Wikipedia talk:Verifiability

  1. Home
  2. Wiki

    Expanding efns in Wikipedia:Verifiability § Responsibility for providing citations

    [edit]

    Hi, I noticed there are many footnotes in the Wikipedia:Verifiability § Responsibility for providing citations section. I'd like to propose removing them and moving the text they contain inside the same section because:

    • They are useful and are relevant to the section, especially efns c, e, and d. B could be explained better, but maybe parts of the footnote should remain hidden.
    • The section is pretty small, so they fit easily.

    The text would remain the same, so this this isn't a content issue. FaviFake (talk) 10:50, 22 August 2025 (UTC)[reply]

    Concur. The information in the endnotes should be in the body. Placing that much substantive content in endnotes was a terrible idea to begin with. --Coolcaesar (talk) 16:07, 22 August 2025 (UTC)[reply]
    Luckily it's not as bad as Exempli gratia, try expanding its footnote ;) FaviFake (talk) 16:13, 22 August 2025 (UTC)[reply]
    Can you work it out in a sandbox and show how you think the end result would look? A lot of the hidden text as written, to me, is quite wordy. While I agree they contain relevant information, having them off to the side (as they are presently) helps maintain focus. We just want to make sure we aren't degrading that in the process. -- GoneIn60 (talk) 16:43, 22 August 2025 (UTC)[reply]
    Yeah, footnotes are frequently unnecessarily wordy. Here's how it might look. I cut down on about 500 bytes and kept everything the footnotes said. FaviFake (talk) 17:25, 22 August 2025 (UTC)[reply]
    The tradeoff is:
    • The shorter the main text (additional material kept in footnotes), the more likely people are to read the main text, but the less likely they are to see the footnote material.
    • The longer the main text (additional material incorporated directly), the less likely people are to read all of the main text, but the more likely they are to see the previously footnoted material.
    I don't personally have a strong opinion about which answer is the right one for this section, but I think that editors should consider whether all of this material is of equal importance. WhatamIdoing (talk) 17:38, 22 August 2025 (UTC)[reply]
    Yes, that's true. I'll just add that this is a policy page; people don't usually read it in their leisure time, but only when they need practcal info on what to do. I think the policy reccomendations in the efns shouldn't be hidden to editors who aren't curious enough to hover over them. FaviFake (talk) 09:30, 23 August 2025 (UTC)[reply]

    One thing to remember is that this is a policy page (and a core one at that). So everything that is in the text of the policy is policy. Moving that large amount of sidebar content would be basically adding about 20 things to policy. IMO any such additions should be decided on individually.North8000 (talk) 14:34, 23 August 2025 (UTC)[reply]

    If "everything that is in the text of the policy is policy", then this isn't adding anything new. These footnotes are already supported by the existing consensus, and this would just allow people to read the entire policy without having to click many blue letters.
    Besides, the footnotes don't mandate anything, they're all mere suggestions. FaviFake (talk) 14:42, 23 August 2025 (UTC)[reply]
    I disagree with several aspects of that. IMO there is a higher level of scrutiny and consensus for what's in the text of the policy than the footnotes. (and the format is more careful and concise). Next, they don't currently mandate anything because they are footnotes. Moving them would convert them into mandates. This might be a good idea for many of them, but IMO they should be individually evaluated in that context. Sincerely, North8000 (talk) 15:12, 23 August 2025
    North8000 Thanks. Do you recommend I create a subsection for each? FaviFake (talk) 16:12, 23 August 2025 (UTC)[reply]
    Some of the footnotes mandate things. See, e.g.,:
    • "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.)."
    Note that word must: it's a mandate. There are two main points that we're trying to communicate here:
    1. BURDEN requires "me" to provide one (1) source that "I" think it's good. "You" don't get to send me on an endless game of WP:FETCH.
    2. Even if "I' provide an undeniably gold-plated reliable source, that still doesn't mean that "my" content needs to stay in this article.
    WhatamIdoing (talk) 18:10, 23 August 2025 (UTC)[reply]
    You're right, I must've missed that! Thanks. FaviFake (talk) 19:20, 23 August 2025 (UTC)[reply]

    Opposed setting this precedent....efns (notes) are mostly used in the same manner as Template:Supplement#History....that is to handle overly detailed information. Many supplemental essays are the result of notes getting too big. They are used for quick explanation of rationale as seen here to explain reasoning or used to give examples that would just crowd protocol text as seen here or as a link to discussions on how we came to this conclusion as seen here.Moxy🍁 15:34, 23 August 2025 (UTC)[reply]

    Thanks. Could you elaborate on which efns in this section only handle overly detailed info? To me, all but "d" (and "b", which doesn't actually originate in that section) seem very relevant. FaviFake (talk) 16:11, 23 August 2025 (UTC)[reply]
    In my view all the notes are summaries of other protocols there to help editors understand context without have to go to another page. The point being - keep focused on the points here without having to read many other pages for understanding. Wikipedia:Content forking/Internal#Policy forksMoxy🍁 16:27, 23 August 2025 (UTC)[reply]

    This section is trying to do two things. It's policy (this is how things shall be) and practice (this is how to get there).

    The text flipflops from policy to practice and back again. The footnotes seem to me like a (somewhat inconsistent) attempt to move the practice guidance to footnotes and leave the policy in the header, but over the course of many years and many edits it's all gone wrong. Needs a full rewrite imo.—S Marshall T/C 19:28, 1 September 2025 (UTC)[reply]

    I wonder if some of this (e.g., WP:Directly supports) could be addressed with a ==Terms and definitions== section. WhatamIdoing (talk) 22:21, 1 September 2025 (UTC)[reply]
    Possibly? Do any terms other than WP:Directly supports need defining?
    I'm not settled on any thoughts at the moment but last night I came up with a very rough draft:
    I've only skimmed it but it looks great!! In case you missed it, i had made a very quick draft too in this comment, if it's of an use. I basically just copypasted and removed some parts, im sure yours is better. FaviFake (talk) 17:52, 2 September 2025 (UTC)[reply]
    A definition of verifiable (as contrasted with cited) would also help a lot of editors.
    Taking the first paragraph from your draft as an example, consider:
    • Each fact or claim must be verifiable. – What's "verifiable" mean?
    • The burden to demonstrate verifiability lies with the editor who adds or restores material. – So I have to do it, but I don't know what "it" is.
    • It is satisfied by providing an inline citation to a reliable source that directly supports the material. – What's "directly" vs (I guess) "indirectly supporting"?
    • The cited source must clearly support the fact or claim as presented in the article. – How is "clearly" different from "directly"?
    Experienced editors like us, although in some ways the only people on the planet who are qualified to write this policy, are always bringing a h-u-g-e amount of background knowledge to these things. We know things that newbies can't even guess at. WhatamIdoing (talk) 00:04, 4 September 2025 (UTC)[reply]

    Well, some of that I'd already addressed in my previous draft, but let's do the rest of it.  :)

    Yes, I know, the WP:WOOKIEE shortcut is very inside baseball.—S Marshall T/C 16:28, 5 September 2025 (UTC)[reply]

    I like the clarity of this approach. I think in a few places the proposed wording goes too far in advising against removal, as compared to the current wording.
    A few specific points:
    • When dealing with facts or claims you think are dubious, a good option is to add a tag... Tagging is helpful when it is likely that a source exists and just needs to be located and added. Dubious unsourced information should be removed. When to remove vs tag is better dealt with under Considerations.
    • Do not overuse this clause. No one is likely to view their own interpretation as overuse.
    • Although policy allows this, removing facts or claims is a relatively drastic action that's best reserved for material you have good reason to doubt. This undermines the core policy in a confusing way. Removing unsourced claims is not drastic; it is routine. It should be done with judgment, but ...material you have good reason to doubt is not the right test. Plausible OR should still be removed.
    • Removing false claims is helpful, but replacing them with verifiable claims is much better. The contrast shouldn't be between false and verifiable.
    Here's a version with some suggested edits.

    --Trystan (talk) 14:04, 6 September 2025 (UTC)[reply]
    "They should not be restored without an inline citation to a reliable source." and sentences like it are problematic (although longstanding, I admit) because they seem to limit the reasons why a sentence or bit of info should be in an article (see NPOV, NOR, BLP, CVIO). We say in other areas like ONUS, that this is not true, but I don't think that is clear enough. I get why this happened, as this is the policy on verifiability, but we should still be more circumspect in such sentences. Alanscottwalker (talk) 13:14, 7 September 2025 (UTC)[reply]
    Further comments and objections are invited? I'm considering editing the policy along these lines now.—S Marshall T/C 07:22, 11 September 2025 (UTC)[reply]
    Similar to Alanscottwalker, I think the text is potentially unhelpful in its coverage of WP:BLUE. Facts can be uncontroversial but not directly relevant. Sourcing can be requested to help demonstrate the relevancy of uncontroversial statements and avoid SYNTH, rather than just because a fact is potentially 'controversial'. I also am unsure if we should keep the note on unreferenced categories, they aren't really cited inline in the same way, so that could read as confusing. CMD (talk) 07:35, 11 September 2025 (UTC)[reply]
    The first consideration you raise has already got its own section at WP:VNOT, so I wouldn't cover that in WP:BURDEN. I agree that the verifiability of categories is confusing.—S Marshall T/C 11:29, 11 September 2025 (UTC)[reply]
    Right, I am suggesting not covering it, which the proposed text above does. CMD (talk) 15:00, 11 September 2025 (UTC)[reply]
    Does it? Where?—S Marshall T/C 15:50, 11 September 2025 (UTC)[reply]
    The new paragraph mentioning the Thames and Paris. CMD (talk) 01:11, 12 September 2025 (UTC)[reply]
    The paragraph says that (wildly oversimplified) if it's BLUE and it's CHALLENGED, then cite it instead of arguing over whether to cite it. I think we all agree to this.
    Maybe uncontroversial is not the right word? "Not one of the four types of content that is required to have an inline citation" is wordier but perhaps more accurate. WhatamIdoing (talk) 01:24, 12 September 2025 (UTC)[reply]
    The wording of blue is "all material that is either challenged...must be cited", rather than suggest arguing with it is an option. Suggesting material "should never need citations", especially when there are other reasons for citations (like those mentioned by Alanscottwalker) is unnecessary, and if the desire is to not cover that in WP:BURDEN the easiest thing to do is not cover it. If there is a desire to cover it, use more direct wording such as "If material has been challenged, a citation should be provided." CMD (talk) 01:31, 12 September 2025 (UTC)[reply]
    If I write, "The Thames is a river in England" without bothering to provide a source, and I'm writing it in River Thames, then someone who uses WP:BURDEN on that is a bloody fool who needs to find a better hobby because they aren't competent to edit an encyclopaedia.
    If I write, "The Thames is a river in England" without bothering to provide a source, and I'm writing it in Mississippi, then someone who uses WP:BURDEN on that is still a bloody fool. Yes, they can legitimately say "That's true, but maybe it doesn't belong in this article," and they're likely right. But that's not about verifiability and we don't need to bend the verifiability policy out of shape to address it. An easily-sourceable statement that's in the wrong place is still a verifiable statement.—S Marshall T/C 14:19, 12 September 2025 (UTC)[reply]
    We do have a certain number of fools...[1] WhatamIdoing (talk) 22:35, 12 September 2025 (UTC)[reply]
    I think that particular instance is more vandalism than foolishness, but yup, we certainly have. It did make me laugh.—S Marshall T/C 01:15, 13 September 2025 (UTC)[reply]
    Why assume they are using BURDEN? That's not a scenario I've seen, but I have seen people respond "it doesn't need to be cited" or similar when the "likely right" issues are raised. CMD (talk) 15:59, 14 September 2025 (UTC)[reply]

    It looks good to me. I do like the "avoid overuse" aspect. The most common overuse is nit-picking the source when sky-is-blue "I don't like it" material is restored. I'd suggest a very thorough (re) review by a thorough expert like S Marshall  :-) be made before putting it in. North8000 (talk) 14:17, 11 September 2025 (UTC)[reply]

