can certainly smear organisations and companies and even political persuasions or ideologies. And a smear can also include an element of truth taken to an inappropriate extreme. Aspects of the statement aren't literally untrue, if you squint a bit, but the implication they make certainly is. It's a common tactic among politically motivated journalists and thus something we can't easily defeat with WP:V. As WAID notes, most of the targets are people and BLP offers us some protection. But plenty other targets too. -- Colin°^Talk 20:46, 12 January 2024 (UTC)

  I never said that, though. I never said that there aren't encyclopedic reasons to exclude something at all, I said quite the opposite. I said there were often in cases where there is no consensus, not encyclopedic reasons furnished to exclude information, but that it was excluded for essentially, aesthetic reasons. In other words, if it were so clearly encyclopedic or not, it should gain consensus for one direction or another. I also think it's a bit odd that you want to remove specifics having said moments ago, "I don't think it is hypothetical at all, speaking from experience." I also have many experiences here and in mine, it's more common that people want to remove information that is encyclopedic, then that people want to add information that isn't encyclopedic, but that there isn't a clear consensus on it.

  One can certainly smear; has someone claimed that one cannot? But not every inclusion of encyclopedic, negative information is a smear. I find it problematic that you think a bigger problem would be the potential damage to the Daily Caller's reputation from true information, versus the problem of pro-Daily Caller sympathetic actors trying to launder the reputation so that people will trust the Daily Caller more in the future. Andre🚐 20:51, 12 January 2024 (UTC)

  Back on topic: The other day, I cleaned up an article about a small business. One of their employees was killed on the job. Should I have included it? Maybe, and maybe not.

  The decision isn't "aesthetic", but it is one that reasonable people could have differed on. On the one hand, it wasn't the business's fault: an apparently legitimate customer placed an order for the purpose of murdering the delivery driver (apparently because dead bodies are easier to rob than living people). On the other hand, for a small business, this was probably one of the biggest crises in their history. It had no notable effect on the rest of the world (e.g., no laws were written, no industry practices were changed), but it probably had a significant effect on the individuals. If someone wrote a book about this business, I'm sure it would be included. But for an encyclopedic summary, rather than a comprehensive explication of everything, I'm not so sure. WhatamIdoing (talk) 23:03, 12 January 2024 (UTC)

  The biggest difference in that example is that we're more predisposed to feel sympathy for the small business due to the circumstances. Why should that incident, no fault of theirs, figure prominently in their encyclopedic summary? But what about the many instances where the business was at fault? The question is whether we inform the reader, and not necessarily whose fault it was. Not knowing more specific information about the consensus, I would hazard a guess that a small group of editors could come to a consensus to exclude it, because it's not very relevant to the topic. But, in a perfect no consensus situation, it would not be easy to determine that. In the case where it is relevant, perhaps because the business actually was at fault for the circumstances, and a group of editors fights to a no consensus result, and that would be excluded? We're now depriving our readers "controversial" knowledge that would potentially influence their view of whether to patronize the business. Potentially giving business to a fraudulent or otherwise unethical business that gamed the system to scrub their bad reviews and negative information. It's naive not to think this happens. Andre🚐 23:32, 12 January 2024 (UTC)

  I don't think that I find consequentialism at such a remote and indirect degree a very compelling argument. We should include or exclude content based on whether we think it's appropriate for an encyclopedia article, rather than whether we think it could make an "undeserving" business earn money. With external links, we say thatWikipedia:We don't care what happens to your website, whether it gets slashdotted or ignored, and I think we should equally not care whether the subject of an otherwise fair, verifiable, and encyclopedic article gets some benefit from that article. We're here to build an encyclopedia for the reader, not to prevent readers from learning about a business that they might care about. WhatamIdoing (talk) 05:06, 13 January 2024 (UTC)

  To clarify, I wasn't advocating an ethical test for determining encyclopedic relevance. My point was that you seemed to be advocating a relevance test based around something like the thought experiment of whether a detail would be in a book or an article about something, and a thought experiment on the business' culpability and the impact to the business of its article, or at least that was my read of your frame. And I'd rather frame this from the perspective of the customer or the reader. The reader being a potential customer of the business or just an interested (or uninterested party). My encyclopedic test is more like, are we informing the reader, and not, would this business like that we're telling readers about this aspect of it. Andre🚐 08:27, 13 January 2024 (UTC)

