This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

I tried to make the role of RS clear in this edit by adding "in reliable sources" (Edit summary: "make clear, as unreliable sources have zero weight, so any comparison is between use in various RS.".)

That edit was reverted by User:Peter Gulutzan in this edit (Edit summary: "Undid Revision as of 23:36, 14 September 2022 by Valjean. This is about prominence of viewpoints, so doesn't depend on whether some editors call it reliable.")

Original version

Neutrality assigns weight to viewpoints in proportion to their prominence.

My version:

Neutrality assigns weight to viewpoints in proportion to their prominence in reliable sources.

Unfortunately, many fringe viewpoints are far more prominent in unreliable sources than in reliable sources, but we do not treat such ideas or unreliable sources as equal to RS and the debunking they receive in the few RS which mention them (see WP:Parity). Even one factual RS outweighs the nonsense spouted by thousands of unreliable sources. Due weight is about disagreements in RS. When there is genuine disagreement, we document both sides and assign due weight according to their prominence in RS.

Differing RS are treated on the same playing field, while reliable vs unreliable sources are on two different playing fields.

IOW, when there is disagreement between a viewpoint in RS and contrasting viewpoints in unreliable sources, we give the viewpoints in RS much more weight, regardless of the "prominence" found when looking at all sources.

Wikipedia is not neutral when it comes to documenting what RS say are falsehoods pushed by unreliable sources. We do not present both sides in a false balance. "Prominence" of viewpoints is not a factor when comparing reliable versus unreliable sources. We show that all opinions are not equal, so the difference between falsehood and facts becomes evident.

What do other editors think? -- Valjean ( talk) ( PING me) 15:52, 15 September 2022 (UTC)

A town's council says "yes", a columnist in the local paper says "no". It happens that the council members are well known and most citizens agree so they're the prominent side. But it happens that some folks on Wikipedia declared the paper is generally reliable so only the columnist's viewpoint could be quoted if somebody decided that's what "reliable source" means. Or if that's not what it means, then Valjean's insertion does the opposite of clarifying, it brings in a criterion that shouldn't be relevant because any source can be regarded as reliable as a source of its own opinion. Peter Gulutzan ( talk) 13:34, 16 September 2022 (UTC)

How do we know the viewpoint of the Town Council? Because it's printed in RS, so both get covered here. This has nothing to do with WP:ABOUTSELF. That always applies in the subject's own article. Even blacklisted sources can be used in such cases, but "due weight" deals with other issues. -- Valjean ( talk) ( PING me) 16:16, 16 September 2022 (UTC)

I am of course assuming for the example that the paper, due to its viewpoint, won't fairly reflect the council's, so their web site + members' blogs + other papers that didn't happen to get the generally-reliable sticker will all be suppressed. And I used the word "quoted" because it's especially a violation of WP:RS to give the council's supposed viewpoint via the paper when the council's "yes" is not sourced to the original. Peter Gulutzan ( talk) 18:04, 16 September 2022 (UTC)

Thanks for the clarification. Then we agree. ABOUTSELF does allow for the POV of the council as the article here would be, per ABOUTSELF, "about themselves, usually in articles about themselves or their activities". WP:ABOUTSELF and WP:RSOPINION are very closely related. -- Valjean ( talk) ( PING me) 18:23, 16 September 2022 (UTC)

It is false that the reason opinions are acceptable is ABOUTSELF, they are acceptable because WP:NOTCENSORED and in the case of the other papers WP:NEWSORG + WP:RSOPINION, and it is false that we agree. Peter Gulutzan ( talk) 14:24, 21 September 2022 (UTC)

RSOPINION allows use to use viewpoints from normally unreliable sources, though when to include of course depends on several factors. Just because someone who is considered a field expert or critical voice makes a viewpoint in an unreliable source doesn't mean we ignore it. EG: if a Republican senator is leading a fight against a bill but their criticism is only covered in depth in a source like Fox News, that's still something to consider for weight.

This is different from fringe views, which we do not want to give coverage to. Masem ( t) 13:38, 16 September 2022 (UTC)

Not sure about that.... An example is Glenn Greenwald. He is only usable for his own opinion in his own article, per ABOUTSELF. If he wants his viewpoints documented anywhere else, then he must be published by RS. So far they have lost all confidence in him. His own colleagues pushed him out after he abandoned good journalistic norms. Our standard is that we use only RS for sourcing, with ABOUTSELF being the only exception. No unreliable source should appear anywhere outside the biography of the author. That's also a BLP matter. You should look at the essay I'm developing: User:Valjean/Essay/Why Wikipedia documents opinions and nonsense#Due weight.

