I stumbled on a very large number of high-speed sequential edits made in just a handful of hours over two months, that seem to be in direct violation of the rules. Dozens of unrelated articles are being stripped of citations to large sets of deprecated sources. It seems to have slipped under the radar. I don't know what to make of it.

Background (optional read): So I was curious about vehicles where people drive on the left vs right hand side of the road, and found our article on such things (which go by so many names), at (oddly) Left-_and_right-hand_traffic. Although it's true worldwide (i.e. in both cases) that the norm is "putting the driver on the side closer to the centre of the road," the article lede portrayed it as true only where one drives on the right. I was going to move that clause out of that sentence and make it neutral, but thought to look at the recent history first. On reviewing the history, I noticed this odd edit https://en.wikipedia.org/w/index.php?title=Left-_and_right-hand_traffic&diff=991172593&oldid=990867476. A while later I realized that the user has a long history and seems adamant about and repeatedly goes and does mass deletes that are in direct violation of the rules. When I tried to find out if there had been consequences, I followed this link from the user's talk page ((October archive, actually): https://en.wikipedia.org/wiki/Wikipedia:Administrators%27_noticeboard/Incidents#Uncivil_behavior_by_User:Horse_Eye's_Back But when I searched the archives for a discussion with that name, I could find none. Now I feel way over my head and in need of help. But the page is full of arguments about such edits, which seem to amount to WP:DE, given the prior discussions. Each section of that short October archive page hints a major pattern of disruption. Then I thought to do the search that is in my opening sentence and saw the machine gun carnage on Oct 11,13,18, Nov 9,10,13,15,28. I'm not saying the sources the user is removing aren't or shouldn't be deprecated. I have no idea. What I do know for sure:

WP:DEPRECATED: "Citations to deprecated sources should not be removed indiscriminately, and each case should be reviewed separately."

WP:V: "Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source. ... When tagging or removing material for lacking an inline citation, please state your concern that it may not be possible to find a published reliable source and the material therefore may not be verifiable.[5] If you think the material is verifiable, you are encouraged to provide an inline citation yourself before considering whether to remove or tag it."

Many of the individual edits I looked at just deleted the citation and replaced it with {{cn}}. These are completely incompatible with the letter and spirit of WP:V. I'm not sure what the norm is for addressing such activity, but perhaps (unless the user self-reverts) a preventative measure is warranted. Over and out.--50.201.195.170 (talk) 04:34, 13 December 2020 (UTC)

There a discussion about this https://en.wikipedia.org/wiki/Wikipedia_talk:Deprecated_sources#Proposed_clarification_of_deprecated_sources_guidelines Selfstudier (talk) 10:49, 13 December 2020 (UTC)

Or: Yup, that's the written rule. We do not follow it. In fact, it's not at all unusual for people to get praised for making it difficult for future editors to figure out which parts of an article originated from unreliable sources. WhatamIdoing (talk) 05:37, 16 December 2020 (UTC)

The male lead was William Holden, not William Shatner. — Preceding unsigned comment added by 64.71.163.28 (talk) 12:22, 21 November 2020 (UTC)

 Not done Wrong location, but Shatner was lead in the The World of Suzie Wong (play) and Holden was lead in The World of Suzie Wong (film) (and have linked the former in the latter). ^~Hydronium^~Hydroxide^~(Talk)~ 13:11, 21 November 2020 (UTC)

You mean Wong location? And size isn't everything. EEng 05:43, 17 December 2020 (UTC)

Could someone attempt to clarify how article 2 in ABOUTSELF should be interpreted in the following scenario? The text is as follows:

Abdulhadi defended herself by noting that on the trip they met with 198 individuals from 89 organizations and that dialogue with controversial figures was an important part of academia.[REF]

[REF] is a letter written by Abdulhadi and published in Jadaliyya [1]. For discussions sake we can assume that Jadaliyya is a questionable source. One editor claims that ABOUTSELF is not fulfilled because the above sentence involve claims about third parties, namely "198 individuals." I disagree for two reasons: 1. The sentence doesn't imply that Abdulhadi met with 198 individuals, only that she claims that she did and 2. The persons ("third parties") Abdulhadi met aren't identified at all in the article.

Otherwise, the sentence "Joe Schmoe said he went to the barber yesterday" would also violate article 2 of ABOUTSELF since it implies something about a third party, namely that a barber cut Joe Schmoe's hair. ImTheIP (talk) 23:59, 12 December 2020 (UTC)

ImTheIP did not ping me here, which is not nice, in particular since their claims of what I said are imprecise. This statement fails WP:ABOUTSELF points:

ABOUTSELF-1, as it is a self-serving explanation.

ABOUTSELF-2, as it involves claims on multiple third parties both those named in her statement and the context of the statement itself.

ABOUTSELF-4, as there are reasonable doubts here.

11Fox11 (talk) 07:29, 13 December 2020 (UTC)

Pretty sure this is a case of "I don't like it".Selfstudier (talk) 11:24, 16 December 2020 (UTC)

Selfstudier, it could also be a case of Wikipedia:Mandy Rice-Davies Applies. We do fairly often include certain kinds of "self-serving" statements, especially of the sort that amount to "The accused says he's innocent". WhatamIdoing (talk) 06:21, 27 December 2020 (UTC)

The text as quoted above is clearly not a claim about third parties as meant by ABOUTSELF #2. It is also not unduly self serving as meant by ABOUTSELF #1. However, the specificity of the numbers may not be due inclusion; and "noted" is a success verb which should generally be avoided, per WP:SAID. - Ryk72 ^talk 08:28, 27 December 2020 (UTC)

I agree with Ryk72's analysis. In reverse order, #4 addresses cases where there is doubt that the ABOUTSELF subject actually said the thing in question. It doesn't appear there are doubts Abdulhadi made this claim. #2 would apply in a case where Abdulhadi was making a specific claim about someone else. This is a generalized claim and doesn't cite any other person so I don't see it applying here. #1 pivots on the word, unduly. The limits of unduly haven't been well defined. However, in almost all cases where a BLP is responding to claims/controversies about them it is not considered unduly. This is in part because there is a strong consensus that those accused should be allowed to say something in their defense. Additionally, this is a case where the BLP is responding to an issue raised by others, vs raising the issue themselves. Springee (talk) 18:18, 28 December 2020 (UTC)

Matthiaspaul added this sentence to "Access to sources" in July 2018: "Rare historical sources may even be available only in special museum collections and archives."

This seems to describe unpublished work, so I'm going to remove it. Material has to be published to count as a source for WP:V and reasonably accessible. We don't define "published" because it's hard to define, but it can't include someone having to get on a flight to Moscow in the hope an archivist will make a document available. See Wikipedia:Archives as sources. SarahSV ^(talk) 06:17, 22 December 2020 (UTC)

Rare, historical sources may have been published well before the modern era. In fact, I agree, "published" doesn't make sense before the invention of the printing press. Now I grant that most of the sources we actually care about will have been reproduced, but there are plenty of moderately to very obscure primary source documents that exist in just a single library. This is WP:V, not WP:OR or WP:N, and very old or obscure things in specialized libraries are indeed appropriate to talk about here. I'd recommend it be reinserted as it stood. Jclemens (talk) 06:38, 22 December 2020 (UTC)

Can you think of an example that a secondary source would not have covered? SarahSV ^(talk) 06:47, 22 December 2020 (UTC)

• Supporting SV's reasoning and action to cut this out of the policy. A rare document exclusively available behind closed doors in a single archive should not be used in a WP:V logic. This dovetails with the other two core content policies. (1. WP:NOR –) Either such exclusive document is described in published reliable secondary sources, in which case these sources should be used in a WP:V logic, or it is not, in which case we should not adopt the document, unmentioned in secondary sources, in a OR approach; (2. WP:NPOV –) How would the WP:BALASP and WP:UNDUE, etc, parts of that policy apply to the exclusive document? The document would have negligible weight as a source when compared to reliable sources published on the subject on which we're building article content. If it contains a view that differs from all other possible sources, that single document would represent an "extreme minority" view, not suitable for adoption in Wikipedia. If its views align with what a substantial part of the reliable sources on the topic say, then, of course these sources should be used for representing that view in Wikipedia. --Francis Schonken (talk) 07:14, 22 December 2020 (UTC)
• I have no difficulty providing examples. Most recently, I tried to help someone find a magazine which is known to contain a review of a book for which an FA is being written. The magazine was published in the early 1970s and so is long out of print. It has not been digitised and it's not the sort of item which will be found in most libraries. I couldn't find it in my own collection but found a second-hand dealer that had a copy. Our readers are obviously not going to be able to easily access a copy either.

As an older example, I made a brief start on the first gardening book written in English – The Profitable Arte of Gardening – which was first published in 1563. I visited the Rare Books Reading Room at the British Library to inspect a first edition but, alas, have not yet taken the topic further.

Even the British Library purges its collection so that it may no longer have copies of many periodicals. For some details, see the American Newspaper Repository, which I also started. I recently got a subscription to the New York Times to get access to its TimesMachine and found this useful when looking for other reviews. But I don't have similar access to other US newspapers and suppose that many issues can be quite difficult to locate and read as they won't have been digitised. See Double Fold for details of why access to originals might be needed.

Other archives which I was lately considering include:

1. the International Guitar Research Archive established by Vahdah Olcott-Bickford
2. records of the Widnes Laboratory held by the National Archives and Imperial War Museum
3. records of Kings Food Markets now held by Rutgers

I could go on as these issues often arise during aggressive attempts to delete the topics in question such as the AfD for the latter case. In such cases, a simple Google search will not suffice and so one has to look around for promising leads. Archives seem quite satisfactory for this and we should not deprecate them.

Andrew🐉(talk) 12:11, 22 December 2020 (UTC)

• Had a closer look at the first two examples:
  1. Re. "magazine ... published in the early 1970s": not an example of a source "available only in special museum collections and archives" (you bought a copy that was available elsewhere), so completely unrelated to the example that was removed.
  2. Re. "The Profitable Arte of Gardening – which was first published in 1563": you got the date of first publication wrong, for which access to the first edition would have been no help whatsoever. Here is access to the title page of that first edition – it carries no date of publication. You'd need a reliable secondary source, like this one, to get the date of first publication right (according to that source first publication was c. 1558, not 1563). The 16th-century book on gardening is at least partially available outside "special museum collections and archives" (some of its early editions even completely), and access to its very first edition is certainly not necessary to write an article on this book (there are sufficient reliable sources that describe the various editions of the book).

The Double Fold argumentation seems rather like a rant: that publication seems far from setting a standard. Looking at your last example (sorry, skipped two), the Rutgers material is available on the internet: the removed example is about "rare historical sources", and the Rutgers web page seems neither rare nor historical, it is a publicly available modern inventory of an archived collection. It lists and describes articles and other items existing in an archive. No document "available only in special museum collections and archives" was mentioned in the AfD discussion, only an inventory that was available outside such collections and archives. --Francis Schonken (talk) 16:50, 22 December 2020 (UTC)

• I support retention for a different reason. It's purpose there is as an example while making the point that "easily accessible" is not a requirement. Makes the point clearly, more clearly than likely replacements. And documents from before the concept of wiki-"publishing" can be appropriate. The wiki-wording is not so infallible that it covers covers such. North8000 (talk) 14:20, 22 December 2020 (UTC)
• Retain - we have editors all over the world. If information is located in a rare (difficult to access) archive, and you do not wish to go through the time and expense of traveling half-way around the world to verify it... consider contacting someone ELSE (someone who lives near the archive) to verify it ON YOUR BEHALF. The requirement is that the information is ABLE to be verified... NOT that any specific user (ie “you”) can do so. Blueboar (talk) 14:42, 22 December 2020 (UTC)
  • Blueboar, re: your revert. This was added in 2018 without discussion or consensus that I can see, even though it arguably changes the nature of this policy. You've now restored it without consensus. Can you give an actual example of the kind of source you have in mind? "Rare historical sources may even be available only in special museum collections and archives." SarahSV ^(talk) 19:40, 22 December 2020 (UTC)
• Retain on the same basis of Blueboar. Further, I would assume as the world becomes more digital, these rare works will have digitized versions made to keep the historical work safe but make the information more readily accessible to all. --Masem (t) 14:49, 22 December 2020 (UTC)
  • Masem, if it's online, it isn't "available only in special museum collections and archives". SarahSV ^(talk) 22:12, 22 December 2020 (UTC)
    • The point is that there are valid sources that are in libraries and museum collections that are not on line YET. Blueboar (talk) 20:48, 23 December 2020 (UTC)
• While it's probably the case that a one of a kind, 15th century text probably isn't DUE, I also support the text in question as there isn't a clear line between "widely accessable" "limited access" and "inaccessible". One of the sources I intend to use (if I get time/motivation) is The Anatomy and Development of the Formula Ford Racecar. This isn't a unobtainable book but a used copy on Amazon goes for almost $200 and library copies are quite rare. Springee (talk) 15:21, 22 December 2020 (UTC)
  • In what universe is that "available only in special museum collections and archives"? --Francis Schonken (talk) 16:52, 22 December 2020 (UTC)
    • Side Note... in most cases, the type of material found in special museum collections and archives will be primary sources. While these primary sources will be of very LIMITED use in WP, that does NOT mean that they are of NO use. Information CAN be taken from (and cited to) rare primary sources - it just has to be done correctly (not easy) - and such information can enrich an article. Blueboar (talk) 17:09, 22 December 2020 (UTC)
      • Yeah, but can you give an example? Something that is both "a rare historical source" and "available only in special museum collections and archives", and was used as a source to verify something in Wikipedia? My point is that we should have a far more practical example. Instead of the removed example, that sends users in wrong directions on several levels, unless they know many much finer points of Wikipedia guidance, I'd rather have nothing unless a far more practical (or at least sufficiently recognisable) example is provided. --Francis Schonken (talk) 17:26, 22 December 2020 (UTC)
      • Blueboar, you need to give an example of a rare historical source available only in special museum collections, etc., and not used and cited by a secondary source that we would use instead. SarahSV ^(talk) 19:46, 22 December 2020 (UTC)

I don’t remember the specifics, but I do remember discussing the use of a document from the US Navy’s archives several years ago. What I remember was that the document had been discussed in secondary sources, but the specific wording was not given in any of those sources. An editor had tracked the original document down in the archives, and wanted the art to reflect the actual wording of the document in question. Our advice was that he could quote the text of the document itself (and cite the document as a primary source to verify the accuracy of the quote).

Now... Let me flip the demand... can you give an example of a citation to a rare source that has actually caused a problem? Blueboar (talk) 20:48, 23 December 2020 (UTC)

I love your example, Blueboar. Thanks for posting it. WhatamIdoing (talk) 06:33, 27 December 2020 (UTC)

• Support removal and send to RfC if necessary. The sentence makes it seem like it's perfectly fine for editors to cite very old historical primary sources like ancient manuscripts instead of the modern scholars who have the necessary expertise to understand and interpret the context of those writings. That is original research and should not be endorsed, and not in WP:V of all places. The preceding line about "a print-only source may be available only through libraries" is much better in covering that there are RS that may be hard to access, of which examples have been given above. Crossroads ^-talk- 20:38, 22 December 2020 (UTC)

Properly done (that is without WP:OR), citing from such old historical and/or primary sources is as appropriate as citing from a modern primary source. I would rather prefer an article, accurate up into the minute details and written neutrally, based on difficult to access but carefully selected, top-relevant and properly cited sources, than an article based on a bunch of freely available online sources of medium quality, mixing up or leaving out some relevant bits. Yes, primary sources are difficult to use, and one needs to be particularly careful to use them properly, but this equally applies to historical and modern sources, no matter if they are difficult to access or not. However, the corresponding do's and don't are covered in other guidelines, and do not matter in regard to WP:V's core message of verifiability.

--Matthiaspaul (talk) 19:55, 7 January 2021 (UTC)

• Retain. One example of an archive are birth and death records maintained by various government archives. While vital records can't be used to verify facts about living people, they can be used for those who have died. Due to the difficulty of connecting a vital record to an individual mentioned in a Wikipedia article, government records are best used to supplement reliable secondary sources. One common issue that comes up is whether a date in a secondary source is in the Julian or Gregorian calendar. The authors of secondary sources often assume that "everybody knows" which calendar should be used in a certain situation, but this is false; it is not intuitively obvious to everybody which calendar the author of the secondary source used. The primary source can often resolve the issue.

In many cases old vital records are only available by visiting a certain government building. Jc3s5h (talk) 21:25, 22 December 2020 (UTC)

• Remove: A government archive of births and deaths, archives of 1970s magazines, and all the other example mentioned above are available to multiple people. A rare historical source (not mentioned in any other reliable source!) that can only be seen by invitation is not suitable as a reference at Wikipedia as it would be WP:OR for an editor to describe what they saw. Johnuniq (talk) 22:02, 22 December 2020 (UTC)
• Retain. This issue isn't in any way limited to unpublished or primary sources. Most of the works that have ever been published are now out of print, even modern print runs of academic works are in the low hundreds, and digitization has so far captured only a fraction of published titles - Google Books has so far digitized only 30% of what it estimates to be the total number of works available to be digitized. Provisions like that being discussed here have been present at WP:RSC for over 10 years. Whether the inclusion of a particular claim/source is DUE needs to be considered for all potential sources, regardless of how easy or not it is to obtain them. Nikkimaria (talk) 23:39, 22 December 2020 (UTC)
  • Nikkimaria, can you give an example of the kind of source you have in mind: "[r]are historical sources ... available only in special museum collections and archives". Note: "available only in". Not available online or for purchase, including used books and journals; not available for loan, including interlibrary loan; and not discussed or cited by secondary sources. Available only in "special museum collections and archives". SarahSV ^(talk) 23:49, 22 December 2020 (UTC)
    • The disputed phrase says nothing about being discussed or cited by other sources; it regards only access to the source itself. Regarding access: if a scholarly work is not a required textbook, a classic, or a popular work, it is less likely to be retained by libraries or available for sale, especially beyond the first few decades after publication. Not to mention the issue of works from outside the US/UK market and/or published in languages other than English, which tend to be even harder to find. Nikkimaria (talk) 03:18, 23 December 2020 (UTC)
• It seems the policy implicitly mixes two different aspects of the term "reliability": trustworthiness and stability (non-volatility). Being a Wikipedian, I know how it is written, and that is why I never trust it. However, I am using it, because it provides good links to the sources that may be a starting point for digging deeper. That is why I think the most important property of Wikipedia is that it provides links to references that can be used to verify the content. Accordingly, the worst thing in Wikipedia is references to some web sites (archives, depositories etc) that ceased to exist many years ago. Those sites could contain reliable and good quality information, but they became dead links, and we will never be able to check if the referenced content was correct or wrong.

In contrast, if just one copy of some old document or book exists in some archive or museum, we do know it exists, we do know it will never disappear like some web archive, youtube video, or newspaper blog, and we do know how get an access to it (no matter how difficult it may be). That is why that source is reliable.