    As @Trystan says, I'm never going to view my use of this clause as "overuse". However, I think it's important for the rest of the community to have something like this, so you can all tell me "No, really: stop being WP:POINTY or WP:DISRUPTIVE or whatever already. It says so right there in the policy that restraint and judgment are needed".
    It's unfortunate that NOTBURO ("Do not follow an overly strict interpretation of the letter of policies without considering their principles") isn't more widely respected these days, but "Don't overuse this" is WP:V's version of NOTBURO. WhatamIdoing (talk) 01:20, 12 September 2025 (UTC)[reply]
    Well, it will be, if I can get consensus to add it. I'm very much hoping that we can coalesce around wording as strong as "do not overuse this clause".
    There are people whose content work needs large-scale removal (not to mention Lugnuts by name... Oh, oops!) but that shouldn't be done under BURDEN. It needs a prior consensus.
    BURDEN is the right tool for small numbers of disputed sentences or paragraphs. It can be, and has been, misused to follow someone around challenging their content work on a very large scale. This is not a pleasant experience for the victim. It feels like griefing.—S Marshall T/C 08:05, 12 September 2025 (UTC)[reply]
    As a watcher of this page who rarely pitches in, I'd like to say that I strongly support a "do not overuse" wording. People aren't run by computer programmes, and such extra words help to make sure we don't forget "common sense". I also disagree with the relativism argument, if I may call it that. People will tend to have highly overlapping understandings about what these words are envisaging. If editors work in good faith then such a wordings do help in practice.--Andrew Lancaster (talk) 10:44, 12 September 2025 (UTC)[reply]
    There are policies to stop POINTY or DISRUPTIVE behaviour, in behaviour policies rather than a content one. -- LCU ActivelyDisinterested «@» °∆t° 14:23, 12 September 2025 (UTC)[reply]
    There are. Editors who're strictly following the letter of a policy page won't know they're being POINTY or DISRUPTIVE, though. We have to tell them.—S Marshall T/C 21:41, 12 September 2025 (UTC)[reply]
    Then we'll need to cover the behaviour issue caused by editors saying they don't have to reference their slop because the other editors are over using policy. -- LCU ActivelyDisinterested «@» °∆t° 22:36, 12 September 2025 (UTC)[reply]
    I agree that this policy, which has been abused in this way by a small number of editors, needs to have a brief statement about this. I don't think that anyone complaining that "Don't overuse this" means "Don't use this on anything I add" is going to be successful. We're looking for a balance:
    • You need to cite four types of content – but you don't need to cite other types of content.
    • You don't need to cite other types of content – but if someone demands a non-required citation, it's faster to cite it "unnecessarily" than to argue about whether it's one of the four types of content.
    • You need to cite what you add – but even if you cite it, it can be removed (e.g., bad content, bad sources, wrong article...)
    • If you didn't cite it, someone else is allowed to remove uncited content – but it's not usually the best thing to do.
    • If you didn't cite it, someone else is allowed to remove uncited content – but large-scale removals are usually the wrong thing to do.
    We have a problem that sounds like this:
    • A: I hereby CHALLENGE all uncited content from here to infinity! I shall blank it all!
    • B: Um, that's POINTY.
    • A: My WP:V policy trumps your little behavioral guideline. I am allowed by the One True™ Policy to blank it all, and you can't stop me!
    That editor needs a note in this policy to rein him in. WhatamIdoing (talk) 22:46, 12 September 2025 (UTC)[reply]
    No that editor should be taken to an appropriate forum for disruptive behaviour, rather than giving one side in an argument a get out of jail card. Why should someone misusing "Don't overuse this" be handled any differently than those who are over using it, and if one is less likely to be censored by the community why should policy be changed in against that community. -- LCU ActivelyDisinterested «@» °∆t° 22:52, 12 September 2025 (UTC)[reply]
    Have you seen any of those discussions? You can only get newbies blocked for blanking large swathes of uncited material. Anyone who knows how to play the game will spend a couple of days at the drama board and leave with nothing worse than a "reminder" that this is supposed to be a collegial place.
    It's not a get out of jail free card for "one side". The point behind the balance is that we're requiring something from both sides (add sources for four types of content; use judgment and restraint), and giving something to both sides (don't argue about whether you 'should' have to cite it; use judgment and restraint). WhatamIdoing (talk) 19:17, 13 September 2025 (UTC)[reply]
    We have a much larger problem of unsourced content that verges on OR, if it isn't actually OR. And many editors who when pushed to show sources get upset they their content has been challenged. -- LCU ActivelyDisinterested «@» °∆t° 22:59, 12 September 2025 (UTC)[reply]
    Technically I guess that's more what NOR is about, but in any case I don't think that including "don't overuse" is in anyway going to distract away from the core messages of these policies. I think no-one is proposing any kind of watering down of those. I think it is wrong to use the term "get of jail" because by all accounts no one is proposing any kind of exception to any rule, or any change at all really? It is just a reminder to be reasonable.--Andrew Lancaster (talk) 14:00, 13 September 2025 (UTC)[reply]
    @ActivelyDisinterested, I'd like you to tell me more about problem of unsourced content that verges on OR. Do you mean that if the article says "Alice Expert is an author", then that verges on OR, but if it says exactly the same thing, but with a little blue clicky number after it, then it doesn't? WhatamIdoing (talk) 19:19, 13 September 2025 (UTC)[reply]
    No I'm mean that there is a lot of content without without sources that appears to be OR. -- LCU ActivelyDisinterested «@» °∆t° 20:15, 13 September 2025 (UTC)[reply]
    Yes there is. I'm not saying "do not use this clause". I'm only saying "do not overuse this clause".
    What I intend is that when you find one or two things in an article that aren't verifiable, you get to remove them. But if you find seven things in an article that aren't verifiable, you either (a) head to the talk page to start a discussion, and while you're doing that, you leave them in; or (b) you head to AfD to start a discussion, and the disputed material stays up until that discussion is closed; or (c) you head to WP:AN/I and start drama about the people responsible for all this unverifiable content, depending on the situation. What wording do you think would have that effect?—S Marshall T/C 21:38, 13 September 2025 (UTC)[reply]
    (a) doesn't seem like a desirable outcome, and (b) only if "seven" means "enough to think the article is a hoax". Nikkimaria (talk) 21:50, 13 September 2025 (UTC)[reply]
    Unfortunately there are entrenched editors who write articles as if they were essays based on their person knowledge, and who dislike "fools" questioning their content. "do not overuse this clause" is going to make matters worse, because it will be misused. Trying to stem one issue by creating another, when both are handled already by behaviour policies, isn't constructive. -- LCU ActivelyDisinterested «@» °∆t° 21:58, 13 September 2025 (UTC)[reply]
    Well, I guess we'll see if anyone agrees with you.—S Marshall T/C 22:52, 13 September 2025 (UTC)[reply]
    I do agree with ActivelyDisinterested. And I think the discussion is getting to be longer than necessary. Yesterday, all my dreams... (talk) 00:41, 14 September 2025 (UTC)[reply]
    Well I thought it better to bring these issues up now, before any full proposal to change long standing policy was put forward. -- LCU ActivelyDisinterested «@» °∆t° 07:18, 14 September 2025 (UTC)[reply]
    Yes, and I think your comment "there is a lot of content without sources that appears to be OR" is correct, and is in fact an understatement. The situation is serious in key articles. Yesterday, all my dreams... (talk) 09:50, 14 September 2025 (UTC)[reply]
    Guidance not to overuse this policy needs to be more concrete than "don't overuse it". Specific guidance on exercising restraint is already in the above drafts, and boils down to: Removing verifiable material just because it doesn't have an inline citation right now is unhelpful. How bold an editor should be in removing uncited material depends on how confident they are it is unverifiable (or subject to other reasons for removal). An editor familiar with the subject and what the sources do and do not say can be very bold about removing material, even large amounts of it. That is a service to Wikipedia, not a disruption. A reader new to a topic should be much more cautious. I don't think a top-level policy can be more specific than suggesting that Whether or how quickly unsourced content should be removed depends on the material and the overall state of the article.--Trystan (talk) 13:56, 14 September 2025 (UTC)[reply]
    I think that some editors struggle with the concept of uncited content being verifiABLE. For them, "verifiable, uncited material" is an oxymoron, because they're thinking verifiable means "supported by the cited source". WhatamIdoing (talk) 18:38, 20 September 2025 (UTC)[reply]

    IMO "over use" is gentle code for "Don't use it for other than it's intended purpose",("intended purpose" = the general intention of wp:ver). This is a common problem. Maybe saying it more directly would be a good thing but either a direct or indirect nudge on that would be a good thing. North8000 (talk) 15:53, 14 September 2025 (UTC)[reply]

    Axiomatic systems

    [edit]

    S Marshall, I should address your comment above that "if you find seven things in an article that aren't verifiable, you either (a) head to the talk page to start a discussion, and while you're doing that, you leave them in; or (b) you head to AfD". Who says there is anyone who bothers to read the talk page? There are now so many, many, many articles that some pages get no attention at all. And in technical cases 99% of Wiki editors know nothing about the subject. I kid you not. I should probably give you an example. Take a look at the talk page and edit history for Axiomatic system. The ONLY user who seemed to know anything was CBM, who is now gone, it seems. The rest are clueless. And most of the content was/is not just OR but reflects a basic lack of knowledge. I have cleaned if up a little, but do not have time to write a book on the subject. The best I can do is delete the nonsense. And this is a "key subject" in logic. There are many more like this. Yesterday, all my dreams... (talk) 10:09, 14 September 2025 (UTC)[reply]

    • With axiomatic system we're lucky, because as well as CBM, it's also been edited by Charles Matthews who, I am quite confident, is far from clueless about axiomatic systems. If we could clone Dr Matthews a few hundred times our maths articles would be a lot better.
      I've long said that articles about maths and logic need special treatment in WP:V because so many of them are, or can be, self-verifying. Look for example at Banach space; my position is that footnote #4 does fully verify the claim it supports. One day, when I'm Ruling Tyrant of Wikipedia and I get to write the policies, articles within the scope of WikiProject Mathematics will have additional considerations and special carve-outs that allow this.
      I fully acknowledge that our maths articles are one of several topic areas where very large numbers of important articles are maintained by an extremely small corpus of tired and overworked people. I don't know how to fix Wikipedia's problems with recruitment and retention.
      I do know that I'm writing policy to help us resolve conflict. We need policy for articles about post-1932 US politics, medical claims, the India-Pakistan conflict, etc. etc. etc. In cases like axiomatic system we don't need policy, because you're the only person editing it, so your editorial judgement prevails and we just have to hope we can trust you.—S Marshall T/C 12:14, 14 September 2025 (UTC)[reply]
      That is so funny. Just funny. Please take a look at the talk page for Formal system and my comment there on June 13, 2025. I pinged Charles Matthews and also suggested that someone should talk him into fixing that page which also needs help from FEMA. Result: nothing doing. As for trusting me on logic, modesty forbids a reply. Let me just note that Matthews did his PhD on number theory, so he probably knows more about that topic than yours truly. But I probably know as much about logic than any current Wiki editor. But no way to verify that, of course. And modesty forbids further... Yesterday, all my dreams... (talk) 12:47, 14 September 2025 (UTC)[reply]
      Non-gender-specific dude — either say who you are or please stop talking about it. Right now it sounds like you'd like people to be intimidated to disagree with you, but you won't specify why we should. Of course, no one should be intimidated to disagree with even Terry Tao; that isn't how Wikipedia works, in theory. But in practice, stop trying to have the best of both worlds. --Trovatore (talk) 21:34, 15 September 2025 (UTC)[reply]
      Yes: there's an awful lot to do, and only one of him, and he only has so much volunteering time. The Augean Stables are going to take forever to clean by hand.—S Marshall T/C 13:03, 14 September 2025 (UTC)[reply]
      The longer and short of it is that ActivelyDisinterested was correct above. Yesterday, all my dreams... (talk) 13:53, 14 September 2025 (UTC)[reply]
      For the use cases you're envisaging, he is. Which tells me that neither AD nor you have ever seen your watchlist get lit up by a griefer who's gone through targeting everything you've ever written with WP:CHALLENGES. You probably don't think that happens; or that when it happens, it's easy to deal with.—S Marshall T/C 14:51, 14 September 2025 (UTC)[reply]
      we will jump off that bridge when we get to it. And I have spent too much time on this, so will move on. Yesterday, all my dreams... (talk) 15:54, 14 September 2025 (UTC)[reply]
      Of course observer bias is unavoidable, I can only assume you haven't seen the difficulties caused by the other issue. Maybe there is an option that covers both without encouraging one bad behaviour in an attempt to discourage another. -- LCU ActivelyDisinterested «@» °∆t° 19:11, 14 September 2025 (UTC)[reply]
      Sigh. That version says nothing that encourages any bad behaviour at all.—S Marshall T/C 21:22, 14 September 2025 (UTC)[reply]
      Sigh. I'm afraid it does. Well when the changes to long standing policy are put forwards for consensus everyone can have a say. -- LCU ActivelyDisinterested «@» °∆t° 22:56, 14 September 2025 (UTC)[reply]

    I wasn't aware of having been pinged, can't say why. A message on my talk page should work.

    I was working on formal system in 2003, which is two decades ago now. Things here have changed greatly. The page now looks like the top level article in a bunch of articles related by summary style (WP:SUMMARY), and I would say that is how it should look. In other words, various sections that deal summarily but usefully with major topics that have their own articles. Possible criticisms are that the summaries are too concise, are too inaccurate, or possibly are inadequately referenced (though being picky about that isn't too helpful, if good referencing is available in the detailed article). So that is how verifiability fits in. Equally important, really, is that the overall organisation of the material serves the purposes of the encyclopedia.

    I was working on axiomatic system in 2004. It does seem to be in worse shape than formal system. Neither does it look much like the top level in articles organised under summary style, nor does it have a clear "logical flow" (ironically enough). Probably the latter criticism is more actionable. Systems of axioms are set up for reasons. Probably the article needs major surgery.