It is an uncomfortable and sad fact that nonsense and lies exist, but it is our job to document all knowledge that is notable enough for mention in a reliable source. Such negative things are often mentioned in fringe and unreliable sources, and sometimes RS document what unreliable sources say. Thus we document fringe topics and viewpoints, but not by linking to unreliable sources. No, we cite the explanations and analyses of the facts from reliable sources as they explain and filter that bogus information from unreliable sources.

If people want us to document their views in other than their own biographies (per WP:ABOUTSELF), they must get them published in RS. Until then, we document the coverage of their views as filtered by RS. Our standard is that we use only RS for sourcing, with ABOUTSELF being the only exception, and then limited to the person's biography. -- Valjean ( talk) ( PING me) 16:16, 16 September 2022 (UTC)

But again, RSOPINION allows for us to consider opinion pieces written in a work otherwise unreliable to be considered reliable for opinion, with consideration for due weight, expertise, etc. The problem tends to be the tunnel vision that opinion pieces can only be pulled from sources that themselves are broadly RS, but we have to consider that there may be non-fringe or non-insignificant minority views that are only captured by reliable opinion pieces in otherwise unreliable sources. Mind you, that is rather exceptional when that may actually occur (I generally agree that all major viewpoints related to a topic will likely be published in sources routinely considered reliable) but we have to make sure that type of case is left open. Masem ( t) 16:34, 16 September 2022 (UTC)

How do you square that interpretation of WP:RSOPINION with NPOV's wording in its first sentence?

"All encyclopedic content on Wikipedia must be written ...all the significant views that have been published by reliable sources on a topic." (Bolding added.)

That bold phrase describes our sourcing limitations. That's what makes us credible, so don't weaken that view. I'm wondering what wording in WP:RSOPINION gives you the idea that unreliable sources are usable outside of the author's own bio? Opinions (even false opinions) in RS are considered from a RS. -- Valjean ( talk) ( PING me) 17:22, 16 September 2022 (UTC)

RSOPINION does not restrict the published source of an opinion otherwise deemed reliable to a source that is normally considered reliable. "Some sources may be considered reliable for statements as to their author's opinion, but not for statements asserted as fact." As long as the opinion fulfills the UNDUE/WEIGHT factor (like a known expert in that field), and we're not talking about opinions related to BLP, there is no issue with using that opinion piece from a publication normally unreliable for fact. That is, the opinion piece on its own is a reliable source for the opinion per RSOPINION, and thus is a reliable source, even if the publication is not. Masem ( t) 17:29, 16 September 2022 (UTC)

You seem to be basing your interpretation on the two words "Some sources" and then trying to distinguish between sources reliable for facts and sources reliable for opinions, but, and that's a BIG "but", both are RS, just for different purposes. There is nothing in that section that implies we can generally use unreliable sources. Although it doesn't mention ABOUTSELF (it should), it does not nullify the fact that opinions published in unreliable sources are covered by ABOUTSELF and BLP. They are the exception that allows use in the author's own bio about themselves and their own POV, not about others. Regardless of any other interpretation, NPOV's first sentence still applies: "that have been published by reliable sources on a topic."

Somewhere in all this is the "primary" and "secondary" source issue, with limitations placed on our use of primary sources. That's where ABOUTSELF comes into play. Even deprecated and blacklisted primary sources can be used for ABOUTSELF purposes. -- Valjean ( talk) ( PING me) 17:47, 16 September 2022 (UTC)

No, ABOUTSELF doesn't limit how we can use opinion from unreliable sources, that is there that restricts any opinion directly related to BLP from unreliable sources to be only from self-published statements. It does not say anything about opinions that are not about BLPs; that's where RSOPINION says that opinions that are from unreliable sources may be used for their opinion (not fact) if they are DUE to include and unrelated to specific BLP. Masem ( t) 18:16, 16 September 2022 (UTC)

Ugh! Now my head hurts. We're getting into the overlap between WP:ABOUTSELF and WP:RSOPINION, which are very closely related. Please keep parsing this, preferably quoting from both. The truth is in there somewhere. -- Valjean ( talk) ( PING me) 18:28, 16 September 2022 (UTC)