• in the 0.001% occurrence of no consensus: if the material is already in leave it, if it not in exclude it. ^Thanks,L3X1 ◊distænt write◊ 16:18, 11 January 2024 (UTC)

  @L3X1, at what point should "already in" be determined? For example if it was in the first version of the article, removed an hour later, and steadily edit-warred over ever since, is that "already in" or "already out"? WhatamIdoing (talk) 17:57, 11 January 2024 (UTC)

  That situation would be already in ^Thanks,L3X1 ◊distænt write◊ 18:28, 11 January 2024 (UTC)

  So, basically, if the material has ever been in the article (i.e., not just a proposal on the talk page), then it should be included? WhatamIdoing (talk) 18:31, 11 January 2024 (UTC)

  if there is no agreed consensus to remove it, yes. ^Thanks,L3X1 ◊distænt write◊ 01:39, 12 January 2024 (UTC)

• My preference is for the question to remain unresolved, to encourage continued discussion and perhaps compromise (too often, we phrase RFCs as “X or Not X” and neglect to consider “would Y resolve this”). However, if we MUST choose, then I lean towards “omit”. Blueboar (talk) 02:27, 12 January 2024 (UTC)
• Include. If it’s misinformation or misleading there should have been other policies to justify its exclusion. If it’s a fact it shouldn’t be excluded just because WP:IDONTLIKEIT (or I don’t know much / I haven’t heard of / I’m not passionate about it).
  There are people with COI who want to include certain information in Wikipedia, but there can also be people with COI who want to have certain information excluded. We need to be aware of both of them. Include only certain information but not the others can be misleading as well.
  If there are really no other policies that can justify exclusion, let our readers choose and decide what to believe by themselves.[1] [2] --Dustfreeworld (talk) 05:10, 13 January 2024 (UTC)
• Reject this RFC. Wikipedia has muddled through for 23 years now without any need for a specific answer to the question posed by this RFC. There is no one-size-fits-all solution to whether to include or exclude contentious material, and it must always be decided through discussion, consensus building, seeking third opinions, and all the usual dispute resolution mechanisms. For us to impose an overarching guideline on top of all of that would IMHO be damaging to the project and a classic example of WP:CREEP. There is no need for anything to change. Cheers  — Amakuru (talk) 18:18, 15 January 2024 (UTC)

  What are editors supposed to do in between failing to find consensus and that magical future when they do? The faction most skilled at edit-warring wins? WhatamIdoing (talk) 18:35, 15 January 2024 (UTC)

We have a bunch of policies for that. Escalating to involve more people via an RFC or noticeboards, for instance. IMHO the absolute worst situation caused by these attempts at "default outcome" policies is when people fail to escalate in that manner. More generally, though, while having long-running disputes that continuously fail to reach consensus is bad, I would argue that a "false consensus" created by policies that dictate a particular outcome and leave people who prefer that outcome with no reason to engage is even worse. Some problems are genuinely difficult and complex and unsettled even in the real world, and therefore are things we have to discuss a lot and revisit regularly. It's bad when policies waste editor time and effort, sure, but if something reaches the point of massive RFCs and still plainly has no consensus, that probably means it's complex and difficult and is deserving of a lot of effort to get it right. A "just say screw it and go with X" policy isn't helpful in those situations. Ultimately, when there's no consensus, what is probably needed is a compromise, rather than one side or the other trying to clobber the other with "we win automatically" policies. --Aquillion (talk) 21:31, 15 January 2024 (UTC)

Yes, we have lots of methods for dispute resolution, but what are editors supposed to do in the meantime, when none of those efforts produce a consensus? Right now, we are telling editors two contradictory defaults. I would like to have either one default or none (none being what @Amakuru seems to lean towards), but we need to stop having two contradictory ones.