In general, I think the policy makes too much stress on one aspect of reliability (trustworthiness) and almost ignores another aspect (non-volatility or stability). I think we should fix it.--Paul Siebert (talk) 00:28, 23 December 2020 (UTC)

Paul, that's not what's meant by "reliable". SarahSV ^(talk) 01:19, 23 December 2020 (UTC)

Merriam-Webster disagrees with you. The second meaning is giving the same result on successive trials, which is exactly what I mean: if we add a reference/link to Wikipedia, we expect that it will not become unavailable in observable future. That is why a single copy of a museum manuscript is more reliable than some youtube video or newspaper blog.--Paul Siebert (talk) 18:31, 23 December 2020 (UTC)

That's an argumentum ad dictionarium. Reliability on Wikipedia is defined at WP:Reliable sources. Crossroads ^-talk- 19:31, 23 December 2020 (UTC)

Thank you for pointing at the article. Incidentally, the message at the top of this talk page says: "This is the talk page for discussing improvements to the Verifiability page." If we are not tot allowed to discuss any changes in WP:V, what are we doing here?--Paul Siebert (talk) 21:34, 23 December 2020 (UTC)

• The objections attempts to try to read this sentence in isolation is just poor reasoning. Don't read sentences of policy in isolation -- nothing takes this sentence out of the overall scheme of policy including dealing with "published" and "primary" and "undue", should they arise, and the context makes clear enough that what is being addressed is potentially using works found in libraries, etc. It's just not a relevant objection to complain you have to go to libraries or archives -- every source is not digitized on the web -- if you want to bring doubt on the source's reliability, complaining it's in a library just does not and cannot do that - such a quibble of having to go to a library is flat out irrelevant diversion. Then too, although this has been in policy in some form for many years, no actual examples of problems have been identified with diffs. Alanscottwalker (talk) 16:33, 23 December 2020 (UTC)

  This was the basis of my "support retention for a different reason" comment above. The context is this is ONLY an example to of the point being made that "easily accessible" is not a requirement.North8000 (talk) 19:00, 23 December 2020 (UTC)

  ... and is also what I was getting at pointing out that such guidance doesn't impact WP:N and WP:OR, where such difficulty might actually matter. Jclemens (talk) 20:05, 23 December 2020 (UTC)

• Retain. In context, this sentence refers only to access. As the following sentence explains (emphasis added): "Do not reject reliable sources just because they are difficult or costly to access." I'll do a bold edit to move that sentence to the foreground. Butwhatdoiknow (talk) 00:00, 24 December 2020 (UTC)
• I don't think this is the most relevant example of the general problem, or a necessary example. I think the most common example is going to use {{Cite sign}}. Either we can require someone to travel to a single, specific, and possibly heavily restricted location – or we can issue a rule against citing signs, roadside markers, and gravestones. It is true that "Rare historical sources may even be available only in special museum collections and archives." It is also true that every single sign in any museum is available only in that particular museum (and may not be permanent), and that we do accept those signs as reliable sources. If we need to record this fact (it's not really a rule) somewhere, then it would fit into Wikipedia:Reliable sources/Cost better than here. WhatamIdoing (talk) 06:31, 27 December 2020 (UTC)
• Let's see... WP:SPS says Self-published expert sources may be considered reliable when produced by an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications. One imagines a blog published by this expert. But what if, instea of a blog, the expert gave an invited talk at an academic symposium, and the record of that talk is a video available only in the library of the sponsoring university? Why would we not accept that when we accept a blog post? EEng 06:58, 27 December 2020 (UTC)

Taking into account that "reliability" means both trustworthiness and stability (see above), some video that is stored just in one place can hardly be a reliable source unless we have a strong reason to expect it will not disappear in observable future. We do have a reason to expect that for items stored in museums, but in your example, it is hard to tell. I think it should be a responsibility of those who add such a source to prove the source is stable, and it it not going to disappear. Again, I am surprised our policy pays so little attention to that aspect.--Paul Siebert (talk) 17:30, 27 December 2020 (UTC)

You're welcome to start a different discussion regarding requiring stability for something to be considered reliable. This is not that discussion. Nikkimaria (talk) 18:26, 27 December 2020 (UTC)

• Retain. In contrast to what Sarah wrote, the sentence does not alter the meaning of the policy, it just illustrates the core point of what WP:V is all about, the principal possibility to verify a statement through reliable references. The requirement for references exists to maintain the very goal of our project to write an encyclopedia containing the sum of all human knowledge, factually correct and presented as neutral as possible. If a reliable source helps to achieve that, it's a good candidate to be used in an article, and if it is easy to access or not simply does not matter in the context of verifiabilty. There are a bunch of editors who erroneously interpret our policy to mean that sources must be online and freely available to everyone (including themselves). While this is certainly desirable it is not a requirement per WP:RSC. Otherwise we could not achieve the goal of the project, as the majority of sources is (unfortunately) not online and not free, and, as a matter of fact, some are even very difficult to access. The sentence illustrates this by giving an example at the opposite end of the range of acceptable references, a reference difficult to inspect because it might be available only at a single location on earth and someone might need to (have someone) actually travel there to inspect it. Of course, if a statement can be sourced through other easier to access sources, this is, in most cases, much preferable, but in areas outside of mainstream interest (in particular historical ones) it is often difficult to find easily accessible sources, so it is important to not rule out some sources beforehand. Ideally, in order to avoid WP:OR such sources would only be used to describe facts rather than opinions, and editors using such sources are encouraged to take the opportunity and use |quote= to quote the relevant parts to make it easier for others to check the facts, but these are details beyond WP:V and subject of other guidelines, which apply to difficult to access sources in exactly the same way as they do to easy to access ones.

--Matthiaspaul (talk) 19:55, 7 January 2021 (UTC)

Wp:ver isn't the place to talk about inclusion/removal decisions, WP:Consensus is the place for that and the place where it is handled much better. The only purpose that the competing fragment of wp:onus in wp:ver serves is a weak attempt to stop the common mis-use of wp:ver which to say that meeting verifiability is a reason or strong force to include the material. (how many times have you seen the edit summary "undo removal of sourced material"?). So the solution is to replace "onus" in wp:ver with "Verifiability is a requirement for inclusion, not a reason for inclusion. See WP:Consensus for the decision-making process regarding addition or removal of material.". And change the target of the wp:onus re-direct to WP:Consensus. That's my proposal, open to tweaking. Sincerely, North8000 (talk) 12:37, 19 October 2020 (UTC)

That seems right, I have been confused about this for a long time. "undo removal of sourced material" is frequent typically because the reason given for removal in the first place often doesn't cut it. In any case that seems more like a consensus issue, solve with an RFC if necessary, that sort of thing.Selfstudier (talk) 13:03, 19 October 2020 (UTC)

I would 3suggest starting by leaving all the language the same but moving the ONUS shortcut to CONSENSUS and using a wikilink like so: The WP:ONUS to achieve consensus for inclusion is on those seeking to include disputed content. Kolya Butternut (talk) 13:09, 19 October 2020 (UTC)

The proposed text, "not a reason for inclusion", was previously added to the policy[2] and there was an RfC which resulted in its removal.[3] S Marshall, you were one of the initiators of the RfC, do you have any new thoughts about this language? And what do you think about my proposal in my previous comment? Kolya Butternut (talk) 13:44, 19 October 2020 (UTC)

• I think there are a lot of facets to this that are hard to summarize in a pithy sentence in a policy, or even a paragraph. I stand by my previous statement that there is, in practice, and ought to be, in theory, a presumption to include reliably-sourced information on Wikipedia, and that if information really is reliably-sourced then it's often more appropriate to move it to a different article than to delete it entirely; but the matter is very nuanced and our wording needs to be hedged with exceptions that can be invoked when we're dealing with controversial topics, biographies of living people, or inexperienced editors. I wonder whether we should start a separate essay about removing reliably-sourced content, with a view to promoting it to a guideline after some testing?—S Marshall T/C 16:58, 19 October 2020 (UTC)

S Marshall You and I have discussed this before. I believe that your objective is that if material is suitable for an enclyclopedia and is compliant with policies (e.g. wp:verifiability) that we should encourage (not discourage) its inclusion somewhere in Wikipedia. I agree with you on that. I would ask you to make a distinction between that and saying that compliance with one of those policies (being reliably sourced) should be a universal argument for a particular inclusion in a particular article.North8000 (talk) 19:15, 21 October 2020 (UTC)

I agree with North on this. But I would go farther. The extent of the world's verified knowledge is many times greater than all the combined encyclopedic narratives we would ever want to publish as articles. DUE WEIGHT and Noteworthiness reduces it by 50-90%. For example, we don't publish the local Atlanta weather on the date of publication of Gone With the Wind, even though it's reported and verified. SPECIFICO talk 20:23, 21 October 2020 (UTC)

Thanks S Marshall, and do you have thoughts about starting with my suggestion to leave the text the same but replace "onus" with WP:ONUS and move the shortcut to CONSENSUS? Kolya Butternut (talk) 17:23, 19 October 2020 (UTC)

• I'll reflect on that, and get back to you in due course.—S Marshall T/C 22:04, 19 October 2020 (UTC)
• I'd advocate the change that I proposed, but changing the target of the redirect this would be a step towards solving the structural/overlap problem. Decision-process issues are naturally the topic of wp:consensus, and decidedly not the topic of wp:ver. North8000 (talk) 18:16, 21 October 2020 (UTC)

• The problem is that these are two distinct and independent policies. We can't link a content policy to a conduct policy. That's a recipe for maximum confusion. SPECIFICO talk 18:27, 21 October 2020 (UTC)

@SPECIFICO:I agree about the separation, actually that is my point. Kolya Butternut's proposal (which I was expressing support for) to change the target of the re-direct is a baby step towards that separation. My proposal at the start of this sub-thread would separate it fully plus solve a lot of other problems. North8000 (talk) 19:04, 21 October 2020 (UTC)

Proposed change

Verifiability does not guarantee inclusion

Shortcuts

• WP:VNOTSUFF
• WP:VNOG

"WP:VNOG" redirects here. For the responsibility to demonstrate verifiability, see WP:BURDEN.

See also: WP:UNDUE, WP:PAGEDECIDE, WP:PRESERVE, WP:SUMMARY, and WP:IINFO

While information must be verifiable to be included in an article, all verifiable information need not be included in an article. Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The WP:ONUS to achieve consensus for inclusion is on those seeking to include disputed content.

Where WP:ONUS would (I assume) be another redirect for WP:NOCONSENSUS (or other section in WP:CONSENSUS). Kolya Butternut (talk) 00:37, 22 October 2020 (UTC)Kolya Butternut (talk) 04:15, 22 October 2020 (UTC)

A content policy can't link to a conduct policy. SPECIFICO talk 00:40, 22 October 2020 (UTC)

SPECIFICO, why is that? (Genuinely curious, I haven't encountered that rule before.) Schazjmd (talk) 00:45, 22 October 2020 (UTC)

It's apples and oranges. They are complementary but not equivalent. Conduct has bright lines, whereas content can change. SPECIFICO talk 00:56, 22 October 2020 (UTC)

WP:Neutral point of view#Achieving neutrality links to WP:Editing policy. Kolya Butternut (talk) 01:13, 22 October 2020 (UTC)

Or phrased more on point: WP:NPOV links to WP:PRESERVE. Kolya Butternut (talk) 02:02, 22 October 2020 (UTC)

Specifico, Kolya is right. There is no such rule. Content policies can and should refer to behavioral rules whenever appropriate. WhatamIdoing (talk) 00:56, 23 October 2020 (UTC)

I did not say they cannot refer to one another. That's why I said they are complementary. But a direct link so as to define one as equivalent to the other is not right. SPECIFICO talk 01:04, 23 October 2020 (UTC)

WP:ONUS already links to WP:PRESERVE using this text: presented instead in a different article. Kolya Butternut (talk) 23:01, 23 October 2020 (UTC)

• Oppose, as it is not needed and also would result in confusion as it changes the meaning of a commonly used link. The current setup of everything is fine and does not need changing, as it works. Don't fix what ain't broke. It also makes no sense: what does "what to do when there is no consensus" have to do with who has the onus to get a consensus for their preferred edit? Last but not least, I oppose this whole ongoing effort, as above and here at WP:CON, which results in making it harder to remove existing material. There are many low-traffic articles and many POV and fringe theory pushers out there. As I said there:
• "Here's a scenario for you: You stumble upon an obscure article on a topic you are familiar with. You notice some content that has sources but that is in some way fringe (synthesis, poor sources, misuse of sources, or some combination thereof), and/or that is undue emphasis on some POV, and remove it. Some editor, either its WP:OWNER or someone who has it on their watchlist but doesn't know the topic well, reverts you because you "removed sourced content" and it looked fine to him, and says that per WP:ONUS (the version you are advocating for right now), you need to get a full-on consensus to get it removed. Yes, there are ways to do so, but they may not get enough attention, and besides, the point is that this alternate version of ONUS made it much harder to do so.
• Wikipedia's quality comes just as much from what we keep out as what we let in. Since your goal seems to result in making it much harder to keep some things out, I have to oppose that."
• As a core content policy at the heart of what we do, any changes here must have a very good reason. Crossroads ^-talk- 03:13, 22 October 2020 (UTC)

This proposal does not change the meaning of policy. Kolya Butternut (talk) 04:11, 22 October 2020 (UTC)

I think the proposal changes the meaning. Maybe that's a good thing – because I believe that right now, ONUS and NOCON are saying exactly the opposite in one situation – but I think it constitutes a change from "those who seek to include disputed content" have to "achieve consensus for inclusion" (or out it goes) to those who seek to change content have to achieve consensus for a change (or else we stick with the previous version). WhatamIdoing (talk) 00:29, 23 October 2020 (UTC)

I agree, and it's an albatross that would thwart article improvement. SPECIFICO talk 00:45, 23 October 2020 (UTC)

WhatamIdoing, I'm not sure I see the difference. NOCON says the same thing as onus only more, that's why it makes sense to redirect the onus wikilink to NOCON. To tweak your words, "those who seek to include disputed content" have to "achieve consensus for inclusion" (or out it goes) to those who seek to change content have to achieve consensus for a change (or else out the edit goes)." In both cases the change is reverted if there is no consensus. Kolya Butternut (talk) 16:04, 23 October 2020 (UTC)

You are changing from a rule in which disputed content gets removed to a rule in which disputed content may be retained. Those are not the same. It might be a good change, and it might make the written rules better align with the community's preferences, but it does still constitute a change. WhatamIdoing (talk) 21:38, 23 October 2020 (UTC)

I am lost. ONUS and NOCON both say that changes to an article must achieve consensus in order for them to stick; it's just that ONUS only speaks to new additions whereas NOCON speaks to all changes. We disagree on the interpretation of ONUS, but the intention of this proposal is not change the policy based on this interpretation. Kolya Butternut (talk) 22:32, 23 October 2020 (UTC)

• Oppose as unnecessary. I also oppose any repointing of WP:ONUS to WP:CONSENSUS or any page besides this one. That shortcut has always applied to insertion of material and I see no good reason to change it. -- Calidum 14:54, 26 October 2020 (UTC)
• Oppose. There's no need to do this, and good reason not to. The standard use of ONUS is to remind people that the burden of proof rests always with those seeking to include disputed content, and has to, because anything else would be a POV-pusher's charter - we rapidly establish that consensus for inclusion exists where removal of long-standing text is obviously capricious. Guy (help! - typo?) 17:43, 28 October 2020 (UTC)

BOLD implementation

To (at least) illustrate the changes:

• WP:V change
• WP:CON change
• WP:ONUS redirect change
• WP:VNOT, new redirect

Kolya Butternut (talk) 21:18, 31 October 2020 (UTC)

All reverted (except for the new redirect). You should not make changes to high-traffic and highly-linked policy pages and the heavily-used WP:ONUS redirect when it is already very clear on this page that there is not a consensus for those exact changes, and that if anything there is a consensus against them. If you were really just illustrating them, you would have self-reverted. And you already illustrated them here anyway. Crossroads ^-talk- 22:03, 31 October 2020 (UTC)

Self-reverting might have been a good idea. Please notice the new section to WP:CON which I hadn't shown before. There is, however, clearly no consensus against the change. There is a fundamental disagreement over interpretation. Kolya Butternut (talk) 22:16, 31 October 2020 (UTC)

Proposed change: "Verifiability is a requirement for inclusion, not a reason for inclusion"

Basically, we have a structural problem (Wp:Ver wading into and duplicating / conflicting with something that is rightly covered in wp:consensus) which has to be in place because it is a partial fix for a widespread problem related to wp:ver (people saying that meeting wp:ver is a reason for inclusion rather than just a requirement for inclusion). And, like always, the big fuzzy Wikipedia system somehow mostly makes it work. I am on board to work really hard at the fix I proposed at the beginning of the "How to quickly fix it all" because it would be an immense clean fix on the structure problems and also a fix on the big "reason for inclusion" problem. If 1-2 people fully support it I'd work harder on that; otherwise I'd probably just let it fade away. I'm getting really worn out on participating in in the flurry of proposals here and on related pages which wouldn't do much either way and plan to sit back rather than weighing in on every one of them. Sincerely, North8000 (talk) 17:01, 23 October 2020 (UTC)

North8000, before you go, could you verify if your suggestion is to completely replace the text of WP:ONUS with this: Verifiability is a requirement for inclusion, not a reason for inclusion. See WP:Consensus for the decision-making process regarding addition or removal of material? It sounds reasonable; I just know that it has been proposed for many years, so I want to hear more from participants in the past discussions in case there may be any subtle consequences from that change that I'm not aware of. Perhaps to address S Marshall's concern, we might want to keep the text certain information...should be...presented instead in a different article in some form. Perhaps: Verifiability is a requirement for information in the encyclopedia, but not a reason for inclusion in an article? Kolya Butternut (talk)

• One concern... there are OTHER policies and guidelines that might mandate removal of material (and which would over-rule even consensus - NPOV, and COPYRIGHT come immediately to mind, but there are others). Any replacement needs to mention this. Blueboar (talk) 20:26, 23 October 2020 (UTC)

Yes, that is my proposal. The sentence that Blueboar has a concern about is there to help make it fly but is superfluous and could be dropped. Or else modified to handle Blueboar's valid concern e.g.: "WP:Consensus and other policies provide guidance on the inclusion/exclusion decision-making process." North8000 (talk) 20:48, 23 October 2020 (UTC)

WP:Editing policy covers those concerns, so we could just add another wikilink. Kolya Butternut (talk) 21:12, 23 October 2020 (UTC)

I'm open to any modifications in that sentence, doubly so that would get people behind the idea. North8000 (talk) 21:22, 23 October 2020 (UTC)

I'm getting a bit lost. I made this a new subsection if that's ok. I added your proposal as it would appear in WP:V; please correct it as necessary or move it to the beginning of this subsection.

Verifiability does not guarantee inclusion

Shortcuts

• WP:VNOTSUFF
• WP:VNOG

"WP:VNOG" redirects here. For the responsibility to demonstrate verifiability, see WP:BURDEN.

See also: WP:UNDUE, WP:PAGEDECIDE, WP:PRESERVE, WP:SUMMARY, and WP:IINFO

Verifiability is a requirement for inclusion, not a reason for inclusion. See WP:Consensus and WP:editing for the decision-making process regarding addition or removal of material.

The WP:ONUS shortcut would redirect to WP:CONSENSUS. Kolya Butternut (talk) 22:58, 23 October 2020 (UTC)

ONUS is not about consensus. It is about NPOV WEIGHT. SPECIFICO talk 00:27, 24 October 2020 (UTC)

If that is true, the proposal links to the editing policy, which includes NPOV and WEIGHT at WP:Editing policy#Try to fix problems. Kolya Butternut (talk) 01:37, 24 October 2020 (UTC)

As I said above, a reference to a complementary subject is not the same as a redirect or definition. This is a dead end. SPECIFICO talk 01:44, 24 October 2020 (UTC)

It sounds like there is some semantic confusion here. If others agree with this please share. Kolya Butternut (talk) 02:00, 24 October 2020 (UTC)

Well, the best solution would be to simply replace the entire section with "Verifiability is a requirement for inclusion, not a reason for inclusion" This solves the issue with wp:ver, and defers to all of the other policies and guidelines to handle everything else. The second sentence is just to try to get support for the change. My thinking is that some people would not immediately take in the structural solution just described and seek to have something in place to fill wp:onus. Sincerely, North8000 (talk) 02:18, 24 October 2020 (UTC)

In other words, since consensus governs most aspects of our editing, there is no reason to mix it up with ONUS. SPECIFICO talk 02:32, 24 October 2020 (UTC)

• Still oppose, per my 03:13, 22 October 2020 (UTC) comment. This is a radical change to our policy regarding inclusion. I don't see that there is necessarily a contradiction with WP:NOCON, which talks about modifying or adding material, not just removing, and which says a "lack of consensus commonly results" - note: not "should result" - in retaining the status quo. And even if there was, that can easily be addressed on the NOCON side, as WhatamIdoing tried to do here, and who was reverted over 3 weeks later by Kolya Butternut. Crossroads ^-talk- 03:40, 24 October 2020 (UTC)

  Three weeks later plus 19 hours I reverted it,[4] when I first noticed the change. I had already expressed disagreement with WhatamIdoing's proposal to change NOCON on October 1st when I said: "you think that the part of WP:NOCON... is wrong, and you want to change it to bring it in line with your interpretation of WP:ONUS. While my thought is that NOCON and WP:ONUS are correct, but that ONUS needs clarification."[5] Kolya Butternut (talk) 04:15, 24 October 2020 (UTC)

  So you are saying the proposal here is a "clarification", but WhatamIdoing's edit was a "change". That has it exactly backwards. She was clarifying NOCON, but the proposal here is a change. Crossroads ^-talk- 04:37, 24 October 2020 (UTC)

  As I said, it's based on our differing interpretations of the policy. And as I said in the edit summary, "The Consensus Required DS should be consistent with WP:CONSENSUS. This has been interpreted in discussions linked from here: [6]" Kolya Butternut (talk) 04:47, 24 October 2020 (UTC)

My own interest is structural. The common wp:ver-related "reason for inclusion" problem which is why wp:onus is in wp:ver, and that wp:onus is a misplaced overlap with wp:consensus which is the proper place for it. Kolya Butternut's main concern seems to be ambiguity in wp:onus but seem open to broader fixes. WhatamIdoing's edit is a good attempt to resolve it but I (possibly as a balance tipper) supported it's reversion because it further entrenches the noted structural problem. To me a partial fix seems obvious. Unless one says that the two policies conflict, just cover and clarify onus at wp:consensus. Sort of like what WhatamIdoing was working on but without referring to wp:ver (onus). Without changing the redirect and without incorporating my proposed change. Those could come later.  :-) Sincerely, North8000 (talk) 15:49, 25 October 2020 (UTC)

Don't know how I missed this: WP:Consensus required. Since this is just a strict version of WP:Consensus, I believe it shows that there is no greater burden to retain text than there is to remove text (besides the normal exceptions per WP:EDIT).

I'm not sure how we could change NOCON (or CONSENSUS generally) to cover and clarify the concept currently described at WP:ONUS without also using the word "onus" and becoming the more appropriate wikilink location, but I'm open to suggestions.

Also notice the use of the word onus here: A corollary is that if you disagree, the onus is on you to say so at WP:Silence and consensus. This sounds like there is a burden-shifting process when it comes to consensus, or generally that everyone must participate in the consensus-building process. Kolya Butternut (talk) 19:29, 25 October 2020 (UTC)

@Kolya Butternut, could you please tell us what a "Consensus Required DS" is, ideally with a link to a page about that? WP:Consensus required discretionary sanctions is a not a thing. So – Nintendo DS? Dal segno? Data science? DS register? Help us out here. WhatamIdoing (talk) 02:21, 26 October 2020 (UTC)

Arbitrary break

When I wrote NOCON, some of it was easy. WP:ELBURDEN could be copied over almost word for word, and nobody cared. But this particular line was added later (and by someone else) because there were pretty significant disagreements. We kept talking about it after I posted the main parts, and I was never satisfied that we'd identified the community's actual best practice in those discussions (which happened on at least two pages, if you want to dive into the archives).

The problem you run into in these discussions is that different groups of editors have different goals.

One group wants to put their thumb on the scale in favor of stability. Once they've gotten something settled into the article, they don't want it to change. This group of editors usually seems to edit in what I'll call 'stable' subject areas (such as famous paintings – subjects that don't change noticeably from one day to the next, but you get tired of arguing about the same thing) or in highly controversial areas (such as the Israeli–Palestinian conflict, or politics in general, for which avoiding edit warring – or even any editing that isn't absolutely necessary, since who knows what tiny, seemingly unimportant change could set off another massive dispute – is an important goal).

The other group of editors is very concerned about accuracy and maintaining our encyclopedic nature. The prototypical examples of this group are editors who deal with self-promotion and paid editing, and editors who work to remove pseudoscience and other garbage. When you're looking at what you think is a WP:PEACOCKy description of an antivaxxer or quack, and a normal discussion doesn't settle it, it's very upsetting to be told that "Wikipedia's policies require" the article to keep what you believe is inappropriate content just because it's been there long enough that the bad content counts as the "status quo". In this context, the QUO-oriented editors are basically saying that since a thousand innocent readers have already been misled, we need to keep misleading the rest of them. You can see why that's a bad idea. These editors cite Jimmy Wales about "no information is better than bad information" and want to remove what they think is bad information.

Now if you check Kolya's edits, you'll see that there are lots of edits to high-volume controversial subjects (e.g., sexual harassment scandals). It's not surprising that someone who edits those articles might sympathize with QUO's "Peace for our time" approach. Many of those articles will be much easier to write several years from now, but for right now, the work is just to minimize screwing up, which is mostly accomplished by minimizing editing. I've worked on some articles that fall into this category; I spent several years explaining to well-meaning editors that when the scholarly sources about Breast cancer awareness differed from their personal beliefs, then the article needed to be based on the scholarly sources. "QUO says I get to keep my long-standing text, so go away" would have been a shortcut to the same end result.

But if you're going to be editing, say, articles about altmed products, then "Get this garbage out of Wikipedia" is the order of the day. Here I'll ping User:Alexbrn, whose average net contribution to Wikipedia articles is actually a negative number. He spends his days protecting Wikipedia articles from garbage. I don't think that we should be putting barriers up that say he can only remove bad content either when that content was recently added to the article, or when he can prove that there is an active consensus in favor of removing garbage.

This is the fundamental problem with what Born2cycle added to NOCON: Either we can live in a world in which I get to keep my long-standing text merely because it's long-standing, without there being an active consensus to keep it, or we can live in a world in which Alex gets to remove long-standing garbage, without there being an active consensus to remove it.

Right now, I believe that ONUS says Alex can remove the garbage, and that NOCON says that he can't (in the absence of an active consensus). We should not have policies that contradict each other. There may be a middle way. Maybe QUO is fine some subjects, but areas plagued by paid editing have a default to remove disputed information, just like we default to removing uncited controversial content about BLPs. But the current system is IMO bad. WhatamIdoing (talk) 03:06, 26 October 2020 (UTC)

Thanks for the ping WAID. I confess I've not ever much paid attention to WP:NOCON. It seems to me there is a lot of redundancy and overlap with the PAGs here. Skimming through the discussion above I think some editors do not fully grasp what consensus is: it's not a democratic vote, but weighted in favour of the PAG-compliant "side". If I alight on some obviously garbage content (for reasons of POV or V, say) and remove it, and some editor argues to retain it, the consensus fully is with me because the only "proper concerns" are with me; the other editor's concerns are improper and thus literally do not count in the assessment of consensus. Thus the current text of NOCON, in setting up special cases for BLP and COPYVIO, betrays a fundamental misunderstanding of consensus. Any obviously policy non-compliant content may be removed immediately from an article because any such removal will be an automatic expression of consensus with the community at large who wrote the PAGs. Of course, in so doing, one has to be confident it is "obviously" so otherwise trouble will ensue! Perhaps one way to patch this up would be to quote WP:CON at WP:NOCON to point out that consensus is only ever a view aligned with with the WP:PAGs and not to be confused with mere local editorial agreement? Alexbrn (talk) 06:43, 26 October 2020 (UTC)

NOCON was meant to be entirely redundant. The idea was to list what was already in "the rules" elsewhere.

We are, in NOCON, talking about what happens in the event that a discussion ends with a "true no consensus". This happens when different editors have different ideas about, e.g., promotionalism or potentially offensive content. WhatamIdoing (talk) 18:26, 26 October 2020 (UTC)

WhatamIdoing, bang on, sir. And this is why ONUS exists and has to rule. If removal is capricious then consensus to include will be achieved before the deadline; in the mean time, mandating inclusion on the basis that nobody noticed it for a while is a POV-pushers' charter. Guy (help! - typo?) 17:48, 28 October 2020 (UTC)

That interpretation makes it easier for POV-pushers to remove text. Kolya Butternut (talk) 20:52, 28 October 2020 (UTC)

Yes. It makes it easier for POV-pushers to remove text, and also easier for our anti-woo editors to remove POV pushing. WhatamIdoing (talk) 23:21, 28 October 2020 (UTC)

POV-pushers can put the onus on editors who want to restore content which had achieved previous consensus (silent or otherwise). It gives more power to pov-pushers than it takes away. If non-neutral content has been in an article for months already and a POV-pusher stonewalls its removal, we can wait a little longer while its opened up to the community for content dispute resolution, and they wouldn't be able to add more POV text while eyes are on the article. If a POV-pusher removes longstanding text they shouldn't be able to keep it out while an RfC is underway, and potentially remove more text during that process. Kolya Butternut (talk) 04:41, 29 October 2020 (UTC)

What strikes me is that this discussion seems adrift of the way Wikipedia actually works, envisaging a situation (between two editors?) where lawyering around a fragment of policy decides whether some piece of article content stays or goes, and where "consensus" is some kind of holy governing factor that exist when some bit of text has gone un-noticed for a few weeks. In reality, if any editor make it a habit to remove bits of content that are legitimate, they will find their career on Wikipedia a short one. More generally, although I came to this discussion cold (via a ping), looking at the background it seems like a meta-dispute carried over from an arbcom case to try and retro-actively win some point from that. Basing policy on individual cases is never a good idea. Are there are wider examples of actual incidents that exemplify what needs to be fixed here? Three or four of those might help clarify things. Alexbrn (talk) 06:34, 29 October 2020 (UTC)

Alexbrn, this whole discussion was precipitated by a dispute about somebody in the entertainment industry, who was involved in a sex scandal. There were lots of changes made, encompassing additions, removals, and other changes, and Kolya – who has repeatedly asked that we not focus on this "actual incident", by the way – reverted them, under the impression that ONUS supported the reversion. ONUS technically does not (for most of it). QUO, BRD, and NOCON do. AFAICT the only thing "wrong" with Kolya's reversion is that the edit summary should have asserted authority on the basis of a different part of our written rules.

Since then, however, I've become convinced that our ruleset is self-contradictory. This is an inherent problem, and should be fixed, even though the practical problem is fairly limited (because it only applies to a limited subset of editing disputes). The reason it hasn't already been fixed is that even among people who understand the problem (i.e., people who don't think that the "inclusion" of disputed content somehow encompasses "removal" of the disputed content), editors are split about whether to narrow QUO to conform to ONUS or to expand ONUS to apply to all edits. WhatamIdoing (talk) 21:45, 1 November 2020 (UTC)

Heck, WhatamIdoing, there is not even agreement among Admins as to what constitutes a "revert". Fortunately, most editors are dedicated to consensus-building regardless of our inability to write a perfect codification of best practices. In the case of the Aziz Ansari article that precipitated this, the discussion did ultimately confirm consensus against the older text. The probem there was that there was no substantive argument in favor of the old text, just that it should be kept because it was old. In Class C articles' coverage of news events, it's to be expected that the passage of time will lead to revision of due weight per NPOV and BLP. SPECIFICO talk 22:03, 1 November 2020 (UTC)

WhatamIdoing, just to add a correction, my edit summary in that content dispute referred to the talk page discussion where I cited WP:V and WP:NPOV. Kolya Butternut (talk) 22:06, 1 November 2020 (UTC)

WhatamIdoing, given the lack of enthusiasm so far here and at the RfC for changing ONUS, I'd say the likely eventual outcome is that your edit here will be reinstated, thus removing any seeming contradiction. In fact, had it not been for this one dispute about that actor's sex scandal, that text would probably still be there now. Crossroads ^-talk- 22:41, 1 November 2020 (UTC)

Crossroads, I agree that edit seemed to clarify without changing the meaning or community understanding of the matter. SPECIFICO talk 23:05, 1 November 2020 (UTC)

But that edit was made because of the discussion which occurred after the content dispute. And that edit was meant to correct a contradiction which we don't all agree exists. Kolya Butternut (talk) 22:48, 1 November 2020 (UTC)

@WhatamIdoing: Thanks for writing that excellent and thorough summary. North8000 (talk) 19:37, 26 October 2020 (UTC)

Next step

The small group here is not going to change these key policies. I suggest posting any proposals at the Village Pump policy page, where a far-reaching question like this will get more attention. SPECIFICO talk 20:52, 23 October 2020 (UTC)

I dunno. This is the talk page of the involved policy and is heavily watched. But either way, unless a smaller group decides on it and gets behind it, it would be doomed as is any significant proposed change under the Wikipedia process. North8000 (talk) 21:02, 23 October 2020 (UTC)

As can be seen above through this whole section and under Wikipedia talk:Verifiability/Archive 69#ONUS, this has already been discussed to death for months and no consensus to change the policy has developed. Crossroads ^-talk- 03:44, 24 October 2020 (UTC)

The previous discussions were about different things. Here, there is still no consensus on whether ONUS applies only to new inclusions, so discussion continues to clarify the policy based on one interpretation or the other. Kolya Butternut (talk) 22:30, 24 October 2020 (UTC)

I think the proposal has merit. Pity there is no easy way to just try something for a time to see whether the opposition fears come to anything.Selfstudier (talk) 13:05, 25 October 2020 (UTC)

Selfstudier, in my experience, it takes editors about two years to notice that a change has been made to the written policies. That makes it hard to try something for a time. WhatamIdoing (talk) 02:24, 26 October 2020 (UTC)

If you want to make a change to this page, it should be discussed here. However, it'd be a good idea to advertise it at a village pump. WhatamIdoing (talk) 02:22, 26 October 2020 (UTC)

WP:Consensus required says removing longstanding text requires consensus

This explanatory supplement illustrates the enforced version of WP:Consensus, which should be consistent with WP:ONUS.

Addition:
Editor1 submits a new addition;
Editor2 reverts, removing the addition;
Editor1 now needs to gain consensus on the article talk page for re-adding the addition.
Removal:
Editor1 removes longstanding portion of text;
Editor2 reverts, re-adding the text;
Editor1 now needs to gain consensus on the article talk page for the re-removal of the text.

Kolya Butternut (talk) 14:48, 26 October 2020 (UTC)

That looks like an attempt to legislate clue, and misunderstands what "consensus" is (it is not just agreement). Sometimes reverting new content is obviously bad; sometimes defending longstanding content is obviously bad. If we elevate processology to look like it enables such obviously bad things, we're not helping. This is given as a "possible scenario" because it's just that. In some cases it would not apply. Alexbrn (talk) 15:08, 26 October 2020 (UTC)

Alexbrn, I don't know what "legislate clue" means; this is an existing explanatory supplement. I'm not sure I understand; the word used is "consensus", and it's not being described or narrowed to "agreement". The normal rules of the editing policy of course still apply. Sometimes reverting new content is obviously bad; sometimes defending longstanding content is obviously bad. The "possible scenario" is not inconsistent with that. Kolya Butternut (talk) 15:35, 26 October 2020 (UTC)

Legislating WP:CLUE is trying to codify the understood subtleties of best practice and in doing so failing to capture them, because they resist codification in something terse and brief like a single Wikipedia policy section. Explanatory supplements of course have to force in the WP:PAGs but to repeat, the problem I'm seeing here is a repeated failure to understand what "consensus" is. Alexbrn (talk) 16:31, 26 October 2020 (UTC)

It seems the above example is missing something like this:

Removal:
Longstanding text was added but never discussed or challenged - lets assume most editors, if asked, would say content is UNDUE;
Editor1 removes longstanding text as UNDUE (passes V but ONUS says doesn't ensure inclusion);
Editor2 reverts, re-adding the text;
What happens next?

My reading of ONUS says Editor2 needs to justify and gain consensus for inclusion. I understand this isn't how it often works but that is my read of what ONUS currently says to do. Springee (talk) 15:17, 26 October 2020 (UTC)

I think that is confusing things. This (WP:Consensus required) is talking about text which has already achieved consensus for inclusion (at least implicitly). Evaluating the strength of the consensus is a separate issue. Kolya Butternut (talk) 15:54, 26 October 2020 (UTC) Kolya Butternut (talk) 19:37, 26 October 2020 (UTC)

There is text on Wikipedia that has no consensus (most extremely, vandalism that has stuck), or consensus in name only (i.e. to be ignored). To repeat, consensus does not just mean agreement. It's good to clear that crap out with vigour! Alexbrn (talk) 16:26, 26 October 2020 (UTC)

I feel like we're talking past each other, so I'll just say again that I am not talking about text which has no consensus, and of course crap can be removed. Kolya Butternut (talk) 17:31, 26 October 2020 (UTC)

I think that @Springee is correct. That scenario is missing. This is not "confusing things"; this is the main point of the dispute that we've spent the last two months discussing. WhatamIdoing (talk) 18:33, 26 October 2020 (UTC)

• The Wikipedia:Consensus required page possibly isn't as clear as it needs to be, but those editing rules only apply to articles that have the "consensus required" restriction applied to them as a discretionary sanction. The page is not intended as a general rule for editing. – bradv🍁 15:39, 26 October 2020 (UTC)

  Bradv, what do you think about removing the supplement tag from that page, and moving it into the ArbCom pages? WhatamIdoing (talk) 18:35, 26 October 2020 (UTC)

RfC: Should we move WP:ONUS to WP:CONSENSUS?

Discussion at Village Pump: WP:VPP#RfC:_Should_we_move_WP:ONUS_to_WP:CONSENSUS? Kolya Butternut (talk) 04:40, 1 November 2020 (UTC)

Discussion carried over from WT:Silence and consensus

• Was closed, WP:Village_pump_(policy)/Archive_162#RfC:_Should_we_move_WP:ONUS_to_WP:CONSENSUS?, overwhelming consensus against. The word “onus” used at WP:Silence is a simple, perfect, use of the word. It does not explain the word, define the word, or expand on the word, and it has nothing directly to do with WP:ONUS. I think it is unfortunate that people try to confuse shortcuts with SINGLESHOUTYWORDS that serve to dumb down discussion. Wherever WP:ONUS points, it is not to redefined wikt:onus. —SmokeyJoe (talk) 22:36, 15 December 2020 (UTC)

  Because that is an Essay carrying no weight or authority, the line in dispute there can remain forever in or out. Nobody much cares. The irony is that its persistence disproves the premise of the page. SPECIFICO talk 23:05, 15 December 2020 (UTC)

  "The onus to achieve consensus for inclusion is on those seeking to include disputed content." The weakest form of consensus is WP:silence. I guess this is the conflation that bothers some? WP:silence does not speak to how to achieve consensus in the case of an objection, except to say that if someone wants to object, they have to say so. The essay claims no weight, it explains reality. What "persistence" do you mean? --SmokeyJoe (talk) 23:24, 15 December 2020 (UTC)

  I'm confused; what are you suggesting bothers some? If an addition is not disputed it achieves consensus through silence. Kolya Butternut (talk) 00:19, 16 December 2020 (UTC)

  Anyway, I agree with you that wherever WP:ONUS points is irrelevant to how we use the word onus. I think if anything, onus is the ideal word to use in more places in WP:CON itself. Kolya Butternut (talk) 00:36, 16 December 2020 (UTC)

  I think it bothers some that there is an essay related to WP:Consensus that uses the word "onus", when they want "onus" to be solely associated with WP:ONUS. I think that some of your properly reverted edits caused this. I guess that there is a wiki-fight somewhere involving getting consensus to add stuff, and it spilled over. --SmokeyJoe (talk) 00:44, 16 December 2020 (UTC)

  I understand the first part. My experience was that my mainspace edits had had silent consensus, so they ideally should have remained pending discussion, or else silent consensus would seem to have no weight at all. Kolya Butternut (talk) 01:03, 16 December 2020 (UTC)

  No, an addition doesn’t get to stay pending a discussion. If reverted, the status quo ante stands pending discussion, or compromise. Per Wikipedia:Don't revert due solely to "no consensus", minimally the revert edit summary should state the substance of the objection. —SmokeyJoe (talk) 01:45, 16 December 2020 (UTC)

  I don't understand that. I would think the longstanding version is the status quo ante version. If longstanding content is treated the same as brand new additions, then I don't understand what the use of silent consensus is. Kolya Butternut (talk) 02:04, 16 December 2020 (UTC)

  When you wrote "my mainspace edits had had silent consensus", I assumed that these are recent edits. Often, I will judge how recent against the number of subsequent edits. Can you point to the edits in question? What is the use of silent consensus? Silent consensus has no "use". It is an ancient concept, well accepted, that agreement may be assumed from people present and able to object who do not. It is very weak, and is broken even by a late objection, so it is not something to "use". --SmokeyJoe (talk) 02:32, 16 December 2020 (UTC)

  The edits I made were in May 2019[7] and September 2019.[8] They were stable until April 2020. Kolya Butternut (talk) 02:53, 16 December 2020 (UTC)

  it achieves consensus through silence

  I think this is a point that editors disagree about. Does silence cause the edit to achieve consensus, or does silence produce something different, like a rebuttable presumption of consensus?

  I think it's the latter: If the edit is made on day #1, and nobody disputes it on day #1, we assume that there is consensus, but that assumption is rebutted as soon as someone actually disputes it, regardless of whether that dispute appears on day #2 or day #200. Silent consensus is not a case of Once saved always saved, and WP:BRD is equally functional on 200-day-old content as it is on 20-day-old content. WhatamIdoing (talk) 05:14, 16 December 2020 (UTC)

  A single voice who comes forward on day 200 to dispute an edit does not immediately outweigh the silent voices of the many. I don't think silent consensus is something to fear; if pseudoscience or fringe theories ended up in an article they will surely be removed once a couple of editors have been alerted. Kolya Butternut (talk) 08:04, 16 December 2020 (UTC)

  it achieves consensus through silence is looking at things wrong. I think is a bad idea to count days, instead count edits to the content in question. If no one has copy edited the content since it was added, that’s recent. If the content has been copy edited by more than ten other editors, then I’d say it has status quo ante. Still, silence is the weakest form of consensus, and once there are substantive argument, the past silence is irrelevant. —SmokeyJoe (talk) 08:36, 16 December 2020 (UTC)

  I don't think the lack of copy edits means the edit should be considered recent, especially for a carefully written short edit. I think there may be different levels of status quo. If someone does a rewrite (not removal) of a section which only had silent consensus, and someone objects to the rewrite, I think the longstanding version should be in place pending an RfC. Everyone thinks their arguments are substantive; I don't know that arguments without consensus would make past silent consensus irrelevant. Kolya Butternut (talk) 10:08, 16 December 2020 (UTC)

  "Carefully written edits" tend to have consensus, and we can usually prove that there is consensus for those edits when the silence is disputed. Consider this scenario:

  • You improve an article.
  • Nobody talks about it (silent assumption of consensus).
  • A long time (however you measure that) later, someone disputes your edit.
  • Everyone jumps on the talk page and supports the edit (written proof of consensus).

  The problem tends to be this scenario:

  • I improve an article (at least, I think it's an improvement, and like most Wikipedia editors, I have an unreasonable amount of confidence in my edits).
  • Nobody talks about it (silent assumption of consensus).
  • A long time (however you measure that) later, someone disputes my edit.
  • During the subsequent discussion, there is no evidence that there actually was – or, more pointfully, that there actually is – any real consensus for my edit. We assumed that there was consensus for it, but even if there might have been consensus for the edit in the past, there isn't consensus now.
  • Since I like my edit, I then declare that the ONUS is on people who want to make a change (but not on me, for making my change!) and that QUO says that since you didn't object when my edit was new, then I get to keep my edit forever, and you have to prove an indisputable consensus to change the part that I edited.

  That last bit is where we see the problem. Once there's evidence that my edit is contentious (e.g., multiple people revert it), then I shouldn't stand on my "QUO rights". I should be working with others to find something that we could all agree to.

  @SmokeyJoe, I don't think that Kolya shares your belief that once there are substantive argument, the past silence is irrelevant. WP:SILENCE says that at the very end, but it is not at all clear on that point up front or what its implications are. We're probably doing a disservice to all of these editors by not putting that in the lead, or providing them with a workflow that explicitly specifies that length of time/attention doesn't matter (because WP:CCC). WhatamIdoing (talk) 17:31, 16 December 2020 (UTC)

  WhatamIdoing, That has been my concern about the one-liner onus cameo that has been in and out of the WP:Silent consensus essay. That line could be misread to say that, if you Mr & Mrs Editors did not object at the time the now-deprecated text was initially added -- then toughies, you failed in your duty to onus and the bad text is now bullet-proof. I am pleased to see that, as I read it, SmokeyJoe seems to agree with (or at least share my concern about) the possibility that onus bit could be misinterpreted in this way. So our disagreement at the Essay page may have been about exposition rather than the message of the essay. The onus to speak up needs to be contextualized by providing some indication as to the time at which any such onus falls upon the editor who disputes the deprecated text. The current objections may be from editors who did not previously see the bad text, or circumsances and sources may have changed. Because it's always the case that, by the time the text is disputed, somebody has exercised such onus, the one-liner in the Essay does not seem helpful or (due to its brevity and ambiguity) descriptive. SPECIFICO talk 19:16, 16 December 2020 (UTC)

  WhatamIdoing, I think we have to distinguish between "silent assumption of consensus" and "silent consensus", i.e., WP:IMPLICITCONSENSUS. WP:Consensus can change is included in WP:Silence and consensus because silent consensus is indeed a form of consensus, however weak. Silent consensus is relative, but I think it makes sense to begin to examine the dynamic of silent consensus with a dispute between just two editors. Anyway, it might be better to not continue this discussion on this page. Kolya Butternut (talk) 20:20, 16 December 2020 (UTC)

  IMPLICITCONSENESUS says that the uncontested edit is only assumed to have consensus, not that it achieved consensus by being uncontested. And that sentence has a link to WP:SILENCE, which means that it's not entirely reasonable to assume that it means something different from SILENCE. Silence – whether you're talking about the essay or about the sentence in the policy – only gives you a rebuttable presumption of consensus. WhatamIdoing (talk) 01:02, 17 December 2020 (UTC)

Post-RfC discussion: ONUS and Consensus

Continued from above discussion

This discussion highlights something I was wondering about not long ago. ONUS and CONSENSUS seem to have a bit of a conflict when it comes to the case of implicit consensus. If new content was added and just ignored for say 6 months would a different editor need consensus to remove? Editor A adds New Fact. 6 months later Editor B removes that content as UNDUE (ONUS assumes WP:V). Editor A restores it citing WP:NOCON which says in the case of no consensus restore to last stable version. Editor B cites ONUS saying the content was never challenged. Essentially we have a 1:1 !vote (fore:against). Will implicit consensus push this into the "consensus" camp? Prior to the ONUS discussions I would have assumed the content stays per NOCON. Now I think ONUS would say this is the first challenge to the material and thus no consensus to retain. However, this could be confusing since I don't see ONUS applying to changes that don't add new content to the article. So, in my view, ONUS doesn't apply to changes like expanding/contracting the lead based on content from the body, changing the order of content, rephrasing text based on existing sources. Springee (talk) 22:24, 16 December 2020 (UTC)

I don't think they actually conflict, but they certainly appear to. I don't think it's a good idea to cite the content policy WP:ONUS for conduct matters (even though everyone does). WP:ONUS includes a sentence which describes how part of WP:CONSENSUS works, which of course is our conduct policy. WP:ONUS/WP:VNOT is best to cite in response to an edit summary such as "Restored well-sourced content". A talk page comment in response might say, "Per ONUS and UNDUE, the content should be removed...."

WP:ONUS says "The onus to achieve consensus for inclusion is on those seeking to include disputed content". Once implicit consensus has been achieved, WP:ONUS (which includes a distillation of part of WP:CONSENUS) no longer applies. Implicit consensus is weak, but at least in the case of a 1:1 dispute the content should remain. Kolya Butternut (talk) 23:13, 16 December 2020 (UTC)

I think that they do conflict, and I think that NOCON is wrong (not 100% wrong, just wrong enough that it needs to be fixed). WhatamIdoing (talk) 00:58, 17 December 2020 (UTC)

Absolutely not. At least when it comes to removing longstanding material, WP:NOCON reflects existing policy and practice, while WP:ONUS patiently does not - at least outside of WP:BLP violations, which are a special case. If you try to remove longstanding text and get reverted, the burden is on you to obtain consensus for your proposed removal, just as with any other contested change to an article, and you could (and people 100% have been) sanctioned for aggressively and repeatedly trying to insist otherwise. The interpretation you're pushing for would have drastically disruptive implications - it would mean that in theory every part of every article that had not been intentionally discussed (which is most of Wikipedia) could be challenged, removed, and left out for the course of an extended discussion - a disruptive process when dealing with longstanding, non-WP:BLP-violating text. The bias in policy when there is a dispute is towards the status quo, not towards removal, since when there is a dispute we need to favor article stability. --Aquillion (talk) 05:35, 17 December 2020 (UTC)

This was all addressed in detail at the RfC as well as in the discussions above. In a nutshell, it is very easy to get a consensus against such bad-faith removals via, at minimum, two or more editors reverting the one removing it (and as you note, such users are often blocked). But a general policy of 'when there is no consensus either way, include' is a red carpet for POV and fringe theory pushers. Just because someone copyedited it a while back doesn't mean that they scrutinized it. Very much of the encyclopedia is underscrutinized and there are very many POV pushers and refspammers out there. Crossroads ^-talk- 04:13, 18 December 2020 (UTC)

• I have been pointing out that problem for a while now, but every effort to modify WP:ONUS to resolve the discrepancy has failed. I don't think it's possible to resolve the discrepancy in favor of WP:ONUS - practice and policy have always been that if you want to remove longstanding text you must obtain consensus for it. Though I do think it's important to make clear that implicit consensus depends on people having seen the text and implicitly endorsed it, ideally by editing the section in question without changing it - something patiently controversial that ended up in an obscure stub with few inbounds that attracts little attention and few edits may not have implicit consensus even a year later. If nobody has seen or reviewed it, there's no implicit consensus. --Aquillion (talk) 05:35, 17 December 2020 (UTC)

  Aquillion, what are your thoughts on the failed RfC? Kolya Butternut (talk) 00:06, 18 December 2020 (UTC)

  Is it a discrepancy with ONUS or with how we have handled NOCON? It certainly is easier to handle things via an assumption that NOCON rules. The only question would be, when is something long standing. A carve out for ONUS may be hard to implement in practice. If new content is added and never discussed is that consensus? What if it is just copy edited but not really changed? Does trying to fix a poorly composed sentence imply you support the actual content or just that you are trying to make something read better. Also, how do we decide which old edits are subjected to ONUS (generally new content) vs aren't (adding a fact from the body to the lead). From a practical POV I see a lot of issues with this. However, I also think it's probably the better policy. We shouldn't just say content that has been around for a while has strong consensus if challenged. Presumably if the content is good others will step up and defend it. If the content is poor and results in a no-consensus discussion then perhaps it really shouldn't have been in the article in the first place. My only concern is how to implement this since general practice has been revert to long standing. Even if we all agree ONUS should be supreme how do you get editors who don't follow these discussions to acknowledge and respect that decision? Springee (talk) 01:04, 18 December 2020 (UTC)

  In my opinion WP:NOCON rules, because unlike WP:ONUS, that is the conduct policy. Kolya Butternut (talk) 01:54, 18 December 2020 (UTC)

• What makes implicit consensus might be worth clarifying a little bit. As I said above, I think that implicit consensus is derived from the number of editors who have viewed or edited a section without removing something (with edits being much stronger evidence of implicit consensus.) Simple age is one factor but is insufficient on its own - a stub with almost no editors can't have implicit consensus no matter how old it gets, say, because implicit consensus depends on the idea that many people have seen it and accepted it. Also, a clearly-present longstanding dispute on something that has never reached the point of implicit consensus before (whether it's a removal that is immediately reverted, a template questioning it, an active RFC or a comment on talk) stops the timer on implicit consensus, of course - "it's been there a year" doesn't produce implicit consensus if people have been constantly objecting to it for that entire year. So the ideal implicit-consensus text is stuff that existed for a long time on an article with high traffic with no objections whatsoever. It can exist with less than that, of course, but the further away you move from each of those points the weaker the implicit consensus becomes. That said I don't think it's a good idea to nail down the rules for things like this too severely, because when you have an unambiguous default outcome it tends to lead to unproductive discussions because one side in a dispute is discouraged from engaging if they know all they need to do is stall and object. The vast majority of text on the wiki is really somewhere between "has a clear unambiguous consensus backing it" and "clearly lacks a consensus, can be removed immediately and a clear affirmative consensus will be required to restore it." And I think that to some extent that ambiguity is good because it encourages people to actually go to talk and hash out their differences rather than saying "WP:ONUS, removing, get consensus if you want to put this back" with no further discussion, or "longstanding text and WP:NOCON; unless you can show consensus, current text stays" with no further explanation. I feel that both those are failure states that need to be discouraged. --Aquillion (talk) 11:40, 19 December 2020 (UTC)

  Aquillion, I like the way your thinking is trending towards seeing consensus as a spectrum instead of a binary condition.

  Your analysis also needs to stipulate that the content in question is sourced, because anything unsourced can be removed under the WP:CHALLENGE rules. Your approach ends up with contradictory outcomes:

  • I contest your change, and it's a new addition – remove until you prove consensus for it
  • I contest your change, and it's unsourced – remove until you add a source for it
  • I contest your change, and it's about a BLP – remove until you prove consensus for it
  • I contest your change, and it's "old", sourced, and not about a BLP – keep until I prove there's consensus against it
  • I contest your change, and you say it's "old" and I say it's not – ???

  This does not seem to me to be a good systemm. WhatamIdoing (talk) 04:12, 27 December 2020 (UTC)

No, no, of course we still have to consider other policies - there are things that aren't subject to consensus or where a WP:LOCALCONSENSUS (implicit or otherwise) can't violate larger consensus, especially the need for reliable sourcing. An implicit consensus is the weakest tier of consensus anyway. Regardless of its age, something that is unequivocally unsourced has to be removed, tagged, or get a source (although there's already plenty of room debate over whether / when to remove on sight or tag unsourced non-WP:BLP stuff.) But there has to be some caution at the intersection between consensus and policies that aren't subject to consensus, because outside of trivial cases like having nothing even purporting to be a source at all, there's usually room for disagreement over whether policies are being followed. I'm sure we've all been in disputes where one person says "the source for this statement doesn't support it, so I'm removing it" and another person says "wtf are you talking about of course it supports it", at which point we're back at having to hash out consensus. I feel the usual presumption is that, in those cases, consensus will respect our other policies; in situations where it doesn't, we have boards like WP:RSN and the like to quickly call for outside opinions. --Aquillion (talk) 15:16, 28 December 2020 (UTC)

Of course we always have to consider other policies. This particular detail doesn't come into play often precisely because so many other policies solve so many problems. But considering other policies doesn't mean that editors will always form a consensus. I remember the fight over which picture to put at the top of Pregnancy. There really was no consensus for months. So did I need a consensus to remove the long-disputed picture (the view of the WP:QUO folks), or a consensus to keep the long-disputed picture (the view of the ONUS folks)? (What I used in the end was something called a "compromise", in which the disputed picture got moved to a more relevant section, where AFAIK it has remained without a single edit war ever since.) WhatamIdoing (talk) 20:47, 31 December 2020 (UTC)

You never need permission to make an edit (removal). If an editor disagrees and cites explicit talk page discussion and consensus for the content, you are not going to undo their reinstatement, right? You're going to argue the case for removal. In a case where a bad picture was unnoticed for whatever period, never discussed, the onus is clearly on the editor who wants to keep it. But if he can't defend it and convince others of his reasoning, then it should stay out -- whether it's challenged the day it went in or 3 years later. There's lots of longstanding bad content all over WP. Anything worth keeping can be defended today just as it should or might have been defended when it first appeared. SPECIFICO talk 23:01, 31 December 2020 (UTC)

When the disputed content has already been the subject of an edit war, then I think I might prefer having written permission in advance. WhatamIdoing (talk) 08:10, 1 January 2021 (UTC)

• I think it isn't worth worrying about. As Springee obliquely implies, the practical state at the matter is that consensus is required to remove longstanding text in most cases, and I don't think there's any serious danger that that is going to change anytime soon. I also think part of the issue is that ONUS' wording seems common-sense (and therefore makes people skeptical of any effort to revise it to match actual practice); people like to picture, in their heads, a scenario where every line of every article is the result of extended discussion and consensus-building. The fact that large parts of the wiki rely on implicit consensus and would fall apart if we were to weaken that isn't as pretty, even if it's a practical reality that no one is likely to meaningfully challenge or change any time soon. The only change I would push for at this point is to have ONUS unambiguously acknowledge the existence of implicit consensus, which would essentially resolve the dispute; but it isn't strictly necessary to do so because implicit consensus is a form of consensus and is therefore sufficient to satisfy WP:ONUS - ie. if someone tries to remove something you can say "sorry, it's been there for two years and none of the numerous editors viewing or editing this page objected, so ONUS is satisfied." Well... that and making it clear that you cannot solely cite WP:ONUS to remove something; an actual, specific objection needs to be articulated before something lacks consensus. But that's a larger discussion that shouldn't really focus on ONUS and which comes more down to WP:IMPLICITCONSENSUS. I would personally require that someone raise an objection specific to text before the "presumption" of implicit consensus is broken - it doesn't have to be a good objection; any objection at all qualifies, even something as simple as "not an improvement." But I believe removing text without an explanation shouldn't be supported, and in my experience the worst parts of ONUS are when people use it that way (ie. removing text under ONUS without giving any further objection, leaving the person who added it with no clear way to begin discussions.) --Aquillion (talk) 11:40, 19 December 2020 (UTC)

I'm curious what the reaction to the RfC would have been if the only change at all was adding this to CONSENSUS: "Once the WP:BURDEN to demonstrate verifiability has been met, the onus to achieve consensus for changes to longstanding content is on those seeking the change." That should make it clear to skeptical editors that there is no change to the policy. (Not that I'm suggesting launching an RfC.) Kolya Butternut (talk) 16:47, 19 December 2020 (UTC)

• Someone moved WP:ONUS away from its original "burden of evidence" location in this policy. It should be moved back. It used to link to "The burden of evidence lies with the editor who adds or restores material." Now it links to this new sentence: "The onus to achieve consensus for inclusion is on those seeking to include disputed content", which is poorly written and not always the case. SarahSV ^(talk) 00:20, 18 December 2020 (UTC)

  JzG made that change in 2014.[9] Just to clarify, the "burden of evidence" is to demonstrate verifiability, not consensus, per WP:BURDEN. Kolya Butternut (talk) 01:49, 18 December 2020 (UTC)

  JzG wrote: "The onus is on the editor seeking to include disputed material, to achieve consensus for its inclusion (the Bold, Revert, Discuss model)." ONUS and BURDEN used to direct to the same section. It was about whether material was correctly sourced. You're expected to cite your source, quote your source if requested, persuade people that the source is appropriate, and so on. SarahSV ^(talk) 02:49, 18 December 2020 (UTC)

• Why is this still being discussed? Watering down ONUS is clearly a non-starter after that RfC, where the numbers were overwhelmingly against. Crossroads ^-talk- 04:13, 18 December 2020 (UTC)

  It's being discussed because, as written, it has nothing to do with this policy. SarahSV ^(talk) 04:35, 18 December 2020 (UTC)

  Crossroads, the RfC would have made no change to the policy. The proposal even made no change to the text of WP:V. The problem is that there is no consensus on the interpretation of WP:V, but also that the proposal was misinterpreted. This will need to be discussed until there is consensus, but there is WP:NORUSH. Kolya Butternut (talk) 04:40, 18 December 2020 (UTC)

  We could replace "Consensus may determine that certain information does not improve an article, and that it should be omitted or presented instead in a different article. The onus to achieve consensus for inclusion is on those seeking to include disputed content" with "See WP:Consensus for the decision-making process regarding addition or removal of material" as was suggested above. We could remove WP:ONUS from the linkbox but maintain the redirect, or perhaps we could move the redirect to an archive and retire it. Kolya Butternut (talk) 04:55, 18 December 2020 (UTC)

  SarahSV, I'm not sure what you mean. That verifiability alone does not guarantee inclusion, and that there has to be consensus as well, is relevant to verifiability. Kolya Butternut, the fact is that it is unlikely that everyone at the RfC simply did not understand you correctly. It does not "need to be discussed" endlessly, and certainly not until a certain editor is personally satisifed. Eventually editors have to accept that consensus is against their idea. No, we could not make any such replacement, as that is literally what was near-unanimously rejected at the RfC. Crossroads ^-talk- 05:04, 18 December 2020 (UTC)

(edit conflict) I was going to say mostly the same thing before the edit conflict, so I'll just add that WP:STICK does apply here as the RFC had "overwhelming consensus against". If someone wants to discuss the nuances of WP:CONLEVEL, how it's a case-by-case thing, and not something readily legislated in to policy, this is not the place. Kingofaces43 (talk) 05:27, 18 December 2020 (UTC)

This discussion is not actually about the proposal in the previous RfC. I've added new subheadings for these discussions. Kolya Butternut (talk) 05:42, 18 December 2020 (UTC)

Everything discussed since the RfC was discussed at the RfC. Crossroads ^-talk- 03:39, 19 December 2020 (UTC)

The RfC close stated "There is an overwhelming consensus against this proposed change, primarily due to concerns that it would undermine verifiability and efforts to remove poorly sourced information." Those concerns had nothing to do with the intention of the RfC, and could have been addressed by changing the proposal to "The onus to achieve consensus for changes to WP:VERIFIABLE longstanding content is on those seeking the change." If you believe that ONUS contradicts NOCON then clearly there is something to discuss, but I am not going to put forward a proposal. Kolya Butternut (talk) 04:19, 19 December 2020 (UTC)

This changed proposal has no meaningful difference from the one discussed and rejected at the RfC. Crossroads ^-talk- 04:52, 19 December 2020 (UTC)

It is also the exact opposite of what ONUS states. It would reverse the established meaning of the policy. SPECIFICO talk 10:17, 19 December 2020 (UTC)

Apologies if this has already been discussed above: from 2008 until 2014, WP:ONUS was directed to the burden of evidence section in V, now called "Responsibility for providing citations", which says: "the burden [of evidence] ... lies with the editor who adds or restores material". ONUS should be redirected back to that. That's what ONUS is about: sourcing, not consensus. That is, if you add or restore something, you're the one who needs to make sure it's supported by an RS.

In May 2014 JzG edited WP:ONUS to redirect it instead to Wikipedia:Verifiability#Verifiability does not guarantee inclusion. In August 2014 he added the following sentence to that section: "The onus is on those seeking to include disputed content, to achieve consensus for its inclusion", which is about consensus and should not be part of the sourcing policy. In September 2014 Kephir moved ONUS to its current position on the policy page and rewrote JzG's sentence to produce the current "The onus to achieve consensus for inclusion is on those seeking to include disputed content." SarahSV ^(talk) 21:11, 28 December 2020 (UTC)

Good overview, and probably a good basis for sorting this out and fixing it. Such things don't even belong in the verifiability policy. And it is a far-too-oversimplified statement mis-statement of the process for inclusion decisions. The only argument that I can think of for not simply deleting it is that in the fuzzy Wikipedia system it puts a bit of a finger on the scale towards exclusion and deleting it might upset the fuzzy balance of things. North8000 (talk) 21:29, 28 December 2020 (UTC)

The part of ONUS that requires an additon be supportable by reliable sources belongs in WP:V. The part that requires consensus that the addition improves the article belongs in the policy that articles be well written. Oh, wait, we don't have a policy that articles be well written. Jc3s5h (talk) 22:37, 28 December 2020 (UTC)

North8000, I agree with everything you say. I haven't been able to find discussion from 2014 showing consensus to make those policy changes. JzG has posted only twice about ONUS to this page, both times in October 2020 in response to this confusion. [10][11] SarahSV ^(talk) 23:06, 28 December 2020 (UTC)

@SlimVirgin:Probably the only way to resolve this is a truly good proposal. Probably take ONUS totally out of verifiability, re-point it to consensus, and add a bit in consensus to preserve a bit of a finger on the the scale (a bit of a higher bar for inclusion) which the current onus material here provides. And avoid the impossible task of trying to define the fuzzy Wikipedia consensus process with a procedural flowchart. North8000 (talk) 00:07, 29 December 2020 (UTC)

An interesting history lesson, I suppose, but if we wanted to see if there was consensus in 2020 for the current text, we could have an RfC at a central location, say, the Village Pump, to see if we should move it to WP:CON and make it about any change. And that is exactly what was done. There was an overwhelming consensus, in 2020, in favor of the text as it now stands. Crossroads ^-talk- 23:15, 28 December 2020 (UTC)

Crossroads, you began editing in earnest in June 2019. What consequences do you believe there would be, based on your experience at the articles you write or maintain, if this sentence were removed from the policy? SarahSV ^(talk) 23:35, 28 December 2020 (UTC)

I explained this at the RfC. Here it is again:

That WP:PROFRINGE material can persist in underscrutinized areas for years, and what I said about how POV pushers behave, is very much based on firsthand experience. Crossroads ^-talk- 00:19, 29 December 2020 (UTC)

That is very close to what I write below. I agree. I would even put more stress on "as much from what we keep out", because purging Wikipedia from poor quality content is even more important.--Paul Siebert (talk) 00:36, 29 December 2020 (UTC)

If I understand all this the argument runs that we must go with what we have now because if not, WP will end up full of rubbish. I don't see this myself, are we saying that because of the polices we have now there is less rubbish in WP than there would have been otherwise? And that these policies in particular will ensure that there is less rubbish in WP in the future? Is there hard evidence for that? If not, I see no harm in trying out the proposal and see what happens, even if it takes a while to see what happens.Selfstudier (talk) 18:59, 29 December 2020 (UTC)

@Selfstudier: No, I don't think that's quite the point. All this short sections says is that the argument "It's Verified, therefore it should be in the article" OR "should not be removed from the article" is not sufficient for inclusion. There's vastly more verified fact in the world than significant verified fact. Verified material must be DUE NPOV content per WP:WEIGHT. That is all this little ONUS section says. It's not about consensus -- every content decision is about policy and consensus. It's just giving us a convenient link to cite in case an editor defends UNDUE text by claiming that Verifiability is sufficient for inclusion. SPECIFICO talk 19:07, 29 December 2020 (UTC)

are we saying that because of the polices we have now there is less rubbish in WP than there would have been otherwise? And that these policies in particular will ensure that there is less rubbish in WP in the future? Is there hard evidence for that? Yes, yes, and reasonably yes: right now Wikipedia has much less crap than it could easily have, and less of a proportion of it than in the past. We shouldn't make it easier for crap to stay, and such an experiment could be very damaging in underscrutinized areas (most of the encyclopedia). What benefits could possibly be brought? See WP:FENCE. Crossroads ^-talk- 19:11, 29 December 2020 (UTC) Updated Crossroads ^-talk- 19:43, 29 December 2020 (UTC)

You are starting with Yes, and I began with No to say the same thing. At any rate, I would go so far as to say that the vast majority of WP pages are full of rubbish or undue narratives. But the vast majority of actively read and edited pages are not. Yes the high quality of our articles on the most important subjects is thanks to, and protected by, NPOV and ONUS. Among other things, it means that a single NPOV editor can clean up an article that's full of biased or trivial content. When the content is appropriate and valid, it will readily gain explicit consensus. We can't be protecting content that was never duly scrutinized by more than a handful of (possibly long-ago) editors. SPECIFICO talk 19:24, 29 December 2020 (UTC)

That's exactly what I am getting at, you say that most of the encyclopedia is underscrutinized, not enough editors, presumably. So that is the problem, not what these policies say or don't say, precisely.Selfstudier (talk) 19:22, 29 December 2020 (UTC)

SPECIFICO, I've added a quote to clarify what I was saying yes to. I presume you said "no" to the general point being made. Selfstudier, we will never have "enough editors". We need to not hinder cleanup by the editors we have. We'll never gain more good editors if POV pushers are running wild and good editors are thwarted by a draconian revised ONUS, anyway. Crossroads ^-talk- 19:43, 29 December 2020 (UTC)

I agree. It is not a matter of not enough. It's that in any open community, a certain number of good faith participants are guided by their enthusiasms rather than by a larger contextualized perspective. SPECIFICO talk 20:01, 29 December 2020 (UTC)

OK, I understand what you are saying, I am not convinced that there is an army of POV pushers hiding somewhere in anticipation of this change going through. I think they are already here and doing it anyway.Selfstudier (talk) 22:40, 29 December 2020 (UTC)

Yeah, but we shouldn't make it easier for them so more of them stick around and they grow in number (plus there's the matter of cleaning up past POV pushing). Crossroads ^-talk- 23:44, 29 December 2020 (UTC)

No army. But look on thousands of articles about video games, rock bands, TV actors etc., and you'll see all kinds of trivia and non-noteworthy analysis. The same thing is all over the articles on current events, social science, and politics. SPECIFICO talk 00:24, 30 December 2020 (UTC)

ONUS and Pfister's destabilisation

It think the core idea of ONUS opponents is that they assume that if some content is "long standing", it is thereby vetted by Wikipedia community, which is seen as some analogy of a conventional peer-review procedure. Although they do not say that, I am 99% sure they imply that. However, there is one huge problem with that assumption. The problem is that Wikipedia rejects common informational routines (so called "Pfister’s destabilization"), and, what it worse, it does that without replacing them with much that could be construed as progressive (read this for more details). There is no such category as "expertise" in Wikipedia, and I know several cases when users who were real experts in some topic were forced to stop editing (or just topic banned) because of their disagreement with other users, who were not experts. Therefore, it would be fundamentally incorrect to see any analogy between a long standing content in Wikipedia and peer-reviewed scientific/scholarly sources. The reason is simple: we (under "we" I mean not real life persons, but Wikipedians) are, by default, not peers. I am not an expert in the topics which I edit here (I am en expert in totally different topics, which I edit only occasionally), and I am sure most of us are not experts too. We are not peers (I mean peers with similar level expertise), and the fact that some text has not been challenged/removed/criticized by us for 10 years does not mean anything. It would be unfair to reject some commonly accepted informational routines when we writing something, and refer to those routines when we are defending what we have written.

In that situation, the only acceptable way would be to capitalise on Wikipedia's strengths: first of all, it is living and constantly changing encyclopedia, and if you are ready to write something, be prepared to defend the text you wrote. Under "defend" I mean not a single act when some consensus is established to keep your edits. I mean a permanent work on watchlisting the article you edited and defending your content from all criticism, new arguments and sources. If you are not ready to do that, be prepared that the content you created will be modified or removed. I think that is absolutely fair, and that is why ONUS must be preserved in its present form or even additionally emphasized.--Paul Siebert (talk) 22:53, 28 December 2020 (UTC)

Do you think that we could write something that would be better than what we have? I think it's possible, and that it might be worth doing. WhatamIdoing (talk) 20:57, 31 December 2020 (UTC)

I think the problem is just that the ONUS section, regardless of what wording or wikilink is used, is a very simple compact message. It says that one cannot justify article text simply, merely, only, on the premise that -- it is Verified in a Reliable Source. All the other stuff about removals, about consensus, about anything else is neither implied nor appropriate in the ONUS section. Consensus is how we do everything. Removals are OK, and then if they're contested we work it out. But there are no perfect articles on Wikipedia at the moment, and our strength is that anyone is welcome to try to improve articles by editing them. That needs to be protected, because that's what has gotten us as far as we have come today. SPECIFICO talk 22:45, 31 December 2020 (UTC)

Yes. I think the procedure is simple and straightforward:

• If some new content is added, ONUS works as usual;

• If some content was added some time ago, and it survived during a consensus building process, then two scenarios are possible for its removal:

1. The arguments presented by a user who removed the text were already presented and rejected during the previous consensus building process. In that case, removal should be reverted (and references to similar arguments presented during the past discussion should be presented upon a request).

2. The arguments presented by a user who removed the text are new, and they were never discussed during the previous consensus building process. In that case, those who want to keep the removed text are expected to address new arguments, and references to some previous consensus are irrelevant.

In both cases, those who want to keep the text are supposed to provide counter-arguments (either references to previous discussions or new facts and sources).

Therefore, the very fact that some content was existing for long time is not supposed to affect the decision about keeping/removal.--Paul Siebert (talk) 17:27, 8 January 2021 (UTC)

Yes, it is absolutely clear and will work very well if a page is well attended or curated by our insufficent number of editors. Else someone with an agenda can come along and remove perfectly good material unchallenged ie the inverse of inserting perfectly bad material unchallenged.Selfstudier (talk) 18:05, 8 January 2021 (UTC)

The #Reliable sources section currently includes at the top:

The middle sentence about <source> tag and syntax highlighting (present since 2013) is irrelevant here and useless, detracting from valid notices before and after it. Further, there is near-zero likelihood that anyone looking for information about syntax highlighting would end up here first and need advice about where to find it. If by some freakish coincidence someone *did* end up here and followed the link, they would find nothing about <source> on the destination page where they did end up, making it a surprising link. In fact, there's been nothing about <source> there since 2010 when tag <source> was deprecated. This redirect advice serves no purpose, makes the top notes harder to understand, and should be removed from the page. Mathglot (talk) 01:12, 3 January 2021 (UTC)

irrelevant here and useless No, it's not. If I were to do something like <source>, I am clearly not referring to reliable sourcing, and you should want people to be able to navigate away from this page so that they can understand what <source> refers to. We're not worried about "people ending up here" by e.g. search or curious wandering, we're worried about people being pointed here.

I don't really understand how I could be more clear. It is not a surprising link because the confusion, if there is any, is fixed by the note itself, when the person clicking WP:SOURCE ends up on this page. He reads "syntaxhighlight aka source" and then if he clicks that link he understands he's going to read about syntaxhighlight, not source.

Your primary worry, that more important hatnote direction is obscured, is easily fixed just by moving the latter statement forward one block. Which I've now done.

And as for source going unmentioned there, it's still in active use, even if deprecated. Maybe when it's been removed from wikitext, and a year or two later, it would make sense to ditch. --Izno (talk) 01:37, 3 January 2021 (UTC)

(edit conflict) You said,

If I were to do something like <[[WP:SOURCE|source]]>, I am clearly not referring to reliable sourcing^{[nowiki tags added]}

You are 100% right. And that is 100% irrelevant, and unhelpful here, and detracts, because it is an imaginary use case, a solution in search of a problem that does not exist in the real world.

Your use case is that of the former Policy page wonk who moved to Pitcairn's Island in 2009, returns in 2021, and his memory is so good, that he actually recalls the <source> tag which disappeared ten years ago, and the first thing he does, is to edit a project page and create the link <[[WP:SOURCE|source]]>. Lucky for him, we anticipated his reentry confusion, and left him this helpful note. I wonder what other pre-2009 features of the encyclopedia we should annotate, in order to help the poor guy out.

The confusion when ending up at Help:Wikitext is exactly the confusion about not finding anything at all relating to the note that got you there. You could "fix" that, by mentioning the ancient, deprecated <source> tag in the syntaxhighlight row of the table, but really, that would just be compounding the confusion. It *is* a surprising link because out of the 278,000 active editors, there aren't more than a handful who will remember anything about the <source> tag, so if they click that link, they will be surprised. (I certainly was, though I've been here since 2006 and forgot the <source> tag; dang!) And I would venture there are zero users who will get to WP:V via the imaginary use case you raised.

Finally, if it's important to cover this vanishingly rare use case, one could easily come up with others; the word source means lots of things; the source disambig page lists many dozens of them (but syntax highlighting is not one of them); it's a highly overloaded term, you could pile up any number of "helplful" things to add to a notice here which are also irrelevant. Maybe one for folks that think it means "source code" and ought to link them to the <code> tag, or perhaps to a project page about Mediawiki software. Or for copying and other from-to operations, people say "from <source> to <dest>" all the time; I've even done that. Maybe we should expand the note to link to the {{copy}} and {{translate}} templates. Should I go on? None of those would be helpful here; all of them are more likely destinations than "syntax highlighting" for the word "source". Linking to syntaxhighlighting here is absurd.

Moving the useless link to last position in the list helps, thanks for that. But it's still completely useless and arbitrary. Remember what the template is for: "Its purpose is to reduce confusion by helping users get to their intended page." On this page, it doesn't help anybody get to the right page (except Pitcairn dude), and it increases confusion. Mathglot (talk) 02:20, 3 January 2021 (UTC)

You are mistaken if you think that because of <source>'s supposed deprecation in 2010 that people are not using it today. Despite its so-called deprecation, the tag continues to function and be used. Here's the tracking category, too. It's only through some effort in the past year (singular) that that has become a wasteland of archives. (See phab:T237267.)

I don't see a reason to reply to your third paragraph. It is nonsense.

Actually, sure, if you think there are other possible confusions, either add them or give those people a potential landing page for all the uses of source that might exist solely in the context of wiki-terms (because I agree, we should have bounds). Could be WP:Source (disambiguation). Could be something else. --Izno (talk) 02:49, 3 January 2021 (UTC)

@Izno, is it a simple enough substitution that the tracking category could be emptied (at least of non-.js pages) by a bot? WhatamIdoing (talk) 21:33, 13 January 2021 (UTC)

I wouldn't bot it without human overview but it's otherwise a trivial find and replace 1-for-1; the problem of course is that (even) the string <source will have false positives in and among the pages on HTML. (And since it's on talk pages, we would need to be careful for the example and help kind of discussion, I suppose.) --Izno (talk) 21:50, 13 January 2021 (UTC)

So it'd be better to handle it as an AWB run than a bot. I wonder if the WP:CHECKWIKI editors have this on their list. WhatamIdoing (talk) 22:42, 13 January 2021 (UTC)

The "policy" concerning "Newspaper and magazine blogs" WP:NEWSBLOG is too vague to be useful.

The policy is as follows:

Some newspapers, magazines, and other news organizations host online columns they call blogs. These may be acceptable sources if the writers are professionals, but use them with caution because blogs may not be subject to the news organization's normal fact-checking process.

The concern with this policy is that it does not define "blog" with any precision. Simply saying a "blog" is an "online column" isn't much of a definition. Yes, it refers to the Wikipedia article Blog, but that is whatever Wikipedia editors decide it might be, whenever they decide to alter it with having the constraints involved when writing a Wikipedia policy.

To put this concern into context, I provide the following scenario:

This vague policy was not at all helpful when I encounter an editor claiming a Washington Post article was not creditable because it was from a "blog".

The article is promoted by the Washington Post on a page called Morning Mix - Stories from all over. It explains itself as "The Washington Post's Morning Mix blog covers stories from all over the nation and world."

The Washington Post calls a collection of links to some of its newspaper articles a "blog". It is an unfortunate selection of a name to define its genre. It isn't a "blog", it doesn't look like a blog, it doesn't smell like a blog, and it does not act like a blog.

The article in question is:

• Elfrink, Tim (August 14, 2020). "'Do you regret at all, all the lying you've done?': A reporter's blunt question to Trump goes unanswered". The Washington Post. Retrieved December 31, 2020.

I am involved with editor @PackMecEng: who claims that because its source was called a "blog" it was not a creditable article. The editor continues to argue this position using only his opinion as support. He particularly dislikes the sentence "Donald Trump has been a prodigious spreader of misinformation" that I want to use in the lead for the article Veracity_of_statements_by_Donald_Trump. He has posed various arguments, but in the end, his final argument is that the article is from a "blog" and is not creditable.

I disagreed and a long, long argument has ensued concerning the news article's creditablity. This discussion can be found at [Talk:Veracity_of_statements_by_Donald_Trump#Sentence_%22Donald_Trump_has_been_a_prodigious_spreader_of_misinformation%22_in_Lead_is_an_issue_for_Editor_PackMecEng].

It would be very useful for the WP:NEWSBLOG policy to define "blog" more clearly.

It seems that such a definition would include, but not be limited to, the key elements of a blog, such as:

Immediate access to readers

Highly interactive

No set deadline or publishing schedule

No fixed length

Relies on comments

More casual in tone

Continuous conversation

Concerning the creditability of this article, I will post this discussion on WP:RSN.

How would one go about improving this policy visa via the definition of what constitutes a "blog"?

Osomite ^hablemos 01:58, 3 January 2021 (UTC)

@Osomite: you mean, link to it from RSN, not post it on RSN right? Mathglot (talk) 02:36, 3 January 2021 (UTC)

@Mathglot:, Thank you for coming to my aid. I have been editing for a long time but in such matters, I am still quite puzzled at times. I have never been involved with RSN before.

I am not sure. A link? I definitely can put a link in a post of RSN.

However, it seems that to correctly request help at RSN there are a few items that need to be added that are not included in this post. If I only link, I would undoubtedly have to then add some more information. It seems that when posting on RSN it would be cleaner if all required information were included on the original post when requesting help to establish the creditability of the Washington Post article.

Could you clear this up for me with more info about doing a link versus a post?

Osomite ^hablemos 02:53, 3 January 2021 (UTC)

@Osomite:, sorry I was unclear, I only meant to have your discussion wherever it makes most sense to have it, but just in one place, rather than fragment it by starting the identical conversation in two places, which might then diverge when some people respond here, and some respond there. Whichever page you decide to place it on, you can just link to it from the other page in a very brief section, possibly with the same section title, and just one sentence or so, saying something like, "Your feedback about <whatever your topic is> would be appreciated at [[PageName#SectionName|this discussion]]"; that's all I meant. Mathglot (talk) 03:02, 3 January 2021 (UTC)

@Osomite, you posted this last week. Were you able to get this dispute resolved, or do you still need help? WhatamIdoing (talk) 21:35, 13 January 2021 (UTC)

@WhatamIdoing,Thanks for asking.

About the edit dispute concerning the use of the sentence from the Washington Post article with PackMecEng, I have let it go. PackMecEng was very argumentatively argumentative and had zero inclination to collaboration. It was strictly PackMecEng's way with no give. Since it was just a single sentence, I left the issue as it was quite uncomfortable and distasteful trying to work with PackMecEng about getting any movement about the use of the sentence.

The issue I posted here about the definition of what is and what is not a "blog" has had no interest. My concern about the policy's ambiguity concerning the lack of an accurate definition of a "blog" still stands. About this, I would appreciate some help. Osomite ^hablemos 02:08, 14 January 2021 (UTC)

I wonder whether there might be an incentive (possibly even a perverse incentive) to avoid a strict definition. If it's strictly defined, then we might draw the line in the wrong place, with the result that Bad Blog is included and Good Blog is excluded. Where it stands now, we can argue it out ...with the downside that it isn't always (or even usually) a pleasant process, as you've seen, and there is no guarantee that the good sources will be accepted and the bad ones won't under this system, either.

AIUI the general rule is that a blog is bad (and specifically, that it falls under Wikipedia:Self-published sources) until proven otherwise, and NEWSBLOG is meant to permit some level of use for the blog-like publications seen at major news websites, on the grounds that the editors still look them over, even if they don't fully fact-check them. (Quite a lot of news stories aren't fact-checked; the "breaking news" headline is meant in part to be a disclaimer.) NEWSBLOG moves the major news blogs move from "blogs are bad" status to "maybe, possibly, perhaps okay" status. In general, the blogs at Washington Post or similar large newspapers would be considered "maybe" usable, taking care not to overstate the claim, but that doesn't necessarily mean that they're usable in any specific article or for any specific claim. (Also, if you're trying to get something into the lead of an article about Trump, anything with "only" one good source behind it might be WP:UNDUE.) WhatamIdoing (talk) 02:39, 14 January 2021 (UTC)

Moved to WT:Cleanup § Template:Source conflict

Template:Source conflict (edit | talk | history | links | watch | logs) FYI, a new template has shown up. -- 70.31.205.108 (talk) 01:22, 9 January 2021 (UTC)

Based on the discussion at WT:Cleanup#Template:Source conflict, I've sent this new template to Wikipedia:Templates for discussion/Log/2021 February 21. WhatamIdoing (talk) 01:45, 21 February 2021 (UTC)

I would like to add this to the section "Accessibility/Access to sources":

I think this recommendation is needed because there are lots of examples where content is put in or has been edited so that it no longer mirrors source material. If the source material is very hard to get at it discourages verification and erodes the quality of wikipedia. Thoughts?Efbrazil (talk) 18:56, 6 April 2021 (UTC)

Such a standard would be unacceptably high barrier to participation, and against current consensus that offline and paywalled sources are being cited in good faith: there are very few instances where a source is entirely inaccessible to a reasonably research-literate user; quotation is subject to manipulation as much as summary; and net such a recommendation would only put a high burden on good-faith contributors, not bad actors. Moreover, it probably would violate WP:Fair use -- because many sentences are based on large sections of sources. The language as stands sets a high enough standard for challenging specific language. Sadads (talk) 19:09, 6 April 2021 (UTC)

How is asking someone to copy/paste a couple sentences for a quote "a significant barrier" or "high burden", particularly compared with the burden of needing to register or pay for a source to verify content?

Quotation is not as easily subject to manipulation as paraphrasing is and will discourage sloppy use of sources. If wiki content is based on large sections of sources it is fair to say that in the reference, although I find that to rare be the case (typically a source will summarize itself in an intro or conclusion). The point is that the reference should make it clear what the connection to source is, since the reader and other editors cannot. Efbrazil (talk) 20:14, 6 April 2021 (UTC)

You can use {{Template:Request quotation}}, I think there's another thing you can do as well but I forgot what it is.Selfstudier (talk) 21:41, 6 April 2021 (UTC)

The standard is that someone can check the information in a source, not that they will be able to do it easily-- this is the historical function of citations in scholarly writing more generally. Of course we should connect with sources that are open where available, but finding an open version or cluttering an article with walls of quotation is not exactly any more useful that not having them -- quotations aren't always useful out of context. As Selfstudier and Femkemilene who you are asking for particulars on in an article offered -- requesting for a contested statement is fine -- but putting that in policy at the level of should or even we recommend trying creates a much higher standard than any other writing environments -- and would be used as yet another barrier to new folks participatin in Wikipedia. Sadads (talk) 22:04, 6 April 2021 (UTC)

I don't think this is needed, but I do think that editors should be aware that if they are sourcing a highly contentious statement that can only be sourced through a paywall or similarly hard-to-access work, quoting the relevant part is absolutely helpful. Mandatory, no. --Masem (t) 22:12, 6 April 2021 (UTC)

There are lots of ways to get information. Ask for help with a particular paper. Ask another editor to send you the relevant section (e-mail is best if it's more than a couple of sentences). Get your free Wikipedia library card. Call your local library's reference desk. Write to the publication's author, if they have public contact information. It can take some effort, but it's not impossible. WhatamIdoing (talk) 04:04, 9 April 2021 (UTC)

I have a dispute at with someone who claims that an article sourced with only non-English sources is equal tot a non-sourced article. He cites this Wikipedia:Verifiability#Non-English sources for this. To my opinion, this policy prefers English-language but does not disallow or ignore them. Am I wrong in this? The Banner talk 14:27, 5 April 2021 (UTC)

I wouldn't think so but I can see some serious issues. First, if the topic isn't discussed in any English language source would it be considered NOTABLE in an English language Wiki entry? I don't think NOTE says anything about language but I would throw that out. V is a bigger concern since it would be hard for most editors to check the content. It's one thing for a source or two but for the entire article that puts a big burden on others. Springee (talk) 14:41, 5 April 2021 (UTC)

The notability rules have zero restrictions about the language of the sources. If having the leading role in three films passes WP:NACTOR, then having the leading role in three films passes NACTOR even if those films are never reviewed in English. WhatamIdoing (talk) 03:46, 9 April 2021 (UTC)

Translated non-English sources are fine, I am having a bit of trouble imagining an article where no English sources at all can be found. What article is that? Selfstudier (talk) 14:42, 5 April 2021 (UTC)

I've seen examples of this. Often, it's a book, film, etc. that only exists in a non-English language. A book of poetry could be quite thoroughly reviewed in the native language, and yet not ever mentioned in English-language media. WhatamIdoing (talk) 03:42, 9 April 2021 (UTC)

• You are correct... while we prefer English-language sources, we absolutely accept reliable sources written in other languages. This is true for establishing both verifiability and notability. That said, it is often difficult for our editors to assess the reliability of non-English sources, so the push back is understandable.

THAT SAID... it seems that this query stems from an edit war at Barn church. So let me address the specifics... there are two sources cited. Both are Dutch websites with a “view in English” option. Thus, the entire debate is premised on flawed argument. These sources qualify as English-language (admittedly, the English version it is a poor machine translation, but we don’t require perfection... the key is that the website itself has an English version).

What is less clear is whether the cited sources are reliable... I would asses one as borderline, and the other as unreliable. Blueboar (talk) 15:22, 5 April 2021 (UTC)

• • To add to what Blueboar is saying: from the point of view of verifiability and notability our sourcing threshold is blind to the language the sources use, just as they are to whether the source is online. From the point of view of maintainability, though, it's a bad sign if we don't have English-language sources. For this article, it's another bad sign that all the sources are in Dutch, but there is no cross-referencing of a Dutch-language WP article on the topic, well-maintained or otherwise. If this were to come to AfD, this absence would bother me. — Charles Stewart (talk) 11:24, 6 April 2021 (UTC)

I have added a link to the Dutch Wikipedia article on the topic (so that issue, at least, has been fixed). User:Banner assures me that scholarly sources that are actually written in English do exist, so... give it some time. Blueboar (talk) 11:52, 6 April 2021 (UTC)

To the above, I would argue that if the central tenet of the article is of extraordinary or contentious claims that are in more obscure foreign sources with no English collaboration, that the lack of sourcable English translation or English sources can be a problem, particularly if a BLP. For example, a foreign businessperson that may only really be notable due to involvement in some scandal - not at the highest level of gov't but say at a city level. We're really want something that's in English to confirm that if the foreign sources are difficult to access and/or translate and would better not to have the article until that happens. --Masem (t) 03:49, 9 April 2021 (UTC)

I don't think we should impose separate standards for non-English sources. If we don't require that claims made in a "difficult to access" English-language source, be pre-verified by a second editor, then we shouldn't do that for a non-English source.

I don't recall ever encountering a non-English source and not being able to either find someone to give me a good translation, or to figure out enough from machine translation the overall gist of the source. Machine translation isn't great (although it's much better than it was ten years ago), but you can usually tell the difference between "Mr Big is arrested" and "School funding is debated" or "This is a good restaurant". WhatamIdoing (talk) 00:06, 10 April 2021 (UTC)

I could use some help here, because I've been working to maintain the integrity of the List of suicides article for many years no, in particular with regard to WP:V. I just discovered something I hadn't noticed or thought of before: There's about 30 citations of Wikisource in the article, and I'm not sure that's right. Does that violate WP:CIRCULAR, or is CIRCULAR only violated when a Wikipedia article is cited in another? Nightscream (talk) 14:40, 6 April 2021 (UTC)

Wikipedia is not a reliable source because it is editor-generated. Wikisource holds published sources that are no longer under copyright, not editor created articles. The citation is actually to the published source, and the Wikisource link is just a convenience link. For example reference 53 was published nearly 2000 years ago. {Seneca the Younger (ca 40 AD). "Chapter X". Of Consolation: To Helvia. Wikisource.) StarryGrandma (talk) 20:13, 6 April 2021 (UTC)

On the general question: It's only a problem with Wikipedia articles. In particular, it's very bad if I write in one Wikipedia article that WhatamIdoing is a very nice person, and then I go to another article and write WhatamIdoing is verifiably a very nice person^[1] with a "citation" back to the first article.

Far more common is that an editor will write something without sources in one article, the article will get copied to another site, and then, years later, a second editor will notice that it's unsourced and innocently find and cite a copy of the Wikipedia article (often a prior version) on a mirror.

The problem with citing the non-Wikisource sister projects is that they are user-generated. WhatamIdoing (talk) 03:53, 9 April 2021 (UTC)

• To summarize:

1. Don’t cite directly to another en.WP article - circular. Instead, see if the information is cited at the other article, and copy that citation.
2. Don’t cite directly to other language WPs - user generated. Instead, see if the information is cited at the other WP and copy that citation.
3. Don’t cite directly to Wikisource - instead cite the original, and LINK to Wikisource as a “convenience link”. Blueboar (talk) 11:45, 9 April 2021 (UTC)

As a related question, what is the best way to handle cases where the sourcing for a claim/statement is in the primary topic article but the claims/statements are used in a different article. For example in Mr BLP's article sources say he is considered a dog-phobe (a negative trait if their ever was one) with all the needed citations. In an article about dog breading laws we say "[Mr BLP], a dog-phobe, said..." with no citations to support the contentious dog-phobe claim. How should such a case be handled? Is "dog-phobe" considered a summary of the primary topic? Do we need multiple citations (we don't call people dog-phobes lightly or with just one or two sources)? Should we just avoid the descriptor? If the topic were banking someone's anti-K9 views certainly could be seen as off topic but if the topic is a new leash law I can see how someone's alleged dog-phobia could be seen as relevant to their POV. Mr BLP's article is linked and the sourcing for the claim is there but is that enough? It is a problem of course if the BLP attributes the claim but the article mentioning Mr BLP puts the claim in wiki voice. At the same time, loading up the other article with citations needed to justify the phobe label seems odd/undue/pointy. Springee (talk) 13:01, 9 April 2021 (UTC)

One thing to consider... we can not guarantee that a cited statement which currently appears in article X (say a BLP) will remain in article X... Article X may undergo a re-write where the statement (and its citation) is omitted.

It is therefore important that if the statement also appears in Article Y, the citation supporting it also appear in article Y. Otherwise the information would end up being unsupported.

In other words, every article needs to stand on its own. Information should be cited in EVERY article in which it appears. So, if the BLP contains a cited claim, and this claim is repeated in some other article, it needs to be cited in BOTH articles. This is not onerous... just copy and paste the citation. Blueboar (talk) 13:41, 9 April 2021 (UTC)

I would add: If you see someone else adding information to article #2 that matches what's already cited in article #1 (and you happen to know that it's in article #1), please don't revert them. This isn't a game of Simon Says, with a goal of sending people back to the start if they don't get it right on the first try. Just copy the refs over, or encourage them to do so. WhatamIdoing (talk) 00:10, 10 April 2021 (UTC)

Springee, maybe Blueboar and WhatamIdoing are way ahead of me but I'm not clear on what your central question is. To me it seems that it might be asking that if the statement about a person is negative to the extent that a BLP requires strong sourcing and attribution to include, do the standards for inclusion and use of the wiki voice get lowered if the article is about a different topic. ? ? North8000 (talk) 19:03, 26 April 2021 (UTC)

Should WP:ECREE have a hatnote pointing to Sagan standard (which has a hatnote pointing to it)? —2d37 (talk) 09:43, 16 June 2021 (UTC)

Probably not. We don't usually point from project-space to mainspace. WhatamIdoing (talk) 22:46, 16 June 2021 (UTC)

For example, a self-published blog entry which states a set of principles which years later are practiced by lots of people. When the practice gets a Wikipedia page, shouldn’t the original source’s statement be included in the Wikipedia article as well as published works in journals?

The scholarly journal articles redefine the principles and get used as the Wikipedia article definition of the topic. The scholarly journal articles refer to the url of the original source.

This will defocus the Wikipedia article from stating the original set of principles as found in the original source.

Should the Wikipedia article state the original principles as well as the scholarly interpretation of them; or just the scholarly interpretation? — Preceding unsigned comment added by 2600:1700:d591:5f10:9db9:1bbc:cb51:65da (talk) 17:22, 25 June 2021 (UTC)

This policy alone does not dictate the answer. And there are too many unknown variables to derive any answer from just your post. So IMO this would need to be discussed at the article in question. North8000 (talk) 18:35, 25 June 2021 (UTC)

Generally, the answer is "no". It's not banned, but we have very little interest in "crediting" the original source. If the article is primarily about the practice, then it would not be unusual to include that post in the Wikipedia:External links section, but we'd generally prefer sources that talk about how important the blog post was, rather than relying on the primary source itself. WhatamIdoing (talk) 23:19, 5 July 2021 (UTC)

BURDEN says "Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source." I take it that "directly" here means "followed by an incline citation"? If so, then there should be one exception to this rule, namely LEADCITE: if content is sourced in the body, then the lead does not require an inline citation, although it is recommended for controversial content. Shouldn't we add and link this caveat to the BURDEN section? Otherwise wikilawyers may delete lead content that is sourced in the body citing BURDEN. ‎⠀Trimton⠀‎‎ 02:53, 30 May 2021 (UTC)

Or does BURDEN (policy) trump LEADCITE (a mere MOS section)? Does this mean MOS:LEADCITE should require inline refs for all lead content? ‎⠀Trimton⠀‎‎ 02:56, 30 May 2021 (UTC)

I don't think "directly" here means "adjacent". It means that the claim itself is supported by a source, and not just implied. But in some cases, such as in controversial articles where claims in the lead are frequently questioned, those statements are in fact backed up by adjacent cites in the lead. – bradv🍁 03:00, 30 May 2021 (UTC)

I would argue that we'd want LEADCITE to overrule BURDEN, but this readlly depends on the quality of the writing of the article. Ideally, if the article and lede are well written, to the point where the highly controversial material is well-documented and sourced in the body (along with clear section headers), and a well-summarized lead, then LEDECITE should be appropriate to avoid citations in the lede - the idea being that it should be very easy to find from the lede where in the body the citations for summarization of contentious statements can be found in the body. (For example arguably the statement at Ishtar (film) that says "Ishtar is one of the worse films ever made" need not be sourced in the lede because I can go to the reception and legacy to find these sources to support this.) This would only leave the required use of citations for direct quotes to be left. The issue that tends to happen is that ledes and bodies often are not well written to be clear where a contentious statement in the lede can be reviewed in more depth within the body, and in such cases it is better to repeat the sourcing in the lede until the opportunity is made to improve this. I will say that it would be inappropriate to apply BURDEN to an unsourced lede statement when its clear its likely repeating information in the body (that just calls for source reuse to fix). --Masem (t) 03:58, 30 May 2021 (UTC)

@Trimton, I've added two more instances of the footnote, so it should be easier to find next time. Please read Wikipedia:Verifiability#cite note-:0-2. As @Bradv said, a reliable source directly supports the content when the reliable source actually contains the claims made in our article. The location of the footnote/link/citation does not matter.

On the more general question, which @Masem hints at, we have never actually written down a rule that says it's sufficient for a given piece of information to be cited a single time, even if it appears multiple times on the page. The closest we come, and one of my favorite examples, is the description of the elephant at WP:REPCITE. WhatamIdoing (talk) 03:59, 7 June 2021 (UTC)

Thanks both :) ‎⠀Trimton⠀‎‎ 05:52, 10 June 2021 (UTC)

I just saw this and was about to comment that MOS (a guideline) cannot "overrule" a policy then saw it was an Admin just making comments. I like an uncluttered lead also as it make even a C-class article look better. "If" a source is in the body and also the lead I will take it out and leave a summaryand maybe even a hidden note. "If" the content is not in the body but referenced in the lead it can be moved per lead right? I mean, the lead is a concise summary of material cited (supposedly) in the body. -- Otr500 (talk) 13:48, 5 July 2021 (UTC)

Admins have no special authority on content matters, just to be clear.

And of course, MOSes can't override a policy, but every time I've seen an apparent conflict, that conflict was due to an extremely broad reading of either the policy or the MOS/essay/guideline. A reading which rarely gets any consensus, I might add.

The example that immediately comes to mind is WP:MANDY, which is an essay about WP:DUE weight. It might be argued that MANDY contradicts our WP:NPOV and WP:BLP policies, because it seems to argue for the exclusion of responses to allegations by a BLP, but the truth is that it's just an essay about how to gauge whether such a response is DUE. If the response is utterly unsurprising and doesn't shed any new light on the allegations, then "Well, they would say that, wouldn't they?" applies to the contents of said refutation. But when a refutation includes new information or manages to otherwise shed new light on the situation, then Mandy Rice-Burroughs would clearly not apply. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 13:59, 6 July 2021 (UTC)

Either way, "directly" refers to the nature of the support and not the location of the cite. North8000 13:56, 5 July 2021‎ (UTC)

In May 2013, following Wikipedia talk:Verifiability/Archive 60#Wikipedia and sources that mirror or use it, Aymatth2 (talk · contribs) made an edit that gave privilege for content related to Wikipedia and sister projects to reference themselves, and this was kept even though this conflicts with other policies and guidelines such as Wikipedia:No original research WP:USERGENERATED. Aymatth2 said

"This proposal started as a simple suggestion that sourcing from Wikipedia (when allowable) should be clearly identified. It turned into a productive review of the entire section that showed various issues, and several editors have given valuable feedback. Despite the title, it was not obvious that the scope is Wikipedia as a whole, not just articles. The section was missing explicit warnings about undue emphasis and original research, and the wording on self-reference and circular sourcing was incorrect. ALT4 tries to address all these problems. Assuming no major objections are raised, I propose to implement ALT4"

As it turns out, if you go to Wikipedia, and before my edits, there are many references to Wikipedia that DON'T clearly state for the reader that is according to Wikipedia policies and guidelines; some statements don't even have proper citation templates. Another bigger problem is that it creates superficial notability to Wikipedia topics that are WP:USERGENERATED (i.e. Wikipedia and sister projects are NOT reliable as a primary source), and even more since it increases the risks of creating circular reporting for Wikipedia and sister projects topics. This is WP:COI. No other topic has privilege like this, only Wikipedia and sister projects are given this privilege.

As such, the paragraph in WP:CIRCULAR which gives privilege to Wikipedia and Wikimedia sister projects shall be removed.Hfnreiwjfd (talk) 21:34, 6 July 2021 (UTC)

• I am not sure that I understand your concern. Any source is a primary source for verifying a statement about the content of that source (purely descriptive statements, direct quotes or close summations, etc). Are you saying that WP is different, and NOT reliable for verifying descriptive statements about its own content, or quotes? Blueboar (talk) 22:14, 6 July 2021 (UTC)

@Blueboar: Wikipedia and sister projects are WP:USERGENERATED, and as such they are NOT reliable primary sources. Because of this, Wikipedia and sister projects were given special privilege to be used as primary source, even if it is unreliable. Is there policy giving such special privilege for other user generated websites such as Everipedia? It doesn't seem so.

Wikipedia community generates its policies and guidelines, so it is in WP:COI when the community gives privilege to Wikipedia and sister projects to reference their user generated policy, guidelines, etc. This results in aspects of Wikipedia and sister projects covered that don't show notability by referencing reliable sources, but with the given privilege they are included.

In practice this privilege is abused and not clearly stated. For example, on Wikipedia it is not clearly stated for the reader that many statements are according to Wikipedia policies, etc. Further, there are sections in Wikipedia that mostly reference its policies and guidelines, which is contrary to the recommendation of WP:RSPRIMARY.

Hfnreiwjfd (talk) 22:44, 6 July 2021 (UTC)

• Self-published sources are legitimate, even if not particularly reliable, as long as the source is spelled out and the information is relevant. E.g.

Company ABC claims they make the best widgets in the world.<ref>https://Company_ABC_website/Puffery</ref> However....

Similarly, in an article about some aspect of Wikipedia, it is legitimate to cite relevant Wikipedia content, as in

According to Wikipedia's Verifiability policy ... <ref>(this page)</ref>

This is similar WP:ABOUTSELF, but WP:CIRCULAR places more constraints on when and how it can be done: only to support a statement about Wikipedia ... follow the policy for primary sources... avoid original research, undue emphasis on Wikipedia's role or views, and inappropriate self-reference ... make it clear the material is sourced from Wikipedia so the reader is aware of the potential bias. Aymatth2 (talk) 22:24, 6 July 2021 (UTC)

• This seems to be related to the Talk:Wikipedia#Wikipedia has unreliable, primary, vested, and general conflict of interest sources and Talk:Wikipedia#Problematic use of sources referencing Wikipedia and/or Wikimedia sister projects discussions. Schazjmd (talk) 22:34, 6 July 2021 (UTC)
  • I agree with Hfnreiwjfd that in general Wikipedia is no more acceptable as a source than Twitter. However, in articles about Wikipedia it is reasonable to cite what Wikipedia says about itself, subject to the various constraints spelled in WP:CIRCULAR. We could of course say that:

    Wikipedia's Verifiability policy says ... <ref>(this page)</ref>

is invalid, but

According to Professor Smith, Wikipedia's Verifiability policy says ... <ref>(Smith's page)</ref>

is valid. That is silly. The underlying problem is that anyone can edit Wikipedia's Verifiability policy, although some edits will be very short-lived. That does not mean an article about Wikipedia cannot mention it, just that mentions have to be hedged as stated in WP:CIRCULAR. Aymatth2 (talk) 13:18, 7 July 2021 (UTC)

A user-generated source most certainly CAN be cited when quoting from (or closely paraphrasing) it. In fact, the source itself is the MOST reliable source possible in that limited situation.

Now, there are very few situations when it is appropriate to quote (or closely paraphrase) a user-generated source (usually this would only occur in the article about that source), but IN those rare instances, we would cite the source itself to verify the quote. Blueboar (talk) 13:50, 7 July 2021 (UTC)

Completely agree. That is why WP:CIRCULAR spells out strict limitations on citing Wikipedia:

• when Wikipedia itself is being discussed in the article ... to support a statement about Wikipedia
• follow the policy for primary sources
• avoid original research, undue emphasis on Wikipedia's role or views, and inappropriate self-reference
• make it clear the material is sourced from Wikipedia so the reader is aware of the potential bias

Obviously an article about Wikipedia or some aspect of Wikipedia should be based primarily on outsider views of the topic as given in reliable independent sources, and that will be the case if editors observe the restrictions spelled out in WP:CIRCULAR. Aymatth2 (talk) 14:53, 7 July 2021 (UTC)

@Aymatth2: @Blueboar: Could then a bold formatting be applied to the excerpts "avoid original research, undue emphasis", and "make it clear the material is sourced from Wikipedia", and edit to the latter "make it clear and stated in prose that the [...]" ? Also, I'd suggest stating the this usage should be limited, like not creating whole paragraphs in that most references are to Wikipedia.Hfnreiwjfd (talk) 21:15, 8 July 2021 (UTC)

I dislike bolding. Books, magazines and serious newspapers very rarely use bolding. It gives a tabloid appearance. Everything in WP:CIRCULAR is relevant, and if an editor follows it they will not give undue emphasis to content sourced from Wikipedia. Aymatth2 (talk) 13:01, 9 July 2021 (UTC)

@Hfnreiwjfd, are you primarily concerned about whether specific articles should exist, or about whether specific claims in certain sentences can be verified in reliable sources? WhatamIdoing (talk) 16:23, 7 July 2021 (UTC)

@WhatamIdoing: Not exactly. I'm concerned that the exception is, in practice, enabling bias, and undue weight like whole paragraphs that mostly reference Wikipedia. I'm also concerned that this is a Wikipedia:Conflict of interest, since it is Wikipedia users who generate its own rules. And thus, that the exception conflicts with other Wikipedia policies.Hfnreiwjfd (talk) 21:15, 8 July 2021 (UTC)

The actual structure of that part of the policy is relevant to the above discussion; it is:

1. A prohibition of using Wikipedia as a source to satisfy WP:VER
2. A conditional waiving of that prohibition when the topic of the text is Wikipedia itself. And the main condition of that waiver is treating it as a primary source, thus invoking all of Wikipedia's rules, cautions, limitations etc. relating to primary sources.

The first thing that arises from the above two combined is that the net effect certainly isn't granting special privilege, in fact it is the opposite. It's Wikipedia prohibiting specifically itself, and then making a narrow conditional exception to that prohibition and so IMO the main premise of the original post is incorrect.

If there is a concern that certain Wikipedia policies and guidelines aren't being followed at a particular article, then that article is the best place to discuss it. Besides being the better choice of venues, that would lead to the much needed specificity on what the stated problems are rather than just receiving an editor's characterization of them.

Finally there is a mention of sources granting notability to topics within an article. If that is referring to Wikipedia:notability, that is a criteria for existence of a separate article, not an attribute or criteria for inclusion of material within an article.

Sincerely, North8000 (talk) 17:26, 7 July 2021 (UTC)

@North8000: Thanks for the detailed comment. This makes sense, mainly in how the articles should be improved to respect the policy. But it is not the opposite, because I'm solely arguing about the exception, not the entire section.Hfnreiwjfd (talk) 21:15, 8 July 2021 (UTC)

Cool. But #2 is inherently linked to and defined by #1. Sincerely, North8000 (talk) 23:07, 8 July 2021 (UTC)

@Hfnreiwjfd, I think you need to be more specific. Is the problem that you raise at Talk:Wikipedia about the fact that Wikipedia editors have written an article about Wikipedia at all, or is the problem that 12 of the 373(!) citations in that article link to a English Wikipedia page/diff instead of to some other website?

If your idea was adopted, what would need to change at that article? WhatamIdoing (talk) 01:30, 9 July 2021 (UTC)

This is a procedural question about WP:ONUS application. Regarding content inclusion, if the number of editors supporting inclusion is larger than those who objected, is ONUS on opposing editors to achieve consensus to exclude? Infinity Knight (talk) 19:44, 19 July 2021 (UTC)

I would say WP:NOTVOTE applies here. DonIago (talk) 20:07, 19 July 2021 (UTC)

A lot depends on the reasons given for the various opinions. I have seen five well reasoned, policy based arguments in opposition to an addition out-weigh ten poorly reasoned “I just like it” arguments for inclusion… and five well reasoned arguments for inclusion out weigh ten “I don’t like it” arguments for exclusion. Usually, however, the correct conclusion is “no consensus”… which means both “sides” in the debate may have to compromise a bit. Blueboar (talk) 21:04, 19 July 2021 (UTC)

This has nothing to do with procedure, it's a content dispute at Mondoweiss where the editor appears to believe that he only needs to say "ONUS" for his opinion to hold sway.Selfstudier (talk) 22:04, 19 July 2021 (UTC)

So would it be fair to say they Expect-o pro-ONUS? Sorry, I'll see myself out... ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 22:14, 19 July 2021 (UTC)

The raw number doesn't matter in an evaluation of consensus. However, an age-old question is: if there is long-standing text in an article which is challenged, should "no consensus" in the subsequent debate result in the inclusion or exclusion of the text? As of the most recent discussion, there is unfortunately no consistent interpretation among the community. -- King of ♥ ♦ ♣ ♠ 01:09, 20 July 2021 (UTC)

I wouldn't say that !vote counts don't matter, but rather that they're not the primary factor in determining consensus. So two well-structured arguments could override 6 poorly-structured objections, but I'd be very suspicious of any close that sided with a tiny minority over an overwhelming majority (think a ration of 30:1, or thereabouts), even if the majority argument was rather weak.

wrt to the ONUS: I'd say that the onus is always on any editor seeking to change an existing consensus, whether that be for inclusion or exclusion. If there is no existing consensus, that's where WP:BRD comes into play. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 14:35, 20 July 2021 (UTC)

That may be the onus with a small o, but it is not ONUS, which is really about NPOV and due WEIGHT of sourced fact and analysis. I would also say that, sadly, the vast majority of polls are closed on a vote count without any deep analysis of the quality or character of the supporting arguments. SPECIFICO talk 18:29, 20 July 2021 (UTC)

That is exactly the disputed point. ONUS, as written, says that it's up to the person who wants to include information to convince everyone else go along with it. The folks who prefer STATUSQUO say that it's up to the person who wants to make a change to convince everyone else to go along with it.

That is, if a (sourced) sentence has been in the article for a long time, then ONUS says you can remove it, and it's up to its supporters to justify its inclusion; in the same situation, STATUSQUO says that the supporters can revert it back in, and its up to you to convince them to remove it.

Put in terms of principles, the question is whether you prefer "stability", even if it's wrong/bad/unencyclopedic, or if you prefer "saying nothing" instead of saying something that might be wrong/bad/unencyclopedic. Different editors have different values. WhatamIdoing (talk) 03:24, 21 July 2021 (UTC)

This is right. I'm saying that my sentiments lie with WP:STATUSQUO because the guidance at WP:ONUS has a glaring hole that permits gaming. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 13:47, 21 July 2021 (UTC)

Yes, of course !vote counts are highly correlated with the end result. What I meant is that the applicability of ONUS has nothing to do with which side has more !votes. -- King of ♥ ♦ ♣ ♠ 02:16, 21 July 2021 (UTC)

I wasn't really disagreeing with you so much as quibbling over a minor point. I can be a pedant sometimes. ᛗᛁᛟᛚᚾᛁᚱPants Tell me all about it. 13:48, 21 July 2021 (UTC)

Some background history

• When different interpretations of policy crop up, it is useful to go back and examine why a particular point was added to the policy in the first place. What was being discussed when it was added to the policy?

In the case of ONUS, the discussion was clearly focused on who had to have consensus with regard to the addition of new (or recently added) information. We really we’re not thinking in terms of removal of material that had been in an article for a while. Back then, the concept of “silent consensus” was fairly well accepted. If material had been in an article for a reasonable period of time, it was assumed that it had been reviewed and thus had consensus. It was the weakest form of consensus, but a consensus nevertheless. This is why ONUS favored status quo. It was to maintain a version that was assumed to have at least tacit consensus.

That said, We did discuss the issue of BLP violations… and that these had to be an exception ONUS (since BLP was considered superior to “silent consensus” and status quo).

History lesson over, I will leave it to others to determine how we should interpret the policy going forward. Blueboar (talk) 13:31, 21 July 2021 (UTC)

I think ONUS helps with cases where we might have content who's weight is questionable but was never really tested. If long standing content was justified when originally added then presumably ONUS has been met. If the content is new then CONSENSUS would apply. However if the content was added a while back but has never been challenged then I think ONUS makes sense. Take a case where something trivial yet verifiable was added to a BLP (Dr Jones has type O blood [RS]) by EditorA. Dr Jones's article gets little traffic so that factoid remains in the article for a few years. Now an EditorB wants to remove it. EditorA can't just say "silent consensus" to keep it. They would have to justify the inclusion. Now if we have EditorA makes a reasonably case for keeping and EditorB makes a reasonable case for removing then I would say silent consensus decides in favor of keep. If EditorB makes a good case for removal while EditorA can't come up with a good reason to keep then consensus goes with the case for removal. In my view this really only comes into play if someone challenges long standing content. The reply argument can't just be, "consensus" or similar. Contrast that with a case where new content is added. An editor could revert the addition as "UNDUE" and it would now be the job of the editor who want's to add to make the case for inclusion, not the reverting editor to make the case for exclusion. The difference being in the ONUS case the editor wishing to retain status quo needs to put forth a reason. In the new content case the editor whishing to make the change needs to put forth a reason. Springee (talk) 16:15, 21 July 2021 (UTC)

I like that interpretation. Reminds me of the decade-long debate over whether WP:NFCCE means that "no consensus" FfDs over non-free images should result in deleting the image or if it is solely meant as guidance for the strength of arguments in favor of keeping an image. -- King of ♥ ♦ ♣ ♠ 21:25, 21 July 2021 (UTC)

What I don't like about any system that depends on "long-standing" is that we get fights over what constitutes "long-standing". A month? A year? A hundred edits to the page? A thousand page views to the article? Nobody knows what long-standing means – except that all content that I want to keep is "long-standing", and all content that I want to remove is "new". WhatamIdoing (talk) 04:55, 22 July 2021 (UTC)

Since this has come up yet again: Wikipedia is vast and has very many underscrutinized and obscure pages. People can and do add material that superficially looks okay, but is actually poorly sourced and/or fringe or undue, and it hangs around for years because that page has no diligent watchers or the ones it does have know little about the topic and didn't look closely. Years later someone removes the junk. We don't want the junk's WP:OWNER, or some passing editor who thinks it's 'removing sourced material', to be justified in reverting to the status quo because the status quo is privileged just by having hung around for a while. I agree with the position articulated earlier that it is better to say nothing than to say something that could be false, especially when falsehoods on Wikipedia tend to reproduce to other sites. Arguments that favor status-quo seem to mainly think of people removing material in bad faith and that ONUS gives them an advantage. It doesn't, because if the material is good, a consensus to keep it will easily be found, even if just two editors (perhaps Huggle patrollers) reverting the remover. Bad faith additions also need to be accounted for, not just bad faith removals. Whatever good reputation we have comes as much as from what we keep out as from what we actually say. Crossroads ^-talk- 04:59, 22 July 2021 (UTC)

I think the timer starts when a plausible source is first added in support of the statement. As long as the source is not level 3 or worse on WP:RSP at the time of addition and isn't totally orthogonal or contradictory to the statement being made, it is considered plausible. So an unsourced statement can be challenged, and in the subsequent discussion if there is no consensus then the previously unsourced content is removed regardless of how long it has stayed up even if sources are found in the course of the discussion. Basically, in this case we consider the sourced content as a whole to be a new addition, even if the prose may still be the same. -- King of ♥ ♦ ♣ ♠ 05:13, 22 July 2021 (UTC)

It isn't necessarily as simple as RS vs. no good source. I speak from experience: material of this sort may have multiple "plausible" sources that superficially meet RS and are vaguely related to the topic, but when examined, the material is WP:Synthesis, and/or sources are actually bad (a predatory journal, say). Crossroads ^-talk- 05:37, 22 July 2021 (UTC)

The way I read it is that a good-faith attempt must have been made support the text with sources. So if there is a dispute over whether the sources support the topic, I feel like long-standing text should remain. -- King of ♥ ♦ ♣ ♠ 06:04, 22 July 2021 (UTC)

That's precisely the issue. What we end up with is dubious (at best) material sticking around and misleading people simply because it managed to evade notice for long enough. Crossroads ^-talk- 03:28, 23 July 2021 (UTC)

Surely the reason for that is the pages in question being poorly attended (by editors) rather than the policy being poorly written?Selfstudier (talk) 11:34, 23 July 2021 (UTC)

If a policy said that "whenever dubious material sticks around long enough simply because it managed to evade notice, then it has to be kept until you can prove that people agree to remove it", then the problem would be in the policy.