RSOPINION is a broad statement that is not specific to the topic area stating "Some sources may be considered reliable for statements as to their author's opinion, but not for statements asserted as fact." For example if Greta Thornburg published an essay discussing her opinion on the state of fighting climate change in her blog, that would be just as reliable under RSOPINION if she published that as an opinion in the NYTimes. In either case, we would consider inclusion based on the nature of Greta's "expertise" under DUE. RSOPINION does call put that this approach cannot be done for material about a BLP, with the only exemption there being for BLPSPS. In other words, RSOPINION wraps around the limits of BLPSPS, but beyond BLP-related opinions, RSOPINION has no concern for what publication an opinion is in (with exception for publications known to alter opinions like Daily Mail). Whether we use those opinions becomes a DUE issue, so in evaluating what are non fringe viewpoints are out there, we may need to be aware of opinions that are primarily limited to publications that aren't reliable for fact but meet RSOPINION. -- Masem ( t) 15:11, 21 September 2022 (UTC)

I think the edit by Valjean was generally valid and descriptive, and I think I disagree with Masem. We've had this same discussion or dispute on other Wikipedia policy pages before, so you can probably guess how I feel about it and my position on it. In fact, reliable sources and more prominent sources should be given more weight. Unreliable sources, to the extent that they are marginally reliable in context, should be used with caution and afforded lesser weight in accordance with, and proportionately to, their prominence and reliability. Prominence, for a source, doesn't just mean that it's numerically common, it means how authoritative and how acceptable the source is, relative to how other RS view the source. So there is indeed some kind of general set of criteria and a kind of a pecking order in that a well-known textbook or a well-known expert author's book, would trump something like an expert author writing a journal article, which would trump a NEWSORG or staff magazine or news or periodical article, which would trump various kinds of other published sources that don't come from a reliable outlet, but of course there are exceptions, for example if Albert Einstein wrote a short magazine article or an op-ed where he also talked about physics and relativity, that might be more reliable and worthy of more BALANCE/DUE/NPOV weight, versus, a long series of academic books published by someone less prominent than Einstein. Andre 🚐 18:34, 16 September 2022 (UTC)

My change and comments above are wholly consistent with the first sentence of this policy:

"All encyclopedic content on Wikipedia must be written from a neutral point of view (NPOV), which means representing fairly, proportionately, and, as far as possible, without editorial bias, all the significant views that have been published by reliable sources on a topic." (Bolding added.)

We base all content on RS, with the exception of ABOUTSELF, which is limited to the biography of the author and documents their own views about themselves, not even about others, per the limitations of BLP. -- Valjean ( talk) ( PING me) 16:24, 16 September 2022 (UTC)

That wording goes back to 13 July 2006 long before what I regard as the start of the idea that a publication source could be unreliable regardless of context i.e. WP:DAILYMAIL1, and the closers of WP:DAILYMAIL1 had to repeatedly explain that the ban didn't apply to Daily Mail's opinions, i.e. the claim that they were only acceptable due to WP:ABOUTSELF was false. So far two editors have disagreed with Valjean's insertion or assertions, and I'm repeating Valjean's question: What do other editors think? Peter Gulutzan ( talk) 18:04, 16 September 2022 (UTC)

I'm surprised that wording wasn't added by myself, as I helped develop the policy. I've been here since 2003, before our first 200,000 articles, and some of my additions and tweaks are still there. It was added by SV. IIRC, we already had a blacklist at that time, so we did have sources that were absolutely unreliable sources. -- Valjean ( talk) ( PING me) 18:34, 16 September 2022 (UTC)

I agree with Valjean and I think I disagree with Peter, at least as pertaining to the question of whether the policy was intended to allow some sources to be categorically disallowed - it was always known consensus to use certain tabloid sources or certainly unreliable news sources for things. This was never doubted AFAIK prior to 2005. Andre 🚐 18:36, 16 September 2022 (UTC)

Valjean's insertion has been re-inserted. I don't see consensus for it but don't intend re-revert. Peter Gulutzan ( talk) 14:24, 21 September 2022 (UTC)

I think the problem here is that we have two definitions of "reliable source". One is whether the source is an "acceptable source" for a specific statement in a specific article (e.g., WP:ABOUTSELF); the other might more properly be called an "ideal source" (e.g., independent, secondary, peer-reviewed, etc.).