Some intractable problems really aren't amenable to consensus. You can't name an article both Gdansk and Danzig, to give one of the oldest examples of this problem. You can't put a particular image both halfway in and halfway out of an article. You can't name someone's spouse and also not mention the spouse's name. Some of these are complex and difficult disputes, but many of them are really quite minor, too. WhatamIdoing (talk) 00:40, 16 January 2024 (UTC)

• I don't think we have a default outcome in these situations. For one thing, there's a lot of other factors to consider and it's rare for something to be a complete void in that regard (eg. the level of consensus the material had formerly; how dangerous it is per policies like WP:BLP, etc.) We have policies for dealing with and resolving longstanding disputes, and few of them really push editors towards "just default to XYZ" as a resolution mechanic. Editors can escalate to larger audiences through RFCs and noticeboards, for instance. My experience with attempts to make and enforce "default outcome" policies is that they tend to do more to derail consensus building - it creates an immediate perverse incentive for people to push discussions towards whatever situation would make their preferred outcome the default, which is very undesirable when that outcome is a lack of consensus. Even WP:BLPRESTORE, which we have to have due to the high risk of harm, has caused this problem at times, as WP:CRYBLP can attest. Consensus-building works best when everyone has an incentive to come to the table, so I think we want to avoid policies and guidelines that set hard default outcomes unless it is absolutely necessary. That said, if we must choose one default, I would go for "include the basic idea in some form, but without any weight given towards any one particular wording". That encourages people who dislike that particular version to go with Wikipedia:Writing for the opponent by coming up with alternate formulations and pushes everyone towards compromise, which is probably what should happen when there's no consensus. Total omission, by comparison, almost always means a lack of any sort of compromise, which isn't the mindset we want people to be in when there's no consensus at the moment - when there's no consensus, the first principle should usually be "chances are neither of you is going to get 100% of what you want." --Aquillion (talk) 21:31, 15 January 2024 (UTC)
• In general, assuming all the conditions described in the hypothetical example apply, and no compromise is possible (I know you already mentioned that no amount of rewriting would resolve the issue—but depending on context, other compromises may yet exist), I would lean towards "whatever the stable version prior to the discussion was". (WP:STATUSQUO specifically indeed refers only to keeping this stable version for the duration of the discussion, but if the No Consensus section of the Consensus Policy can be trusted to be reflective of actual practice, this is already the common result post-no consensus outcomes.) When both sides hold equally strong, valid positions that do not run counter to existing policies, other considerations than the merits of the arguments need to be weighed to come to a decision. While it's situational and exceptions do exist, typically stability and not wasting volunteer time are among the next-strongest considerations. (As also evidenced by how many parts of en.wiki have this codified into The Way Things Are Done. ENGVAR, keeping pre-existing date order, not messing around with MOS style variations or reference style, COSMETICBOT, and so on.) I do however think the Consensus Policy has it right: describe what's typically done, but don't prescribe a hard default. AddWittyNameHere 03:32, 16 January 2024 (UTC)
• Reject this RFC Editing and discussion is the way forward in all situations. So two editors disagree there is WP:3O, still no resolution then have a RFC. For what this will solve it will create a thousand times more problems. This maybe stated as For one specific edge case, but that's not how any result will be used. For all it's good intentions it will be used by those wishing to include "alternative knowledge" as a way to retain the content, and create endless timesinks to remove it. These and more was expressed in my comments before this RFC. -- LCU ActivelyDisinterested «@» °∆t° 21:11, 16 January 2024 (UTC)

  When two editors reach no consensus, there is WP:3O.

  When the RFC reaches no consensus (or perhaps multiple discussions and RFCs), there is... what else?

  You can't leave the fact halfway in and halfway out while editors prepare for RFC #6. What happens in the meantime, while one group is demanding that you remove the fact per Policy A and the other group is demanding that you include the fact per Policy B – and both groups are right about what their preferred policies say? WhatamIdoing (talk) 22:16, 16 January 2024 (UTC)

  As I said previously this is asking a question about an extreme edge case when any result won't be used in an edge case. It's all well and good asking hypothetical questions, but the actual impacts are what matter. -- LCU ActivelyDisinterested «@» °∆t° 22:52, 16 January 2024 (UTC)

  If we can figure out what happens in this edge case, we can figure out how to resolve the existing conflict in policy, which has caused "actual impacts". WhatamIdoing (talk) 01:50, 17 January 2024 (UTC)

  No we can't because as I said in the discussion coming into this your are asking the wrong question. You are asking a question about what to do at the end of consensus building, but the discussion it stems from is about the start of consensus building. The two situations are not the same. The discussion this stems from is what to do before any discussion, third opinion, or RFC has happened. So any outcome from this isn't applicable to that discussion, because you carefully crafted your question to the point it no longer applies. -- LCU ActivelyDisinterested «@» °∆t° 19:42, 25 January 2024 (UTC)