    Personally, I think everything written here before 2007 needs to be rewritten. Charles Matthews (talk) 21:32, 14 September 2025 (UTC)[reply]

    Yes, most of those articles need to be rewritten. And world hunger needs to end. But we both know that neither is going to happen in the next decade. The fact that the Wikipedia model has so far failed with respect to technical articles is not supposed to be mentioned here. But it is a fact. So let us not pretend that discussions on verifiability here will affect that. Those discussions assume that plenty of people care. In the technical articles that assumption fails, obviously. All that the few of us can do is to remove the most obvious errors. In the articles on logic the most glaring errors are/were based on the fact that the amateurs who wrote the material did not understand the difference between provability and satisfiabilty. I have removed some of those but there are plenty more. Yesterday, all my dreams... (talk) 09:17, 15 September 2025 (UTC)[reply]
    Given the incremental and ameliorative nature of Wikipedia editing, I don't see that the rhetoric is called for. Charles Matthews (talk) 06:44, 16 September 2025 (UTC)[reply]

    Next iteration

    [edit]
    Unpacked and rewritten
    Responsibility for providing citations Basics
    All content must be verifiable. The burden to demonstrate verifiability lies with the editor who adds or restores material, and it is satisfied by providing an inline citation to a reliable source that directly supports (A source "directly supports" a given piece of material if the information is present explicitly in the source, so that using this source to support the material is not a violation of Wikipedia:No original research. The location of any citation—including whether one is present in the article at all—is unrelated to whether a source directly supports the material. For questions about where and how to place citations, see Wikipedia:Citing sources, Wikipedia:Manual of Style/Lead section § Citations, etc.) the contribution. 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 cited source must clearly support the material as presented in the article. Cite the source clearly, ideally giving page number(s)—though sometimes a section, chapter, or other division may be appropriate instead; see Wikipedia:Citing sources for details of how to do this.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Whether or how quickly material should be removed for lacking an inline citation to a reliable source depends on the material and the overall state of the article. Consider adding a citation needed tag as an interim step to removing unsourced material, to allow references to be added. When an article contains a lot of uncited information, it may be impractical to add specific {{citation needed}} tags. Consider then tagging a section with {{unreferenced section}}, or the article with the applicable of either {{unreferenced}} or {{more citations needed}}. For a disputed category, you may use {{unreferenced category}}. For a disambiguation page, consider asking for a citation on the talk page. When tagging or removing material for lacking an inline citation, state your concern that it may not be possible to find a published reliable source, and the material therefore may not be verifiable. When tagging or removing such material, please communicate your reasons why. Some editors object to others making frequent and large-scale deletions of unsourced information, especially if unaccompanied by other efforts to improve the material. Do not concentrate only on material of a particular point of view, as that may appear to be a contravention of Wikipedia:Neutral point of view. Also, check to see whether the material is sourced to a citation elsewhere on the page. For all these reasons, it is advisable to clearly communicate that you have a considered reason to believe the material in question cannot be verified. If you think the material is verifiable, you are encouraged to provide an inline citation yourself before removing or tagging it.

    Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living people or existing groups, and do not move it to the talk page. You should also be aware of how Wikipedia:Biographies of living persons also applies to groups.

    Policy

    Each fact or claim must be verifiable. The burden to prove verifiability lies with the editor who adds or restores material.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living people, and do not move it to the talk page. You should also be aware of how Wikipedia:Biographies of living persons also applies to groups.

    Practice

    A fact or claim is "verifiable" if a reliable source that supports it exists. Facts and claims can be verifiable even if there is no citation for them in the article at the moment.

    The burden to prove verifiability is satisfied by providing an inline citation to a reliable source that directly supports the material. The cited source must clearly support the fact or claim as presented in the article.

    A source "directly supports" a fact or claim if that fact or claim is explicitly written in the source. For questions about where and how to place citations, see Wikipedia:Citing sources and Wikipedia:Manual of Style/Lead section § Citations.

    Once an editor has cited a source they believe, in good faith, is sufficient, any editor who later removes the material must explain. State specific problems that would justify its removal. If necessary, all editors are then expected to help achieve consensus. Problems should be fixed before the material is added back.

    Removing text is sometimes mistaken for vandalism. To avoid this, when using WP:CHALLENGE, best practice is to say in an edit summary and/or talk page note that you have removed the text because you don't think it's verifiable unless you are removing text that makes an unsourced, negative claim about a living person, in which case best practice is to remove it and say in an edit summary or talk page note that you have removed it using WP:BLP.

    {{policy shortcut}} WP:WOOKIEE

    Some claims are so uncontroversial that they should never need citations. ("The Thames is a river in England". "Paris is the capital of France.") But if good faith dispute exists about whether a claim is uncontroversial, it is better to add a citation than to argue about whether a citation is needed.

    Considerations

    Removing verifiable material just because it doesn't have an inline citation right now is unhelpful. Removing unverifiable claims is helpful, but replacing them with verifiable claims is much better.

    Before removing text, always check to see whether the disputed material is sourced to a citation elsewhere on the page.

    Where you feel the need to remove large quantities of material, it's best to start a discussion first.

    Instead of removing text, consider adding a tag, which allows other editors time to find references. When an article contains a lot of uncited facts or claims, it may be impractical to add specific {{citation needed}} tags. Consider tagging a section with {{unreferenced section}}, or the article with {{unreferenced}} or {{more citations needed}}. For a disputed category, you may use {{unreferenced category}}. For a disambiguation page, consider asking for a citation on the talk page.

    Misuse

    Do not use WP:CHALLENGE to selectively remove material to favour a particular point of view. Always make sure the article is neutral.

    Do not follow an editor around using WP:CHALLENGE on things they've written, and particularly, do not follow an editor from article to article doing this. Always be mindful of harassment and hounding.

    Do not use WP:CHALLENGE to gut an article that you have already nominated for deletion. In some circumstances you might use WP:BLP to do this.

    This variant tries to do an end run around the stonewalling about a "do not overuse this" clause by being more specific about what constitutes misuse.—S Marshall T/C 12:17, 17 September 2025 (UTC)[reply]

    I wonder whether the "burden" language could be removed. "If you want to add or restore material that requires an inline citation, you need to cite a source yourself, instead of expecting others to do that work for you"?
    The word exists, as in "A fact or claim is "verifiable" if a reliable source that supports it exists", seems to confuse people. Some people (correctly) say "A source exists – in my library!" Others say "No source exists – see, there are no little blue clicky numbers in this whole article!" Perhaps this would be clearer: "A fact or claim is "verifiable" if any reliable source has been published that says the same thing." WhatamIdoing (talk) 19:18, 20 September 2025 (UTC)[reply]

    We might be stuck with the "burden" terminology because WP:BURDEN has become such a meme? I agree that "exists" is likely to be misunderstood and this paragraph needs changing.

    Before Revised
    A fact or claim is "verifiable" if a reliable source that supports it exists. Facts and claims can be verifiable even if there is no citation for them in the article at the moment. A fact or claim is "verifiable" if a reliable source that supports it could be cited, whether or not there is a citation for it in the article at the moment.

    And also

    Before Revised
    {{policy shortcut}} WP:WOOKIEE

    Some claims are so uncontroversial that they should never need citations. ("The Thames is a river in England". "Paris is the capital of France.") But if good faith dispute exists about whether a claim is uncontroversial, it is better to add a citation than to argue about whether a citation is needed.

    		{{policy shortcut}} WP:WOOKIEE

    Some claims are so uncontroversial that they should never need citations. ("The Thames is a river in England". "Paris is the capital of France.") But if there is good faith dispute about whether a claim is uncontroversial, it is better to add a citation than to argue about whether a citation is needed.

    How about this?—S Marshall T/C 21:16, 20 September 2025 (UTC)[reply]

    "Challenge", as you have framed it, as only in contra to the person's good faith belief, is a poor indicator because there are many good reasons to challenge content beyond verifiable and even within the verifiability rubric, these are suggested in the footnote c: "the source is unreliable; the source does not support the claim; undue emphasis; unencyclopedic content; etc." Alanscottwalker (talk) 14:53, 21 September 2025 (UTC)[reply]
    No, the proposed text allows WP:CHALLENGE for all those reasons. The constraint is that the challenger must "state specific problems that would justify [the disputed content's] removal." I haven't tried to enumerate them and I suggest we don't.—S Marshall T/C 15:00, 21 September 2025 (UTC)[reply]
    That's the problem I pointed to, the way you have framed it is limiting. 'Burden', as a catchword is more narrowly focused on 'provide your source', but the challenges don't end there. Burden is minimally met (and its yardstick is personal perception), the challenge is multifaceted and its operation is communal. Alanscottwalker (talk) 15:21, 21 September 2025 (UTC)[reply]
    That's not a change. It's how it works now.—S Marshall T/C 19:40, 21 September 2025 (UTC)[reply]
    You appear to be trying to not listen. Your proposal changes the words in the direction that I find objectionable. You may disagree with my objection, but don't pretend your proposal is not trying to change how this reads. Or perhaps you are unaware, that your use of CHALLENGE completely merges it with BURDEN, indeed makes the section cite itself to explain itself (just follow the links), but I'm saying they can't merge and they should not, as your proposal would have it. Alanscottwalker (talk) 20:49, 21 September 2025 (UTC)[reply]
    I'm trying to listen, and to overlook the levels of aggression and hostility that you're displaying, but I don't find your objection intelligible.
    Proposed text Policy shortcut
    Each fact or claim must be verifiable. The burden to prove verifiability lies with the editor who adds or restores material. This is WP:BURDEN.
    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source. This is WP:CHALLENGE
    They aren't merged.—S Marshall T/C 21:48, 21 September 2025 (UTC)[reply]
    Aggression? What? Is it that you somehow did not understand when I said your use of challenge limits challenge? What you just now quote limits challenge. -- Alanscottwalker (talk) 21:54, 21 September 2025 (UTC)[reply]
    Can you explain exactly how, please. Explain it like I'm ten.—S Marshall T/C 22:10, 21 September 2025 (UTC)[reply]
    Let's start off socratically. Are burden and challenge two different things? Are they raised by different users? Is answering a challenge done solely by fulfilling burden?
    My answers are yes, yes, and, no, which I explained above as 'Burden', as a catchword is more narrowly focused on 'provide your source', but the challenges don't end there. Burden is minimally met (and its yardstick is personal perception), the challenge is multifaceted and its operation is communal. And, there are many good reasons to challenge content beyond verifiable and even within the verifiability rubric, these are suggested in the footnote c of the present policy: "the source is unreliable; the source does not support the claim; undue emphasis; unencyclopedic content; etc.
    Your text makes challenge the mirror of burden, which is fulfilled by meeting the limited burden, thus they merge. But they should be distinct. Alanscottwalker (talk) 22:24, 21 September 2025 (UTC)[reply]
    WP:CHALLENGE should probably redirect to a different section. I see WP:BURDEN as a subset of CHALLENGE. Blueboar (talk) 23:08, 21 September 2025 (UTC)[reply]
    On my screen, challenge and burden are shortcuts that point to the same section of the same policy. The plain meaning of that policy section is that you can remove text that doesn't have a citation and nobody can put it back in without a citation. If someone puts it back in with a citation, they've met the CHALLENGE and satisfied the BURDEN so you don't get to take it out again, but if you're not happy with their citation, you take it to the talk page and everyone tries to achieve consensus. That's all it says, so that's all that my revision says.
    It does not say anything else.
    You can remove text for other reasons. You can remove it for being a copyvio or for being an unsourced negative statement about a living person or violating NOT or being a SYNTH or for many other reasons, but those are not CHALLENGE situations, so supplying a citation does not counter them. We do not want to make them CHALLENGE situations because we do not want to make supplying a citation into a counter for them.—S Marshall T/C 23:43, 21 September 2025 (UTC)[reply]
    CHALLENGE is currently pointing to the same section as BURDEN… but I don’t think it should (since you can challenge material for other reasons than just Verifiability). It’s easy to point a redirect elsewhere. Blueboar (talk) 00:38, 22 September 2025 (UTC)[reply]
    Well no, CHALLENGE is not on the page, at all, although it does secretly point to that section. Nor should we now readily introduce a word in policy that takes on a narrow peculiar specialized meaning when capitalized. And no, to fulfill BURDEN, all you need to do is cite a source that you personally believe in good faith is reliable and supports. The thrust of BURDEN is to get a source the community can examine and judge, so we can decide what to do next. Giving the source, is the 'to check' reason for verifiability, and thus sources are the irreducible control for almost all our content policy operations. You can't even judge cvio without a source. Alanscottwalker (talk) 00:51, 22 September 2025 (UTC)[reply]

    Folks, this policy is phrased in a pretty obscure way and I'm pleased that my attempt to rewrite it in plain English is shining a light on all the ways it is currently being interpreted. That's a good thing.

    When we're done writing what it says in plain English, there will be scope to discuss what it should say. But let's get the plain English version of what it says right now sorted first?

    I want to be very clear that my rewrite is not changing the rather narrow meaning of WP:CHALLENGE or WP:BURDEN as currently written, and these objections are seeking to add desired changes to the rewrite.—S Marshall T/C 07:36, 22 September 2025 (UTC)[reply]

    1) Well, your use of challenge is not plain language, as we have said.
    2) Nor is the current challenge as narrow as your proposal, as I have pointed out, the current version suggests multiple reasons for challenge (which is plain language meaning for challenge). Alanscottwalker (talk) 11:39, 22 September 2025 (UTC)[reply]
    S Marshall - I accept that your intent is not to change meanings… but I have to agree with Alan that your proposed language does in fact change meanings. Blueboar (talk) 12:12, 22 September 2025 (UTC)[reply]
    In what way, specifically?
    What I need from you two is either (preferably) a suggested way to phrase my rewrite, or (failing that) a clear explanation of what's wrong with what I've written that will lead to something actionable. What I'm getting is a recondite disquisition on the philosophical, socratic nature of verifiability, which while obviously fascinating doesn't do anything to improve the policy wording.—S Marshall T/C 12:40, 22 September 2025 (UTC)[reply]
    I'm having trouble understanding this (fair warning: I'm tired and hungry):
    @Alanscottwalker, a WP:CHALLENGE, in its simplest form, means I removed uncited text because it's uncited (and suspected of being WP:Glossary#unverifiable). After all, as the lead says, we're talking about "material whose verifiability has been challenged", not "material whose presence in the article has been disputed for any of a wide variety of reasons, most of which have nothing to do with verifiability".
    All that stuff in footnote c ("the source is unreliable; the source does not support the claim; undue emphasis; unencyclopedic content...") is about removing content for non-WP:CHALLENGE reasons. I wrote that footnote because of the people who were playing endless games of WP:FETCH, and who thought that if they couldn't CHALLENGE already-cited (including poorly cited) content, then nothing could ever be removed from any Wikipedia article again.
    This is a WP:CHALLENGE:
    • Alice adds something she thinks doesn't need an inline citation.
    • Bob reverts it with an edit summary of WP:CHALLENGED.
    • If Alice (or anyone else) wants to revert Bob, they have to add an inline citation when they restore the material.
    This is not a WP:CHALLENGE:
    • Alice adds something she thinks doesn't need an inline citation.
    • Bob reverts it, claiming that it's off topic.
    This starts as a WP:CHALLENGE but ends up being something different:
    • Alice adds something she thinks doesn't need an inline citation.
    • Bob reverts it with an edit summary of WP:CHALLENGED.
    • Alice reverts Bob and adds an inline citation.
    • Bob reverts it again, claiming that Alice's addition is bad grammar. Or a copyvio. Or anything at all, except a (false) claim it's still uncited.
    WhatamIdoing (talk) 01:39, 23 September 2025 (UTC)[reply]
    Well, if I recall correctly I wrote part of note c, and it was in part because the reasons for removal are many and in part because the standard given for fulling burden is internal (what they believe in good faith) to the one providing the source, but the reasoning in response needs to be external from another user.
    Regardless, in its simplest form, things are challenged by removal (which does not mean there are not more involved ways to challenge like taking it to the talk page, etc.). Again, CHALLENGE is not in overt usage in the current policy, although it does in secret anchor "challenge" to the BURDEN section, which in the current policy gives multiple reasons for removal. Which only makes sense because any challenge is to how an article reads or a topic is presented (for which there can be many good reasons). So, no, I can't agree with any of this fussiness about CHALLENGE, or more correctly, challenge, being so limited.
    I can't think of any good reasons, we would want to promote arguments about, 'NO!, you did not challenge, according to some peculiar limited meaning!' or 'NO! you did not use the magic words or the correct fomulairy for a challenge!' etc. If you want to say, 'no, you had no good reasons to challenge', simply say that, and that will cover it perfectly, and does not require a limited meaning. -- Alanscottwalker (talk) 05:39, 23 September 2025 (UTC)[reply]
    Oh, right, so we've got "challenge" in the WAID sense and "challenge" in the ASW sense and you two have meant different things by the word for all this time. That's actually simple, then. I see how I can finesse this. Rewrite to follow later.—S Marshall T/C 09:05, 23 September 2025 (UTC)[reply]
    • Let’s cut through the Wikilawyering… there are many reasons why someone might challenge a statement: it could be that the challenger thinks the statement is not Neutral. It could be that the challenger thinks the statement is Original Research. It could be that the challenger thinks it isn’t verifiable. Etc. etc.

    Unfortunately, challengers are not always clear as to why they are challenging (even when they leave an edit summary or a comment on the talk page). However, clear or not, the first step in resolving any challenge is for those who want the statement to remain to add an inline citation. That may or may not end the challenge, but it gives everyone a base for further discussion. Blueboar (talk) 12:22, 23 September 2025 (UTC)[reply]

    Iteration #3

    [edit]
    Unpacked and rewritten
    Responsibility for providing citations Basics
    All content must be verifiable. The burden to demonstrate verifiability lies with the editor who adds or restores material, and it is satisfied by providing an inline citation to a reliable source that directly supports (A source "directly supports" a given piece of material if the information is present explicitly in the source, so that using this source to support the material is not a violation of Wikipedia:No original research. The location of any citation—including whether one is present in the article at all—is unrelated to whether a source directly supports the material. For questions about where and how to place citations, see Wikipedia:Citing sources, Wikipedia:Manual of Style/Lead section § Citations, etc.) the contribution. 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 cited source must clearly support the material as presented in the article. Cite the source clearly, ideally giving page number(s)—though sometimes a section, chapter, or other division may be appropriate instead; see Wikipedia:Citing sources for details of how to do this.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Whether or how quickly material should be removed for lacking an inline citation to a reliable source depends on the material and the overall state of the article. Consider adding a citation needed tag as an interim step to removing unsourced material, to allow references to be added. When an article contains a lot of uncited information, it may be impractical to add specific {{citation needed}} tags. Consider then tagging a section with {{unreferenced section}}, or the article with the applicable of either {{unreferenced}} or {{more citations needed}}. For a disputed category, you may use {{unreferenced category}}. For a disambiguation page, consider asking for a citation on the talk page. When tagging or removing material for lacking an inline citation, state your concern that it may not be possible to find a published reliable source, and the material therefore may not be verifiable. When tagging or removing such material, please communicate your reasons why. Some editors object to others making frequent and large-scale deletions of unsourced information, especially if unaccompanied by other efforts to improve the material. Do not concentrate only on material of a particular point of view, as that may appear to be a contravention of Wikipedia:Neutral point of view. Also, check to see whether the material is sourced to a citation elsewhere on the page. For all these reasons, it is advisable to clearly communicate that you have a considered reason to believe the material in question cannot be verified. If you think the material is verifiable, you are encouraged to provide an inline citation yourself before removing or tagging it.

    Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living people or existing groups, and do not move it to the talk page. You should also be aware of how Wikipedia:Biographies of living persons also applies to groups.

    Policy

    Each fact or claim must be verifiable. The burden to prove verifiability lies with the editor who adds or restores material.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living people, and do not move it to the talk page. You should also be aware of how Wikipedia:Biographies of living persons also applies to groups.

    Practice

    A fact or claim is "verifiable" if a reliable source that supports it could be cited, whether or not there is a citation for it in the article at the moment.

    The burden to prove verifiability is satisfied by providing an inline citation to a reliable source that directly supports the material. The cited source must clearly support the fact or claim as presented in the article.

    A source "directly supports" a fact or claim if that fact or claim is explicitly written in the source. For questions about where and how to place citations, see Wikipedia:Citing sources and Wikipedia:Manual of Style/Lead section § Citations.

    Once an editor has cited a source they believe, in good faith, is sufficient, any editor who later removes the material must explain. State specific problems that would justify its removal. If necessary, all editors are then expected to help achieve consensus. Problems should be fixed before the material is added back.

    Removing text is sometimes mistaken for vandalism. To avoid this, when using WP:BURDEN, best practice is to say in an edit summary and/or talk page note that you have removed the text because you don't think it's verifiable.

    Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note giving all the reasons.

    {{policy shortcut}} WP:WOOKIEE

    Some claims are so uncontroversial that they should never need citations. ("The Thames is a river in England". "Paris is the capital of France.") But if there is good faith dispute about whether a claim is uncontroversial, it is better to add a citation than to argue about whether a citation is needed.

    Considerations

    Removing verifiable material just because it doesn't have an inline citation right now is unhelpful. Removing unverifiable claims is helpful, but replacing them with verifiable claims is much better.

    Before removing text, always check to see whether the disputed material is sourced to a citation elsewhere on the page.

    Where you feel the need to remove large quantities of material, it's best to start a discussion first.

    Instead of removing text, consider adding a tag, which allows other editors time to find references. When an article contains a lot of uncited facts or claims, it may be impractical to add specific {{citation needed}} tags. Consider tagging a section with {{unreferenced section}}, or the article with {{unreferenced}} or {{more citations needed}}. For a disputed category, you may use {{unreferenced category}}. For a disambiguation page, consider asking for a citation on the talk page.

    Misuse

    Do not use WP:BURDEN to selectively remove material to favour a particular point of view. Always make sure the article is neutral.

    Do not follow an editor around using WP:BURDEN on things they've written, and particularly, do not follow an editor from article to article doing this. Be mindful of harassment and hounding.

    Do not use WP:BURDEN to gut an article that you have already nominated for deletion. In some circumstances you might use other policies (particularly WP:BLP) to do this.

    578 words 597 words (NB: too long)

    The shortcut WP:CHALLENGE would still point to this section, as it does at the moment, until we get consensus otherwise.—S Marshall T/C 12:35, 23 September 2025 (UTC)[reply]

    Can I ask, that we change:
    "Removing unverifiable claims is helpful, but replacing them with verifiable claims is much better."
    With
    "Removing unverifiable claims is helpful, but depending on the context, replacing them with policy compliant claims is much better."
    Verifiability is not the only reason you should include something, and many times, trimmed articles are better, and sometimes 'less is more'. Alanscottwalker (talk) 11:26, 24 September 2025 (UTC)[reply]
    To save wordcount, as an alternative, can we move WP:VNOT including WP:ONUS to directly beneath WP:BURDEN?—S Marshall T/C 12:59, 24 September 2025 (UTC)[reply]
    Perhaps, but it remains, often cutting is the best thing to do for our prolix articles, so it is not "much better" to do something else. Alanscottwalker (talk) 13:55, 28 September 2025 (UTC)[reply]
    Not necessarily opposition, but a caution: I think word count is less important than clarity. If it takes more words to explain our policies, so be it. That said, I would agree that we could make the policy clearer. Blueboar (talk) 12:45, 29 September 2025 (UTC)[reply]

    Closing in on a final version

    [edit]
    Unpacked and rewritten
    Responsibility for providing citations Basics
    All content must be verifiable. The burden to demonstrate verifiability lies with the editor who adds or restores material, and it is satisfied by providing an inline citation to a reliable source that directly supports (A source "directly supports" a given piece of material if the information is present explicitly in the source, so that using this source to support the material is not a violation of Wikipedia:No original research. The location of any citation—including whether one is present in the article at all—is unrelated to whether a source directly supports the material. For questions about where and how to place citations, see Wikipedia:Citing sources, Wikipedia:Manual of Style/Lead section § Citations, etc.) the contribution. 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 cited source must clearly support the material as presented in the article. Cite the source clearly, ideally giving page number(s)—though sometimes a section, chapter, or other division may be appropriate instead; see Wikipedia:Citing sources for details of how to do this.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Whether or how quickly material should be removed for lacking an inline citation to a reliable source depends on the material and the overall state of the article. Consider adding a citation needed tag as an interim step to removing unsourced material, to allow references to be added. When an article contains a lot of uncited information, it may be impractical to add specific {{citation needed}} tags. Consider then tagging a section with {{unreferenced section}}, or the article with the applicable of either {{unreferenced}} or {{more citations needed}}. For a disputed category, you may use {{unreferenced category}}. For a disambiguation page, consider asking for a citation on the talk page. When tagging or removing material for lacking an inline citation, state your concern that it may not be possible to find a published reliable source, and the material therefore may not be verifiable. When tagging or removing such material, please communicate your reasons why. Some editors object to others making frequent and large-scale deletions of unsourced information, especially if unaccompanied by other efforts to improve the material. Do not concentrate only on material of a particular point of view, as that may appear to be a contravention of Wikipedia:Neutral point of view. Also, check to see whether the material is sourced to a citation elsewhere on the page. For all these reasons, it is advisable to clearly communicate that you have a considered reason to believe the material in question cannot be verified. If you think the material is verifiable, you are encouraged to provide an inline citation yourself before removing or tagging it.

    Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living people or existing groups, and do not move it to the talk page. You should also be aware of how Wikipedia:Biographies of living persons also applies to groups.

    Policy

    Each fact or claim must be verifiable. The burden to prove verifiability lies with the editor who adds or restores material.

    Facts or claims without an inline citation to a reliable source that directly supports them may be removed. They should not be restored without an inline citation to a reliable source.

    Never leave unsourced or poorly-sourced material that could damage the reputation of living people in an article. Do not move it to the talk page.

    Practice

    A fact or claim is "verifiable" if a reliable source that supports it could be cited, even if there is no citation for it in the article at the moment.

    The burden to prove verifiability is satisfied by adding an inline citation to a reliable source that directly supports the material. The cited source must clearly support the fact or claim as presented in the article.

    A source "directly supports" a fact or claim if that fact or claim is explicitly written in the source. For questions about citations, see Wikipedia:Citing sources.

    Once a good faith editor has cited a source that they believe is sufficient, any editor who later removes the material must explain. State specific problems that would justify its removal. If necessary, other editors should help reach consensus. Problems should be fixed before the material is added back.

    Removing text is sometimes mistaken for vandalism. To avoid this, when using WP:BURDEN, best practice is to say in an edit summary and/or talk page note that you have removed the text because you don't think it's verifiable.

    Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note giving all the reasons.

    {{policy shortcut}} WP:WOOKIEE

    Some claims are so uncontroversial that they should never need citations. ("The Thames is a river in England". "Paris is the capital of France.") But if there is good faith dispute about whether a claim is uncontroversial, it is better to add a citation than to argue about whether a citation is needed.

    Considerations

    Removing verifiable material purely because it doesn't have an inline citation right now is unhelpful, and can be disruptive. Removing unverifiable claims is helpful, but replacing them with well-sourced claims—where relevant and WP:DUE—is much better.

    Before you remove text, check whether the disputed material is sourced to a citation elsewhere on the page.

    Where you propose to remove large quantities of material, it's best to start a discussion first.

    Misuse

    Do not use WP:BURDEN to remove material selectively to skew the article towards a particular point of view. Always make sure the article is neutral.

    Do not follow an editor from article to article using WP:BURDEN on their contributions. Be mindful of harassment and hounding. WP:BURDEN is for dealing with content problems. If you feel there is a problem with an editor, contact an administrator.

    Do not use WP:BURDEN to gut an article that you have already nominated for deletion. In some circumstances you might use other policies (particularly WP:BLP) to do this.

    578 words 516 words ☺👍︎
    Consequential amendments

    The shortcut WP:CHALLENGE would still point to this section, as it does at the moment, until we get consensus otherwise.—S Marshall T/C 14:29, 1 October 2025 (UTC)[reply]

    The second two part of misuse could still use work. Following an editor to correct mistakes they are making is allowed and not harassment, if an editor is spamming articles with controversial unsourced content then reverting them with BURDEN would be appropriate. So the wording needs to be watered down. The third part mixes deletion and content which are different subjects, just because a subject is not ale doesn't mean that all the content in an article is due for inclusion. There is no reason why improving an article can't include removing someone unsourced stream of consciousness, even if the topic itself is notable -- LCU ActivelyDisinterested «@» °∆t° 14:51, 1 October 2025 (UTC)[reply]
    What this tells me is which kind of weirdo you've met.
    BURDEN interacts with weirdoes in two ways. Firstly, there's the kind of weirdo you mean. That person who's going around adding stream-of-consciousness stuff to articles about not ale subjects (did you mean notable?)
    Secondly, there's the kind of weirdo I mean. Those people who follow you around to "correct" all the "mistakes" they think you're making. I suspect this means you've never had the pleasure of dealing with a, err, "fan".
    Neither of those weirdoes is a BURDEN situation.
    BURDEN is for dealing with content. You find an unsourced statement, you can't verify it, so you tag or remove it depending on how certain you are. It's not for dealing with conduct.
    If you find an editor who's adding stream-of-consciousness material to an article, then it isn't your job to revert them. What you should do is find a sysop and ask them to intervene. We have sysops who're elected to deal with conduct issues. You're not one.
    If you ever have a "fan" who's going around vexatiously tagging or removing everything you've ever written, then it isn't your job to add citations for all that content. Do not reward low-effort behaviour by doing a huge quantity of hard work because they'll only go on to target someone else. Find a sysop and ask them to intervene.
    BURDEN is for content. If your problem is editor conduct, sysop or AN/I.—S Marshall T/C 17:56, 1 October 2025 (UTC)[reply]
    I've picked up a fan or two, thankfully they're not very active. But I never mentioned conduct. Checking other edits by someone who has made a mistake is not harassment and is allowed, and that could include situations that involve BURDEN. To say it's not allowed is just wrong -- LCU ActivelyDisinterested «@» °∆t° 18:58, 1 October 2025 (UTC)[reply]
    Checking their edits is allowed and nothing in this proposal would prevent that. What I'm proposing to advise against is mass-reverting, mass-tagging or mass-removing across several articles, which are behaviours already prohibited by the Universal Code of Conduct, paragraphs 3.1 and 3.3 (and I'm still cross about the imposition of those paragraphs on the community without a vote, but this is where we are now).
    If someone's edits need mass-reversion or mass-removal, then non-sysops should ask a sysop or start a discussion on a drama board, and not take it on themselves to begin a reversion crusade against another editor.—S Marshall T/C 19:14, 1 October 2025 (UTC)[reply]
    It's specifically says "Do not follow an editor from article to article using WP:BURDEN on their contributions." That's not a proposal against doing something, it's saying that you can't do it. There is no policy against undoing bad edits, and those could include situations where BURDEN applies. That would not be against the UCC, that would only apply if it was harassment and just undoing an edit is not immediately harassment.
    If anyone needs to seek sysop help it's someone being harassed, not someone making a change to content. Reverting an edit is normal editing practice, no matter how many edits it involves and so is not a matter that admins have jurisdiction over. -- LCU ActivelyDisinterested «@» °∆t° 22:28, 1 October 2025 (UTC)[reply]
    This word "proposal" is creating confusion. I'm proposing a rewrite; and the rewrite that I'm proposing says you can't follow people from article to article reverting or deleting their content.
    There is no policy against undoing bad edits, but the wording I suggest wouldn't stop you undoing a bad edit.
    What it would stop is WP:FAIT. Arbcom has rightly decided that you may not make a large number of edits that are difficult to reverse. There were good reasons why Arbcom decided that.
    WP:BURDEN edits are difficult to reverse. Finding sources is work.
    Therefore you may not use that power excessively, and you may not use it to target another editor.
    This is not a novel rule that I'm inventing. It's longstanding guidance.
    If you want to target an editor, ask a sysop or go to the drama boards. Do not take it on yourself to follow them around.—S Marshall T/C 23:11, 1 October 2025 (UTC)[reply]
    Maybe it needs to say just that: "Mass-reverting, mass-tagging or mass-removing across several articles may constitute harassment or disruptive editing". WhatamIdoing (talk) 02:49, 3 October 2025 (UTC)[reply]
    I agree with ActivelyDisinterested's comments on the second paragraph of Misuse, and I also think the third paragraph is too broad. If an article nominated for deletion is kept, that is based on the availability of sources, and not the state of the article. If an editor then thoroughly reviews those available sources and removes content they can't find a source for, that would be constructive. I think both the second and third paragraphs are trying to make summary judgments about too-specific situations that will depend on context.
    I really like something Blueboar said above, and think it could be incorporated: Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note giving all the reasons. However, the first step in resolving any challenge is for those who want the statement to remain to add an inline citation. Other challenges to the content may arise once a source is provided for review.--Trystan (talk) 14:51, 2 October 2025 (UTC)[reply]
    I want to rearrange from this:
    Policy
    (1) Each fact or claim must be verifiable. (2) The burden to prove verifiability lies with the editor who adds or restores material.
    (3) Facts or claims without an inline citation to a reliable source that directly supports them may be removed. (4) They should not be restored without an inline citation to a reliable source.
    (5) Never leave unsourced or poorly-sourced material that could damage the reputation of living people in an article. Do not move it to the talk page.
    Practice
    (6) A fact or claim is "verifiable" if a reliable source that supports it could be cited, even if there is no citation for it in the article at the moment.
    (7) The burden to prove verifiability is satisfied by adding an inline citation to a reliable source that directly supports the material. (8) The cited source must clearly support the fact or claim as presented in the article.
    to:
    1 6 2 7 8 3 4 5
    The most important part is that Sentence 6 is not a "practice". It's a definition and belongs to the "policy" subsection. WhatamIdoing (talk) 02:46, 3 October 2025 (UTC)[reply]

    Removing text

    [edit]

    There are several reasons why someone might remove text, including but not limited to:

    1. Because they think the text is unverifiable.
    2. Because it's an unsourced negative claim about a living person.
    3. Because it's a copyright violation.
    4. Because it's probably verifiable but it's in the wrong article.
    5. Because it's probably verifiable but it's UNDUE.
    6. Because it's written by an editor with a history of problematic contributions.

    WP:BURDEN is only concerned with case #1. All the other stuff is true, and needs to be said, and the place to say it is in the editing policy, and specifically in WP:CANTFIX. Not here.—S Marshall T/C 17:35, 2 October 2025 (UTC)[reply]

    In the interest of keeping BURDEN focused, perhaps the sentences Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note giving all the reasons. could come out. Many potential reasons for removal (WP:COPYVIO, WP:UNDUE, WP:PROPORTION, WP:SYNTH) cannot be meaningfully evaluated until a source has been provided, so perhaps it is a more complicated issue than can be addressed concisely in BURDEN.--Trystan (talk) 18:31, 2 October 2025 (UTC)[reply]
    Also, nobody does that "best practice" in practice. It's easier to demand that the editor WP:FETCH a source than to give them a laundry list of complex rules from the very beginning. WhatamIdoing (talk) 02:51, 3 October 2025 (UTC)[reply]
    I agree that games of FETCH are commonplace and that this is a powerful tool for POV pushers and crusaders who want to prevent articles being made neutral.—S Marshall T/C 07:47, 3 October 2025 (UTC)[reply]
    I have more often encountered the reverse: An editor adds unsourced content to make the article say Exactly What They Think It Should Say (EWTTISS). When challenged for a source, they cherry pick one that kind of, if you squint, supports the EWTTISS content. When challenged that the material is UNDUE, they accuse the other editor of moving goalposts.
    But the best version of any article comes from editors who start by reviewing a bunch of sources first and then neutrally describing what they say. Sometimes you can wrestle EWTTISS content into an approximation of that, but more often we end up paving the cow path.--Trystan (talk) 14:00, 3 October 2025 (UTC)[reply]
    There's a fine line at play here with what's mentioned above about moving ONUS, etc. under BURDEN (but not merging those sections). As long as BURDEN, ONUS, etc. aren't being merged it's not a problem, but WP:ONUS is definitely more of a multi-factor thing that can also invoke what you mention in 5 or other areas. Ultimately if there's a major issue identified regardless of which, then removing text and needing to get consensus under ONUS makes sense.
    If you're just zeroing on the role of BURDEN though, then it makes sense to say it only has to do with verifiability of text. Just thought it would be good to get that out there so wires aren't crossed on the subjects. KoA (talk) 18:59, 2 October 2025 (UTC)[reply]
    I would agree with this. BURDEN is about the needs to source contested content, while ONUS is stating that just because something can be verified doesn't mean it must be included. BURDEN is just about verification, while ONUS is about verification not trumping BLP and other concerns. -- LCU ActivelyDisinterested «@» °∆t° 10:02, 3 October 2025 (UTC)[reply]
    Right. So given everything that's been said here, we might actually be coalescing on a proposal to revise the editing policy as well.—S Marshall T/C 10:23, 3 October 2025 (UTC)[reply]
    The only part currently in BURDEN that's not V related is the BLP warning. I'm not against keeping that, or something similar, but nothing else is needed. Maybe something could be added to ONUS to clarify that verification is only a concern among many, it would be a better fit there.
    Having something on BLP could be helpful as there are many BLP concerns that are stricter than just V. Similarly both WP:SPS and WP:PRIMARY have cautions around BLP. -- LCU ActivelyDisinterested «@» °∆t° 10:39, 3 October 2025 (UTC)[reply]
    Just a nit-pick… I think that people are using “ONUS” when they mean “VNOT”. While both point to the same section, each shortcut points to a different part of that section. I think there is a solid consensus on VNOT, while the consensus on ONUS is a bit more… mixed. Blueboar (talk) 12:24, 3 October 2025 (UTC)[reply]
    I agree: This sounds like VNOT. ONUS is "the other guy always has to start the discussion". VNOT is "I don't care how many sources you have, that ___ doesn't belong in this article". WhatamIdoing (talk) 20:04, 4 October 2025 (UTC)[reply]
    I think trying to make it so boxed in is problematic. Let's say you don't understand an uncited sentence or why it is there, and you remove it as "uncited". Coming down on you as not using the magic words, or a completely narrow rubric, does not make the article any better. After all, why would a source exist for un-understandable information. The remedy in any event is the same, someone provides the source that fits it into the topic and conforms the wording to that source. You may still not understand it, but at least now you have been given the tools to try to, and to analyse it under other policies and guidelines. -- Alanscottwalker (talk) 16:04, 4 October 2025 (UTC)[reply]
    Please don't go around deleting things you don't understand. Please.—S Marshall T/C 17:26, 4 October 2025 (UTC)[reply]
    What an odd request, but I think I must decline. Perhaps you have a different view of what understanding may entail. But I see no reason to abdicate editorial functions. It has always been part of the editors job to make articles understandable. (Remember when someone suggested, articles should be made understandable to the teenager in Delhi.) Alanscottwalker (talk) 17:52, 4 October 2025 (UTC)[reply]
    They should be, but it's a long jump from there to removing text you don't understand. If you don't understand something, then instead of deleting it, please ask someone who does.—S Marshall T/C 18:23, 4 October 2025 (UTC)[reply]
    You appear to be working under a different assumption about what is and isn't understood. Moreover, shortening articles can make them more understandable. And editors regularly make edits without talking to someone else, that is not going to change. -- Alanscottwalker (talk) 18:34, 4 October 2025 (UTC)[reply]
    One doesn't want an editor to say they can't understand an article on a technical subject, and so to blank the whole thing, or any individual sentences.
    One does want editors to say that they can't understand the nonsense some vandals stick in articles, and so to blank the nonsense.
    I don't think that "I do/don't understand this" is a useful model for describing desirable edits to articles. WhatamIdoing (talk) 20:27, 4 October 2025 (UTC)[reply]
    My op did not suggest they don't understand the topic, it was limited to a sentence. A sentence does not a topic make, usually. Sure, we expect editors to be competent, perhaps we assume them to be, but if they are overly incompetent that is usually addressed by behavioral processes. Understanding by the reader is the purpose of our sentences, it is the major thing we want in what is written in our articles. -- Alanscottwalker (talk) 20:42, 4 October 2025 (UTC)[reply]
    I do know what you mean. I would love to be able to go through every word ever written by QuackGuru and remove the bits I don't understand. He wrote a lot of content, and a great deal of it was incoherent, semi-comprehensible word salad. I don't think he is or ever was competent to edit, particularly in the medical articles he usually inhabited, and I have cause to be very suspicious of his text. But with each individual sentence, there's a good chance it's verifiable. (WP:CHERRYPICKING is prominent among his issues.)
    But the problem with removing uncited text is that there are lots of reasons why you might not understand something. It could be perfectly accurate and verifiable, but written by someone with such limited English that you can't parse their intended meaning. It could have been perfectly accurate and verifiable until a key clause got cut when someone fat-fingered their edit. Or it could be that you're a humanities graduate who's stumbled into an article on probability calculus, or a science graduate who's stumbled into an article on the trochaic septenarius.
    We obviously want editors to be able to remove nonsense vandalism without a whole lot of process, but policy cannot say, imply, or suggest that any editor is free to remove content they don't understand. Insert the slippery slope analogy of your choice here.
    What makes this even more complex is that some things can be hard to understand, false, unverifiable, and still shouldn't be removed. We have an article on bigfoot, for example. Or consider the text:
    "Monsieur! , donc Dieu existe–repondez!"
    French: Sir! , therefore God exists–reply!
    This is incomprehensible in lots of ways. It's in a foreign language; it's not true; it doesn't follow; and Leonhard Euler didn't say it to Denis Diderot. But there have been academic papers (plural) about the fact that Leonhard Euler didn't say it to Denis Diderot.—S Marshall T/C 21:39, 4 October 2025 (UTC)[reply]
    What the? But it is not that hard. If an editor in their considered judgement thinks this does not make sense, they should feel free to edit it, including edit it out. We can't assume the first editor was competent but the second editor is not (on the other hand, with good sourcing, which yes, I understand is in part domain knowledge, more assurance is given to other good editors, to shrug and say 'ok, yeah, maybe' let that slide). As I have said before, I think, perhaps the worst problem with text not clearly following sources, is the 'almost right' or 'kind of right' text, but to follow sources we need the sources (say where you got it). Then too, I have silently asked, as you probably have silently asked, while reading this or that article, 'has anyone ever read this?' Alanscottwalker (talk) 22:42, 4 October 2025 (UTC)[reply]