(Technically, the STATUSQUO page is an essay, but the ideas explained in it do represent actual practice for a non-trivial proportion of editors.) WhatamIdoing (talk) 17:16, 23 July 2021 (UTC)

I don't think the ONUS policy is poorly written, but in any case our policy has to account for the fact that most articles are poorly scrutinized, rather than treat that as a side issue. Crossroads ^-talk- 04:57, 24 July 2021 (UTC)

@King of Hearts, if the timer starts when a plausible source is added, what are the units on your timer? Calendar days, editor activity, page views, something else? WhatamIdoing (talk) 06:10, 22 July 2021 (UTC)

A combination of factors. For Donald Trump a couple of months would be enough. For an obscure article edited by 3 people maybe a few years. -- King of ♥ ♦ ♣ ♠ 06:22, 22 July 2021 (UTC)

So everything I want to remove is "new", and everything I want to keep is "long-standing", right? WhatamIdoing (talk) 06:36, 22 July 2021 (UTC)

The case at hand (not really a case imo) arose by virtue of adding things new which were supported by a majority of editors, if then an editor still disagrees citing ONUS (which can look like "Idontlikeit" without any further explanation), then isn't dispute resolution (RFC) the usual way to deal with that in order to draw in outside editors to the discussion? Selfstudier (talk) 09:07, 22 July 2021‎ (UTC)

In a simple case, ONUS has only a temporary effect:

• Alice boldly adds something (sourced)
• Bob removes it because ONUS
• Carol, Carl, and Chris quickly show up to tell Bob that there's a consensus for putting it in the article (or a clear consensus for removing it).
• Alice and Bob both go along with the decision of the group.

The difficulties are what to do if only Alice and Bob engage in a discussion; what to do during the discussion (which could go on for weeks or even months – ONUS says to take it out [if we're talking about an addition of new material, but not, say, a copyedit or rearranging the page], but the essay STATUSQUO says to leave it in if it's been there long enough but to take it out if it's new, and the essay BRD says for Alice not to care what state the article is in during the discussion), and what to do if there is no consensus. WhatamIdoing (talk) 15:29, 22 July 2021 (UTC)

We need to keep in mind that ONUS assumes WP:V. If something is said that doesn't have sourcing then ONUS doesn't apply. It applies in cases where we can verify the claim (Mr Smith has type O blood [source]) but there is a debate about WEIGHT. What ONUS says is even if the content has been around a long while, that by itself doesn't mean is has weight. Springee (talk) 13:35, 22 July 2021 (UTC)

Don't forget WP:LOCALCONSENSUS in this discussion. Quite a lot of editors think that pretty much anything can be done if there is a consensus for it. That is not true. A plain reading of LOCALCONSENSUS is that the editors of an article cannot agree to allow an article to violate a policy. For example, if a controversial statement has no source or is not supported by the given sources, it should be removed regardless of how many editors like it. Zero^talk 13:41, 22 July 2021 (UTC)

This is why ONUS was placed where it was. A lot of content debates center on whether a given source is reliable or not… which gives newer editors the impression that reliability is the sole determinant for inclusion. However, there are many policies and guidelines that determine whether, when and how information is included or excluded. ALL have to be met.

Also, while a policy based consensus is best, sometimes the issue isn’t a question of policy, but editorial judgement. Is the information too trivial to mention? Is it better presented in some other article? Etc.

We do intentionally “err on the side of omission” if a compromise can not be reached. While there are rules that tell us what is prohibited, there are no rules telling us that we MUST include everything that is not prohibited. That is up to consensus. Blueboar (talk) 14:16, 22 July 2021 (UTC)

Some editors "err on the side of omission". Other editors prefer to "err on the side of stability". Every claim that we need to follow STATUSQUO when the inclusion of "long-standing" material has been challenged (and many of the claims that BRD must follow) is an editor indicating that stability is more important. WhatamIdoing (talk) 15:26, 22 July 2021 (UTC)

WP:Onus really doesn't belong here except that it is weak substitute for the related statement that DOES belong here which is "Verifiability is a requirement for inclusion not a reason for inclusion" The operative part, the last sentence "The onus to achieve consensus for inclusion is on those seeking to include disputed content. " doubly doesn't belong here and is far too short to give guidance as wp:consensus does. It includes / hints at some important concepts but is far too short and misplaced to give real guidance or to being interpreted as the above attempts to do. So the big and real solution is to replace the whole "onus" section with (both title and content) "Verifiability is a requirement for inclusion not a reason for inclusion" and move the last operative sentence to WP:Consensus (I finished/refactored this after Selfstudier's post) North8000 (talk) 14:08, 22 July 2021 (UTC)

[12]:- Per COVID-19 GS (now included under DS): "Editors are reminded that the onus is on the editor seeking to include disputed content to achieve consensus for its inclusion. Any content or source removed in good faith and citing a credible policy-based rationale should not be reinstated without prior consensus on the article's talk page." Selfstudier (talk) 14:03, 22 July 2021 (UTC)

North, I think that your bolded statement is too high-falutin' to be effective. "You don't get to include something just because it's verifiable" or "It is not enough for material to be verifiable. All content kept in articles must comply with all policies, especially WP:DUE and WP:NOT" might be clearer. WhatamIdoing (talk) 15:37, 22 July 2021 (UTC)

• What isn’t being discussed (yet) is the situation where there isn’t a policy issue in play… but simply a question as to whether to include or omit some bit of material from an article - purely on editorial judgement grounds.

For example: Historians do know what Richard III had for breakfast the morning of the battle of Bosworth (there is a primary source, and one or two very detailed biographies do mention it)… but is it worth mentioning somewhere in WP? If so, which article? I could see a legitimate debate on that. So is the onus to “get consensus” on those who want to include, or on those who want to omit? At the moment, we say it’s on those who want to include. Blueboar (talk) 16:15, 22 July 2021 (UTC)

I agree, including on what the real question is. Besides lobbying for the unrelated change, my post was to point out that the answer can't be found in one misplaced sentence in wp:ver. North8000 (talk) 16:31, 22 July 2021 (UTC)

@Blueboar, your example does involve a policy issue. Specifically, whether to include his breakfast menu is a question of WP:DUE. WhatamIdoing (talk) 17:14, 23 July 2021 (UTC)

Meh… WP:DUE is discussing the prevalence of opinions, and not the triviality of facts. I would agree that the concept that underlies DUE could apply, but the policy itself does not. Blueboar (talk) 17:41, 23 July 2021 (UTC)

I think many of us say DUE when we should be saying WP:PROPORTION Springee (talk) 17:54, 23 July 2021 (UTC)

There's a long and interesting discussion on WT:OR about this. The idea has emerged that, to an extent, correct mathematics is self-verifying. For example, our article on Fourier analysis contains:

This is cited to a footnote. The footnote isn't a source, it says:

I understand that to a proper mathematician, this does verify the claim being made, although D. Lazard has criticized it as something that "should not be done" (I think because it requires more advanced mathematics to follow than the rest of the article does).

This appears to allow for the possibility that editors performing a piece of routine arithmetic per WP:CALC may be able to verify it by using a footnote to show their working, rather than citing a source. What do we think?—S Marshall T/C 12:36, 14 June 2021 (UTC)

Is this ever going to be necessary? WhatamIdoing (talk) 04:53, 15 June 2021 (UTC)

It does come up, yes. First time for me was in 2010, when I calculated the luminosity class value of a star and was surprised to learn I wasn't allowed to (see Talk:HIP 56948). For other examples where this idea might have helped, see Wikipedia talk:No original research/Archive 54#Calculations using well known mathematics is not original research, Wikipedia talk:No original research/Archive 57#Possible enhancement of WP:CALC, Wikipedia talk:No original research/Archive 58#Disagreement over WP:CALC, Wikipedia talk:No original research/Archive 61#Elaboration on "routine calculations".—S Marshall T/C 09:01, 15 June 2021 (UTC)

@S Marshall, I meant: Can you think of a single time when an article really needed this, and we couldn't source also it to an external source, instead of hoping that the editor got the calculation right? Is there, to use the example in the Fourier analysis, no published source anywhere in the world that says something about coefficients being "proportional to samples of S( f ) at discrete intervals of 1/P" anywhere in it? WhatamIdoing (talk) 22:46, 16 June 2021 (UTC)

Thanks for asking me that question. I'm sorry for writing such a long reply: it's because you've made me think. I see a lot of facets to this.

In that particular case, yes, I'm confident that there will be a published source. The issue that I'm raising is of Wikipedians performing calculations, though. I think there are several use cases to consider.

First, there's the level that S Marshall thinks is obviously OK, such as converting one unit of measurement to another. Let's imagine I'm compiling an article on agriculture. Some of my sources give land areas in hectares, and others give land areas in acres. I find a good source that says: In 2020, 1,265,000,000 hectares of wheat were planted in England. I write: In 2020, 1,265,000,000 hectares (3,125,000,000 acres) of wheat were planted in England. That kind of calculation is self-verifying, right? I don't need a separate source for the acreage? (If not then we need to send {{convert}} to MFD, although the discussions I've already linked show that some Wikipedians genuinely object to other Wikipedians doing the kind of arithmetic that's routinely taught to pre-teens.)

Second, there's the level that S Marshall thinks is obviously not OK, such as the tables in COVID-19 pandemic cases, which don't clarify that these are self-reported figures given by different countries that use different methodologies. (Iniquitously the WHO page doesn't, either, and I don't know why we don't prefer Johns Hopkins, although I'm avoiding the temptation to start editing COVID-19 pages, because that way lies madness.) Anyway, the problem is that readers are likely to draw comparisons between numbers that aren't actually comparable.

Thirdly, there are the things that S Marshall thinks are edge cases, such as the calculation I did in 2010 to identify a star as a main sequence star (mentioned above). There was, in 2010, no source that said the star was on the main sequence, but we did have its relative magnitude and distance. I derived its absolute magnitude using the standard equation: .

We don't have clear rules about Wikipedians doing sums. This creates a tyranny of ignorance, where someone who can't follow high school arithmetic is allowed to create a crushingly huge burden of work for someone who can. We also don't have clear rules about how we use statistics, because (by and large) our sources don't either, and I think that's a problem too.—S Marshall T/C 11:05, 17 June 2021 (UTC)

I'd love to have you editing pandemic-related articles. I think anyone who reads WP:PROSELINE and is willing to take a big hatchet to the accumulated cruft and trivia (do we really need to know the names, cities, dates of diagnosis, etc., for so many people?) for an hour can do a significant service to the world. Pick the article for any location you would like to visit some day, or whatever scheme you'd like; you can hardly find any article that wouldn't be helped by wholesale removal of whatever you think will seem like obvious trivia in ten years.

Now about the subject at hand: I think that the edge cases involving standard and fairly simply equations are probably okay. I'd say that if a few folks at Wikipedia:WikiProject Mathematics understand it without needing more than a couple of minutes to check that you copied the formula correctly and plugged in relevant numbers, then it's probably okay. You don't need to be able to do the calculation yourself in that case, because you can put the formula into a fancy calculator to get the result. You just need it to be simple enough to have no doubts that you've put the right numbers in the right spots in the formula.

This likely means no statistics (at least, nothing beyond very simple statistics), because just understanding the data set's limitations is going to take more than a couple of minutes. (Need proof? Look at this report, and think about how many wrong claims a good-faith editor could make from it. I gave up on trying to use any of the numbers. Also, please write me an article about Multiplicative effect on odds.) WhatamIdoing (talk) 04:27, 19 June 2021 (UTC)

I think one problem with "doing things with numbers" isn't so much that people get the maths wrong, but that they don't really appreciate what the number in their source actually is. So they end up saying things about it, or assuming things about it, that aren't true. One example is confusing an estimate with an actual count, or failing to know how precisely to present a number in article text. That {{Convert}} spends a lot of effort on rounding, so that we convert 3 metres to 10 feet and not 9.8425 feet, shows the issues are more than just doing the sums correctly. -- Colin°^Talk 09:51, 19 June 2021 (UTC)

WAID, I've stared at that report in despair. I think that with scientific or statistical reports and analyses, the encyclopaedist's role is to summarize the report's conclusions (and, preferably, only where the conclusions are reached by qualified people and published by other, editorially independent people). I definitely don't think editors should be performing their own interpretations of the data set. In the case of that report, I feel that it's the telling lack of a conclusions section that makes it unusable.

Of course, you could make the case that in the star example cited above, I did interpret the data set. I'd say that was a trivial and unambiguous exercise, because we routinely do report the star's absolute luminosity class and the data were all to hand -- the fact that the original paper didn't mention it is likely because, to their intended audience, the matter was incredibly obvious. I'm not trying to say that I should be allowed to do sums and other people shouldn't, although it would certainly be convenient if that was the rule.

Thank you for the kind invitation to edit COVID-19 articles. In the past, I have dipped my toe into editing in areas covered by WikiProject Medicine, and I've found those the waters a little too hot for my taste. I'm retired now, and I'm trying to relax!

I might look into Multiplicative effect on odds, though.—S Marshall T/C 20:15, 20 June 2021 (UTC)

Like most things in Wikipedia, usually multiple factors get weighed together. I think that one variable is determining the need or value for putting such into the article and I'm thinking that this is starting to move into rare situations. With the general principle that we really don't want editor-creations/interpretations in articles. But that the most obvious calculations done just for article clarity are not editor-creations / editor interpretations. Where you start moving into areas where ever-smaller circles of people with higher levels of expertise are needed for the calc to be obvious vs. an editor creation/interpretation, you are moving away from that concept. Also keep in mind is that people who read an article are typically there to learn from it, i.e. are there because they have less expertise. And a core concept is that challenged material needs to be directly supported by a source, not a situation of "you can't demand that because it might be something that us experts know is obvious". Not that that is exhibited by you or your proposal; it can even be a creation in the mind of the reader. North8000 (talk) 12:21, 15 June 2021 (UTC)

• To me, the clearest and most obvious case for this is articles about hard mathematics. The reader's understanding is typically enhanced by a worked example, and for non-trivial calculations, I don't think it's right to plagiarize a textbook. Mathematics articles tend to contain worked examples devised by Wikipedians and checked by other Wikipedians. In some cases to carry out these checks requires mathematical literacy. Take Fourier inversion theorem, an article whose only inline citation is to a wordpress blog, but to a proper mathematician, is self-verifying.—S Marshall T/C 12:43, 15 June 2021 (UTC)

Are there cases on those where the policy has become a problem? I ask this not for the usual reasons but because what usually works to sort out things on (non-political) articles is things that operate in the gray area (which most things do- very little content could withstand a full assault by a wiki-lawyer) but are never challenged unless there is a genuine concern about the veracity of the material. North8000 (talk) 11:21, 17 June 2021 (UTC)

Well, whether it's "a problem" is in the eye of the beholder, but there are periodic disagreements about whether it's permissible for an editor to do basic sums. The discussion on WT:NOR, which prompted me to raise this, arose on Ingenuity (helicopter) from when an editor converted an Earth date (3 April 2021) to a Mars date (sol 43 of the mission). Another editor objected because, and I quote exactly, I don't know how to do it (and his opinion was that most other people wouldn't either). After some discussion in which I explained how it was done, he said: I think it would be a good idea to provide a footnote, or some type of non-article text, analogous to what S Marshall depicted above, which shows either the ET-to-Earth conversion formula itself, or step-by-step instructions to perform the calculations. I understood this as a suggestion that an arithmetical calculation could become "verifiable", if you like, by showing your working: and I think that could be right, but I also thought it was a matter for WT:V. So here I am.—S Marshall T/C 13:13, 17 June 2021 (UTC)

I'd generally like us to move towards checkability as our verification criterion and not sourceability. It would solve this issue, but the big problem is that moving in this direction magnifies the threat of WP:SYNTH. For mathematics this problem is much reduced, since we are generally working within a recognised axiomatistion and correctness of such arguments is a mechanical matter that can be checked by computer for much, possibly most, mathematics. Accordingly, editing in maths articles is generally much more tolerant of deviation from sources: it is commonplace that proofs given in maths articles are quite liberal adaptations of published proofs, and this causes no problems. I don't like the footnote argument since it is a nontrivial identity, which generally should come with some sort of hint and not be relegated to a footnote, but not having a source for a mechanically verifiable equation I don't think is fundamentally problematic. — Charles Stewart (talk) 06:27, 8 July 2021 (UTC)

My problem with math equations is that anything past the basics can't be verified or understood by the typical reader, because they can't do the math problems themselves. Let alone is there a good way (that I'm aware of) to paste anything past a simple equation into a search engine so someone can find a source that discusses it in a more reasonable manor then Wikipedia. Whereas neither are issues if equations are summarized into understandable English. Good examples of the problems are the various Wikipedia articles on Quantum mechanics. Which are all extremely heavy in un-referenced math equations that can't be understand by anyone without a high level math degree, but don't present the information in plain well sourced English. So allowing for unverified math equations does come at a cost. --Adamant1 (talk) 03:23, 31 July 2021 (UTC)

I am trying to make an article and it says WordPress isn't allowed? That's where they all posted because they didn't have personal sites!

Trakaplex (talk) 01:09, 16 August 2021‎ (UTC)

In general, and with some exceptions, anything that is self-published (the person who write it is the person who posted it on the internet) is not a good source. WhatamIdoing (talk) 23:46, 18 August 2021 (UTC)

I have found discussions but no Wikipedia policy or consensus on using obituaries as sources. If someone can enlighten me, please do. In the absence of consensus or policy, I would like to make strides here in hopes of the establishment of Wikipedia policy discouraging obituaries as sources. While obits meet the low standard of being published, they lose the battle of verifiability, as described in this article regarding questionable sources: "Questionable sources are those that have a poor reputation for checking the facts, lack meaningful editorial oversight, or have an apparent conflict of interest." Obits hit on all three. As a retired newspaper editor, I can confirm what many know - that obits are provided by funeral homes which get them from the family or friends of the deceased. There is no oversight. Few newspapers have obit editors on staff as they did long ago. Occasionally, a newspaper will publish an obit like a editorial if the deceased is a prominent person. Rarely, there is a reporter's byline. Unfortunately, it is not possible for Wikipedia to discern if any editorial oversight has been given to the information. Obits are paid advertisements, funded by family and/or friends advertising the passing of their loved ones. While obits are successful in their mission of providing some context of a person's life and giving those mourning an outlet for their grief, they are wholly unreliable as a source for an encyclopedia. I hope if it has not done so, that Wikipedia will establish a policy that disallows obits as reliable sources. If it already has, then never mind. God bless and happy editing! MarydaleEd (talk) 05:34, 30 August 2021 (UTC)

Hi, @MarydaleEd. I think you've got a handle on the situation, namely that it depends on which kind of obit you're talking about. If it's a news article written by a journalist because someone died, then that's a great source. These may be less common than they used to be, but they do still exist. If it's something that the family paid to publish, you should probably treat it approximately the same as you would treat social media posts by family members, or a paid advertisement from a business. That is, they're not likely to be completely wrong, but they're not Wikipedia:Independent sources, either. WhatamIdoing (talk) 00:24, 4 September 2021 (UTC)

Its content is determined by previously published information rather than the beliefs or experiences of editors.

Its content is determined by previously published information rather than the beliefs or experiences of editors, also known as contributors.

Moved to where it belongs: Wikipedia talk:Wikipedians#contributor=editor=Wikipedian. Lembit Staan (talk) 21:51, 12 September 2021 (UTC)