  I don't think I understand the distinction you're drawing here. One of the policies is explicitly about what ecitors should do when, "at the end of consensus building", there is no consensus. Why would me asking about what to do "at the end of consensus building" efforts not be relevant for figuring out what to do about a policy statement that only addresses actions taken at the end of the discussion(s)? WhatamIdoing (talk) 03:51, 26 January 2024 (UTC)

  The answer to the specific question you asked is that the situation before the disagreement should stand. If the content was pre-existing then it should be retained, if the content is being added it should be excluded (if after a well attended RFC the community can't even agree if the content is verifiable, I doubt it should be in the article). If, as is often the case in such situations, the fail to find consensus reflects a wider disagreement amongst sources then both sides should be included in the article. But again the answer is pointless to the discussion it stems from. -- LCU ActivelyDisinterested «@» °∆t° 19:47, 25 January 2024 (UTC)

  So you would make WP:QUO permanent (not merely applying during discussion, which is what it says), and you would delete WP:ONUS altogether, since it indicates removal. WhatamIdoing (talk) 03:52, 26 January 2024 (UTC)

  What would you recommend when there really isn't any status-quo option (e.g., an article that is only a few weeks old, and the contested material has been edit-warred over so steadily that neither side can claim to be 'winning')? WhatamIdoing (talk) 03:54, 26 January 2024 (UTC)

  No, absolutely and completely not. You can't just ignore my previous comment and come up with something I specifically didn't say. The question of QUO vs ONUS is one of how consent building starts, you RFC is specifically about what to do if at the end of that process there is still no consensus. Nothing coming from this effects the previous discussion at all. Again as I said in the discussion before this RFC and just above (and you ignored then, and have ignored again), you have crafted a question that has nothing not do with the discussion it comes from and so it will have no impact on that discussion. -- LCU ActivelyDisinterested «@» °∆t° 19:00, 26 January 2024 (UTC)

  AFAICT "If the content was pre-existing then it should be retained" is 100% incompatible with ONUS. Do you have a different understanding of what ONUS means for whether we retain pre-existing material for which "those seeking to include disputed content" have "The responsibility for achieving consensus for inclusion" and have failed to achieve that consensus? WhatamIdoing (talk) 19:04, 26 January 2024 (UTC)

  If you want to have an RFC about ONUS vs QUO, then have an RFC about ONUS vs QUO not some aside that you made up. If that issue isn't addressed then there is nothing to discuss because this is just an aside that doesn't address the issue.
  And no I don't agree with you assessment, in fact that disagreement is the basis of the discussion this comes from. The fact it isn't resolved by your RFC question just shows thr issues with your RFC question. -- LCU ActivelyDisinterested «@» °∆t° 19:10, 26 January 2024 (UTC)

  Tell me how to reconcile these two. How exactly do I retain the long-standing content after failing to achieve consensus to include it? If you don't think it's possible for me to understand, then imagine that you're trying to convince someone like @Bon courage instead of me. WhatamIdoing (talk) 19:42, 26 January 2024 (UTC)

  O this again! I remember this issue and IMO it's true WAID has identified an issue, and that the community has failed to (or doesn't want to grapple with) it. I've seen some interesting incidences recently of pointy uses of WP:ONUS where an editor removes something obviously well-sourced, claims no-consensus, and then relies on the tiresomeness of going through process of establishing consensus to get their way. Bon courage (talk) 20:11, 26 January 2024 (UTC)

  But that is my point. There are editors behaving disruptively by requiring other editors to get consensus for an edit, but this question is about what to do after consensus isn't achieved. It's the wrong question.
  Editors shouldn't have to get permission to edit, it's against BOLD and the spirit of Wikipedia. The issue isn't ONUS, it's distuptive behaviour. In the same way that QUO isn't the problem with editors WP:STONEWALLING. This won't solve the behavioural issue it will just make legitimate situations where ONUS applies a massive timesink for the community.
  There is an issue, but this is the wrong question and wrong RFC to fix it. It has been specifically designed to highlight only one small facet of a larger problem. -- LCU ActivelyDisinterested «@» °∆t° 22:20, 26 January 2024 (UTC)