    Another rewrite

    [edit]

    Page diff between existing and new text

    Draft is here: User:S Marshall/Policy rewrite 2025. Note that this reorganizes the policy to address some of the concerns above, and (hopefully uncontroversially) rewrites a couple of the clunkiest sentences in other parts of the policy. I do think this touches on WP:PRESERVE and WP:CANTFIX, so probably leads to consequential rewrites of WP:EP as well. It's worthwhile, though.—S Marshall T/C 13:19, 3 October 2025 (UTC)[reply]

    Without reading the previous discussion fully, and with thanks for putting the effort into a rewrite, some feedback
    • I don't understand the reference to wookie. This might be introducing more inside baseball into policy.
    • "To avoid this, when using WP:BURDEN, best practice is to say in an edit summary and/or talk page note that you have removed the text because you don't think it's verifiable." --> Usually, verifiability is not the main reason you delete uncited text. It's usually because it's unhelpful in a combination (unorganised, partially duplicative, not quite neutral). This puts too much of a burden on people cleaning up articles. Maybe the 'when using BURDEN' implies it's purely about verifiability, but then still, I'd like to say that we should reflect current practice. That is: content is removed when people are not sure it's verifiability (like 50/50), not when they believe it's unverifiable.
    • Where you propose to remove large quantities of material, it's best to start a discussion first. --> I don't think this reflects current practice. Usually when there are large quantities of unsourced text, it's from single editor and there are more problems with it than WP:V. Let's weaken this to: 'consider starting a discussion first'
    • 'Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note giving all the reasons.' -> Again puts more burden on people cleaning up articles. Unless it's been recently added by an account who is likely to read the talk page, adding talk page notes for obvious things all the time is disruptive as it pollutes watchlists.
    It seems the new text is 300 words longer than the old. Is that a necessary expansion? —Femke 🐦 (talk) 07:20, 4 October 2025 (UTC)[reply]
    • Thank you for this feedback. The wookiee reference is quite inside baseball and might well have to go.
    • The business about explaining when you remove text is not novel. It's longstanding policy, but is buried so deep in the wordy gobbledygook of paragraph #4 of WP:BURDEN that even longstanding admins like yourself don't know it's there. Once I've got it rewritten in plain English we'll have an opportunity to discuss whether it should be there.
    • The business about removing large quantities of material has its roots cases like Lugnuts, who generated tens of thousands of biographies all sourced to online databases published by, err, who, exactly? As you'll likely recall, mass use of WP:BURDEN on those articles is highly drama-genic. Personally, I would love to go through mass-removing unsourced text from the WP:LUGSTUBS. But the consensus is that I should talk first, so policy needs to reflect that.
    • The perfect policy would make cleaning up after vandals and idiots very easy, and going on deletionist crusades against popular editors very hard. Wording suggestions to achieve that are welcome?
    • Yes, this text is longer. Mostly that's because it brings up footnotes into the body text, but partly it's because of growing consensus on this talk page about adding the new section about removing text. I'm happy to carry on working to reduce word count.—S Marshall T/C 09:11, 4 October 2025 (UTC)[reply]
      • I do know it's there, and I also know it does not reflect how people actually edit. I came to this talk page because I wanted to ensure this text of the policy reflects actual practice. By rewriting something into understandable English, you do make it more prominent.
      • The ambiguity may be in 'large-scale'. Removing half an article because it's unsourced and questionable should happen boldly. Doing this with fanfare on the talk page, repeatedly, can be disruptive. When we go to massive scales, a prior-consensus based approach is more suitable.
      • The more policies the text violates, the less need there is to put anything on talk. Because it'll be bloody obvious why it's deleted. My preference is that we leave this sentence out.
      I'm happy to see the text on reasons for removing cruft. Let's see if we can adopt a one-in one-out principle and keep the overall text roughly the same? (I've done some housekeeping in the policy which loses some words at least). —Femke 🐦 (talk) 09:23, 4 October 2025 (UTC)[reply]
      S Marshall's Law of Editing Disputes reads: When you're arguing about content, use the talk page.
      Policy is mainly concerned with resolving disputes. Most content removals are uncontroversial, and nobody's going to drag you to AN/I for doing it, so policy never gets involved; but if you're removing text and editors object, then both you and they need to go to the talk page. Not edit summaries. Not each other's talk pages. The actual talk page about the content that's actually disputed. And not anywhere else.
      In an editing dispute you need to give your reasons in full. Citing policy is helpful because it lets you explain a lot of thought with a few characters.
      BURDEN and ONUS combine to give the removing editor the upper hand, and we've seen a number of examples of removing editors who stonewall and won't use the talk page because ignoring it helps them win the dispute. I think it's right that in these disputes, policy should force the removing editor to be the one who starts the talk page discussion.
      I'm very aligned with you on the need to keep the policy as brief as possible.—S Marshall T/C 13:56, 4 October 2025 (UTC)[reply]
      Of course when there's a dispute, go to the talk page. Almost always when I'm removing uncited text, I'm removing text that's been placed there ages ago, by folks long gone from Wikipedia. Maybe that's different for others, but in these cases, distracting people will talk page messages is not respectful of other people's time, and I've seen it lead to conflict when someone does it repeatedly. While policy is mostly cited in disputes, it's read and internalised for normal editing as well, as should be written such that it works outside of disputes too. —Femke 🐦 (talk) 14:09, 4 October 2025 (UTC)[reply]
      Well, let's imagine that. Imagine if policy said: You can remove anything without a citation and you don't have to give a reason. If someone objects, the onus is on them to achieve consensus to include it. Do you think that's good policy?—S Marshall T/C 14:41, 4 October 2025 (UTC)[reply]
      Nope, I'm arguing that usually an edit summary suffices with your reasoning. Only when someone objects do you work together on finding a consensus. —Femke 🐦 (talk) 14:45, 4 October 2025 (UTC)[reply]
      That's fair. Is it worth the extra word count to make it Some text violates more than one policy. In that case, when removing it, best practice is to add a talk page note [ins]or edit summary[/ins] giving all the reasons?—S Marshall T/C 17:23, 4 October 2025 (UTC)[reply]
      Even there, I'm don't think that's best practice. I will often delete copyvio that also breaches WP:V, neutrality, and SUMMARY, but will be kind to fellow editors and just name the one or two most relevant policies. The key is that you're expected to explain. How much really depends on the situation, with only a few where best practice is to be somewhat exhaustive. —Femke 🐦 (talk) 19:04, 4 October 2025 (UTC)[reply]
    • If we're going to keep WOOKIEE (which I'd like to), then @Blueboar needs to write Wikipedia:Let the Wookiee win, so we can link it in the sentence at the end.
    • @Femke, I disagree that accurate explanations for edits put "too much of a burden on people cleaning up articles". If you're removing text that would need to be removed even if a reliable source is provided, then you should type something like copyedit or correction instead of uncited in the edit summary. If you don't want to spare anyone's feelings, then Rm bad covers a lot of territory in just six keystrokes. BURDEN is supposed to be entirely about verifiability. If you want to remove content for any other reason, then give another reason. BURDEN means "WP:FETCH me a source, or you can't have your content back". It should not be used when you're thinking "What a hopeless mess."
    • The "business about removing large quantities of material" has nothing to do with LUGNUTS, whose articles almost never had large quantities in the first place, and so couldn't possibly have large quantities removed from them. It has everything to do with edits like this one, in which about two-thirds of an article, apparently all good information – including sentences like "A subsequent section explains common variations", which never needs a citation – was blanked because an editor spammed fact tags throughout, and then, since no editor dropped everything to supply him with 85(!) inline citations, blanked everything uncited 30 days later.
    WhatamIdoing (talk) 20:51, 4 October 2025 (UTC)[reply]
    Faced with a hopeless mess of an edit, you have four potential options for a revert summary: none at all, "rm bad", "rm unsourced, undue, promotional, unencyclopedic, poorly written, etc etc etc", or "rm unsourced". Each has drawbacks. "rm bad" is easy, and may be accurate, but also doesn't give a good-faith editor any insight into why it's bad; it's a shortcut for a reverter, but probably not any shorter or easier for the adder than the FETCH scenario. Nikkimaria (talk) 22:21, 4 October 2025 (UTC)[reply]
    I'm not arguing against providing accurate edit summaries, I'm arguing against prescribing what that good edit summary should look like. It's rarely helpful to give a long laundry list of all the subtle mistakes someone has made. You want to note those things that are most important to remedy, while perhaps helping with other PAGs once someone has fixed the key issues. So rm unsourced and WP:UNDUE is often a better edit summary than rm unsourced, undue, unencyclopedic, poorly written. That edit summary might reflect all the problems, but comes across as mean. —Femke 🐦 (talk) 07:44, 5 October 2025 (UTC)[reply]
    I think the key point is this: Don't use an edit summary of (only) Rm per WP:BURDEN if you'd re-revert even if a reliable source is added. That's mean to the good-faith editor ("but you said all I had to do was cite one reliable source!") and stretches BURDEN beyond its narrow remit. WhatamIdoing (talk) 19:24, 7 October 2025 (UTC)[reply]
    But an editor asking for a source under BURDEN should keep an open mind as to what sourcing is going to be provided. If an editor familiar with a topic suspects something is unverifiable, a cherry-picked source might establish it as minimally verifiable, but still UNDUE, while a few strong sources would establish it as DUE. The challenging editor can't meaningfully evaluate that until the editor supporting the material shares their sources. The editor restoring the material should be corrected from any misapprehension that the sourcing they add won't itself be subject to scrutiny.--Trystan (talk) 13:59, 10 October 2025 (UTC)[reply]
    Sometimes that's true, but the more common problem, especially in WP:CTOPS and WP:FRINGE subjects, is that the editor asking for a source under BURDEN knows full well that it's UNDUE. "Oh, you want your nonsense about HIV not causing AIDS? Well, gimme a source..." WhatamIdoing (talk) 16:01, 14 October 2025 (UTC)[reply]
    Looking at the diff, the policy has been edited recently, and some of the changes are uncontroversial (e.g., rearranging). To simplify the diff for an eventual RFC, I think we should boldly make the smaller/uncontroversial changes in both the policy and the draft. WhatamIdoing (talk) 20:34, 4 October 2025 (UTC)[reply]
    Feel free. Although the draft is in my userspace, I trust you (plural, meaning all of you) to edit it.—S Marshall T/C 23:10, 4 October 2025 (UTC)[reply]

    Yet another rewrite

    [edit]

    Now up at User:S Marshall/Policy rewrite 2025.—S Marshall T/C 12:20, 7 October 2025 (UTC)[reply]

    • Misuse #3 looks better, I fear it will still be misused but editors misusing policy isn't something new.
      Misuse #2 needs work. It currently says that if there is unverifiable or just plain wrong content in an article you can't remove it if you've nominated an article for AfD. That's just a weird rule to make up.
      Misuse #4 has similar problems, if someone one mass adds unverifiable content to an article this rule says you can't remove it. So fait accompli addition can't be undone, because removing it would be fait accompli.
      I like 'Other reasons for removing text', but feel it should be under, or reworded to be part of, 'Verifiability does not guarantee inclusion' as that seems more appropriate.
      Are there any other areas that have been changed, would it be possible to have a side by side of the changes? -- LCU ActivelyDisinterested «@» °∆t° 13:58, 7 October 2025 (UTC)[reply]
    Here you go.—S Marshall T/C 14:36, 7 October 2025 (UTC) < Link to Diff page. Butwhatdoiknow (talk) 15:22, 7 October 2025 (UTC)[reply]
    I support all the other minor changes, they all serve to simplify and clarify. -- LCU ActivelyDisinterested «@» °∆t° 15:50, 7 October 2025 (UTC)[reply]
    I made some of the minor changes. I didn't make some because they're not minor, others because I prefer the old wording, and still others because I just decided to stop. But the diff should be a bit simpler now, and maybe someone else would like to have a turn at either fixing User:S Marshall/Policy rewrite 2025 to match WP:V or WP:V to match it. WhatamIdoing (talk) 21:38, 7 October 2025 (UTC)[reply]
    I'm wondering if the language in the 'Misuse' section should move alway from "Do not use". Take for instance Misuse #1 this is a prohibition from using BURDEN to push a POV, which should already be covered by policies on conduct. Also someone using BURDEN could, in good faith, remove content in a way that means the article is no longer compliant with NPOV. It would be more useful to have a statement that when you remove content from an article, because of verification issues, you should take care not to skew the article.
    Similar issue apply with the other Misuse entries. For instance "don't harash people" isn't a verification issue, but "Take care that following editors to correct verification issues can be seen as (?may feel like?) harassment" could be a useful warning. Usually "should not" is used instead of "must not". -- LCU ActivelyDisinterested «@» °∆t° 16:33, 7 October 2025 (UTC)[reply]
    Policies and guidelines should use "do not" and "must not" whenever the prohibition is nearly universal. For example, if you blank most of an article and then take it to AFD, you might get accused of Wikipedia:Gaming the system, and you might make it harder for editors to find sources (e.g., by removing uncited information that could prompt better keyword searches, or that might suggest the existence of particular places to look for sources). That's bad behavior that will make AFD less fair. So we say: Don't do that. "Should" language in a context like that ends up sounding like "It might be nice if you didn't behave in unfair and drama-producing ways, but really, being honest in your interactions with the rest of the community is an optional nicety, and especially if you think you can get away with it, then blanking most of the article might increase your chance of WP:WINNING the AFD battle, because there's usually someone who will vote against any WP:UGLY stub just because it's short". WhatamIdoing (talk) 22:40, 7 October 2025 (UTC)[reply]
    Gutting an article before nominating it for deletion is a pain in the arse for AfD contributors. Circumstances exist where it should happen, but just being unsourced (without more) isn't it. You certainly can and should gut it if it's a hoax, an attack piece, an advertisement, a close paraphrase or plagiarism. You shouldn't gut it if it's just a piece of well-meant but poorly-sourced drivel, and editors who do that are usually overexcited.—S Marshall T/C 09:01, 8 October 2025 (UTC)[reply]
    But as I've said these shouldn't be prohibitions, and make no sense as prohibitions. -- LCU ActivelyDisinterested «@» °∆t° 16:57, 12 October 2025 (UTC)[reply]
    Just to check the obvious: Do you think there really are situations when an editor actually should use BURDEN as a tool for violating the Wikipedia:Harassment policy? Like, you need to harass someone out of the community, and BURDEN would give you plausible deniability and therefore a way to wriggle out of sanctions against yourself, and you can't get that terrible person out through other ways, but (you believe) the community will overall be better without them, so the ends (getting rid of a Bad™ person) justify the means (harassing someone)? WhatamIdoing (talk) 17:10, 14 October 2025 (UTC)[reply]
    As a different rhetorical question could you tell me a situation where that's a verification issue and not a behavioural one, or should we add an addendum to every piece of policy and guidance that you shouldn't use it for harassment? That's also, since it was cut back, the least problematic of these new prohibitions. It would be a good idea to take these to VPP to ensure the community does support them. -- LCU ActivelyDisinterested «@» °∆t° 19:54, 15 October 2025 (UTC)[reply]
    We've seen BURDEN used in ways that target individual editors and that make them feel harassed enough to complain about harassment. We haven't seen, e.g., the Wikipedia:Editing policy or even most other parts of this policy used to harass people. Therefore, while we might not need to "add an addendum to every piece of policy and guidance that you shouldn't use it for harassment", it's at least plausible that we would benefit from doing it for this particular point in this particular policy. WhatamIdoing (talk) 03:08, 16 October 2025 (UTC)[reply]
    I agree with AD that there's a big asymmetry in Wikipedia where you're allowed to disruptively mass-add poor quality content but mass-removing it makes people very cross. Certainly in my experience when someone has mass-added a load of nonsense that shouldn't be there, getting rid of it is quite needlessly painful.
    In writing that, I was trying to describe practice as I've observed it, rather than practice as I want it to be. On reflection I would personally prefer it if it was easier to mass-remove content that never had consensus to add in the firstplace.
    What are people's thoughts? Does that bullet need to go?—S Marshall T/C 09:11, 8 October 2025 (UTC)[reply]
    For me, that's the main point, conlevel when added, did it just slip through relatively unnoticed (bit like ONUS ha). Selfstudier (talk) 09:17, 8 October 2025 (UTC)[reply]
    Right. It should not be hard to edit, including edit out. This is a wiki, and no OWNership aloud, improvement welcome. I think there may have been a time (a long time ago) when there may have been a 'we want anything' written vibe, but maybe that was when we did not know what we were, or had, or wanted. Well now, we want good, policy compliant text. And, expertise welcome, but that is expertise that proves itself on the page, so that others will stand up and protect that (or not mess with that), because it is written and presented in such a way that shows/demonstrates itself worth protecting. -- Alanscottwalker (talk) 22:10, 12 October 2025 (UTC)[reply]
    I wonder how many editors these days are familiar with what the Wikipedia:Editing policy say about the value of adding vs removing information:
    Wikipedia summarizes accepted knowledge. As a rule, the more accepted knowledge it contains, the better. Please be bold and add content summarizing accepted knowledge, but be particularly cautious about removing sourced content.
    This has been official written policy for 15 years, and yes, it was duly discussed on the policy's talk page. I don't think that a cavalier or reckless approach to removing good/encyclopedic "accepted knowledge", especially over a WP:SURMOUNTABLE problem like nobody having yet added sources, especially when those sources could be found easily, is consistent with this policy. On the opposite end of the spectrum, keeping non-encyclopedic stuff (including but not limited to: gossip, trivia, errors, excessive/non-concise detail, advertisements, spam, misinformation, bad jokes, persuasive writing, creative writing, things made up one day...) is also not consistent with that policy (or any other).
    If you're interested in how policies change over time, then it can be fun to look back on a policy as it stood on the day you created your account. Alan, for you, that would be this version of the Editing policy. This sentence particularly stands out to me: "Rather than being horrified by this ugliness, we should rejoice in its potential, and have faith that the editing process will turn it into brilliant prose." WhatamIdoing (talk) 17:54, 14 October 2025 (UTC)[reply]
    We should see nothing wrong with being 'horrified' (or 'loved', for that matter), and never should have. It means someone actually cares about what's there. And since no one has proposed being 'cavalier' or reckless that's a strawman. The only thing 'cavalier' would be the idea that just because it's written there, it can't be touched, ignore it, some process will happen. Alanscottwalker (talk) 21:29, 14 October 2025 (UTC)[reply]
    Sorry I still object to this and codnt see that there's consensus for making the change, so I'm going to revert the latest changes. -- LCU ActivelyDisinterested «@» °∆t° 16:55, 12 October 2025 (UTC)[reply]
    All right. I've offered you the chance to edit the draft, I've asked you for your suggested wording, and I'm getting nothing. I'm not willing to play a game of "bring me a better rewrite" forever. If you'd like to suggest revisions that work for you, go ahead. Failing that I'll just proceed to RfC.—S Marshall T/C 21:41, 12 October 2025 (UTC)[reply]
    I've offered my side earlier, my suggestion is to not to included de the Do not" language as it's logicaally inconsistent and that we should "should not" language. The response has been that "do not" must be used! If you're unwilling to take onboard my suggestions is one thing but to claim I haven'y made any is false. I didn't realise you would be rushing to make the change without further community involvement, especially as you creating new policy. -- LCU ActivelyDisinterested «@» °∆t° 22:17, 12 October 2025 (UTC)[reply]
    There's obviously been massive amounts of community involvement. These are some of the longest talk page threads we've ever had on WT:V, and everyone interested has had plenty of opportunity for input.
    You are the lone remaining objector and your remaining objection is poorly thought through. Please allow the policy to be edited.—S Marshall T/C 07:53, 13 October 2025 (UTC)[reply]
    I'm sorry I don't see the consensus you do during these discussions. What your doing is WP:CREEP, you should find a better consensus before enforcing new restrictions on editors. -- LCU ActivelyDisinterested «@» °∆t° 14:00, 13 October 2025 (UTC)[reply]
    The sticking point really is the misuse section, which I could agree to if it didn't create new prohibitions. What about implementation the other parts and taking the misuse section to VPP to see if the community does support it's addition? -- LCU ActivelyDisinterested «@» °∆t° 13:07, 14 October 2025 (UTC)[reply]
    • Do you see any other editor actively agreeing with you? (If not, then User:Guy Macon/One against many might apply.)
    • What makes you believe these are new prohibitions? Being a new prohibition would mean, e.g., that you believe that harassment or hounding via BURDEN is actually totally 'legal' and accepted right now.
    WhatamIdoing (talk) 17:58, 14 October 2025 (UTC)[reply]
    I never said it was 'legal' to harass anyone using any particular part of any policy. That's an entirely different question than adding a new prohibition to this policy that this policy can't be used for harassment. This is the policy for variability, not the one for editors behaviour. -- LCU ActivelyDisinterested «@» °∆t° 19:58, 15 October 2025 (UTC)[reply]
    ActivelyDisinterested does have a point here. We do have to guard against CREAP, and avoid the temptation to overload this policy with instructions that would be more appropriately covered in other p&gs. Blueboar (talk) 20:30, 15 October 2025 (UTC)[reply]
    Everyone editing this page ought to know that you can't harass people. That's been behavioural policy for nearly twenty years. I'm sure we can all think of examples of bad faith users hounding people with reversions and removals and citation needed tags, and that's not new either. What's changed recently is that the anti-harassment rule has become part of the Universal Code of Conduct. If you'd like to see how the UCoC is being used in practice, please look at this utterly soul-destroying page.
    (Yes, that's largely trolls and wikipedia-haters trying to weaponize the UCoC against us. During the implementation phase I did tell the WMF loud and clear that the UCoC shouldn't be imposed on the en.wiki community until we'd voted to ratify it, but they just couldn't quite seem to hear that. Next time I'll try typing slower.)
    The UCoC means we have to have clearer documentation about how our behavioural policies and our content policies interact. There are people who will try to trick you into following them around removing their edits so they can complain about you. We have to place that documentation where good faith content contributors will see it and remember to think about it. It's no good putting it in a behavioural policy because so many of the good faith content contributors who'll get caught up in this don't look at behavioural policies.
    The very brief and succinct pointer that we're adding is fully justified in the circumstances.—S Marshall T/C 22:37, 15 October 2025 (UTC)[reply]
    What? The problem you're describing in the UCoC context is the trickster, not the good-faith contributor. The good-faith contributor is not helped by handing the trickster another cudgel. Nikkimaria (talk) 23:19, 15 October 2025 (UTC)[reply]
    But I'm not. I'm pointing at the cudgel the trickster is already holding and saying: "Be careful of that!"—S Marshall T/C 23:56, 15 October 2025 (UTC)[reply]
    I don't agree with that interpretation. In my view, we'd be introducing something new for a bad-faith "victim" to point to, without the nuance that the original behavioural policies have. Nikkimaria (talk) 00:04, 16 October 2025 (UTC)[reply]
    The relevant bit of the UCoC, which is section 3.3, paragraph 1, is not nuanced at all. It's horribly simple.—S Marshall T/C 00:45, 16 October 2025 (UTC)[reply]
    AD, you didn't say that it was 'legal' to harass people, but you did say that this would "create new prohibitions" against harassing editors. If it's already 'illegal' to harass editors, then it's not a "new" prohibition, right? I'd understand you saying "This sentence needlessly duplicates the existing prohibition on this behavior, as seen in the WP:Harassment policy, so we can omit it", but your objection was that it was creating a new prohibition, i.e., prohibiting something that is not currently prohibited. WhatamIdoing (talk) 03:10, 16 October 2025 (UTC)[reply]
    So, is the proposal to put in this policy, "Do not use this policy for WP:HARRASSMENT"? Maybe that's ok, but it should be in all the policies then. But I think perhaps you're not being that clear, and the problem starts with suggesting an argument that relying on this specific policy is harassment. And that would be quite a problem because asking for sources to be identified is essential to most of what we do. -- Alanscottwalker (talk) 11:55, 16 October 2025 (UTC)[reply]
    Or maybe it's okay, but should only be in those parts of our policies/guidelines/advice pages that have a clear history of being misused for harassment.
    It would make more sense to say, e.g., that WP:EL has a history of being misused to violate WP:LINKVIO, so mention LINKVIO in WP:EL (but not HARASS, because editors tend not to make that mistake in ELs); that WP:CHALLENGE has a history of being misused to violate WP:HARASS, so mention HARASS in CHALLENGE (but not LINKVIO, because editors tend not to make that mistake in CHALLENGEs), etc., than to say that all pages everywhere must mention everything because someone might use Page A to violate Rule B. WhatamIdoing (talk) 20:19, 16 October 2025 (UTC)[reply]
    Since WP:EL is no policy, that's beside the point. Indeed, you appear to have missed the question and the follow-on point, entirely. If you are not proposing a clear statement in this policy (ie, "Do not use this policy for WP:HARASSMENT"), then it is an unclear. And that's a problem for the reasons I set out. Alanscottwalker (talk) 02:04, 17 October 2025 (UTC)[reply]
    That one specific entry would add, rather than create, a new prohibition in this policy. Quibbling over the exact wording of a comment isn't helpful, my original comment was about the whole section and so may not have been specific enough about one particular detail. -- LCU ActivelyDisinterested «@» °∆t° 13:14, 16 October 2025 (UTC)[reply]
    [edit]

    @Femke:, the sentence that you removed in this edit was added as a result of Wikipedia_talk:Verifiability/Archive_86#citations_-_Are_we_over_thinking_this?_New_proposal. Perhaps the discussion on whether it's needed should be revived? Schazjmd (talk) 21:36, 3 October 2025 (UTC)[reply]

    I've never seen this come up myself, so I feel it's WP:CREEP. The previous paragraph already implies this:
    Each fact or claim in an article must be verifiable. Additionally, four types of information must be accompanied by an inline citation to a reliable source that directly supports the material: —Femke 🐦 (talk) 06:53, 4 October 2025 (UTC)[reply]
    @Femke, would you still feel that the same implication were present, if that paragraph included an accurate definition of verifiABLE?
    "Each fact or claim in an article must be verifiable. A fact or claim is "verifiable" if a reliable source that supports it could be cited, even if there is no citation for it in the article at the moment. Additionally, four types of information must be accompanied by an inline citation to a reliable source that directly supports the material..."
    Part of the problem is lists. Imagine that I create a List of presidents of the United States, and all it says is an introductory sentence or two and the names of the presidents (as the first non-redirect revision of that page actually did):
    etc. Someone comes along and complains that there isn't a single Wikipedia:Inline citation supporting any of these links. I say instead that anyone who clicks the link will find ample sources, and if the complainer wants the sources copied over, then the complainer can click through to the linked articles and copy them himself. What do you say? WhatamIdoing (talk) 20:00, 4 October 2025 (UTC)[reply]
    So, this wouldn't be captured by the sentence I removed, right? It says "When material that needs an inline citation appears in two or more articles, an inline citation is needed in each." As it's not BLP, nor likely to be challenged, this doesn't help somebody who wants a nicely-cited list.
    For material that does fall under the four bullet points, fair point. —Femke 🐦 (talk) 20:45, 4 October 2025 (UTC)[reply]
    I think this would be captured, because as soon as someone complains that there isn't a source listed, it's been CHALLENGED and therefore has a 100% chance of being Wikipedia:Likely to be challenged. That's two of the four bullet points.
    The immediate impetus for this addition was editors discovering that the List of common misconceptions had been reformatted to hide almost all the refs in subpages. You could easily click through to find all the refs, but this is not good enough. WhatamIdoing (talk) 20:58, 4 October 2025 (UTC)[reply]
    I have come across this situation frequently. I even came across a case where someone was arguing that a bit of information was cited “in the other” article, and didn’t realize that the “other article” had been completely re-written… and in the process both the information and its citation had been cut. That meant that the information was no longer cited in any article. Blueboar (talk) 21:24, 4 October 2025 (UTC)[reply]
    I'm convinced. I'll look elsewhere in my quest to make our PAGs a bit shorter and easier to digest. —Femke 🐦 (talk) 21:30, 4 October 2025 (UTC)[reply]
    I wish you lots of luck with that, but I'm reminded about a teacher discussing the length of California's Education Code. He said, "Every line in there is because somebody had a problem." I suspect that the same is true for our policies and guidelines: Every duplication is because someone expected to find something "here", but it was "there", so they copied it "here". Every instance of verbosity is because someone didn't understand the shorter version. Every time we spell out something that should go without saying, it's because someone needed it spelled out. WhatamIdoing (talk) 21:57, 4 October 2025 (UTC)[reply]
    It's certainly a trade-off. The longer text gets, the more likely it is it will only be partially read or digested. But some repetition is useful. So far, when dealing with repetition, I've mostly removed things that are very close to each other, where the repetition is more likely to confuse (as it's not clear they link to the same page), rather than clarify. Small steps. —Femke 🐦 (talk) 22:07, 4 October 2025 (UTC)[reply]
    I agree. WhatamIdoing (talk) 22:18, 4 October 2025 (UTC)[reply]

    Fun

    [edit]

    I've been wondering whether a few colored boxes with at-a-glance summaries might help. For example, the SPS section might have a note about Einstein's blog:

    Self-published sources
    Shortcuts
    Further information: Wikipedia:Biographies of living persons § Avoid self-published sources, Wikipedia:List of companies engaged in the self-publishing business, and Wikipedia:Identifying and using self-published works
    "WP:WORDPRESS" redirects here. For the use of Wordpress as a source, see WP:RSPWORDPRESS.
    Einstein's blog

    If Albert Einstein had a blog about physics, you could cite it for facts about physics.

    Anyone can create a personal web page, self-publish a book, or claim to be an expert. Self-published material, such as books, patents, newsletters, personal websites, open wikis, personal or group blogs (as distinguished from newsblogs, above), content farms, podcasts, Internet forum postings, and social media postings, are largely not acceptable as sources. Self-published sources may be considered reliable if published by an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications. Be careful when using such sources: if the information in question is suitable for inclusion, someone else will likely have published it in independent, reliable sources. Never use self-published sources as third-party sources about living people, even if the author is an expert, well-known professional researcher, or writer.

    I think we could write several of these, and perhaps they would improve comprehension.

    Let the Wookiee win

    It's usually faster to find and cite a source than to argue about whether one is really needed

    What do you think? WhatamIdoing (talk) 22:28, 7 October 2025 (UTC)[reply]

    “Information” vs “facts and claims”

    [edit]

    Why the switch in language? (No real objection, just want to understand the thinking). Blueboar (talk) 00:47, 14 October 2025 (UTC)[reply]

    Apparently you're referring to this edit by WhatamIdoing. It looks bad to me, as I think of "information" as something that can be false and "fact" as something that by definition isn't. If we concede in advance that it's a fact then anything that says it is ipso facto reliable. Peter Gulutzan (talk) 01:19, 14 October 2025 (UTC)[reply]
    I don't feel strongly about this change. I changed it to "facts and claims" because that language is being used in the proposal above, and I thought it might be slightly less susceptible to wikilawyering ("I'm not adding information. I'm just adding a sentence about what he claimed").
    Peter, fact also means a statement whose accuracy can be determined, even if it is determined to be inaccurate. It is possible in that model to distinguish true facts (the kind you have in mind) from false facts (e.g., an honest mistake). This contrasts with claims, whose accuracy may not actually be determinable (e.g., whether a painting is 'art', whether this theory is plausible). Between the two, I think they cover everything that belongs in a Wikipedia article. WhatamIdoing (talk) 04:02, 14 October 2025 (UTC)[reply]
    Thanks Waid… I understand the intent now. Neutral on which is better. Blueboar (talk) 13:00, 14 October 2025 (UTC)[reply]
    I don't think this is a good or natural distinction, as evidenced in part by the set phrase "factual claims". Rollinginhisgrave (talk | contributions) 13:11, 15 October 2025 (UTC)[reply]
    Agreed. A fact is a verified claim that is no longer just an opinion. In principle, it's like a scientific theory, which, in principle, can be changed upon the presentation of better evidence, but otherwise, in daily speech, a "fact" is not just a "claim" subject to question without good evidence. It's already been through the questioning process and found to be true. So, to word it better: "fact also means a statement whose accuracy can be HAS ALREADY BEEN determined" -- Valjean (talk) (PING me) 02:06, 17 October 2025 (UTC)[reply]
    Not always. Consider: "The next US presidential election will be in 2028". This is a statement of fact and not a statement of opinion. But as it is a statement of fact about the future, it's impossible to have already determined its accuracy. There's a tiny chance that something will arise between now and then to make that forward-looking fact become inaccurate, but that doesn't make the date of the election an opinion (judgment, viewpoint, personal preference, etc.). The date of the event is still a fact. WhatamIdoing (talk) 02:43, 17 October 2025 (UTC)[reply]
    This sounds like Justified True Belief and the Gettier Problem. Is it a good idea to need to have such philosophical debate to understand a policy? To avoid having to put things in such metaphorical boxes is likely a reason why such a broad single term such as "information" was used. -- Alanscottwalker (talk) 11:37, 18 October 2025 (UTC)[reply]
    • I proposed the change after dealing with an editor who thought that individual words could fail verification if they didn't come from a source. Other cases, like Rlevse's, also suggest that some editors think you should build articles by copy pasting from sources, and I wanted wording that was clearer on what needs to be verifiable. Facts or claims, not words or sentences.—S Marshall T/C 13:34, 14 October 2025 (UTC)[reply]
    Got it. Again, I am neutral on the change. I just wanted to understand the thinking behind it. Thanks. Blueboar (talk) 13:53, 14 October 2025 (UTC)[reply]
    I guess I'm not seeing how it improves or really clarifies anything, and it loses some concision while becoming wordier to the point of feeling slightly legalese to me. Information includes facts and claims (and anything else really), so it seems like it's becomes more restrictive by making it a list instead. The long-standing version gets the point across really well that basically whatever (nonrestrictive) may be verifiable does not equate to automatic inclusion. KoA (talk) 15:31, 14 October 2025 (UTC)[reply]
    This being a wiki, neutral changes shouldn't be reverted, even if it's not an improvement in the eye of every beholder. But if you think it's actually harmful (e.g., due to loss of concision or legalese), then that's a different kettle of fish. As I said above, I thought it might be slightly better, but I don't feel it's harmful.
    @S Marshall, for the problem you mention, it might be more effective to add a sentence that directly says that editors are required to use Wikipedia:Use our own words because copyvios, close paraphrasing, and plagiarism are prohibited. We've had enough editors over the years claim that using synonyms violate verifiability that it's worth addressing directly. WhatamIdoing (talk) 15:56, 14 October 2025 (UTC)[reply]
    The first thing I noticed was how many times "facts and claims" was being repeated throughout the section to the point I did worry a bit about readability vs just plainly saying something, so yeah I didn't really view it as just a wash between versions, but also that there's enough question on talk for a policy text change that it's better to hold the change back. Things could change here and get explicit consensus too, but better to be on the safe side at this point. KoA (talk) 16:14, 14 October 2025 (UTC)[reply]
    I'd vote for waiting for someone to come up with something better. WhatamIdoing (talk) 16:25, 14 October 2025 (UTC)[reply]
    FWIW, both "information" and "facts or claims" (or a variant, like "fact or claim" or "facts and claims") appear several times throughout WP:V. "Information" is used more often, but the only latter is used in "This page in a nutshell." It also already appeared in the second sentence of the existing version of WP:VNOT. Another word that appears a lot in WP:V for this kind of thing is "content". FactOrOpinion (talk) 16:58, 14 October 2025 (UTC)[reply]
    "Material" gets used about twice as often as "information". WhatamIdoing (talk) 18:10, 14 October 2025 (UTC)[reply]
    KoA, this has been extensively discussed above. In our discussion, we wondered whether there were any examples of things that are "information", and have been tagged with {{cn}}, but are not "facts or claims". We've come up with:
    a) Individual words or sentences
    b) Pictures, photos and diagrams
    c) Sound and video files
    We've agreed that what needs to be cited is facts or claims, not "information".—S Marshall T/C 17:39, 14 October 2025 (UTC)[reply]
    I guess I'm not seeing what you're referring to outside of this section, and I have this page on my watchlist, so I do check in on some of the proposed edits when I can. I do have to say though that it's getting hard to follow who is suggesting what when though when going through the large talk section currently at the top of the page.
    With that said, your example you give in your top-level comment is a case where I'd say be wary of a certain sense of WP:INSTRUCTIONCREEP. We can't instruct our way out of people trying to wiki-lawyer, and this change really wouldn't have fixed that. In that case, it would be better to link to guidance on how we paraphrase sources. If more information may be useful on guidance about paraphrasing, a new sentence might be more helpful in clarification. I'm not sure if it would belong in this section of the policy or not though. KoA (talk) 22:13, 14 October 2025 (UTC)[reply]

    It's all interleaved with the extensive discussion in archives 86 and 87, which refers to the VPP thread that started this. Since you clearly haven't followed the discussion, would you please allow the policy to be edited? We don't have to satisfy you personally of the need for this. You're not the WP:V gatekeeper.—S Marshall T/C 23:03, 14 October 2025 (UTC)[reply]

    I think we have bigger fish to fry at the moment. Maybe we could leave this small change for another month? WhatamIdoing (talk) 01:58, 15 October 2025 (UTC)[reply]
    I thank KoA for reverting WhatamIdoing's bad without-consensus policy change. Peter Gulutzan (talk) 12:52, 20 October 2025 (UTC)[reply]
    Perhaps you should also thank ActivelyDisinterested for reverting a much larger change made by S Marshall that included this. WP:PARTR is generally considered good advice for figuring out whether someone objects to all of a large change, or only to part of it. WhatamIdoing (talk) 22:38, 20 October 2025 (UTC)[reply]