If you are starting an article, and you expect to have many sources available, you should normally focus on the "ideal sources". Then you go through the sources you have, and you balance the various viewpoints. This works extremely well on subjects like American Civil War, and it works fairly well on subjects like Cancer.

It works less well on other subjects, like scandal-ridden celebrities and commercial products. You might, for example, discover that all (i.e., both) modern scholars agree that the blue-green widgets are the best widgets ever created and that it's a crying shame that the manufacturer discontinued them. But the manufacturer has a POV on this, too, and that POV isn't guaranteed to be represented in the "ideal sources".

The solution, I think, is to apply some common sense. You can take the information from "ideal sources" and add information from "acceptable sources", but you don't want to add just any old statement with a source that is reliable for the specific statement. A blog post from a kid talking about how his parents ramble on and on about the wonders of blue-green widgets is technically reliable for a statement that "Some Kid published a blog post about his parents' love for blue-green widgets", but we don't want to include that (unless some special circumstance is relevant). A blog post from the manufacturer about their reasons for discontinuing it probably should be included, so the article can say something like "The manufacturer discontinued blue-green widgets due to poor sales". WhatamIdoing ( talk) 07:51, 15 November 2022 (UTC)

So do you approve or disapprove of adding "in reliable sources"? Peter Gulutzan ( talk) 14:38, 15 November 2022 (UTC)

Is there an example when this caused a problem? I think it's covered by WP:PST. PaulT2022 ( talk) 23:21, 15 November 2022 (UTC)

Do we have anything in our manuals of style about this? It came up in discussing with a new user the use of the word "Holy" for religious texts. Thanks. Doug Weller talk 14:09, 16 November 2022 (UTC)

I don't remember MoS status off hand but very long and substantial history of discussion exists around the issue, previously I had referred some where. Bookku ( talk) 15:44, 16 November 2022 (UTC)

• Title Jesus Christ was moved to Jesus on 4 February 2005‎ I have not come across discussion in that month. But previous discussions can be found @:
• Talk:Jesus/Archive 1
• Talk:Jesus/Archive 9 2004

Bookku ( talk) 09:46, 18 November 2022 (UTC)

@ Doug Weller, MOS:GOD uses mainly Jesus ( "Christ", "Jesus Christ", "Jesus the Christ", and "Jesus of Nazareth" are redirects). Thinker78 (talk) 01:21, 17 November 2022 (UTC)

@ Thinker78 Thanks. I see nothing specific to my issue so I've raised it at the talk page. This really needs guidance. Doug Weller talk 08:59, 17 November 2022 (UTC)

@ Doug Weller, there is some info in the talk page of Jesus that you can add to said discussion, for example, under the FAQ, "Q1: What should this article be named?". Cheers! Thinker78 (talk) 00:48, 18 November 2022 (UTC)

I'm not sure that we need something on this. We don't seem to have trouble resolving these disputes; we just have to take a few minutes to explain it to the occasional newbie, and we'd have to do that anyway, because Wikipedia:Nobody reads the directions. If this became an increasing problem, then Wikipedia:Manual of Style/Words to watch might be a suitable place to add a "don't call things holy" statement. WhatamIdoing ( talk) 04:24, 18 November 2022 (UTC)

This might be a bit of a tangent, but we might need to explain what an opinion is. There seems to be an idea that all labels are opinions. We keep having editors claim that there's no such thing as right-wing politicians; there are only politicians about whom sources have expressed the strictly subjective and possibly wrong opinion that they're right-wing.

There is a debate in philosophy about whether there's such a thing as a "subjective fact". I think that "That hurt me!" is what's intended, because it's a claim that could be true or false (and so a candidate for being a fact), but it's also a claim that depends on perspective and perception (and so subjective). But this disconnect sometimes goes further than that; it is basically a claim that only information that is precise, discrete, and concrete deserves a place in an encyclopedia. This makes it difficult to describe people (who are complex and contradictory) and subjects that we don't know much about. WhatamIdoing ( talk) 14:57, 18 November 2022 (UTC)

• A lot depends on context. Using “Christ” (in running text) will be appropriate in some articles, and inappropriate in others. It isn’t something we can make a one-size-fits-all rule about. