Wikipedia:Requests for comment/Deprecated and unreliable sources

Several prior discussions have indicated editor concerns with elements of deprecation. Currently no guideline describes deprecation and all that it entails. There is:

• a section within WP:RS (Wikipedia:Reliable_sources#Deprecated sources),
• the information page WP:DEPS: Wikipedia:Deprecated sources,
• another information page WP:DEPREC: Wikipedia:Reliable_sources/Perennial_sources#Deprecated,
• and the definitions and information within the WP:DAILYMAIL1 and WP:DAILYMAIL2 RfCs (the first source to be 'deprecated', see History section).

Related discussions at ANI:

• Wikipedia:Administrators'_noticeboard/IncidentArchive1087#Can we decide what the heck "deprecation" means, or alternately, use a different word? leading to this RFC.
• closed ANI discussion Wikipedia:Administrators'_noticeboard/IncidentArchive1087#User:David Gerard violating Wikipedia:Deprecated_sources#Acceptable uses of deprecated sources (of which the immediately above referred to disussion was a part).

Open questions

Purpose of deprecation

What is the meaning and purpose of deprecation? How is this distinct to general unreliability?

Discussion (Purpose of deprecation)

It seems reasonable. I would add for deprecated, "self-published or user-generated content authored by established subject-matter experts is not acceptable." That is the key distinction between generally unreliable and deprecated. If a source is deprecated, we cannot be confident that the article was written by the named author or that it has not been altered. Private Eye for example had a diary supposedly written by then UK PM John Major. TFD (talk) 14:53, 30 December 2021 (UTC)

I don't think this (If a source is deprecated, we cannot be confident that the article was written by the named author or that it has not been altered) is the (or even a) consideration when deprecating a source. If a source is altering someone else's content, or falsely writing that someone else authored it, then it's probably Spam Blacklist eligible. I'm not even sure the Daily Mail does that. Private Eye is a satirical magazine; I'd have to see the article in question, but I doubt someone reading the entry would seriously think John Major wrote it. ProcrastinatingReader (talk) 14:56, 30 December 2021 (UTC)

I wouldn't say it's a specific consideration in deprecation discussions, but it's definitely relevant afterwards. Once we've established that a source doesn't faithfully report facts in general, that necessarily applies to any individual facts, like ​who authored a particular article - and similarly, they might take excessive editorial liberties for the sake of clicks, and so on. If there are deprecated sources that we do trust to that degree, I suppose we could make a further distinction between deprecation and the spam blacklist, but I'm not sure that there's a useful one to be made, and that might fall outside the remit of the spam blacklist regardless unless we redefine it. Sunrise (talk) 00:07, 31 December 2021 (UTC)

I don't see that this clearly differentiates why we should have "unreliable", "deprecate" and "blacklist". I get there is a big difference between a source that is unreliable because it "reliably" reports the claims of "crap" experts as fact vs a source that falsely attributes claims to an expert. One is unreliable because the messages they report aren't worth reporting while the other might actually be asking good and important questions but is changing the answers provided. My understanding is the reason the DM was deprecated vs just "unreliable" was editors couldn't trust that quotes etc weren't altered. While certainly there is a difference between how editors will be allowed to use an unreliable vs deprecated source, this doesn't make it clear why a source should be deprecated vs marked as just unreliable. Honestly, most of the debates on the topic come across as, "deprecate because a lot of editors don't like it". I will note this comment clearly hits on the Criteria for deprecation section below. Springee (talk) 04:55, 2 January 2022 (UTC)

Current definitions

Generally unreliable: Editors show consensus that the source is questionable in most cases. The source may lack an editorial team, have a poor reputation for fact-checking, fail to correct errors, be self-published, or present user-generated content. Outside exceptional circumstances, the source should normally not be used, and it should never be used for information about a living person. Even in cases where the source may be valid, it is usually better to find a more reliable source instead. If no such source exists, that may suggest that the information is inaccurate. The source may still be used for uncontroversial self-descriptions, and self-published or user-generated content authored by established subject-matter experts is also acceptable.

Deprecated: There is community consensus from a request for comment to deprecate the source. The source is considered generally unreliable, and use of the source is generally prohibited. Despite this, the source may be used for uncontroversial self-descriptions, although reliable secondary sources are still preferred. An edit filter, 869 (hist · log), may be in place to warn editors who attempt to cite the source as a reference in articles. The warning message can be dismissed. Edits that trigger the filter are tagged.

Name

Are "deprecation" and "deprecated sources" the best terms for the concept described in section above?

Discussion (Name)

At WP:V, "source" includes three possibilities, publisher, work and author. The words "unreliable" and "deprecated" are not present, "reliable" and "questionable" appear plus a pointer to WP:RS, where "unreliable" does appear ("No source is 'always reliable' or 'always unreliable' for everything.") and where there is a section for "deprecated sources" (as mentioned in the introduction above). — Preceding unsigned comment added by Selfstudier (talk • contribs)

• Sorry if this is the wrong place, but I should note the description above for deprecated -- "The source is considered generally unreliable, and use of the source is generally prohibited" -- differs substantially from what the linked page actually says, "Deprecated sources are highly questionable sources that editors are discouraged from citing in articles, because they fail the reliable sources guideline in nearly all circumstances" (emphasis mine in both cases). The latter is more in line with the dictionary definition of deprecate [1] [2] and I think we should strive to use words for what they actually mean, not our own internal definition. (I should also note the definition of deprecated given at Wikipedia:Reliable_sources#Deprecated_sources differs somewhat from the other two already cited above.) Calidum 17:31, 30 December 2021 (UTC)

I wouldn't think of them as being different definitions, but rather as different ways of saying essentially the same thing. A general prohibition is a type of discouragement, and the rest of WP:DEPS (including the followup "in nearly all circumstances") should make it clear what form the discouragement takes. Likewise, the third example you mention from WP:RS is simply focused on what that means in practice ("they should not be used, unless there is a specific consensus to do so"). I suppose if you have to pick one to be "official", the language "generally prohibited" is directly taken from the summary of consensus in DAILYMAIL1. Sunrise (talk) 00:07, 31 December 2021 (UTC)

@Sunrise, I think there is a big difference between "generally prohibited" and "discouraged". We discourage a lot of permissible things (e.g., using unnecessary technical jargon in articles, citing journal articles without providing a URL/doi/PMID/other link [usually, a bot adds the link later], adding only a bare URL in an article because you don't know how to format a citation properly, creating articles that are obviously verifiable but for which you didn't post any sources in the first draft...). See also Wikipedia:Strongly Discouraged. In the "generally prohibited" category, I'd put copyvios and circumventing the spam filters. Those are things you normally must not do, i.e., "generally prohibited". WhatamIdoing (talk) 03:44, 2 January 2022 (UTC)

I would agree that it could be interpreted that way if comparing the two phrases in isolation. However, I think the context (including, but not limited to, the second half of the same sentence) makes it fairly clear which interpretation is being used. Sunrise (talk) 05:58, 4 January 2022 (UTC)

When we use words that "in isolation" mean something different from what we intend, editors will come to different conclusions about what is actually meant. This is an avoidable source of disputes. WhatamIdoing (talk) 02:50, 8 January 2022 (UTC)

• This requires an answer to the "purpose" question first. If the purpose of deprecation is general prohibition, I'd like to see it called "prohibited publishers" or "banned sources" or similar, rather than the WikiSpeak definition of "deprecated" (which is rather out of touch with the dictionary definition, and besides is software lingo). This also helps editors at RSN be more clear on what they're voting for. I don't think sources with some history of fabrication but a lot of history of good publication should be 'prohibited'. It requires volunteer judgement, for sure, but 'generally unreliable' seems more accurate in such a case. ProcrastinatingReader (talk) 19:33, 31 December 2021 (UTC)

Criteria for deprecation

What is the criteria for a source being 'deprecated' (as opposed to being 'generally unreliable')? What kinds of evidence should editors look for to make this determination?

Discussion (Criteria for deprecation)

• One thing I wanted to get at when creating this RfC is inconsistency in accepted evidence. Consider Independent Press Standards Organisation results. In the RfC for The Mail on Sunday [3] editors decided to ignore IPSO outcomes. David Gerard (I'm choosing his comment because he was the one of the only ones to address the 'IPSO argument for reliability' head-on) wrote Because IPSO complaints are not the be-all and end-all of whether a source should be deprecated in Wikipedia, and IPSO is widely regarded as a captured regulator. In other RfCs, we decide IPSO outcomes are highly relevant for deprecation and consider them heavily, sometimes use them as a sole basis for deprecation. This seems like an inconsistent approach to me. I'd appreciate specific guidance on what kinds of evidence, in what kind of quantity, constitutes enough evidence for deprecation. I appreicate it's not an algorithm, but with over 40 deprecated sources now, I think we're at least able to write up roughly what kinds of evidence is considered convincing. Right now, it's somewhat a free-for-all. ProcrastinatingReader (talk) 23:35, 29 December 2021 (UTC)
• Consider IMPRESS as well? Selfstudier (talk) 11:47, 30 December 2021 (UTC)
• To my mind, we are over-using “deprecation”. It should be reserved for those few sources that go beyond “general unreliability”. Blueboar (talk) 13:13, 30 December 2021 (UTC)
• I would expect a relevant reliable source that says the publication routinely publishes false information that is not corrected. That would not be a problem with hate groups and conspiracy theory websites. TFD (talk) 14:55, 30 December 2021 (UTC)

  So then what's the criteria for general unreliability? Because I thought that was the requirement for a source to be considered GUR. ProcrastinatingReader (talk) 23:03, 30 December 2021 (UTC)

  My impression for GUNREL is that the standard is basically "whoever showed up that time didn't like it". I have seen editors claim unreliability because (e.g.,) the news source seemed "biased", meaning that it chose to run certain stories but not others, or that it presented the views of one side by not the other side, and that viewpoint is not the one familiar to the editors in the discussion. To pick two completely made-up random country names, "China" and "the US", a news source may be called "GUNREL" for everything if mostly-non-Chinese editors feel that it has a pattern of reporting what China's politicians say that China's government's viewpoint is on a particular dispute between China and the US, without giving what they feel is adequate airtime to what the US's politicians say that the US government's viewpoint is on that same dispute. The problem IMO is the "for everything" part; IMO such sources should be considered highly reliable for what they actually report (e.g., the title of the person being quoted; the quoted statement; China's view at that point in time). WhatamIdoing (talk) 03:56, 2 January 2022 (UTC)

I also question, at least partially, the entire foundation of deprecation. For example, and I am not saying FrontPageMag should be cited for anything ever, but seven people at RSN in one discussion is enough to decide that a source may not be used on Wikipedia at all? How does that make sense? There are currently 5 RFCs on RSN right now. When did that stop being a place to discuss if a given source is reliable in a given context? It has turned in to a means to attempt to banish sources across Wikipedia based on essentially click-bait (they published this thing I dont like this one time). This should be reserved for truly exceptional circumstances. I cant imagine any professional publication saying, except in extremely rare circumstances (and 5 times right now is not rare), that such and such website should never be used for any purpose regardless of the circumstances or particulars. nableezy - 10:24, 4 January 2022 (UTC)

I whole heartedly agree with this concern. I think this is a problem that is born out of the RSP list. In the past editors were a lot more likely to use RSN to argue *this* article is reliable/unreliable for a particular claim. With the advent of the RSP list the objective seems to be to secure a ranking as that ranking now carries a lot of assumed value. This is fine with some of the big sources we have discussed. The rankings for things like Fox News, CNN, Daily Beast etc are likely to be rather stable since we have a lot of discussions and a lot of editors who participated over time. However, for a smaller source we might have only a few editors who weigh in. After that it's easy for someone to add a source to the RSP list based on the opinions of just a few editors in one or two discussions. In the case of deprecation, if we can't get at least say 10 editors (no idea what a good number would be) to weigh in how can we say there was enough input to apply such a drastic restriction to a source? This is why I'm concerned that we don't have a clear definition of when/why a source should be deprecated. Even if we agree what to do with deprecated sources (lock them up!) if we can't agree on the crime we moved away from reason. Springee (talk) 13:00, 4 January 2022 (UTC)

I think that "a lot of discussions and a lot of editors who participated over time" is a key point. It would be useful to have a central archive for subjects that have been discussed repeatedly. It is not so useful to spend time on major discussions that could be handled as simple questions. WhatamIdoing (talk) 03:05, 8 January 2022 (UTC)

"What kinds of evidence should editors look for to make this determination?" is a great question. In a recent, well-attended deprecation RfC, the closer discounted evidence presented that the source was unreliable. They reasoned that since the choice was between "generally unreliable" and "deprecated", evidence of unreliability was not relevant. My view is that there is "generally unreliable" covers a broad spectrum of bad sources, and that evidence that a source is on the stinky end of the shit stick should be a part of deprecation proceedings. Firefangledfeathers 06:40, 5 January 2022 (UTC)

WP:QUESTIONABLE (RS) "Questionable sources are those with a poor reputation for checking the facts or with no editorial oversight....blah

WP:QS (V) Questionable sources are those that have a poor reputation for checking the facts, lack meaningful editorial oversight, or have an apparent conflict of interest....blah

Both versions are similar, idk why they are not identical.

Is gunrel intended to be different from this? (Def above says "Editors show consensus that the source is questionable in most cases."

And what is it exactly that turns questionable/gu into "stinky", fabrication? And? Selfstudier (talk) 11:28, 5 January 2022 (UTC)

+1 to "When did [RSN] stop being a place to discuss if a given source is reliable in a given context?"

If we want to have a vote-out-stinky-sources-wholesale page, let's have that page, but let's not have those discussions dominating RSN, which used to be a useful and interesting noticeboard. WhatamIdoing (talk) 03:03, 8 January 2022 (UTC)

Valid uses of deprecated sources

• Does/should deprecation apply to publishers and/or authors and/or single works?
• What is the difference in valid use between generally unreliable vs deprecated sources?
• When is it valid to cite an article in a deprecated and/or generally unreliable publication? (is it valid to cite an established subject matter expert who is published in a deprecated publication?)

Discussion (Valid uses of deprecated sources)

Wikipedia:Reliable_sources/Noticeboard#De-deprecate CounterPunch and Wikipedia:Reliable sources/Noticeboard#RFC: Counterpunch discuss the third point above. If consensus were possible that articles by experts a la WP:SPS may be cited even if published in a deprecated source, that would seem to be the simplest approach? The ONUS would remain with wouldbe citers to justify inclusion, probably it ought to be discussed first at a talk page. Selfstudier (talk) 12:09, 30 December 2021 (UTC)

I really oppose this if so called expert choose depreciated source to publish then we assume more respectable outlets didn't want to publish their views so their views are WP:UNDUE. Depreciated source should be used only WP:ABOUTSELF but only on the page source for example Daily Mail could be used only on Daily Mail page and nowhere esle Shrike (talk) 13:10, 30 December 2021 (UTC)

If the consensus is instead like that then the criteria for deprecation assume more importance, in other words people may be less likely to consider deprecation in the first instance since expert opinions are allowed for "merely" unreliable sources. Also WP:SPS stipulates that expert designation requires "work in the relevant field has previously been published by reliable, independent publications." Selfstudier (talk) 13:29, 30 December 2021 (UTC)

That assumption is based on what exactly? nableezy - 15:26, 30 December 2021 (UTC)

If we allow experts' articles in deprecated sources to be used, then there is no difference with generally unreliable. It then isn't a useful distinction for editors, but merely a collective opinion of Wikipedia editors. Experts may not have the same view of which sources are deprecated and may allow The Daily Mail or CounterPunch to publish their articles, although I expect they would draw the line at Alex Jones. But I wouldn't use an experts' article published in a generally unreliable source because it would lack weight. Weight after all is determined by coverage in reliable sources. TFD (talk) 15:03, 30 December 2021 (UTC)

The question is if an expert writes something in the area of their academic expertise is that a reliable source regardless of where it is published. Because if it is a reliable source, then it does have (some) weight. There are also examples of columns by experts in deprecated sources being cited repeatedly by other reliable sources. Can somebody argue that such a source does not have any weight? nableezy - 15:25, 30 December 2021 (UTC)

At that point the weight would come from the coverage in the reliable sources instead, and that coverage is what we would use for article content. Even if all they do is repeat the same information, it will have gone through their editorial and fact-checking processes, which makes it different from the same information published in the deprecated source. More generally, I suppose this could be a valid candidate for an exception to deprecation, but naturally consensus would need to be established in favor of it before it could be used. In addition, discussion of such an exception would have to specifically focus on information that is not reported in the reliable sources, as otherwise we could just use those sources instead. Sunrise (talk) 00:07, 31 December 2021 (UTC)

The language at WP:RS, "..should not be used, unless there is a specific consensus to do so." does suggest that exemptions are possible (beginning with local?).Selfstudier (talk) 10:54, 31 December 2021 (UTC)

My 3¢:

• Does/should deprecation apply to publishers and/or authors and/or single works?
  • I think it isn't pointful to declare any single work (e.g., a single book, a single news article [including a multi-part series]) to be "deprecated" or "generally unreliable". Individual publications should use the ordinary process of comparing the individual work against the specific material to be included in an article, and determining whether (a) that work is reliable for that material and (b) that material meets both DUE and WP:BALASP for that Wikipedia article.
• What is the difference in valid use between generally unreliable vs deprecated sources?
  • "Deprecated" should be considered closer to the spam blacklist, including the possibility of "whitelisting" an individual use. It should not be considered "banned". "Generally unreliable" should be considered more like saying that this is not a great source, but if you know what you're doing, and you are writing very carefully, it may be perfectly reliable for the purpose. Consider, e.g., the difference between writing "There is no good scientific evidence for using parachutes when jumping from an aircraft" and "The BMJ ran an article describing the evidence levels for parachute use", or the difference between "Chris Cultic is the Grand High Master of the Universe" and "The religious organization issued a press release saying that they had adopted a new title for its founder, 'Grand High Master of the Universe'."
• When is it valid to cite an article in a deprecated and/or generally unreliable publication? (is it valid to cite an established subject matter expert who is published in a deprecated publication?)
  • There are and should be different rules for the two categories. In both cases, it can be valid. The only exception is if the publication has a documented history of lying about who wrote their articles (e.g., writing articles and claiming that a Nobel prize winner wrote it – at the borderline for this, I might include wholesale copying of Wikipedia articles).
  • The main thing that needs to be determined is what to do if someone cites a source that you think is inappropriate to support content that you think is otherwise appropriate. Should you blank the good content? Blank the poor source but leave the good content? Go to the trouble of finding a better source yourself? Tag the source in the hope that someone else will do the work that you don't want to do yourself?