  “Editors shouldn’t have to get permission to edit”… true. But… remember that removing information that some other editor has added is just as much “an edit” as adding information was. Neither edit requires permission. However, if there is a dispute over these two edits, then discussion and consensus determines which will prevail. What WAID is asking is: What if there is no consensus? Blueboar (talk) 22:52, 26 January 2024 (UTC)

  Then if there is going to be an RFC about ONUS vs QUO, let's have an RFC about ONUS Vs QUO. Not an RFC designed to highlight one part of the disagreement, that goes to great lengths to mention neither. Things should be done in the light not in shadow. -- LCU ActivelyDisinterested «@» °∆t° 14:42, 27 January 2024 (UTC)

  Not exactly. WP:QUO doesn't apply after the discussion's over (go read it...). This problem touches small pieces of multiple policies, and which ones might benefit from changes depend on what editors believe is the best practices.

  For example, if editors decided to take a hard-line exclude-everything-unless-agreed approach, then we should be reconsidering the second sentence in Wikipedia:Editing policy#Adding information to Wikipedia. We might similarly update WP:PSCI to say not only that pseudoscientific views should not be given undue weight, but also to say that they shouldn't be mentioned at all unless there is a positive consensus to do so. We might understand WP:CALC more strictly, requiring a slightly higher level of agreement to include it. We might expand upon the Jimmy Wales' line in WP:DUE to point out that when there is a dispute about whether a POV is "a minority" or "a tiny minority", the default is to exclude.

  And, of course, if the best practice is inclusion, then it might be a different set of policies and a different set of changes altogether. For example, ONUS would get a re-write to clarify that it's only a procedural point about who has to start the discussion and has no bearing on what happens if the discussion fails to produce a consensus. WhatamIdoing (talk) 19:12, 27 January 2024 (UTC)

  And as I said, and as your comment has mentioned, the disagreement is about the starting point of consensus - when editors misuse a policy to force a long-winded consensus building discussion. But this RFC question is very specifically as about what should be done if that consensus building discussion fails. -- LCU ActivelyDisinterested «@» °∆t° 14:47, 27 January 2024 (UTC)

  It sounds like you see ONUS as meaning "the person who wants to include the material has to start a discussion", and saying nothing about what happens if that discussion does not produce a result (either way). WhatamIdoing (talk) 18:39, 27 January 2024 (UTC)

• Inclusion. In this rare and very specific situation the goal of providing information should prevail. - Butwhatdoiknow (talk) 21:51, 16 January 2024 (UTC)
• We should consider what is most useful to our readers. More information is probably better. Other ways to go might be third opinion, widen the RFC to a more central board, put up with a slow revert war, and possibly block the perpetrators from the article in question. Graeme Bartlett (talk) 05:02, 19 January 2024 (UTC)

  I'm curious that a few editors have suggested more information is probably better. But in the scenario posted, an equal number of editors have apparently argued equally persuasively (or unpersuasive) that the article would be better without the information. For example, they consider it misleading, or in the wrong place or insufficient weight to be appropriate to mention. Not all "information" is better. For example, "information" that Keir Starmer "failed to prosecute Jimmy Savile", as claimed by the then prime minister Boris Johnson, lead to death threats and a persistent false belief that Starmer actually defended Savile, that polls show impacts whether people might vote Labour. While that "information" is easily dismissable because we do (fortunately) have reliable sources arguing it is a misleading half truth, there are far more examples where, for example, politically motivated press run with stories that are misleading, and more neutral sources simply don't mention the nonsense. We end up in this position where it is left to editors to work out whether on balance the information is encyclopaedic and neutral, and often if both sides are equally represented, we have a stalemate. I don't think it is safe to assume, if editors argue about inclusion, that inclusion is the wisest default for our readers. -- Colin°^Talk 09:10, 19 January 2024 (UTC)

• There is no black-white solution; it depends on specific scenarios. Further, Wikipedia has existing policies and procedures for resolving disagreements. Path2space (talk) 19:33, 20 January 2024 (UTC)

  @Path2space, we currently have one policy that says to remove the content unless there is consensus to keep it, and another that says to keep the content unless there is a consensus to remove it. Which policy do you think editors should be following? WhatamIdoing (talk) 05:28, 21 January 2024 (UTC)

I'd bother to disagree but will follow your instruction above: "Don't tell us what this or that policy says ...". Peter Gulutzan (talk) 18:25, 21 January 2024 (UTC)

WAID, you haven't expressed the conflict here as accurately as you did elsewhere. I'm not sure keep/remove are the right words to use in the policy examples.

One policy is exclude unless there is consensus to include, it doesn't matter if the disputed text is a new addition or longstanding text that is only now being disputed. But the other policy is to restore to the previously existing text unless there is consensus to change it, it doesn't matter if the change is addition, removal or modification. -- Colin°^Talk 09:44, 25 January 2024 (UTC)

Which, under the specific circumstance of "long-standing material" with no consensus, means that wikilawyers and other editors cite policy #1 (as if it were the only one that exists) when we want to remove it, and policy #2 (equally as if it were the only one that exists or matters) when we personally want to keep the material in question. Unless you know the policies as well as the top 1% (or top 0.1%), you have very little chance of discovering that you're being manipulated by the editor who claims that "the policy" requires things to go my way. There is a conflict; we need to know what editors believe the best practice is before we can figure out how to remove the conflict. WhatamIdoing (talk) 15:29, 25 January 2024 (UTC)

I don't disagree with that, just your choice of "keep/remove" to describe both policies. We have a policy ONUS that allows text to be "removed" if there isn't consensus to include it, but that same policy might also be invoked if someone had merely suggested the addition on a talk page (so there's nothing to remove), or when it had already been removed prior to discussion. That policy can be used to argue against putting it back or adding in the first place. The other policy, NOCON, could similarly be cited to "remove" recently added text, or "keep" recently removed text. Both policies can be used to "remove". And both policies can, with such ambiguous terms, be used to "keep", where "keep" might mean "keep the existing text that you removed" or "keep it how it was before you added your text".

Best to keep the language clear and unrelated to what edits are needed to bring it about. One policy is on the include/exclude axis wrt disputed text, and another is on the new/old axis wrt disputed text. What edits are required to achieve these results is a separate matter. -- Colin°^Talk 08:42, 26 January 2024 (UTC)

• There is a “third option” that lies between inclusion and exclusion… which is “temporarily retain the disputed material… BUT… tag it with an appropriate tag so readers know it under discussion.” Perhaps we need to explore this option a bit more. Blueboar (talk) 13:29, 25 January 2024 (UTC)

  I don't think that "option" is really in the spirit of WAID's question unless your "temporarily" is until the heat death of the universe. There is already WP:STATUSQUO, which is an essay that advises a slightly different approach, which is to go back to what existed before the dispute, whether that means restoring text that got removed or removing text that got added or un-rewording text. Maybe that's what you meant anyway. The point here is we tagged the dispute, we discussed it, and we still haven't found consensus. Then what?

  If you think that behavioural advice, for what to do during a debate, is worth reconsidering, then probably the first step is to discuss on the essay talk page. But that advice isn't in conflict with any of our no-consensus P&G. -- Colin°^Talk 15:18, 25 January 2024 (UTC)

  The other difficulties I foresee with adopting that as the solution are:

  • We have Wikipedia:No disclaimers in articles, and "an appropriate tag so readers know" would count as a type of disclaimer.
  • All of those tags operate under the principle that when the discussion stops, the tag can be removed, and we can't realistically expect editors to endlessly discuss the same thing over and over and over for months/years/decades/the rest of their lives.

  WhatamIdoing (talk) 15:37, 25 January 2024 (UTC)

• Another solution: Instruct closers to NOT close with a simple “No Consensus” but with either: “No Consensus - Keep pending further discussion” or “No Consensus - Omit pending further discussion”.

In other words, the closer has to make a decision … but one that acknowledges the lack of consensus and the need for further discussion… a decision that can intentionally be easily overturned if/when consensus actually forms.