WhatamIdoing (talk) 04:18, 2 January 2022 (UTC)

How do show that someone said what you are quoting or relating it if they wrote it in e.g a column in an unreliable / deprecated source? The citation tools warns against this, but there needs to be some reference to check/proof. Is there a method for this, or is it a case of justifying each inclusion? ITellComputerYes (talk) 11:47, 5 September 2024 (UTC)

Who is an expert?

There is another piece to this puzzle - the question of who qualifies as a “subject matter expert”? When we wrote the SME exemption, we were really thinking of academics who had previously published in respected journals … but that concept seems to have been extended beyond academia… especially to media columnists who have previously published their opinion/analysis in reliable news outlets. Have we taken the concept of SME too far? Blueboar (talk) 15:35, 30 December 2021 (UTC)

Wikipedia talk:Verifiability#WP:SELFPUB and the definition of a subject-matterexpert. is a current discussion.Selfstudier (talk) 15:44, 30 December 2021 (UTC)

Ah… thanks for the link. Blueboar (talk) 16:24, 30 December 2021 (UTC)

Removal of deprecated or unreliable sources

At present, the insertion of new references to deprecated sources is slowed down through the use of edit filters and bots, which is relatively uncontentious. However, how should existing references to deprecated and unreliable sources be dealt with?

The ultimate trade-off is between speed in removal vs judgement on a long-term solution. That is, is it acceptable to indiscriminately remove references to deprecated/unreliable sources and replace with {{cn}} tags? Or should each reference be judged on its own merits, and either a new reference found to support the underlying text, or the text removed along with the ref? Or something else?

Discussion (Removal of deprecated or unreliable sources)

Concrete cases of removal at sight according to a deprecation notice

David Gerard removed from Edward Said (a) Edward Saïd, Defamation, Revisionist Style Archived 10 December 2002 at the Wayback Machine, CounterPunch, 1999. Accessed 7 February 2010 and (b) David Price, "How the FBI Spied on Edward Said," Archived 16 January 2006 at the Wayback Machine CounterPunch 13 January 2006, accessed 15 January 2006. 2021}} and (c)Cockburn, Alexander (2006-01-12). "The FBI and Edward Said". The Nation. Retrieved 2021-12-19.. (c)’s removal was incomprehensible since the Nation is not deprecated. He replaced all with a cn tag. David Gerard removed from History of Gaza Roy, Sara. "The Economy of Gaza". Counter Punch. Archived from the original on 2008-12-29. Retrieved 2009-01-19.. Sara Roy is perhaps the leading world authority on the economy of Gaza. David Gerard removed from the article on Nuaman Levy, Gideon (27 December 2005). "Dusty Trail to Death". CounterPunch. Retrieved 17 January 2016.. Levy does thorough interviews with numerous Palestinians, is an award winning Israel investigative journalist, and a regular contributor to Haaretz, a mainstream Israeli newspaper. He noted that Bedouin of the Ta'amreh clan were numbered among its inhabitants, a datum hard to source elsewhere. David Gerard removed Robert Richter War Hero or War Criminal? 4 October 2008. Robert Richter was political director for CBS News from 1965 to 1968, when he became an award winning documentary filmmaker. He interviewed Telford Taylor and recalls the problems he had in broadcasting the material. Now the passage has a cn tag David Gerard removed from the philosopher Ted Honderich’s wikibio (a) a piece by the author himself (a) T. Honderich, On Being Persona Non Grata to Palestinians Too CounterPunch, March (2005) "Archived copy". Archived from the original on 23 August 2007. Retrieved 2007-08-30.{{cite web}}: CS1 maint: archived copy as title (link) concerning a legal case he won when defamed as an antisemite, and (b) Alexander Cockburn’s comment on the trial, Alexander Cockburn, World-Famous Philosopher Honderich Hit with "Anti-Semite" Slur in Germany; Habermas and Suhrkamp Cut and Run, CounterPunch, 13 August 2003. Nishidani (talk) 22:26, 27 December 2021 (UTC) Amigao has expunged at sight a CounterPunch article co-authored by Neve Gordon and David Dean Shulman, leading scholars in their field. See here for an argument as to why this removal deprives wikipedia of top experts in that field, merely because they chose to publish in what has now become a generally deprecated source David persists after disagreement with his erasure was noted here on the talk page. Cf this excision, and a further stubborn revert here. David is clearly taking deprecation to mean there can be zero exceptions to the ruling, and that the expunger need not even stop to think about the quality of sources otherwise automatically expunged. Nishidani (talk) 21:57, 31 December 2021 (UTC) David here even expunges as non-RS an impugnable RS (Norman Finkelstein, Beyond Chutzpah)published by the University of California Press, as he removes the two CounterPunch articles that summarize the contents of that RS, written by the same author who contributed an epilogue to Finkelstein's book. He removes it because it is close to the CP articles, failing to even note that Beyond Chutzpah, elsewhere on that wiki page, remains an uncontested RS source.Nishidani (talk) 18:55, 2 January 2022 (UTC) Gerard first removes an essay from counterpunch at Yinon Plan and when I restore the content with an alternative source for the same article,he reverts even that with the edit summary:’ (revert unexplained re-addition of deprecated source in controversial editing area without summary).’ This is one more proof that (a) he doesn’t care to cast about for an alternative source, which I found in 20 seconds, (b) he doesn’t examine what other editors are doing in fixing the damage his crusading erase-at-sight rampage through articles causes. Gerard removes two obituaries] to Raul Hilberg, the greatest Holocaust historian, because the distinguished scholars Michael Neumann and Norman Finkelstein who wrote them, friends of Hilberg, posted their commemorations in Counterpunch. They are the source, not CounterPunch, and are irreplaceable. He persisted when challenged by reverting again with the edit summary Rv deprecated content. did they really say nothing outside this? The content is not deprecated, the venue is, and the point that these are obituaries by noted scholars who knew Hilberg, one of the greatest historians of all time. Worse you can easily find Neumann at another RS site, but it is only partial and directs readers to the full Counterpunch article. He didn't check round, but threw the burden on people who, if they notice, actually do spend time checking the net to resolve these issues. Gerard eviscerates several sources from on the basis of WP:NOTRESUME at Jeff Halper including sources in the Jerusalem Post and Al Jazeera. The practical solution was to read those extra further reading sources and summarize them for the article, including the work he wrote for Counterpunch, since Halper is reliable for his own views, regardless of where he chooses to write. Nope. Too much work? Gerard at Issam Sartawi removes a key testimony from a venerated Israeli insider, Uri Avnery, who personally knew Sartawi. Who are we protecting readers from, CounterPunch or Avnery a presumed liar? Gerard expunges a personal anecdote about Robert Fisk by Richard Falk, again a legal scholar of world standing. Is Falk's personal recollection (attributed) unreliable simply because he was interviewed for CounterPunch?Nishidani (talk) 17:31, 5 January 2022 (UTC) David Gerard removed mention of Haifa Street helicopter incident from the article together with a reference to Counterpunch. Alaexis¿question? 07:46, 28 December 2021 (UTC) sorry but what has that to do with figuring out the policy on deprecation? Can we please not drag the PIA debate into here again? Mvbaron (talk) 22:48, 27 December 2021 (UTC) Ignore the fact it is IP, for Nishidani those would be ones he notices in particular but other editors might have different experts that they look to. We can put it one of those boxes if you like.Selfstudier (talk) 22:55, 27 December 2021 (UTC) I just fear it will completely muddy the waters here and we will accomplish nothing but a drawn out re-hash of the same we already have :( Mvbaron (talk) 23:02, 27 December 2021 (UTC) It has everything to do with figuring out the policy on deprecation. These are illustrations of what one interpretation, current, on deprecation does to articles. You cannot analyse policy in the abstract. You have to look at how it functions in practice under certain conditions. Nishidani (talk) 22:53, 27 December 2021 (UTC) And zeroing in upon the cases where removals were maybe, maybe debatable is an excellent way to bias the sample and thereby draw skewed conclusions about how the policy functions in practice. XOR'easter (talk) 23:01, 27 December 2021 (UTC)

• I don't think the "ultimate trade-off" is between speed and judgment. Or, rather, "speed" and "judgment" are both so subjectively perceived that complaining about choosing one over the other is destined to be fruitless. Removals will always be too fast for somebody and too slow for somebody else. Scattered examples of debatable removals will always loom large, because nobody is going to gather statistics on what the proportion of questionable removals actually is — and if they did, anecdotes would still win out over data. The underlying issue is the sheer number of deprecated links still lurking about. XOR'easter (talk) 23:33, 27 December 2021 (UTC)

I think that the concern is a bit overblown here. Your average removal of deprecated source looks like this. Does anyone really think that Robert Jungk did not continue working as a journalist after the war? Btw in this case the {{cn}} was added, which is the right thing to do.

I support removal of CP and other deprecated sources in most cases but let's not pretend that every single reference to them is hugely damaging and justifies extreme measures to remove it. Alaexis_¿question? 08:32, 28 December 2021 (UTC)

Removing every CP ref on sight, without even examining it, is what led to this discussion. By the way, 'pretend' in the sense of claim, is something of a Frenchism or Italianism, isn't it? I don't think pretence in the traditional sense has anything to do with the frenzy of elisions seen recently. It was one editor's honest conviction that deprecation demands such mechanical weeding out. Nishidani (talk) 22:31, 28 December 2021 (UTC)

After some thought it would be nice to see something along the lines of grandfathering deprecated sources. That is, in the immediate days after a source has been made deprecated, it would be inappropriate to remove such in a rapid fire bot like fashion, but in a year or so after that point, it should be okay. This is prevalent that the deprecation is well announced (VPP in addition to RSN or RSP and that editors should be away that the onus is on them to replace or remove deprecated sources before this de facto grandfathering period is up. After that year, then no editor should be able to complain about rapid fire removals. This balances the avoidance of mass disruption across a large number of articles in the wake of a deprecatuon determination, and actually achieving that deprecation in a reasonable time frame. And if a source warrants a faster removal (eg DM and BLP) that should be established during the deprecation debate. --Masem (t) 23:01, 29 December 2021 (UTC)

Thats only part of the problem, the other part is the unwillingness to engage in discussion about individual sources in a deprecated source that may or should be used. Is "its deprecated" enough to remove a source when evidence is offered to suggest it is reliable? nableezy - 23:07, 29 December 2021 (UTC)

I think that question (how to handle one specific case of a deprecated source), but I will add that what propose is a system when the onus initially is on those seeking removal to justify it (even if that means leaving a cn behind when no replacement sources can be found), to the long term onus on thise wanting to keep a cite to a deprecated source to justify its usage. --Masem (t) 00:11, 30 December 2021 (UTC)

As Aquillion pointed out to you when you proposed the same thing at WP:VPP a couple of years ago, an RFC that directly contradicts policy in this manner would be invalid - David Gerard (talk) 00:59, 30 December 2021 (UTC)

I have the impression that selecting specific events like the above, when it's a rather standard procedure to tag or remove unreliable sources, may smell of WP:POINT and not be very convincing. I too have removed and tagged some, there are others but an example I immediately remember was creationist homeschooling textbooks by DI imprints used in WP (liberaly mixing preaching editorials everywhere including math). Those did not require specific consensus deprecation and better textbooks were easy to use. When a source does, isn't it an even better indication that consensus is against their general use by current policy? Context-specific requests can, and are posted at RSN and often the answer is that in some specific cases it may be usable (also by existing policy, like limited WP:ABOUTSELF). Also by policy, material only covered in unreliable sources and that cannot be written about from the point of view of better analysis is generally considered WP:UNDUE. When it's a WP:BLPRS issue, even generally reliable sources that are not considered by all of the community to be independently published, are routinely removed (an example being Skeptoid that some editors target even if it's technically not necessarily published by the guest and the article is borderline BLP, i.e. it has been retargetted to a particular claim more notable than its proponent; I could, too, cherry-pick many examples of source removal). It seems to me that policy is already good enough to resolve most situations and that when unsure, case-specific discussion can take place. —PaleoNeonate – 03:37, 30 December 2021 (UTC)

I forgot this but it's important and beyond deprecation: guides like RSP can also be more flexible than yes/no and specify circumstances where a source is infamous, or can be used, depending on its reputation and the conclusions made at RSN. A great example is Fox News that hasn't been deprecated even if the shows and editorial stance are unreliable. —PaleoNeonate – 03:50, 30 December 2021 (UTC)

I went through the VPP from 2019 (#150) but can't find that discussion that you are refering to (at least, where Aquillion is said to have responded), but 1) policy is not law 2) my point is simply to minimize disruption in the wake of a source being labeled "deprecated", and minimizing disruption is also part of policy. Once a source is marked deprecated, the end goal is to eliminate all uses of deprecated sources outside the limited exception set (which is part of this larger discussion but not important here). But particularly with a source that has had wide use before deprecation was applied can't just be stripped out of WP instantaneously without potentially harming the work. (When we blacklist a source, on the other hand, we've established that removal is a higher priority than minimizing disruption). Hence having a brief grandfather period before mass remove of deprecated sources would be the normal way that disruption is minimized while still driving to meet the end-goals of WP:V and other policy. That's why it is important to understand and or establish the distinction between deprecated from blacklisting. --Masem (t) 15:46, 30 December 2021 (UTC)

I support the "grandfathering" idea, I think removal of "questionable" sources should be somehow subject to a (semi) automatic procedure that allows time for exceptions to be argued while getting shot of the rest in a reasonable time frame.Selfstudier (talk) 15:57, 30 December 2021 (UTC)

@Selfstudier, WP:QUESTIONABLE sources are officially permitted, so they shouldn't be subject to any automatic or time-based removal process. A "questionable" source is any source that at least one of these four qualities:

• expressing views that are widely acknowledged as extremist,
• promotional in nature,
• relies heavily on rumors
• personal opinions

Think about that second item in particular, and how often we cite a subject's official website. If we remove questionable sources, then we can't cite: any manufacturer's website about their products (Special:LinkSearch tells me that there are 80K links to microsoft.com on wiki – you volunteering to remove those?); any book/film/author website; any politician's campaign website; basically any press releases at all; and the list goes on and on. WhatamIdoing (talk) 04:44, 2 January 2022 (UTC)

At Wikipedia:Village_pump_(proposals)/Archive_164#Discussion_on_Proposal_3 - Aquillion points out that "'We want to deprecate this source' does not mean 'we want to provide special protections for existing usages of this source'", in fact it means the opposite. Your proposal is to provide more protections for sources explicitly deprecated by a broad general RFC than are provided for merely unreliable sources, which can be removed summarily.

He also notes: "WP:RS is core policy and not subject to consensus; therefore, you can always remove an unreliable source on sight with the reason of 'unreliable source', no matter what, without exception" - though actually WP:RS is a guideline included by reference in WP:V, which enforces that provision. An RFC can't actually find against that, and an RFC proposing to do so would be invalid. Your proposal would require a policy change to enforce, or at the least an RFC to alter all previous deprecation RFCs - David Gerard (talk) 19:02, 30 December 2021 (UTC)

What prior, formal RFCs (not just proposals discussed at VPP or elsewhere, but actual RFCs) have there been on the nature of what deprecated sources mean and how they should be handled. We as a community have talked around the matter for a few years, but this is the first actual RFC that I'm aware of to try to nail down those facets. So there would not be any alteration of prior results. Also, I would point out that there's a difference here of what I'm talking about, specifically a grandfathering period after a source becomes deprecated via an RFC, and not the issue of dealing with long-term deprecated sources as that link above was more focused on. Long-term, I tend to agree we can be more indiscriminate about removing deprecated sources well after the source has been established as deprecated ,but the short-term disruption -as well as wider notification that a source is deprecated and will be removed indiscriminately by X date - is what needs to be handled better. That's not really anything discussed in that prior VPP page. --Masem (t) 02:16, 31 December 2021 (UTC)

Perhaps some stats on the usage levels of deprecated sources might be useful here? If it's an issue, I think it would be fine to establish something like: starting X months after deprecation, deprecated sources can be removed indiscriminately, except where there is consensus to retain it, to be indicated by a specific "Acceptable use of deprecated source" template. The challenge would be in making sure that doesn't get interpreted as a way to protect those sources for X months, but if those sources are being left in articles regardless then perhaps it would still be an improvement over the current system. Sunrise (talk) 00:07, 31 December 2021 (UTC)

This is sneaking in Masem's assumption that the current removals are "indiscriminate", which is something he's consistently failed to show to the satisfaction of others. Remember that he's been trying to put this plan for special protection of deprecated sources into place for a couple of years now, and it's failed to achieve consensus every time - David Gerard (talk) 12:25, 31 December 2021 (UTC)

I have shown on several pages and the hatted examples above that your reverts are indiscriminate. A fresh one, among many today, is instanced here, with the edit summary:'Rv repeated deliberate addition of deprecated sources to wikipedia - do not do this').

This revert took place after David was notified on the talk page that the policy he cites allows for exceptions. Nishidani (talk) 13:56, 31 December 2021 (UTC)

Oh please. Two or more removals per minute does not allow time for considered judgement (which would simply involve a very basic search for replacement sources in many cases – something that has almost never been done) so they cannot be described as anything other than indiscriminate. _wjemather^{please leave a message...} 16:37, 31 December 2021 (UTC)

Once more, indiscriminate, and when contested, indiscriminately restoring his excision in the face of a talk page challenge. The only argument is 'deprecation' means I can removed everything, apparently. See here, here and here persists in removing Uri Avnery as a source for Amos Kenan, though the two were intimately acquainted and this is an obituary written by an insider. He does so by not responding on point on the talk page and against two editors who disagree with his reading of deprecation policy. Encyclopedias are written by people who generally have extensive knowledge of the specific topic, not by editors who have a strong but partisan reading of one policy that appears to give warrant to consecutive blind erasures regardless of content familiarity or expertise. Nishidani (talk) 16:46, 31 December 2021 (UTC)

Another reason to automate the process, thereby avoiding such accusations.Selfstudier (talk) 12:32, 31 December 2021 (UTC)

The problem is that you have to determine when you have "<content> <citation to deprecated source>" if the content needs to stay (now with a CN behind it) or go, and no bot will be able to do that. But once we're in a mode of doing that indiscriminately (after grandfathering period), it only takes a few seconds for a user to make that determination -eg what DG has been doing reflects the rate and outcome I would expect once we're past grandfathering. I don't think we want all deprecated sources replaced with CNs, given that this would not be appropriate on BLPs, for example. --Masem (t) 18:36, 31 December 2021 (UTC)

What I had in mind (if it is possible) is that all "questionable" sources (I am not going to distinguish between unreliable and deprecated because I don't really see that much difference apart from the SPS aspect) get tagged (could even cat them, maybe) with auto removal after a time period unless (talk page discussion/something). So people get a chance to make the argument and if they don't, it's gone.Selfstudier (talk) 18:44, 31 December 2021 (UTC)

Using a bot to tag articles creates the problem that tags would never get removed, which is also a problem as identified in these discussions on deprecation. This is different from humanly deciding that removing a deprecated source but likely with a truthful statement that just needs a CN or even keeping a deprecated citation tagged appropriately, compared to outright removal. That all sstill requires a human review. --Masem (t) 18:33, 2 January 2022 (UTC)

It is not possible to auto remove within a timeframe? I didn't know that, that's no use then, if there is no time pressure to remove then I agree with you, they will just sit there. Selfstudier (talk) 18:39, 2 January 2022 (UTC)

This is what we see in practice. Tagging bad sources results in nothing happening, often for years. Removing them (either bad cite and claim, or replacing with {{cn}} sometimes results in a replacement, but vastly more often they were always bad and stay gone. Functionally, "why don't you just tag them first/instead" is a way to not do anything about the bad sourcing - David Gerard (talk) 20:44, 2 January 2022 (UTC)

You are describing problems that are the direct result of indiscriminate mass removals. If WP:BURDEN and WP:PRESERVE were being followed, these things would hardly ever happen. _wjemather^{please leave a message...} 21:46, 2 January 2022 (UTC)

Special protection for deprecated sources is a bad idea and not worth further discussion IMO, although I can see where the idea comes from. Thing is, these removals would not be an issue (regardless of speed) if there were actual checking being done on removal, and (per the close of WP:DAILYMAIL1: Volunteers are encouraged to review them, and remove/replace them as appropriate.) an actual 'review' was being done and the 'appropriate' action was being taken. Thing is, many policies break down when applied indiscriminately and at scale, because almost all of them require actual human judgement, and I think that's where the dispute with your style of removal comes in, because you place more value in having an immediate disclaimer that the content is effectively uncited, whereas other volunteers place more value in retaining the encyclopaedic content if possible (or having the content removed otherwise). Regardless, I don't think the solution to this problem is a set of bureaucratic rules which add more protections for deprecated sources. ProcrastinatingReader (talk) 13:41, 31 December 2021 (UTC)

What the DM close should have done (and what we've failed to do on all such deprecation RFCs since) is to set a grandfathering period. Using the comp.sci. meaning, users are usually given fair warning how soon a deprecated feature will no longer be supported as to find alternatives, and that's the same thing we should be doing. We should have had consensus set a reasonably time frame for editors to remove/replace them, but we didn't, and now we're at the conflict being for any now-deprecated source "they must go ASAP" versus "giving indefinite time to be fixed, which often means never". Deprecated sources should be removed and not allowed to linger, but we should be accounting for the fact that until the RFC that marked them deprecated they were used in good faith as valid sources, and thus giving a reasonable amount of time for human review to try to remove/replace them without disrupting other content. And to me, six months after an RFC on deprecation of a source is closed and announced to VPP/CENT is more than sufficient time before we go from "careful, human review" of each instance to indiscriminate removal, as well as consist when we have had to grandfather other content due to cchanges in core policy (eg when NFC was required by the WMF in 2008 for example). --Masem (t) 18:36, 31 December 2021 (UTC)

Masem. As a content editor I'm still puzzled. David Dean Shulman is a scholar of international standing, who has written 2 academic books on the dispossession of Bedouin in the South Hebron Hills area, of which he has perhaps a unique knowledge, since his Sabbath for 20 years has consisted in being present there to prevent their harassment. Since he has an unparalleled knowledge of innumerable incidents, why should he be deprecated as a reliable source because when not publishing for the New York Review of Books on some incidents, or Haaretz on others, he chooses CounterPunch for some details, whose wiki reputation as a genocide-promoting, Holocaust-denying, conspiracy-mongering antisemitic website is unknown to him (as it is to peon tillers of the wiki patch like myself who have read it (selectively for the abundant quality articles it carries) for almost two decades), or if known, would be probably dismissed as risible, since like so many other Jewish scholars, he publishes there. I can live with a general deprecation ruling, but not one that allows of no exceptions and endorses instantaneous erasures of anything regardless of the authors' global reputation. There is, in such a case, no alternative source to substitute for the CounterPunch material, because these incidents are rarely covered by mainstream sources (and most mainstream sources are written by journalists with, unlike Shulman, no direct knowledge of these incidents). Nishidani (talk) 18:55, 31 December 2021 (UTC)

PR asks (in effect) what is the solution? Dedeprecating CP isn't winning at RSN although it has some support. I suggested SPS exception or some sort of autoreview process. DG (and Amigao) want it all gone pronto. No front runner yet, any more ideas? Selfstudier (talk) 19:04, 31 December 2021 (UTC)

One can forget predictably partisan scumbags like myself in these deliberations. People with a far better record of impartial and authoritative commentary exist among that significant minority, meaning that my perception that we have created a serious problem with deprecation as warrant for automatic erasure is not solipsistic but shared by saner wikipedians. A a notable number of high quality and irrepeatable sources of encyclopedic value are being junked or trashed (the Daily Mail analogy is sand-in-the-eyes, scores of top-ranking scholars do not contribute to it). Since most commenting here are policy wonks (I'm out of my depth) some collegial solution that addresses the nature of those policy grey zones is required. Nishidani (talk) 21:50, 31 December 2021 (UTC)

What the DM close should have done (and what we've failed to do on all such deprecation RFCs since) Yes, but (a) it didn't do that (b) your attempts for literally years now to make it mean that have consistently been rejected. You could mount an RFC to retrospectively make all deprecation RFCs mean that, but it absolutely doesn't mean it now, and going on and on and on as if it really should mean that is just you refusing to accept that broad general consensus has been against you on this point. At this stage, do you think you could craft such a proposal as an RFC with a chance of passing? - David Gerard (talk) 00:12, 1 January 2022 (UTC)

Is this about the correct construal of Holy Writ/fundamentalist approaches to the literal sense of the Founding Fathers or about problem-solving (what do we do when a deprecated source regularly publishes articles by scholars/researchers/journalists of distinction. Deprecate or define exceptions more lucidly?) If any issue or crux arises is to be talked down by simply saying 'it is written', then we are behaving like theologians, for whom the answer to anything is already inscribed in traditional sacred texts (policy) and not problem-solving analysts. Nishidani (talk) 09:37, 1 January 2022 (UTC)

Based on past deprecation discussions and currently what's in here, an RFC to establish a grandfathering approach for a limited period after a source becomes deprecated would actually have a good chance of passing. But before starting an RFC would be to see what this whole discussion (on the other areas) comes out to be. The broad consensus has been split, not outright against this. --Masem (t) 18:35, 2 January 2022 (UTC)

• Two ideas (ideally done together):
  1. Upon deprecation, a bot will change all usages of deprecated sources to include {{deprecated source}} or similar, which functionally acts as a disclaimer and behaves equivalent to {{citation needed}}. However, this has the effect of still leaving the link to the source present, which will mean deprecated source usages are categorised separately and the article can be easily accessed to aid a volunteer who actually wants to 'review/replace' the reference. It might look something like ^{[1][unreliable source]} and adding a category Category:Articles with references to CounterPunch.
  2. Codify in the hypothetical guideline page the 'handling deprecated references' guidance from WP:DAILYMAIL1: Volunteers are encouraged to review them, and remove/replace them as appropriate. Specifically, clarify what "review" and "remove/replace as appropriate" means, and institute a de-facto WP:BEFORE-like requirement as exists for AfD. Something like Existing references to deprecated sources may be handled by any editor. Editors reviewing existing citations should review whether the reference is appropriate in the context of its usage. If it is, it should remain. If it isn't, the volunteer should see if the reference can be replaced with another source. If a source cannot be found, either leave as-is for another volunteer to deal with, or remove the content supported by the ref along with the ref. In the case of a dispute about whether the reference is appropriate in the context of its usage, editors should seek WP:Dispute resolution and discuss on the talk or another appropriate venue, giving consideration to WP:REPUTABLE and considering the criteria individually. (subject to word-smithing) This makes a deprecated reference (and its removal) functionally identical to a {{citation needed}} (and their removal). After all, you wouldn't just remove a {{citation needed}} tag without either adding a reference or removing the underlying content, that'd just be absurd. You would only remove the content and tag in line with the provisions in WP:BURDEN. This offers good guidance to editors, and makes the expectations more transparent for ANI to deal with individual cases of mishandling.
  ProcrastinatingReader (talk) 16:52, 1 January 2022 (UTC)

  Support in principle. I would submit for consideration that the "hypothetical guideline page" should amalgamate both unreliable and deprecated under the rubric "questionable" as per V resulting in a substantial simplification of existing procedures. In other words, unreliable sources lose the SPS exception, apparently the only meaningful distinction between the two, unless editors make the argument per the outlined procedure (as potentially wordsmithed).Selfstudier (talk) 22:17, 1 January 2022 (UTC)

  That ("amalgamate both unreliable and deprecated under the rubric "questionable") doesn't work.

  First, no source is always unreliable; see the FAQ at the top of WT:V. There is no "SPS exception" for an unreliable source; if the source is unreliable for the material it's meant to be supporting, then being self-published by an expert does not change the fact that it is unreliable for that material.

  Second, "questionable" is its own jargon, and the use of questionable sources is expressly permitted "per V" (also "per RS", where I think you'll find the slightly clearer of the two explanations). We shouldn't conflate "questionable", "unreliable", and "deprecated".

  Third, "deprecated" includes sources that we deem to have altogether too strong of a "reputation for checking the facts" (to make sure that they line up with the official version) and far too much "meaningful editorial oversight" (to make sure that the censors don't object), so they don't really fit the definition of "questionable". Perhaps the solution there is to re-write the definition; it happens that our idea of reliable sources also includes sources that do absolutely no fact-checking at all (peer review is not fact-checking) and sources that have obvious conflicts of interest (e.g., almost every single academic journal article ever published about any pharmaceutical drug during the ~30-year-long span between the initial discovery and the date when the last patent finally expires).

  The more I read about the active disputes, the more I wonder whether the problem lies less in differing definitions of the terms and more in a lack of clarity about editors' behavior. Is it okay to mass-remove sources from articles? Exactly how much WP:BEFORE-style effort do I need to go to? Engage the brain fully, or just enough to make sure that I'm not blanking dailymail[.]co[.]uk from the article about Daily Mail itself? Is it better for Wikipedia to have uncited information with no idea of its provenance, or is it better to have a flag on it that says ^{[better source needed]} so that editors will know where that potentially dubious material came from? WhatamIdoing (talk) 05:57, 2 January 2022 (UTC)

  I get where you are coming from but the editor behavior is actually being driven by the definitions because everyone can find a WP:something to support what they are trying to do. The definitions need clarifying at a minimum and personally I still think simplification would be a good idea (although I am not sure if RS RFCs going 1,3,1,3... is an improvement over 1,4,1,4..).Selfstudier (talk) 10:40, 2 January 2022 (UTC)

  This also seems to contradict policy, in that it attempts to make special rules to preserve unreliable sources. This is quite apart from the bit where you're proposing something to apply to those sources found to be the worst sources, worse than merely unreliable sources.

Look, think of it this way: if you see a clearly garbage source in an article, can you just remove it? Yes, you can. It's frankly implausible to make it harder to remove the sources that a broad general consensus has found to be garbage than to remove ordinary garbage sources. If your proposal does that, it's ill-thought-through. Deprecated sources are those that are known-bad, and there are no grounds to assume a priori that they are not - David Gerard (talk) 23:44, 1 January 2022 (UTC)

The aim isn't really to make it harder. I agree with you that proposals to add novel restrictions on the removal of deprecated sources are untenable, hence why I disagree with Masem's proposal and particularly any kind of 'time limit' proposal. AFAIK there's no novel restriction in either of my ideas. Plus, #2 is exactly how you would handle statements that currently have no reference and just a {{citation needed}} tag (or no tag at all); indeed WP:BURDEN says more or less the same if one reads the full section.

The aim is to make clear the process for removal and that WP:DR should be sought in disputes. The specifics of deprecated sources being not entirely agreed on, it's not immediately clear what grounds editors would have to challenge the removal in WP:DR, but I think other sections in this RfC (namely: 'purpose of deprecation', 'criteria for deprecation', and most importantly 'valid uses of deprecated sources') need to be answered first to determine such grounds. ProcrastinatingReader (talk) 00:27, 2 January 2022 (UTC)

So long as "it's deprecated" is not grounds to remove source reinserted and a good faith discussion about any particular source can take place to examine any individual source on its own merits, I dont have a problem with removing sources deprecated. So I agree, valid uses of deprecated sources is the biggest thing to be decided. nableezy - 00:33, 2 January 2022 (UTC)

There's also the problem of how to signal to editors who are less than fully informed that it's an acceptable use. You've probably seen this happen before: Someone changes an article; there's a discussion that results in a consensus. Then someone else changes the same thing, and everyone yells at the latest editor for not having magically guessed that there had been a discussion about that exact thing.

As an example, we had a couple of years of edit-warring over whether the WP:LEADIMAGE for Pregnancy needed to have a naked woman in it. Editors (often less-experienced editors) were regularly surprised by the rather WP:GRATUITOUS nudity at the top of the article, so they would remove it, and they were regularly reverted by a couple of guys claiming WP:NOTCENSORED. It took a couple of RFCs to settle the matter, and there is no more edit warring over the images now, but I still wouldn't expect any editor to know that these prior discussions took place (except for the participants in the discussion, and even some of them might have forgotten by now).

If you are just working your way down Special:LinkSearch or using WP:AWB, hitting hundreds or thousands of articles, you can't possibly know that there was a discussion about this particular URL being used in this particular article. WhatamIdoing (talk) 06:13, 2 January 2022 (UTC)

I think that generally, in an encyclopedia, editors who are less than fully informed should refrain from editing. nableezy - 22:16, 2 January 2022 (UTC)

If inexperienced editors refrain from editing, no editors become experienced and Wikipedia dies. I think the onus should be on those seeking to include deprecated material rather than on those seeking to remove, but it's good to give inexperienced editors as much guidance as possible. BobFromBrockley (talk) 19:02, 3 January 2022 (UTC)

If I had said editors not experienced that might make sense. I said editors not informed however. As in if you have no understanding of the topic or the source, maybe dont remove it? nableezy - 15:19, 4 January 2022 (UTC)

OK, that I agree with (but extended-confirmed doesn't mean informed, and we don't really have a mechanism for identifying who is informed, except using community consensus to correct un-informed editing). BobFromBrockley (talk) 12:19, 5 January 2022 (UTC)

While noting that the process for new inclusions differs somewhat from the process of removing old ones, the onus is of course with those arguing for inclusion but that is the case for every single source not only deprecated sources. Selfstudier (talk) 19:07, 3 January 2022 (UTC)

If we decide that editors who are less than fully informed should refrain from editing, then we all need to quit now. Nobody is fully informed about everything, and most of us are not fully informed about anything. WhatamIdoing (talk) 04:05, 8 January 2022 (UTC)

The issue I see is that this appears to put the onus on the editors seeking to remove the depreciated source, rather than on those seeking to restore it - that onus has already been met by the depreciation. I think it would be better to state that once a depreciated source has been removed, it should not be restored without consensus. BilledMammal (talk) 03:52, 2 January 2022 (UTC)

When I started the RfC I intended for the questions to be answered sequentially. I do partially think it's a bit silly to answer this question without having the foundation gained from the previous ones. To the point, any response I can give to you will be based on my (unwritten) opinions to the previous questions, rather than anything we've established. I did comment on this to try re-focus the discussion on something actionable, but I still think it's more ideal to answer the issues sequentially so we're not talking over each other or starting from different foundations. ProcrastinatingReader (talk) 04:13, 2 January 2022 (UTC)

I did have a similar thought when writing that reply, as answering this question without an answer to the others will not resolve the issues that started this discussion, but I am still working through my position on the other questions, and I felt it was important to state my position here, both because we have a clear proposal, and because I see the onus being against those who wish to retain as a fundamental aspect of deprecation that should not be changed even if we clarify other aspects.

To expand on why I see it this way, it is because we need a practical method to address broad historic use of inappropriate sources, and to switch the onus would stop deprecation being that practical method. BilledMammal (talk) 04:28, 2 January 2022 (UTC)

@BilledMammal, since you are, compared to most of this in the discussion right now, a relatively new editor, I think it might be informative for the rest of us if you explained what you mean by "inappropriate sources". Most of us remember when the idea of a list like Wikipedia:Reliable sources/Perennial sources was laughed at as pointless, because reliability depended so much on the specific material in the article that you were supporting. But I wonder if you've "grown up", as it were, with a model in which the article content is less relevant than something else (the publisher's reputation, maybe?). What do you think makes a source "inappropriate"? WhatamIdoing (talk) 06:22, 2 January 2022 (UTC)

Nothing more than that "inappropriate sources" are those whose publisher has been deprecated by consensus. What should be deprecated (and thus made inappropriate) is a question I am struggling to answer, but I have some thoughts that I'll go and post in the relevant section. BilledMammal (talk) 06:54, 2 January 2022 (UTC)

Unfortunately, I still don't have a definition of "inappropriate sources" or "Generally unreliable" sources, but I am not sure we need to produce one here; the consensus process of determining whether a source is "Generally unreliable" appears to be functioning fine, with the issues being related to the lack of ability to distinguish between sources, and because rather than including nuance in the discussion and summary, we appear to use the easy method of deciding whether we think a source is "bad" or "really bad" and classify the first as "Generally unreliable" and the second as "Deprecated". BilledMammal (talk) 15:46, 2 January 2022 (UTC)

@BilledMammal, your description of "the easy method" fits with my impression, too. The RSP process results in far more emphasis on picking a red–yellow–green color than on producing nuanced understandings. Sometimes nuance isn't necessary – as an example of this, I remember someone copy-pasting the contents of an AIDS denialist website into articles; it was both factually wrong and a copyvio, neither of which require careful consideration of details – but for most news media, nuance should be the main goal. WhatamIdoing (talk) 05:34, 8 January 2022 (UTC)

I said something similar a long ago, I think of the first times this blew up. In general, it's fine to remove a detail sourced to a deprecated source if an editor feels this is justified, just as you can with anything unsourced. I think doing it en-masse is likely to be seen as a problem just as deciding you're going to remove anything tagged with CN throughout Wikipedia but that's mostly a separate issue which frankly I don't care about. But it's a major mistake to remove deprecated sources and replace them with CN. It's making things worse not better. While deprecated sources may be completely unreliable, they still help provide details such as a rough time frame or maybe other related details in the deprecated source which may help find a reliable source.

As a far better alternative, personally I suggested the better souce needed tag back then but if editors feel this isn't clear enough then we can make a new tag e.g. that suggested by ProcrastinatingReader. As I think I also suggested, if editors really want, we can even make the tag hide the source so it doesn't appear to readers. Alternatively editors could put the source into a hidden comment although this isn't ideal since someone trying to source the detail may look for a source before editing. Whatever there are surely so many solutions that don't involve damaging Wikipedia by making it harder for editors to improve it, which is what we're doing when we remove a deprecated source and replace it with an CN tag.

P.S. Why do people keep saying that we replace unreliable sources with the CN tag? There's nothing in WP:RS that supports this practice and just like replacing deprecated sources with a CN tag, it's a bad idea unless you certain the source provides no useful information. Sometimes where the source is so bad, it may be better to remove it and bring it to the talk page but frankly this is rare, mostly it's better just to tag it. The exception would be where the source itself is blacklist or harmful e.g. may have malware.

P.P.S. I admit, I've probably personally removed some forums and other non RS and replaced them with CN well generally the fact redirect, in the past. I don't think recently but I'm not sure. Most of the time, I shouldn't have done so. I do appreciate in some cases it's questionable if the source provides any useful information for finding another source that isn't obvious from what we say in our article so I acknowlede in some cases we're probably not making things worse, however as this requires careful analysis and you could easily be wrong, it just seems like a bad idea in general when there are much better options available.

Nil Einne (talk) 09:07, 3 January 2022 (UTC)

There are multiple possible actions - replacement of the bad cite, remove the claim and the bad cite, remove the bad cite but not the claim - and advocates who will insist that any of these must be required. But every case is unique - David Gerard (talk) 15:13, 3 January 2022 (UTC)

The fact that every case is unique is why removal of deprecated sources should be human-based and not done by a bot or made through bot-like, rapid fire actions at least without prior consensus that supposed the use of bot or bot-like actions (per WP:MEATBOT) I would figure with my Google-fu that judging if there's another source out there or if the claim has some potential to be valid but needs a better source to affirm would take 2-3 minutes for a given citation, and that's assuming its a topic within the 21st century. The result could still be outright removal, or leaving a CN behind, or replacing the source but there's no way I can do that in less than a minute per cite. Now, as I proposed, after 6 months from a formal announcement, then we can talk about using high-speed bots or tools to remove or handle such cites without the same human-review care. (Personally, creating a special CN tag that includes in hidden wikitext the original citation, and flagging said page to note its past use of deprecated sources, would be a good step). --Masem (t) 01:31, 4 January 2022 (UTC)

Your 2-3 minutes claim is completely made up, and not at all supported by the near-uniform badness of deprecated links - and mainly shows that you complain about deprecated links being removed while having spent negligible time going through the backlogs yourself. Have a look through my edit summaries for the past couple of days - I took a cue from Bobfrombrockley and made each summary much more explanatory. If you can claim with a straight face that this editing is indistinguishable from a bot, then you have the programming skills to radically advance Wikipedia bot editing, and you only need to take care not to tell the bot what a good idea paperclips are - David Gerard (talk) 23:48, 4 January 2022 (UTC)

The ones that I see of late are removing opinions from Counterpunch under the claim of UNDUE. That's extremely subjective (RSOPINION can still apply to deprecated sources, but it becomes a matter of discussion if that person's opinion is UNDUE or not, not simply because it was in a deprecated source (it depends on the nature and reasoning for deprecation), and of course, making your own decision on the outcome of this page's very discussion, which is basically not appropriate per WP:FAIT. But even those that are removing deprecated sources to claimed facts, I know that it takes a bit of time to do a skim of results from Google, Google Books, and Google Scholar, and that's at least a 2-3 minute process just to satisfy that a replacement sources is not immediately obvious (eg on this diff [4], it took me about 1.5 minutes to locate this as a replacement source from Deutsche Welle, at least for the first part of the removal). The rate your edits are going and lack of any attempt to find replacements (instead just more interested in removal) is overtly disruptive, particularly given wholly separate issue about possible sock involvement in its deprecation discussion. What you are doing is again against what ArbCom has said not to do in WP:FAIT, regardless if at the end of the day the goal is to remove all uses of deprecated sources. There are far too many open questions on the process that you should absolutely not be continuing in the mode you are in until the issues raised on this page are resolved. --Masem (t) 05:26, 5 January 2022 (UTC)

I get your point Masem but the Arafat poisoning example is actually a good instance of the system working isn't it? David Gerard deleted the cite but left the claim, which looks plausible on the face of it, and added a CN tag; another editor then provided a reliable source; overall the encyclopedia was improved by a stronger source being used instead of a poor one. The problem is only when the whole claim is deleted too quickly, but most instances of that happening seem to be defensible (and can be corrected by other editors watching the pages, if we work together collegially). BobFromBrockley (talk) 12:19, 5 January 2022 (UTC)

I mean, I am not denying that the ultimate net result (where a deprecated source is replaced with a higher quality source) is right. However, if we go off the DM RFC, the key is that deprecated sources should be replaced or removed, and there is no sign that David is making any good faith effort to find replacement sources, creating the onus on others to fix up when they are just removed (removing material completely or leaving a CN), which is a MEATBOT problem. And this is still within days of Counterpunch, having been previous assumed a reliable source in good faith across numerous articles it seems, being declared at RSN to be deprecated (note: not widely announced to the rest of WP), giving no time for editors on affected pages to try to find replacements first. That's a FAIT problem and is disruptive to the work. Now, if David wanted to run a community effort to stamp out deprecated sources, both announcing it widely (VPP/CENT) and seeking consensus to make sure this was what the community wanted (without any grandfathering period), then hey, that's all good. But David's taken on their own interpretion of the results of the various deprecated RFCs to imply they have to be removed immediately (no such urgency was mentioned in any RFC outside of DM + BLP) and is against the current PAG pages related to deprecated sources (what we're here on this page to resolve), even while the community is trying to resolve that. You can be right as to the end goal, as to eliminate deprecated sources, but the methods of getting there can be against standard practice or community expectations, and that's the issue here is David's practice, not the end goal. --Masem (t) 13:11, 5 January 2022 (UTC)

While deprecated sources may be completely unreliable, they still help provide details such as a rough time frame or maybe other related details in the deprecated source which may help find a reliable source. Exactly. I don't see the obsession with removing links. It's not like visiting the website will give you a virus or something. There's the same unreliable content in both scenarios, but in one it's easier to detect the source of unreliability and easier to find a replacement. It hurts my head thinking about how some folks are doing removals currently, because it just seems quite clear to me that this approach doesn't make sense regardless of what your POV is on the source in question. ProcrastinatingReader (talk) 18:49, 3 January 2022 (UTC)

There has been at least one unique situation where deprecated links needed to go post-haste - that being Daily Mail for BLP, but that was clearly defined during the DM RFCs leaving no question of rushing through to remove them. But most of the other deprecations RFCs have not discussed speed or urgency for removal. I have to agree with David on that if there's no deadline set, these will linger forever, and we want to promote a means to replace/removed, but unless we have a DM BLP situation, urgency to remove is not established. Hence why I've suggested this grandfathering as to actually create an explicit default timeframe for that purpose. --Masem (t) 02:37, 4 January 2022 (UTC)

I agree with @ProcrastinatingReader about the verifiability problems created by removing sources. Content with a bad source (preferably a tagged bad source) behind it is easier to evaluate than content with no source. WhatamIdoing (talk) 05:44, 8 January 2022 (UTC)

Definition of "Self-published sources"

In cases where the creator and the publisher are the same, a work is always considered self-published. In cases where they are different a work is considered self-published if the publisher has minimal direct or indirect control over the content, such that work is the independent product of the author and can be solely attributed to them, rather than having aspects that were influenced by the publisher. There is a presumption that publishers with an editorial process has more than minimal control over content.

Is this a suitable definition for whether a source is self-published, and if not how can it be improved? For additional information, including the reasoning for the specific wording, see this section, though most of information there, including how it might be integrated in the RSN process, is beyond the scope of this sub-discussion. BilledMammal (talk) 17:22, 2 January 2022 (UTC)

While the beginning is very clear, I think that such that work is the independent product of the author and can be solely attributed to them, rather than having aspects that were influenced by the publisher. There is a presumption that publishers with an editorial process has more than minimal control over content. is unnecessary and adds a lot of confusion.

• such that work is the independent product of the author and can be solely attributed to them Isn't that the case for all works, though? Even those that have been reputably published? If I cite something from the New England Journal of Medicine, the work is still the independent product of the authors and can be solely attributed to them. The NEJM is only the publisher.
• rather than having aspects that were influenced by the publisher This very absolute phrasing contrasts with "minimal" at the beginning. If you have minimal editorial oversight (limited, for instance, to copy-editing) some "aspects" are "affected", but the work may still be classified as a self-published source.
• There is a presumption that publishers with an editorial process has more than minimal control over content. This is not a very useful presumption, because it shifts the initial problem ("minimal control") to the question of what constitutes "editorial process". It's also a non-sequitur because "editorial process" is by definition a more substantial review than "minimal control". So we don't really gain anything by adding a new concept. I think that the concept of "minimal control" is sufficient. JBchrch talk 00:20, 4 January 2022 (UTC)

In a prior separate discussion related to SPS, I've suggested that when we talk SPS, we're distinguishing SPS from sources where there is an entity involved that edits, reviews, validates, and greenlights publishing of the work (either they publish directly, or give the go-ahead to the author), and subsequently handles post-publishing matters (redaction, etc.). This is more than just a having a simple copyedit review before the user publishes (a mechanism that Forbes Contributors use as best we know). This was intended to be clear that an SPS is not necessary a work published by the creator on a site owned by that creator - eg this expands SPS to include YouTube, Forbes Contributors, and Medium, as well as op-eds, short correspondence in most peer-review journals, and open-source or predatory journals. Whereas, with a peer-reviewed journal, there's the whole peer-review process and board of editors that provide the editorial oversight, equivalent to the editorial desk at NYTimes. There was some agreement but the discussion (wasn't a formal RFC) never got to a conclusion. --Masem (t) 02:32, 4 January 2022 (UTC)

@BilledMammal, I think your definition might be confusing "platform" or "printer" with "publisher". The publisher is the person/organization that decides whether this source is going to be made available to the public.

The source doesn't need to have "aspects that were influenced by the publisher" to have been properly published. Similarly, merely having aspects that were influenced doesn't mean that it wasn't self-published. You can hire a designer and an editor from a vanity press. They'll definitely "influence aspects", if you get what you paid for. But you're still the one who decided that it was going to be published, and that makes it self-published, no matter how many others influence your source. WhatamIdoing (talk) 05:50, 8 January 2022 (UTC)

Discussion

This is related to the RFC on CounterPunch and honestly this is the heart of what I was trying to get at with it. We are somehow treating "editorial control" as a detriment compared to actually self-published sources, and that goes against the entire hierarchy of sources we have here. Unless, and only unless, there is some evidence that the "editorial control" involves misrepresenting what a given person has written, then a self-published source is the very lowest rung in our source quality hierarchy. And every exception to self-published sources should exist to any other source, absent that evidence of misrepresentation in which you cannot even trust that the person actually wrote what the source says they wrote. I still do not understand how the counterargument makes any sense. That if somebody had posted the article they put up on a deprecated source on their personal blog it would somehow be a more acceptable source? Seriously, can anybody explain the logic of this, assuming there is no argument that the deprecated source has ever modified a piece submitted by the author? nableezy - 00:38, 4 January 2022 (UTC)

The basis that I am working from is the notion that if this is not the same article that would have been published without the influence of the publisher, then it is not self-published, with the exception of copy-editing changes. The reason for this is that the publisher, even if they don't misrepresent the work, has an influence on it, and if the publisher is not reliable then that means the work is not. For instance, an article published in History will be heavily influenced by History's editorial preferences, and so even if the author would be acceptable as an SPS if they published on Medium, they are not acceptable as an SPS if they published on History.

To meet this, and taking into account the comments by User:Masem and User:JBchrch, I propose the following altered text for consideration: In cases where the creator and the publisher are the same, a work is always considered self-published. In cases where they are different a work is considered self-published if the publisher has minimal direct control or indirect influence over the content. BilledMammal (talk) 05:17, 4 January 2022 (UTC)

That isnt addressing my point. There is no "influence of the publisher" in any source in which they do not actually "edit" the material. And even if that were the case, why would it matter? Why would it matter if I, video games history professor and widely cited scholar on the mythology of Halo, wrote something for a deprecated source or posted the same column on my personal blog? Exactly how is my blog more reliable? Nobody is saying that it is self-published, and I would appreciate if my actual argument was addressed, not one I did not make. Mine is, given that self-published is the lowest rung on the ladder of reliability, every exception made for that applies to any other rung. Given, again, absence of any indication that the source cannot be trusted to accurately reflect the named authors words. Why exactly would my posting a blog article be citeable but my posting in a source, lets call it UpperCut, that has been deprecated not be? nableezy - 10:14, 4 January 2022 (UTC)

The reason is because the deprecated source has influence on the work. We know that the SME is suitable for use when writing without that influence, but we can not assume that they are when they are writing with that influence. BilledMammal (talk) 10:44, 4 January 2022 (UTC)

What exactly is that based on? That is, to me, completely irrational and illogical. What influence does a publication have on a scholar writing a piece published under their own name? Where is there any evidence of any influence at all. And, crucially, how would that impact the reliability of the scholar. nableezy - 15:29, 4 January 2022 (UTC)

The fact that works are not independent of the publisher, per current policy. And the nature of influence can vary, but it does exist; scholar writing an article about the Holodomor for RT is going to be limited in what they can say, and so the reliability of the piece will reflect RT's reliability, rather than the reliability of the scholar. To be clear, the reliability of the scholar isn't influenced, the reliability of the work is. BilledMammal (talk) 01:07, 5 January 2022 (UTC)

This argument doesnt make any sense to me. Yes, things are not independent of one another, but you seem to be taking "deprecated publisher" as a negative as a opposed to "self-published". But that is not true if the pieces are not being modified, if the author is putting their name to a work it is their reliability at issue. The different parts that affect reliability, author, publisher, the work itself, those are cumulative. Think of it as a math problem, with each part adding to the other. And the lowest value for publisher, 0, is self-published. What WP:RS says is Reliable sources may be published materials with a reliable publication process, authors who are regarded as authoritative in relation to the subject, or both. These qualifications should be demonstrable to other people. That or at the end there, means the sentence should be read as that either the author or publisher being reliable may make a source reliable, or, obviously ideally, both. Deprecated should mean that the value of the publisher is 0, and in cases like the Daily Mail where the publisher is responsible for the actual output more so than any named author, that may render a source closer to the blacklist. But for things like opinion pieces or works by experts, that just makes them equivalent to being self-published. nableezy - 02:57, 5 January 2022 (UTC)

I disagree that the lowest value is zero; I believe the lowest value is negative. I think we can agree on this in the case of publications like the Daily Mail, where we know that works are directly modified, but I also believe this is true in cases where the work is not directly modified, as might be the case in my Holodomor example above - but I don't think we are going to agree here. BilledMammal (talk) 03:12, 5 January 2022 (UTC)

I think Ive been clear where there is evidence of manipulation everything else gets tossed out. But see for example how WP:RS treats predatory journals: The lack of reliable peer review implies that articles in such journals should be treated similarly to self-published sources. It is saying that when you have no faith in the publisher, treat it as though it is self-published and evaluate the author. nableezy - 03:13, 5 January 2022 (UTC)

I see it as saying that when the publisher doesn't do anything except publish (when the publisher has minimal direct control or indirect influence over the content) we can treat them as a self-published source. This typically applies to predatory journals, but not to RT.

I would also note in practical terms that I don't believe a proposal along the lines of what you are advocating has a snowballs chance of passing; even the more restricted version that I am advocating has opposition on the grounds that it is not restricted enough. BilledMammal (talk) 03:26, 5 January 2022 (UTC)

Take that a little further, how would we determine the amount and nature of this influence? Selfstudier (talk) 11:01, 4 January 2022 (UTC)

RT has enough influence that they are not a SPS; Medium is the opposite. Drawing a line between the two is difficult, but we will need to do that, and I believe the line above is a reasonable way to do it. The nature of the influence is simply our current assessment of the sources reliability. BilledMammal (talk) 11:59, 4 January 2022 (UTC)

Oppose - Self-published means self published. A work is self published if the author is the publisher, anything else doesn't make sense. It's not up to us to judge how "minimal" or not the editorial process is. If someone publishes an article, a chapter or an essay in a medium that s more than a blog, then it's no longer self-published, even more so if the publisher chooses which works to publish or not. --Mvbaron (talk) 13:10, 4 January 2022 (UTC)

I thought someone would say that :) Which takes us right back around to the CP rfc, "treat as" SPS which can only mean that it is not SPS but we still want to treat it as if it were in order to keep access to expert opinions. Of course, if the principle were accepted for CP, why would it not be for all others. Selfstudier (talk) 13:26, 4 January 2022 (UTC)

My problem with that definition is when I imagine a hypothetical publisher that selectively publishes unmodified submissions - I can't see how this isn't usable as a self-published source. BilledMammal (talk) 14:28, 4 January 2022 (UTC)

You don't have to imagine a hypothetical publisher that selectively publishes unmodified submissions. This really does happen. The New Yorker was famous for it. You will find that magazine listed at RSP as reliable, with a note about its "robust fact-checking process". Many old-school magazines accept unsolicited submissions, pick the best of the bunch, and run them with no more than copyediting. They don't need to do more, because they are already looking at finished work, and there is so much good content available that if yours isn't good enough, they can pick something else. The publisher's role in this model isn't to influence what you write. The publisher's role is to pick and choose the things that they want to present to the world. WhatamIdoing (talk) 06:04, 8 January 2022 (UTC)

That's a good example, but I can't think of how to address it in our policies - and we need to. BilledMammal (talk) 03:46, 12 January 2022 (UTC)

Question I think we all agree that if a person publishes a thing on their personal website that is self published. What if Honda publishes a claim on their website? I'm going to assume that Honda, like many large companies, has some sort of PR group that reviews content before publishing. Is that self published? A person within Honda writes a story about the new Honda Civic. It's almost certainly going to be reviewed before publication so does that make Honda a RS publisher? How should we handle publications that, while certainly edited by someone, are not independently edited or have a clear vested interest in the content of the material they publish? Note, this question is not specifically related to Deprecation but since the question is being asked. Springee (talk) 14:09, 4 January 2022 (UTC)

Supposedly that falls under questionable source (per WhatamIdoing, promotional). Then gunrel and dep are carve outs from questionable with their own specific characteristics, although I remain unconvinced of an essential difference between these two. I could understand "fabrication" as a differentiator but the list is presently extended to include all sorts of other factors besides. Selfstudier (talk) 14:41, 4 January 2022 (UTC)

I think this question highlights that material that originates from an organization may not have a single author, and doesn't operate in the same manner as traditional publishers. I suppose the question comes down to can an organization be an "author"? And if so, wouldn't that make material authored by an organization and published by that same organization be self-published? --Kyohyi (talk) 17:04, 5 January 2022 (UTC)

An organization can be an author (both in law, and in practice), and whatever a corporation writes for itself and makes available to the world itself is self-published. Self-published sources can be reliable. Honda is assumed to be a reliable source for what Honda says about itself (including about its employees, who are not considered "third parties" within the meaning of BLPSPS). WhatamIdoing (talk) 06:06, 8 January 2022 (UTC)

General discussion

History

• First use: The initial remarks of this page say Daily Mail was "the first source to be 'deprecated'". In fact in Daily Mail RfC (closed on 2017-02-08) the word "deprecate" only appears in a suggestion that some anti-Daily-Mail !votes should be deprecated. On 2017-11-17 the "Deprecated publisher" template was created. On 2018-02-08 an RfC re Fox News mentioned deprecating, but it wasn't. On 2018-08-30 InfoWars RfC was closed with no mention of "deprecate". On 2018-09-25 Breitbart RfC was closed, which appears to be the first time the word is used by an RfC closer, so it would be correct to say Breitbart was "the first source to be 'deprecated'". On 2018-12-17 the WP:DEPS essay was created which incorrectly said Daily Mail was "the first source to be deprecated" (later the word "formally" was added but that doesn't save it). On 2020-01-09 Omer Benjakob in Haaretz said Daily Mail was the first, which unfortunately is the source for the cite in the Daily Mail article. However, the WP:DEPS statement that the Daily Mail RfC was a landmark decision is not obviously false. Peter Gulutzan (talk) 19:49, 27 December 2021 (UTC)

  Wasn't Daily Mail still the first to be 'deprecated' (as in, given the treatment we now describe as 'deprecation', even though this was not the term used at the time)? It seems in the Breitbart RfC, Guy/JzG just gave what happened to Daily Mail a name ("deprecate") and it stuck, but that RfC didn't define the concept AFAICS; WP:DAILYMAIL1 and WP:DAILYMAIL2 did. ProcrastinatingReader (talk) 23:29, 29 December 2021 (UTC)

  It's justifiable sometimes to apply a term that wasn't used at the time, like a Proleptic Gregorian calendar, provided we accept first that it means exactly the same thing that was meant at the time. But isn't that assuming, in advance, that the definition is strictly what WP:DAILYMAIL1 closers said, so we're done here? Peter Gulutzan (talk) 00:46, 30 December 2021 (UTC)

  It is by definition the same thing, isn’t it? The opening statement wrote: Should Breitbart be deprecated as a source in the same was as WP:DAILYMAIL which defines the meaning of what he meant. That RfC might’ve coined the word, but it didn’t define any of the concept. ProcrastinatingReader (talk) 02:57, 30 December 2021 (UTC)

  Actually "deprecate" was coined long before Guy|JzG, but even A new meaning of 'deprecate' doesn't resemble what WP:DAILYMAIL1 closers said, so it's not what anyone could have meant then using a dictionary. Instead it's been called a "term of art". Sunrise called it that in June 2019 but I might have missed an earlier use. Peter Gulutzan (talk) 15:32, 30 December 2021 (UTC) Additional note: at the time of that Guy|JzG quote, which was before the decision to redirect, WP:DAILYMAIL pointed to the Daily Mail RfC. Peter Gulutzan (talk) 17:42, 30 December 2021 (UTC)

DAILYMAIL1 doesn't specifically use the term "deprecate", but it set the initial precedents for what would later come to be called deprecation. This includes the precedent that consensus found it possible for a source to be "generally prohibited". However, the precise meaning of "generally prohibited" was left open to discussion, and it was the discussions over the following years that established the consensus towards a strict interpretation. That was the same period that the concept was applied to other sources as well. I don't know if I was the first person to specifically call it a "term of art", but I think it was clear by that point that the word "deprecate" had taken on WP-specific meanings that did not fully match the dictionary definition. For instance, this is the revision of DEPS from the time that I made that comment. Sunrise (talk) 00:07, 31 December 2021 (UTC)

• Exception for old articles: The Daily Mail closers said it "may have been more reliable historically" and two of them participated without objecting when an exception was accepted for a 1940 Daily Mail use. A closer of a June 2020 RfC accepted the claim that "Daily Mail published a faked version of its own front page" but only changed WP:DEPSWP:RSP. Peter Gulutzan (talk) 18:27, 29 December 2021 (UTC)
  • It also changed WP:RSP, and was in fact to do so - David Gerard (talk) 22:52, 29 December 2021 (UTC)
    • Indeed it was WP:RSP, I made a correction.Peter Gulutzan (talk) 23:27, 29 December 2021 (UTC)
• Edit filter for all: The Daily Mail closers suggested an edit filter to warn about use "as a reference". One of them said in a follow-up thread "Note that in the close of the Daily Mail RFC, we intended the edit filter to warn editors, but not disallow edits. This was a very intentional decision, and I hope the wording reflects that clearly." However, in April 2019 discussion and an RfC determined adjusting the edit filter wording to reflect results of each RfC separately would be difficult so Edit Filter 869 can't mention all possible exceptions or extra restrictions. The current edit filter text is here, most was written by Guy|JzG in August 2019 and May 2020. National Enquirer was excluded. Peter Gulutzan (talk) 21:15, 30 December 2021 (UTC)
• Exception for opinions and/or aboutself: The February 2017 Daily Mail RfC close said "it could make sense to cite it as a primary source if it is the subject of discussion", which could be interpreted as: WP:ABOUTSELF applies. The idea that WP:ABOUTSELF is the only important exception -- so not opinions in general -- seems to have first appeared on WP:RSN in November 2018 in an RfC re World Net Daily, the idea was rejected with the closer saying that opinions in general are allowed but this might have been superseded when World Net Daily was deprecated. In July 2019 the only-if-aboutself idea came up again and was rejected by a Daily Mail closer saying (as other Daily Mail closers had already said) that non-about-self opinions are acceptable too though one might wonder why. In December 2021 the only-if-aboutself idea came up again and was rejected by the Breitbart closer, who said that the closing remarks themselves were intended to be clear that opinions in general are allowed. For some other sites (Occupy Democrats, Crunchbase, Epoch Times) the closer remarks suggest some opinion would be allowed, and about 18 say something like "as in the 2017 RfC of the Daily Mail" without being explicit about that corollary. However, for at least four sites (VDARE, Taki's, FrontPage, bestgore), and some fake news sites, closer remarks seem stricter. Peter Gulutzan (talk) 21:59, 31 December 2021 (UTC)

  That's interesting. It's not as clear cut as it might be. Would you say that there is a trend to a stricter interpretation over time? Selfstudier (talk) 09:46, 1 January 2022 (UTC)

  I can't say that because the sample size is small and might be skewed (I don't know how many RfCs failed). Peter Gulutzan (talk) 18:31, 1 January 2022 (UTC)

  I believe that @Iridescent has a story to tell about citing the Daily Mail as a primary source about itself, and having someone declare that the February 2017 Daily Mail RfC prohibited editors from citing it.

  Self, the general rule is that all rules get applied mindlessly. WP:Policy writing is hard. You can write a policy to defend the goal against wikilawyers, POV pushers, people with an axe to grind, etc. (e.g., by deploying a lot of WP:BRADSPEAK), but what really helps is if editors routinely resist overly aggressive interpretation of "the rules" in their everyday editing. WhatamIdoing (talk) 06:31, 2 January 2022 (UTC)

  It was here; even though it was sourced to an impeccable academic source specifically discussing the Mail story, someone tried to claim that the RFC meant we could neither mention the existence of the Mail story, nor link to it to allow readers to see for themselves what the story in question had been. ‑ Iridescent 08:51, 2 January 2022 (UTC)

  Erm, that’s correct. The removed text does not fall under ABOUTSELF… Mvbaron (talk) 09:15, 2 January 2022 (UTC)

  Earlier Iridescent refuted a different someone, me. Peter Gulutzan (talk) 15:17, 2 January 2022 (UTC)

  ) Selfstudier (talk) 15:23, 2 January 2022 (UTC)

  +1 that’s a very good analysis! It seems that from the original daily mail closers at least 3/5 allowed for opinion pieces, and subsequent closers sometimes allow it and sometimes are stricter. What this means, I guess, is that there's a default assumption that deprecation allows for more than ABOUTSELF, but that default can be overridden by consensus. I wonder what the point of deprecation is then - it doesn't seem to entail a stronger prohibition than that for generally unreliable sources. The straightforward policy-RFC question would be "Are opinion pieces/pieces by established subject matter experts in deprecated sources allowed"? Mvbaron (talk) 10:22, 1 January 2022 (UTC)

  That would be the generalization of the Counterpoint RFC running at the moment. Idk whether it is better to run the general question or press on with the attempt to make this guideline which covers a bit more than just the SPS question as well as their being no consensus in here on that as yet.Selfstudier (talk) 12:55, 1 January 2022 (UTC)

  not sure it is, the CP RFC is (a) about counterpunch and (b) at least ambiguously phrased and about self-publishing. In my opinion, we definitely need the general RFC. And I thought that we need multiple RFCs on the guideline here anyway. Mvbaron (talk) 13:09, 1 January 2022 (UTC)

  If there is enough consensus to draw up a draft guideline, then all the points can be addressed in one RFC on the guideline itself. There does seem to be a consensus that we need a guideline but not what it should say.Selfstudier (talk) 13:27, 1 January 2022 (UTC)

  One massive RFC may not be the most effective way to find agreement. WhatamIdoing (talk) 06:32, 2 January 2022 (UTC)

  That's perhaps true, it's also true that we are not in agreement here either just in terms of even drafting such an RFC so the alternative? Given that there are currently running open RFCs addressing lesser matters equally without agreement (not to mention the ANI discussion).Selfstudier (talk) 10:25, 2 January 2022 (UTC)

  Just as a practical example of whether RSOPINION should allow for deprecated sources: in the UK, many of the top recognized film/television critics are those that work for the tabloids of the UK, including the DM, Mirror, etc. I've postulated that while we'd not use these critics for films produced outside the UK or that generally have an international showing, it would be odd not to use them for their RSOPINION on Brit-made works, as long as consensus agrees that individual critics here are recognized as those experts. I havent followed up on this much as most of these DM reviews were used on TV episodes and wthin the TV wikiproject, they are backing off on articles where there are "routine" reviews and little else to go beyond that as requiring standalone, so there is less a need to get these as RSOPINION, but I am certain there are other cases out there that involve deprecated sources and authors of opinion pieces that are considered appropriate experts to include without being UNUDE on a topic (eg this seems to be an issue with Counterpunch). --Masem (t) 16:16, 6 January 2022 (UTC)

I think that is the core issue of recent worries. Of 70,000 articles, we are talking about the retention of 200 or so written by top scholars in their fields, who unfortunately do not perceive that venue in the way a majority of the 30 odd editors who voted for option 4 did. I can handle deprecation, even though I think that the deprecation process was wildly subjective in skewing that webzine as toxically anti-Semitic, conspiracy-mongering, anti-Israeli, genocide-promoting, and Holocaust-denying on the basis of a handful of mostly unexamined diffs. What worries me is automatism in any form of decision-making - 'it is written ergo . .' is not what modernity is about. That we are questioning the use of Norman Finkelstein or Michael Neumann, who are both sons of Jews who either survived the Holocaust or were tipped off to flee before they were nabbed, as source for its greatest historian, whom they knew, i.e. Raul Hilberg, is witheringly uncommonsensical, and extremely doctrinaire in its insistence that there can be no exceptions. Fear, taboos, rote judgment from precedent should never trump the exercise of careful context-embedded analysis in writing encyclopedic material. To the contrary, across-the-board application of an 'exclude-on-sight-rule' is dangerous bait for our laziness. Your point about British film critics is also well-taken. Deprecation is fine, as long as it is clear that we don't write off eminent journalists and scholars from having their opinions known on this encyclopedia simply because they don't share our rather random decisions about what is reliable or not.Nishidani (talk) 18:27, 6 January 2022 (UTC)

+1 about keeping scholars who write for the public. That's what scholars should do.

I am not a fan of the idea that if you work for the Daily Mail, then everything you write, even for more reputable publishers, should be banned. I think this is an especially silly idea for pop culture reviews. We don't expect any fact-checking in a film review (or any other opinion piece). A fact-checker just can't do anything with a sentence like "This film has one of the most thrilling car chase scenes since The French Connection in 1971 – five stars!" WhatamIdoing (talk) 06:17, 8 January 2022 (UTC)

• WP:DEPS says that deprecation is just 'general unreliability + edit filter + autorevert', which doesn't seem correct. Consider the entry for National Enquirer at WP:RSP: In the 2019 RfC, there was weak consensus to deprecate the National Enquirer as a source, but no consensus to create an edit filter to warn editors against using the publication. It seems current 'deprecation' is just 'super unreliability' (in intent), although in practice the quality of discussion varies so much, particularly in fact-finding, that I don't think the discussions even conclusively establish 'super unreliability'. ProcrastinatingReader (talk) 03:33, 5 January 2022 (UTC)

I'm not even sure how we distinguish between "unreliable" and "super unreliable"; I think I've said this elsewhere, but I worry that it only serves to make users lazy and results in those who believe it can be used for certain topics !voting "unreliable", rather than specifying what those topics are so that they can be discussed and recorded as exemptions. BilledMammal (talk) 03:37, 5 January 2022 (UTC)

Unstructured discussion and unintended proposals

Removed as too long and not helpful BilledMammal (talk) 03:28, 12 January 2022 (UTC) BilledMammal (talk) 15:40, 2 January 2022 (UTC)

What is this though? A continuation of this RFC? An RFCbefore a new one? A basis for a proposal? What do you expect editors to do as regards the above, that's what I am asking? Selfstudier (talk) 15:58, 2 January 2022 (UTC)

If I understood your comments elsewhere correctly, this discussion is an attempt to generate a draft proposal for an RFC, and I suppose it is an attempt to move forwards with that by providing specifics as a starting point - I suppose what I expect editors to do is see if the method of considering the issues that RSP is intended to address helps them understand what they believe needs to be done, and from there possibly engage with the ideas that I found to be a natural consequence of considering the issue in that light. If you believe it is premature or otherwise inappropriate, feel free to collapse. BilledMammal (talk) 16:17, 2 January 2022 (UTC)

Idk whether the proposal is good or not, I read it but haven't considered it in detail. It's a lot to take on board in one go. Does it all have to be considered at once or can it be broken into pieces. I ask because you can see from the other discussion that there has not been agreement even on matters that can be straightforwardly expressed in a sentence or two. For instance, your definition of a self published source is one such, why not see if you can get some agreement on that part? Or the part about deprecation being time limited. Or some other part.Selfstudier (talk) 16:29, 2 January 2022 (UTC)

While each aspect is intended to complement other parts of the change, most of the proposals can function independently. I'll look at breaking them out, starting with the SPS one. BilledMammal (talk) 17:08, 2 January 2022 (UTC)

Here is a concrete current question Diff. So CP is deprecated and it has to go, right? Selfstudier (talk) 16:35, 2 January 2022 (UTC)

Under what is proposed above, yes. It will not immediately alter the status of CounterPunch or any other source, as I believe any such alteration would derail the discussion - because of this I am also uncertain about the inclusion of the expiry date, as while the functional change to the status will be minimal it may appear to be significant and that alone may be enough to cause issues.

What it will do is open a door for CounterPunch to be considered a SPS, if it can be demonstrated that the influence of CounterPunch on the content of the works published by them is minimal. BilledMammal (talk) 17:08, 2 January 2022 (UTC)

And it's gone Diff Selfstudier (talk) 17:47, 2 January 2022 (UTC) Back again:) Selfstudier (talk) 18:46, 2 January 2022 (UTC)

Ah, I made the same mistake overlooking the "Beyond Chutzpah" source; with that it is appropriate to keep, both under the above proposal and the current situation, as the only statement sourced to CounterPunch is an WP:ABOUTSELF. What I wrote above remains true in general though. BilledMammal (talk) 00:52, 3 January 2022 (UTC)

Reopening the closed

The discussion that prompted the creation of this discussion, mentioned at the top here, Wikipedia:Administrators' noticeboard/Incidents#Can we decide what the heck "deprecation" means, or alternately, use a different word?, was closed on 26 December last by Ymblanter with comment

This might be a useful discussion, but it should occur either at Wikipedia talk:Reliable sources or at Wikipedia:Village pump (policy). I am closing it here and suggest that whoever wants to proceed to move it to a more appropriate venue.

Here it is, straight copy paste, I propose we continue this discussion (and the "voting" as well).Selfstudier (talk) 15:50, 3 January 2022 (UTC)

Should anyone want to reply to anything in here, then make sure to ping whoever made the comment you are replying to (outside of the archive).Selfstudier (talk) 22:54, 3 January 2022 (UTC)

@Selfstudier: I think the close was trying to say the point of the discussion was useful to continue, not the specific text itself. The ideas were incorporated into the questions above (e.g. JPxG's criticism of the terminology "deprecation"). The discussion below is too large and unfocused, however. I suggest deleting this section; the discussion is already getting a bit unfocused and cluttered, and focus is necessary to get to any kind of outcome and not just end up going round in circles chasing our tails. ProcrastinatingReader (talk) 17:29, 3 January 2022 (UTC)

That's why I bought it over, we are already chasing our tails, at least there was some signs of agreement in this discussion, especially early on, more than there is presently in the other. The closer thought it was potentially useful, let's see if anyone wants to continue, if not then we can close it, with the other in all likelihood following on behind I'm sad to say. Selfstudier (talk) 17:41, 3 January 2022 (UTC)

There is one thing below that caught my eye reading the below through again, the idea "....may be as simple as formally adopting Wikipedia:Deprecated sources". Maybe that's the way to go to get a more focused discussion.Selfstudier (talk) 17:46, 3 January 2022 (UTC)

Wikipedia:Deprecated sources is largely procedural (and literally an 'information page') and doesn't provide guidance on any of the large issues IMO (some of the outstanding questions above). It's not a solution to these problems. ProcrastinatingReader (talk) 18:02, 3 January 2022 (UTC)

I know it isn't a solution, if it was we wouldn't need to do any of this, would we? I am trying to address your concern over focus. Perhaps think of it as a clotheshorse on which we might proceed to hang some clothes. I know the ..writing policy is a hard thing.. but I wouldn't even dream of trying to write a guideline following the above discussion, it would be a complete waste of time.Selfstudier (talk) 18:10, 3 January 2022 (UTC)

The discussion below is a part of the ANI discussion Wikipedia:Administrators'_noticeboard/Incidents#User:David Gerard violating Wikipedia:Deprecated sources#Acceptable uses of deprecated sources closed on 4 January 2021 referred to at top.

Wikipedia:Administrators' noticeboard/Incidents#Can we decide what the heck "deprecation" means, or alternately, use a different word?

Start Copy I have noticed a bunch of people here seem to have completely different ideas of what it means for a source to be deprecated. The word itself, in a literal sense, means "to ward off by prayer" (deprecari). In colloquial usage, it can mean anything from "disliked" to "strongly and officially advised against". In programming, a "deprecated" feature or method generally means one that you're advised not to use when writing new code. Sometimes this is because a better or more secure feature has been introduced, sometimes this is because supporting the deprecated feature is an inconvenient timesink, and sometimes this is because a haphazard system is being streamlined into something simpler. In some of these cases, it makes sense to go through old code and rip out every instance of the function (say you're upgrading a system from Python 2.6 to 3.5 and a bunch of the old shit will literally stop working). In other cases, the situation is more lenient (legacy code will continue to run fine but it's a good idea to use the better thing if you are writing new stuff). At any rate, the fact that something's "deprecated" doesn't make any definitive case for what action you should take regarding it. People saying that it "literally" means one thing or the other are... well, it literally means to avert disaster by appealing to the gods, so I don't think we are talking about literal definitions here. I think that this discussion (which we've had several iterations of by now) would go a lot better if we came up with some clarifying language for what it meant, or perhaps used a different word, like "blacklisted", "forbidden", "censored"... or, alternatively, "non-recommended", "superseded", "obsolete" or "not very good". jp×g 22:38, 23 December 2021 (UTC) Support The above is a policy disagreement and community should resolve policy by discussing policy, not the behavior of two editors with entrenched positions.Slywriter (talk) 22:45, 23 December 2021 (UTC) Relevant discussions: Wikipedia talk:Reliable sources § Deprecated sources section of WP:RS; Wikipedia:Village pump (proposals)/Archive 164 § On the use of deprecated sources JBchrch talk 22:47, 23 December 2021 (UTC) Note: I've undone @ScottishFinnishRadish:'s NAC on this section because it has only been a couple of days and there've been a variety of rumblings about different actions. I'm not confident that any consensus will emergy, but it seems too soon for a definitive close. jp×g 23:01, 23 December 2021 (UTC) Frankly, I don't think that swapping out the terminology will stop people from getting upset that their favorite sources are being removed, which is what most of these arguments typically arise from. We can bikeshed the jargon all we want, but the underlying psychology will remain the same. XOR'easter (talk) 23:04, 23 December 2021 (UTC) I am not sure that this is an accurate summary of what's going on. It seems to me like, in this discussion as well as the previous ones linked, there is a concrete disagreement about what "deprecation" actually means. That is to say, we are uncertain of what actions editors actually have consensus to carry out based on an RfC closing as "deprecate". If everyone who commented on the RfC supported "discourage its use and remove it if bettter sources exist", and people actually editing the encyclopedia are interpreting it as "remove at all costs wherever found", there is a problem, and the actions are not supported by consensus. Conversely, if RfC commenters agreed on "go through Wikipedia with a chainsaw and rip this website out of every page you find it on", and editors are interpreting this as "we ought to reduce the use of this source somewhat", this is also a problem. As for "favorite" sources, I can't speak for anyone else, but I don't know how relevant it is. My personal opinion of Vice, for example, is that it has gone utterly to the dogs in the last few years, but I'd still object to someone removing it from hundreds of articles if I didn't think there was consensus for its removal. I think the psychology here is more that people disagree on an issue of fact, and some people think it is one way, whereas some people think it is another way. jp×g 23:36, 23 December 2021 (UTC) I can only call it as I see it, and that's my take-away from many, many arguments on Talk pages, at RSN, etc. XOR'easter (talk) 23:48, 23 December 2021 (UTC) I've explained elsewhere that given WP has computer-savvy user bent, that the choice of using "deprecation" for sources reviewed in an RFC (like Daily Mail) may have been a poor choice due to the fact that in comp sci, deprecation is more a warning that such material will no longer be supported and should be removed in time. Indeed, the reading of WP:DEPS supports this concept and the issue with these removals is that they violate that principle, treating the sources as blacklisted and thus can be removed without worrying about the mess left behind. That said, I am all for a discussion to be clear if we can support "deprecated" as a lower rating of a source below "generally unreliable" but not as low as "blacklisted" and if we need another level for sources like The Daily Mail, Breitbart, or RT that are to be avoided outside ABOUTSELF circumstances. Once that's clarified, the past RNS RFCS on specific sources should be reviewed just to know how to classify them. --Masem (t) 00:00, 24 December 2021 (UTC) From reading the archived discussion on this topic linked by JBchrch, it seems like there's never been a clear consensus that deprecated sources should be a priori treated differently from other sources in an article. signed, Rosguill talk 23:24, 23 December 2021 (UTC) Indeed. And as such, this ANI will end like all the others; inconclusively. I don't really think David Gerard's removals are in line with deprecation as defined by the DM1 RfC. They're indiscriminate. But DG has been quite clear that he won't stop, and there's enough policy ambiguity for the community not to step in on DG individually. I think the way forward is to construct an actual guideline for deprecated sources. I would do it either through drafting phases to construct the proposed guideline page, and then RfC, or as a two-part RfC; the first, to create a Wikipedia:Deprecated sources guideline on the relatively uncontentious parts and then a second part to deal with more contentious parts. I think the key issues that need to be addressed are: the guidance on removing sources after they've been deprecated, and what it takes for a source to be deprecated (as distinct to the spam blacklist or just being generally unreliable), specifically what kinds of evidence. Finally, some formal clarification on accepted usages of deprecated sources. I think this is urgently needed, because deprecation is inconsistent. While we're at it, a more accurate term like Wikipedia:Discouraged sources might be better than software lingo like 'deprecated'. ProcrastinatingReader (talk) 23:31, 23 December 2021 (UTC) The fact, though, is that there is currently no protection for unreliable sources - for the most part, it is completely normal and acceptable to go through and remove unreliable sources just like depreciated ones. There's perhaps a somewhat higher expectation that you'll be cautious, search for reasonable replacement sources, avoid removing text for which it's reasonably likely an acceptable source could be found, etc., but part of the reason depreciation was created was because the Daily Mail, as an unreliable source, was supposed to be getting phased out after the numerous discussions agreed on that point; and that was happening far too slowly (in fact, its numbers kept increasing) because people kept adding new citations to it. Every proposal I've seen to slow the removal of depreciated sources has seemed ass-backwards to me because it would set the bar for removing depreciated sources higher than the (currently nonexistent) bar for removing unreliable sources. --Aquillion (talk) 08:23, 24 December 2021 (UTC) Yes, this, exactly. Building up an extra barrier around deprecated sources, like by requiring an intermediate step with a {{better citation needed}} tag and a waiting period, just makes it harder to remove deliberate disinformation and state propaganda from our encyclopedia. XOR'easter (talk) 20:46, 24 December 2021 (UTC) That isnt what is happening here. This is removing an exceptional source (see Sara Roy). This along with a whole bunch of other careless mistakes, is removing an exceptional source (see David Price (anthropologist)). Stop pretending that what is happening here is the removal of deliberate disinformation and state propaganda, it is a fabrication. nableezy - 20:52, 24 December 2021 (UTC) Deprecated sources do indeed include deliberate disinformation and state propaganda. Putting extra regulations on how deprecated sources should be removed is protecting exactly that kind of material. XOR'easter (talk) 04:04, 25 December 2021 (UTC) The sources being removed as deprecated that are under discussion here do not include deliberate disinformation and state propaganda. They include actual scholars writing in the area of their academic expertise. It is entirely fabricated that anybody is arguing for protection of deliberate disinformation and state propaganda. If you are going to apply this label to a whole host of things you cant just justify your actions based on the most extreme part of a wide ranging set. nableezy - 04:50, 25 December 2021 (UTC) No, people are talking about DG's entire project of removing deprecated sources. And, indeed, they are coming to judgments based on the most extreme part of a wide ranging set, i.e., a few examples out of thousands where he maybe, maybe, messed up. XOR'easter (talk) 05:11, 25 December 2021 (UTC) As I've said repeatedly, if you object to that you ought to be challenging the depreciation of the source directly (ie. start a new RFC.) Yes, there are occasional exceptions, but exceptions are exceptions because they're, well, exceptional, rather than being the sort of thing the source publishes regularly, and as far as I can tell you've argued repeatedly (and the crux of your objection here) that Counterpunch generally publishes stuff that is reliable because of who the author is. If there are generally applicable exceptions that allow a source to be used, then the source can't even be called generally unreliable and isn't suitable for depreciation, so that's an argument you ought to make in an RFC about the reliability of the source. But right now, if there were a consensus that the high quality of authors there kept it from being generally unreliable, then it would be yellow at RSP and not red. --Aquillion (talk) 08:19, 25 December 2021 (UTC) Support and establish a best-practices standard operating procedure for deprecation. Perhaps we need a separate label and category of sources that are actually prohibited and should be mass-removed. BD2412 T 23:38, 23 December 2021 (UTC) Support The term deprecation is a computing/IT term, for something that was in use and is no longer, meaning there was a period when it was functional, correct and useful but it is no longer, so it out of use. It is really the wrong term, and shouldn't be changed. We need something much much more accurate and instantly recognisable. They have always been junk sources. For example, they're has never a time when the Daily Mail wasn't junk, except perhaps during WW2. Infowars is slightly different, its almost disinformation and was never anything else. So the new term needs some flexibilty and be recognisable. All the disinformation that is on the go, since we are the truth. I hope that helps. Its late. scope_creepTalk 02:03, 24 December 2021 (UTC) That is not quite correct. In computing, a deprecated feature is one that should be avoided for new projects but which still works and is still supported although it probably will be removed in a year or two. It is not necessarily "out of use". Johnuniq (talk) 02:48, 24 December 2021 (UTC) Support This may be as simple as formally adopting Wikipedia:Deprecated sources, although the Acceptable Uses section may need a bit of a cleanup as it contains patently absurd statements such as "editors are also expected to use common sense and act to improve the encyclopedia." Deprecation is something that we just sort of started doing after the Daily Mail RfC; the definition varies depending on how each RfC was closed and which little caveats the closer decided to include. I think that in general the intent of each close is the same, for example WP:ABOUTSELF is applied fairly consistent across the board even though some closers mentioned it and others didn't. A set definition would help clear up any confusion and wikilawyering. –dlthewave ☎ 05:30, 24 December 2021 (UTC) Comment I don't much like the concept of deprecation at all. A lot of the problem comes from the deeper issue of what a "source" is. When a magazine publishes an article, is the magazine the source or is the article the source? Unfortunately, the official answer is "yes" and this enables people who don't want an article to be cited to attack a weaker point instead. Rather then arguing directly that an article is unreliable, they attack the publisher on the grounds that it published other articles which everyone agrees are unreliable and nobody would consider citing. "The magazine published crap article A, so we will deprecate the magazine and now you can't cite article B even though it is authored by a highly respected expert in the field." Since nobody would even dream of citing article A, the motivation must be to eliminate article B. So deprecation becomes a convenient tool. (I believe that is an accurate description of the current case.) The community is perfectly capable of deciding "article A is crap and we won't cite it, while article B was written by a subject expert and is citable". Zerotalk 07:20, 24 December 2021 (UTC) The contrary argument would be that the fact that crap article A was published at all means that the source has such low standards for fact checking that we can't trust anything they publish. Even experts need to be peer-reviewed. Mlb96 (talk) 17:15, 24 December 2021 (UTC) That makes no sense, if the expert had published it on toilet paper we could use it but because he has published it in CP we can't? CP don't fact check anyway.Selfstudier (talk) 18:05, 24 December 2021 (UTC) Oppose, at least without a more specific proposal. The original purpose of depreciation, as I recall it, was that the Daily Mail, despite repeated and clear consensuses that it was unreliable, continued to be used across much of the encyclopedia, in part because of new people adding citations to it. The current terminology and implementation of depreciation has largely resolved that problem; while there may be some individual sources whose categorization or usages are worth quibbling over, overall, depreciation is working. Wikipedia's sourcing since we depreciated the Daily Mail has generally improved sharply in quality, and in fact we've gotten significant coverage from outside sources as being one of the few places that managed to find a way to deal with the era of "fake news", despite being a user-generated encyclopedia. There might be room for a few refinements or clarifications around the edges, but I'm completely opposed to anything that would substantially change the terminology or the way we handle them, per WP:DONTFIXIT. --Aquillion (talk) 08:23, 24 December 2021 (UTC) Neutral - WP:DEPREC lists the differences pretty well: use of the source is generally prohibited. But I don't really understand what the point of deprecation is if we can go on and use a deprecated publisher for more than ABOUTSELF (contra DEPREC). If we allow deprecated to be used for more than ABOUTSELF, then there is no difference to generally unreliable sources. Per WP:NOTRELIABLE: Questionable sources should be used only as sources for material on themselves, i.e. generally unreliable sources already have the highest bar for acceptance (but here we may be more lenient with e.g. texts by experts). If - contra WP:DEPREC - deprecation is allowed for more than ABOUTSELF, then we might as well get rid of it altogether. The clarification RFC needs to be about what the difference between generally unreliabe and deprecated is - if we don't want to follow WP:DEPREC --Mvbaron (talk) 08:29, 24 December 2021 (UTC) Oppose making any decision here; it would probably be better to use different language altogether that cannot be misinterpreted, either by good or bad-faith actors, however discussion on the way forward should be at an appropriate forum, not ANI. Accordingly, I suggest closing this sub-thread. wjematherplease leave a message... 10:12, 24 December 2021 (UTC) You know, that looks like what some of the "support people are saying. We need to make a decision, but not here. Doug Weller talk 10:36, 24 December 2021 (UTC) How about creating an RfC subpage (Wikipedia:Requests for comment/Deprecated sources)? Clearly some brainstorming is needed and there are various questions and concerns, so maybe that’s a good place to start ironing out details before a guideline proposal is put forward? ProcrastinatingReader (talk) 12:06, 24 December 2021 (UTC) MvBaron has a valid point I think. The RFC product should spell out the difference between generally unreliable and deprecated (including how they are dealt with) so maybe title it Deprecated and unreliable sources.Selfstudier (talk) 12:25, 24 December 2021 (UTC) @Selfstudier: I created a draft RfC before I saw this, but the "Criteria for deprecation" section was intended to address this issue (as I agree it is an unresolved problem). ProcrastinatingReader (talk) 12:44, 24 December 2021 (UTC) Yeah, to be clear, what I propose is that an RfC be opened somewhere else, perhaps a dedicated subpage (and advertised on WP:CENT) -- not that we try to draft a new guideline on the fly in the middle of an AN/I thread, which would be a grotesque shitshow. It's probably worth noting that in this very subsection about the ambiguity of "deprecation", there are eight instances where someone said "depreciation" instead... jp×g 13:36, 24 December 2021 (UTC) I think it is, I added the draft to cent just now. maybe some notifys on pump, V etc?Selfstudier (talk) 13:43, 24 December 2021 (UTC) I removed it for a bit. I want to get some thoughts first to make sure it's the right structure and we're asking the right questions before it goes live, since it's poorer form to modify a live RfC that's in-progress. ProcrastinatingReader (talk) 13:49, 24 December 2021 (UTC) You marked it DRAFT? I'll go remove it again :) Selfstudier (talk) 13:56, 24 December 2021 (UTC) Oh, you did already, lol. I thought it was fine as was but OK, if you want a pre pre RFC, we can do that:)Selfstudier (talk) 13:59, 24 December 2021 (UTC) Oppose per the WP:DONTFIXIT argument of Aquillion. What we're seeing are the inevitable edge cases of (a) a very small set of cases where reasonable people can disagree on the best course of action, and (b) wall-of-texting in favor of sources that some people really, really don't want removed. These are not sufficient reason to tamper with mechanisms that help keep Wikipedia a significantly less awful place than most of the Internet. XOR'easter (talk) 20:44, 24 December 2021 (UTC) Ultimately the issue is a mixture of policy, and whenever people attempt to clarify what the consensus on the policy is, certain editors derail the discussion with bludgeoning/battlegrounding/etc (see Wikipedia_talk:Deprecated_sources#Proposed_clarification_of_deprecated_sources_guidelines, for instance). Heck, the last time we were here (Wikipedia:Administrators'_noticeboard/IncidentArchive1066#Indiscriminate_removal_of_deprecated_sources) we had an admin closing the discussion after 4 hours (ultimately reversed). It's valid for the consensus to be "all is fine", although I doubt that's actually what the consensus is per the reasons we discussed last time this was at ANI, but we can never actually have that discussion. So no progress can be made either on the behavioural front, or the policy front. Tbh I agree with nableezy's idea of just taking it to ArbCom, since I'm at a loss for ideas at this point; the behavioural element prevents progress. ProcrastinatingReader (talk) 01:42, 25 December 2021 (UTC) Strenuously disagree that there are behavioral issues here. The issue, to me, is this. Before the depreciation RFC, IIRC, we had something like 12,000 cites to the Daily Mail, and similar numbers for many other high-profile depreciated sources. Now we have something like 14. I don't think it's reasonable to suggest that there are enough places the Daily Mail could legitimately have been cited under WP:V to get us anywhere remotely close to those 12,000 citations. That means that anyone who wants to change how we handle or enforce depreciation needs to answer two questions - first, do they agree that we ultimately needed to drastically cut the number of citations to a source like the Daily Mail, and that 12,000 citations to it was almost certainly indicative that of many violations of WP:V / WP:RS? And second, if they intend to slow down or prevent mass removals, what's their alternate route to get us to those low double-digit numbers for sources like that? Because it feels to me like people are beating around the bush of those fundamental questions; if someone thinks it would still be acceptable for us to cite the Daily Mail 12,000 times, then in my view they're fundamentally challenging either the consensus that it's generally unreliable (not just the depreciation; we should not be citing a source like that so heavily), or they are fundamentally challenging WP:V. Either way, focusing on DG is a distraction because I don't think anyone can articulate a way to get from 12,000 citations to 14 without it looking, basically, like what he's been doing; to me, depreciation was an agreement that we needed to drastically cut those 12,000 citations, and DG's actions have mostly been a good-faith implementation of that consensus. --Aquillion (talk) 03:12, 25 December 2021 (UTC) I doubt anyone is saying it was okay in the long-term for WP to be citing DM 12,000 times post-RFC. But, outside of BLP articles, there was never any deadline suggested by the RFCs nor in general WP:DEADLINE that those DM cites had to be removed post-haste. Because some had been in place for years, it would be reasonable to develop a consensus-based grandfathering practice (as done in most similar situations) to give editors the chance to remove and replace the DM cites with more reliable ones or remove material otherwise unsourcable over, say, a six month period, after which it would have been 100% fair game for David or others to strip out DM cites without impunity. (This again is commonly an approach taken with "deprecation" in computer science and other areas) That's a non-disruptive approach to deal with long-standing content, and standard practice whenever we have changed a content policy or guideline that would affect a fairly large number of articles. But the issue stems from David taking it on themselves to strip DM citations without trying to seek alternate sources or non-disruptive remedies, which is basically against WP:FAIT. Of course, I will assert too that we have a consensus-disagreement on what deprecation means and we need to resolve that first, but that David continues to remove sources in a disruptive manner is still a problem. They may be doing the right thing per WP:V and other policies, but the method of doing it is causing problems, and we have blocked editors for doing that in the past. --Masem (t) 03:25, 25 December 2021 (UTC) Long-standing content doesn't become good just because it's long-standing. Should we dally with removing a hoax if it was extant for ten years? I'm sorry, but this just sounds like imposing a rule that the worse a source is, the harder it should be to remove it. In my view, the disruption is the existence of deprecated sources in Wikipedia articles in the first place. That is what degrades the quality of the encyclopedia. The root cause of the issue isn't DG taking the mission on himself, it's that nobody had taken it up before. But whatever; this kind of bullshit is why I'm probably quitting soon. XOR'easter (talk) 04:11, 25 December 2021 (UTC) It is not that the content was good that it was long-standing, but that many of these deprecated citations have been in place for years, and we should give editors some reasonable amount of time to do whatever corrections they can with the change in standing of the source. Remember that before the DM RFCs, it should have been taken that the use of DM as a citation was done in good faith on the idea it was a reliable source; the sudden change to make DM deprecated should not be invalidating the past good-faith assumptions that editors were adding appropriate content. If there was a community-set need to have these removed as quickly as possible, that would have been a result of the RFCs, but the only situation on that is DM on BLP being an absolute no-no. Every time in the past where a content policy or guideline has changed in a manner that affects many hundreds+ articles where the change cannot be done by a bot, we have always used some type of grandfathering approach to give time to transition and avoid outright disruption. Same here: given that most of the DM cites prior to the RFC were added in good faith, we should be giving time in good faith to fix them, and, as per DEPS, not wholesale removal or disruption. The ultimate goal is to remove the DM links outside the few ABOUTSELF allowances, but we should not be massively disrupting article content created in good faith to get there, and that's the behavioral problem here, particularly as David is well aware these actions are contentious with some editors and that there's motions to resolve PAG in a way to be clear what should be done. (To wit, any DM links added after the RFCs can be presumed to be done in bad faith and can be removed on sight, but that's not what is basically being talked here). --Masem (t) 04:20, 25 December 2021 (UTC) ...and we should give editors some reasonable amount of time to do whatever corrections they can with the change in standing of the source. On average, the rate at which such corrections actually happen is never. The bad content sits until an editor's hand is forced. If there's a better way to force those hands, well, good luck finding it, because this kind of time-and-energy-wasting drama that throws a shield over bad content and provides covering fire for trolls has just about succeeded in getting me to stop caring. XOR'easter (talk) 04:30, 25 December 2021 (UTC) Yes, this. Anything cited to unreliable sources, anywhere on Wikipedia, should eventually be fixed; this isn't some new controversial statement, this is core policy. There was broad agreement that the Daily Mail was unreliable for a long time before it was depreciated, and in all that time, progress at reducing our reliance on it was nonexistent. Arguing "well, it's generally unreliable, but you can't remove it too fast" amounts to either challenging WP:V or challenging the consensus of its general unreliability. I also disagree with the premise that the removals are disruptive - as I say below, I feel a {{cn}} tag is generally preferable to a citation to an unreliable source, since it warns the reader that the text is unverified and encourages anyone reading it to either verify it or, if they decide it can't be verified, rework or remove it. It is better to fix it completely, but a cn tag is generally an improvement - arguing otherwise, again, means challenging the consensus that depreciated the source. --Aquillion (talk) 04:36, 25 December 2021 (UTC) That's why I've said that this is a limited period for grandfathering, something like six months. After six months, any links to DM or RT or whatever source would then be in the clear to be removed without having to supply an alternate source or the like (that would even mean not having to leave a cn). Such grandfathering is standard practice when a change of PAG affects long-standing content, and not considered to be disruptive nor forcing hands (as long as the grandfathering is announced at places like VPP and CENT) No one seems to be asking for never removing these citations, just that the means to remove them should be handled in a non-disruptive way. And it is important to stress that there is no deadline to fix sourcing, outside of BLP-related content. Grandfathering like this is a balance of that lack of deadline with the need to remove deprecated sources in a reasonable timely manner. --Masem (t) 16:53, 25 December 2021 (UTC) Excellent. Same principle can be applied to GU sources as well, which are essentially the same thing as deprec. Write a script that flags them all for x months auto deletion if still extant. All the expert opinions will be gone as well, though, and then perhaps people will be a little less inclined to class sources such as CP as GU in future.Selfstudier (talk) 17:03, 25 December 2021 (UTC) The reason those are called "generally unreliable" is that there are still cases they can be, and thus should require human review before removal (particularly when several GUs are only for specific topics, like Fox for politics and climate change, or Rolling Stone for politics). "Deprecated" are where ultimate we want no links at all to those sites outside ABOUTSELF or where other factors come into play, and thus there's less need for human review of each instance. --Masem (t) 17:08, 25 December 2021 (UTC) I understand that but I do not see any practical difference, that's essentially what we are asking now for deprec. Like I said, I don't object to mass removal if that's what the community really wants. Be careful what you wish for applies.Selfstudier (talk) 17:16, 25 December 2021 (UTC) There's a massive difference between generally unreliable sources and those deprecated. In general, because "generally unreliable" may be reliable in some context, non-human removal is a problem. Deprecated sources are known to only be allowed in very limited cases, and thus, ultimately, should be placed on editing blacklists (as to warn editors) and should be removed wholesale by bots - but only after giving editors a chance to reticify their use. There's no such rush to remove those considerd generally unreliable. --Masem (t) 17:30, 25 December 2021 (UTC) I don't think there is a rush to remove either and neither do you, thus grandfathering suggestion. But that same grandfathering equally allows human review of unreliable sources and so I maintain my view that there is no practical difference. We can play with x, 3 months for deprec, 6 months for gu, or whatever.Selfstudier (talk) 17:51, 25 December 2021 (UTC) Somehow we always end up back at the Daily Mail, but just to discuss your example situation: He replaced 12,000 citations to Daily Mail with {{cn}} tags, while leaving the content in, and not checking if the content was true or false (presumably, otherwise he would've either removed/replaced it or added a better citation). So, content cited to an unreliable source known for disinformation went from a tracking link of Daily Mail citations and is now lost within the millions of articles within the general "unsourced content" tracking category. The behaviour is fundamentally incomprehensible. ProcrastinatingReader (talk) 04:04, 25 December 2021 (UTC) Actually, he's cut out the content completely in, e.g., [5][6][7][8][9][10][11], etc. XOR'easter (talk) 04:17, 25 December 2021 (UTC) Its almost like those who remove deprecated sources do in fact apply editorial judgement and add {{cn}} tags for content that looks benign or more likely to be verifiable and cut out the content which appear promotional, extraordinary, etc and are less likely to be verifiable. Tayi Arajakate Talk 05:47, 25 December 2021 (UTC) This is not benign? A person is not a reliable source for who they are married to? nableezy - 06:16, 25 December 2021 (UTC) Depends on the family. :-P But, more seriously, if the only source that has paid attention to a fact is deprecated, then including it is almost always WP:UNDUE. XOR'easter (talk) 06:23, 25 December 2021 (UTC) You really think it is UNDUE weight to include in a biography of a person that he is married? Do you not see the circular logic here? And do you not agree that WP:ABOUTSELF links are definitionally reliable for information on themselves? nableezy - 06:57, 25 December 2021 (UTC) Uh, yes, if reliable sources don't discuss a person's marriage, we don't need to talk about it. Many academic biographies exist because their subjects pass WP:PROF and we can write about their work, but information about their families is basically non-existent. XOR'easter (talk) 16:18, 25 December 2021 (UTC) That's a dead link to a deprecated source on a BLP, effectively unsourced. Tayi Arajakate Talk 06:24, 25 December 2021 (UTC) Um, look at the history of the article. See the link fixed. Dead links have never ever meant unsourced. See WP:DEADREF. nableezy - 06:57, 25 December 2021 (UTC) The point was that there were too many red flags in that case. It wasn't removed after you fixed the link, so what's the issue? Its use is also dicey even in this state, it complies with WP:ABOUTSELF only if you consider Counterpunch to be equivalent to a blog source and consider the article to be entirely authored by them. This is not clear at all, the last paragraph, "Debbie Dupre Quigley is an oncology nurse. She and her husband Bill Quigley, who is a law professor at Loyola University New Orleans, spent four nights and five days in a hospital in New Orleans before they were evacuated. They can be reached at ...", reads like a statement from the website about the authors. This is also a BLP so it probably should be removed, at the least, till the new discussion at RSN on whether to treat it as an SPS is concluded. Tayi Arajakate Talk 07:28, 25 December 2021 (UTC) lol wow. Both people are co-authors of the piece. And no, a dead link is not a red flag, and no, that is not why it has not been re-removed. It was not re-removed because another editor restored it as clearly permitted by WP:ABOUTSELF. The issue is despite your contention that David is discerning and only removing content that is not mundane that he is indeed removing basic biographical facts about people entirely. nableezy - 07:33, 25 December 2021 (UTC) Yeah they are the co-authors and the ending bit reads like a blurb about the authors from the website, not unlike those present in newspaper op-eds. It's not a red flag solely because of the dead link but because it is also from a deprecated source and is in a BLP, not to mention the ending part of the same sentence is literally unsourced which doesn't bolster confidence. When one is going through a list of articles and encounters something like this, they would most likely identify it as a poorly sourced BLP and will be inclined towards removal. In the end, even basic biographical facts in BLPs need to be properly sourced, treating them as not mundane and removing the entire sentence is still reasonable. Tayi Arajakate Talk 08:53, 25 December 2021 (UTC) I feel that having a {{cn}} tag is, on the whole, generally better than having an unreliable source. It alerts the reader that the cited text is not verified, and it increases the chance that someone reading the article (who probably has at least some interest in the topic) will edit it to add a citation. A CN tag itself is, already, a warning to the reader that "this text may be false, since it lacks a citation"; without that, anyone skimming the article is unlikely to notice that the citation is to an unreliable source, and will therefore take it at face value. I don't think he just indiscriminately replaced every single citation, but the fact is that it required moving quickly and making a lot of changes, because with the sheer number of citations a source that requires the step of depreciation can accumulate, doing it slowly will (given the limited number of people actually interested in that cleanup) not get anywhere in an appreciable amount of time. I'm not seeing any of the vague alternatives people are suggesting as workable - the Daily Mail was widely-agreed to be generally unreliable for years, many people repeatedly pointed out that we were citing it too many times, and while there were token efforts to replace some of them nobody made a dent in it until it was depreciated and efforts were stepped up. Again, I feel that anyone who objects to DG's methods needs an actually practical proposal for how we can drastically trim the usage of an unreliable source in a reasonable timeframe - preferably a demonstrated one (since it's easy to say "oh, let's just go slowly and replace them bit by bit", which is the one method we know did not work for years on end.) --Aquillion (talk) 04:30, 25 December 2021 (UTC) Again, I feel that anyone who objects to DG's methods needs an actually practical proposal for how we can drastically trim the usage of an unreliable source in a reasonable timeframe Use a bot and replace all these citations with a variant of {{better source needed}}, or [unreliable source]? That way you don't lose the link, and you can still track that it's a DM cite so a volunteer who actually wants to review the cite can do so, and there's still a warning to readers (not that the purpose of cleanup tags is to be a warning, supposedly). ProcrastinatingReader (talk) 14:55, 25 December 2021 (UTC) We don't need an editor to do what a bot can do. If all we need or want is strip all the sources David Gerard and the others are surplus to requirements.Selfstudier (talk) 15:03, 25 December 2021 (UTC) Then instead of Daily Mail and Russia Today links sitting around for years, we'd have {{better source needed}} or {{deprecated inline}} tags sitting around for years, and we'd still be pointing readers to the Daily Mail and Russia Today. That doesn't really give a reader a clue what's wrong, and it doesn't really offer editors any help, either, since the links to those sources could already be found via in-text searching. And requiring such a step is instruction creep that acts to protect the worst "sources". If someone wants to make a bot that runs around tagging footnotes with {{deprecated inline}}, that's fine, I guess, but it seems to me like wallpapering over the fundamental problem. XOR'easter (talk) 16:12, 25 December 2021 (UTC) I think you misread what I said there. I have no objection to a bot tossing (indiscriminately) all the sources if that's what the community agreed to. Btw, what do you think is the difference between an unreliable source and a deprecated source? Selfstudier (talk) 16:19, 25 December 2021 (UTC) Sorry, I was replying to the comment above yours. I'm not sure what your question is getting at; I mean, I could quote Wikipedia:Reliable_sources/Perennial_sources#Legend, which seems to summarize things fairly well? XOR'easter (talk) 16:27, 25 December 2021 (UTC) generally unreliable, then follows a list basically repeating the same things y'all keep saying for deprecated. Then for deprecated it says "The source is considered "generally unreliable". I get it, a deprecated source is a generally unreliable source. The only difference in practice is that there are editors going around removing the source en masse. To repeat what I said just before, if the community agreed to that, then fine, script a bot and get on with it and do all the generally unreliable ones while you are at it because they are the same thing. Selfstudier (talk) 16:47, 25 December 2021 (UTC) For deprecated, it says The source is considered generally unreliable, and use of the source is generally prohibited (emphasis added). The difference in practice is that deprecation is a harsher judgment. Any discussion at RSN that concludes in a deprecation indicates this. XOR'easter (talk) 05:35, 26 December 2021 (UTC) Close this subthread, this is a distraction, that too with a largely superficial focus on name changing. ANI isn't the place to make policy suggestions. Tayi Arajakate Talk 22:46, 24 December 2021 (UTC) No, I think this is a useful discussion that should not be closed yet, and I largely agree with Aquillion. A "citation needed" tag is better than a link to a deprecated source for plausible assertions. Contentious assertions cited to deprecated sources should be removed entirely. Who gets to decide what is plausible and what is contentious? Individual editors acting in good faith. In the end, the core content policy of Verifiabilty reigns supreme, and we should never use deprecated sources in an attempt to verify contentious assertions. Wikipedia editors should not be expected or required to function as a "de facto" editorial board for deprecated sources, determining which of their output is reliable, and which isn't. That's a path (one of many) to madness. Cullen328 (talk) 07:46, 25 December 2021 (UTC) This is an unfocused subthread at this point, but yeah essentially agree with what you and Aquillion are saying and I have said the same in the main thread. Tayi Arajakate Talk 09:07, 25 December 2021 (UTC) I don’t know David Gerard from a bar of soap. Checking, I see he is a highly productive, indeed valuable editor, with a good many here who recognize and respect his work, on solid grounds. They take it that his recent actions must be okay. He's generally a good editor in the round. He barged into an area widely regarded as one of the most difficult thematic zones , and in a few days achieved more than what more than a hundred socks managed in decades, impoverishing numerous articles by erasing important sources by ranked authorities. His warrant for this was a finickly stringent one-sided reading (contested by many highly experienced editors here) of deprecation with regard to anything sourced to CounterPunch. Over 15 years, on a rough calculation, I’ve read 1 article a week in it and of these, perhaps I’ve cited a score or two for the I/P area. The criterion I use is status of authorship in the field, competence in the subject matter, etc. Most of the articles are of no encyclopedic value: but, as I listed, over 54 scholars and writers of recognized standing choose to write occasionally for it. So, in summing up, let me quote some remarks here. WP:V's general requirement that sources be assessed based on context. Rosquill. You cannot demand that a deprecated source be inspected in the manner of a reliable source -because it's a deprecated source. It's presumed bad. David Gerard. Rational-Wiki, which Gerard is ironically heavily invested in, ought to be more than deprecated due to being an open wiki that has tolerated two people who have been the subject of WMF trust and safety office actions on Wikipedia using R-W to post further attacks on Wikipedians including doxxing and threats of off-wiki attacks, with one of those being a former board member there. PCHS-NJROTC. "fully agrees" is doing all the work in your statement. David Gerard David in this erase CP activism is being lazy. He doesn't care to work on contextual merits. Editing Wikipedia for encyclopedic ends means reading numerous sources for background perspective and content and, when we have borderline cases, closely evaluating the quality of the contribution in terms of its author’s scholarly or professional competence to see whether a general rule about deprecation or even non-mainstream sources has, case by case, grounds for exceptions to retain and use or not. That is laborious, requires deep familiarity with the topic, and careful judgement in context. Editors who, like David, just jump at deprecation listings to zoom through wikipedia erasing at sight the source used are examples of energetic laziness when they do this kind of mechanical weeding. They admit they don’t feel obliged to read the source they erase. For, by virtue of deprecation, they can ‘presume’ it's bad. I can understand it with the Daily Mail. But major scholars don't write for that rag: they do for CP. How does this work, this carelessness? Well, to cite just one example,in the deprecation RfC, Lord Swag set forth a diff-rich j’accuse list of ‘proofs’ CounterPunch approved genocide, holocaust denial, antisemitism. Patently dopey. It was froth, and I ignored it, expecting editors to check the tirade’s 'evidence' as I had. No, actually many editors quoted with approvgal Swag’s swag of pseudo proofs. Then Gerard chimed in and cited the evidence mustered in the original RfC (where Swag’s material dominated) as proof that CounterPunch merited deprecation. So, I sat down and analysed Swag’s influential ‘case’. Result? Pure trumpery. David just ‘presumed’ at a glance Swag’s evidence was cogent, rather than a mugged up heap of misdirections. Exactly what is he doing with his energetic removalist indifference to the quality of what he is erasing? What you appear to get in short is a practice of (a) not needing to know the topic area and the history it deals with (b) indifference to checking what you elide: suffice that it is deprecated, ergo weed out on sight (c) ignoring all the ambiguities of deprecation (as many editors have noted) (d) taking blindly on trust, without scrutinizing the diffs, what colleagues write. Everything is based on appearance, trust in those you trust, distrust of those you don’t know. The result is serious damage to the encyclopedia, since David can’t recognize a notable name, a scholar of major standing in the field and stop to reconsider and stay the itchy trigger figure. The impression is of hyperactivism whose main effect, regardless of his intentions, which I have no doubt are genuinely sincere, is to ratchet up an indeed impressive edit count, whatever the collateral damage might be to the ambitions of wikipedia to achieve encyclopedic ends, i.e. comprehensive scholarly coverage. Encyclopedias are not only a congeries of articles requiring bot-like checks, monitoring etc: the content is mostly written by people who take the trouble and effort to spend sometimes hours on each particular edit, checking any potential author’s competence and background, reading up several other sources to see if the claim or viewpoint is fringe or not, and examining all these things in context. David’s approach - insouciant to the efforts of content editors- thinks none of this is necessary. There is a law, it allows no exceptions, erase at sight, and snub talk pages where those who differ with his ultramontane legalism, and actually read the topic closely, give solid reasons, case by case, for retaining an reference within the framework of the broader wikipedia guidance principles. The solution is simple. Ask him, in the light of serious concerns at the collateral damage his mechanical rampage of elisions is causing, to desist.Nishidani (talk) 14:33, 25 December 2021 (UTC) The above screed is laboring under the various personal attacks ("puritanical", "lazy", "hyperactivism", etc.) used to describe David Gerard. I'd suggest withdrawing this, Nishidani. — The Hand That Feeds You:Bite 21:18, 25 December 2021 (UTC) I like many other editors dislike sloppiness, disattentiveness, especially if melded to zeal. By writing screed you are saying that the evidence provided (justifying those adjectives) needn’t be examined or answered. This is precisely what happened with Swag’s evidence and DG’s acceptance of it. No evidence given was checked or examined. If you dislike the adjectives, then I’ll replace thjem with ‘stringent/fundamen talist’, ‘otiose’ and ‘over-energetic’, but the substance of my documen tation is there. Ignore it by all mean s. Much of the original RfC for deprecation consisted of editors ignoring any significant control on diffs, and opinionizing instead, and I get the feeling the same unempirical impressionism will win the day here as well. Rather than look into the substance, one challenges the tone all too often. The tone innocuously reflects exasperation at the amount of work controlling sources takes for serious content editors, all evaporated by rapid mechanical rollbacking sight unseen, which we see here. That's very I/P-ish. A content dispute is 'resolved' by ignoring the content dispute and complaining about manners, even at AE/ANI.Nishidani (talk) 21:35, 25 December 2021 (UTC) If you've created a situation where an interview with Edward Said is considered unreliable for use as attributed support for the views of Edward Said you've created a really stupid situation. This also illuminates how wooden and childish many of the rules are around "reliable sources." In the specific case of Counterpunch while in recent years it has published a fair amount of, in my opinion, batshittery, it is not a hoax generator. If they publish an article under the name of a scholar or researcher, or claim an interview with someone of note, the claim of authorship should be treated as ironclad reliable. Then, as with many journals of opinion, you really need to consider what era of the publication an article is from if there are concerns about reliability or slant. But that would require an immersion in both general epistemological questions and a particular field, the ebb and flow of its controversies over time, and recent scholarship. Aint nobody got time for that. This is Wikipedia, after all.Dan Murphy (talk) 17:35, 25 December 2021 (UTC) Support. Wholesale removal of citations published in mass-circulation mainstream publications, particularly citations dealing with the arts, such as reviews of films, stage performances, books, art gallery exhibits, museums, etc, especially when such reviews are otherwise unavailable or difficult to find, is unacceptable and harms Wikipedia and its users without any tangible benefit in improving Wikipedia's reliability. I would also support the establishment of a review board where complaints can be submitted regarding indiscriminate deletion of specific reliable citations, even if such citations come from publications that have been accused of recent unreliability, but have a vast archive of valuable historical reporting. —Roman Spinner (talk • contribs) 10:00, 26 December 2021 (UTC) We definitely do not need another notice board. Either this board or RSN should suffice. Doug Weller talk 10:06, 26 December 2021 (UTC) Close thread and start a new discussion at a more appropriate noticeboard. This is the Administrators noticeboard/Incidents and this thread is not about taking Admin action. Doug Weller talk 10:08, 26 December 2021 (UTC) End copy