It doesn’t really matter Which decision the closer makes. Some closers will usually lean towards “keep” and others will usually lean towards “omit”… but that’s Ok. It is understood that such closures are temporary in nature - telling the disputing editors what to do pending an actual consensus (eventually) forming. Blueboar (talk) 15:45, 27 January 2024 (UTC)

I like this solution, and it might be particularly helpful for the fairly common situation in which there isn't a a true consensus, but the discussion leans a little to one side. WhatamIdoing (talk) 19:15, 27 January 2024 (UTC)

• Exclusion. We have a duty to get the facts right. If there's doubt, then err on the side of caution. Without knowing what the actual dispute is, various factors could apply. The "MOS:RETAIN-like argument" someone was thinking of above is WP:EDITCON plus WP:PRESERVE. If the material's been in there a long while, there's a stronger argument to retain it with RS, than otherwise (and the principle to work with other editors to retain some kind of version of it, a compromise both sides can live with); but it's not an overwhelming strength. If the material is not properly sourced and can't be to the satisfaction of about half the concerned editors, then it should be removed per WP:V policy. If there are RS about evenly split for the factuality of the claim, then what we usually do, pretty much must do, is indicate that there's real-world dispute about it, and cite sources for both views (sometimes also middle-of-the-road stances), and give them WP:DUEWEIGHT. But if none of these factors are strongly in play for some reason, then I would default to excluding the claim until a better consensus arises for including it.  — SMcCandlish ☏ ¢ 😼  10:03, 28 January 2024 (UTC)

  @SMcCandlish, would you also default to exclusion if the question had nothing to do with accuracy and verifiability? What if the question is just whether this (good) content really belongs in this particular article? WhatamIdoing (talk) 17:54, 28 January 2024 (UTC)

  For example: we know what King Richard III had for breakfast the morning before the Battle of Bosworth. It is information that 500 page biographies do mention … but, I could see editors disagreeing on whether it merits inclusion in our bio article. It is 100% verifiable but also somewhat trivial. At a RFC, there would be no real policy reason to exclude… but I could easily see it resulting in a “No Consensus”, with half of the editors saying it is extraneous and should be cut, while the other half think it interesting enough to include. The project is not really “harmed” no matter what we do.

  Do I think we need a “default” to resolve this dispute? I would say “no”, we don’t… all we need is temporary resolution” at the article to prevent edit warring. It can always be overturned by later consensus. Blueboar (talk) 18:57, 28 January 2024 (UTC)

  Fwiw, at the moment we don't even have that detail in our long article Battle of Bosworth, where we have detail on the sleeping place and his mood, let alone his bio. I think any argument to include it in the bio would quickly conclude it could go in the battle article instead. But a concrete, quick, example is helpful. Johnbod (talk) 19:19, 28 January 2024 (UTC)

  It could also be something that people feel very strongly about, like the proper spelling of a placename. Will Wikipedia endorse the spelling that suggests "my" cultural or political connection to that place, or the one that suggests "yours"? We've had multiple fights over the placenames. WhatamIdoing (talk) 20:12, 28 January 2024 (UTC)

  Normally a clear case where both (or all) plausible versions should be included, and the differences explained. Any argument then moves to the one used for the article title. That's not an inclusion/exclusion issue. No one I think wanted to wholly exclude Danzig. Johnbod (talk) 20:27, 28 January 2024 (UTC)

  To respond to that block of commentary all at once: In general, yes, I would default to the exclusion, though WhatamIdoing's scenario was vague. That "What if .... sentence" suggests that the material in question might be better moved to another article, and that's often the case. I was doing a bunch of that yesterday. On Blueboar's royal breakfast, I would concur generally with Johnbod that it might move to the battle article instead of the main bio, but a good argument could be made per WP:NOT#INDISCRIMINATE to exclude it entirely, since it's not particularly relevant to the battle, or to any other encyclopedic interest either (Richard III would likely have eaten between 2 and 4 times per day every day, and no one cares what each meal was). Placename and similar conflicts: The solution is to put the title at the WP:COMMONNAME whether some people won't like that or not; it's a policy, is defensible, and usually produces a certain result. But in the article, give both (or all three or whatever) names back-to-back in the lead. We do this for reader benefit (MOS:BOLDSYN, "Am I at the right article? Why was I redirected here?"), but it also reduces a lot of editorial conflict. Even which comes first and should be the default in our running text is already determined by the article title in most circumstances. If there will be conflict, it should usually be about the article title, and COMMONNAME proof is mostly a matter of checking usage in modern English-language sources and thus resolvable one way or another in most cases.  — SMcCandlish ☏ ¢ 😼  22:15, 28 January 2024 (UTC)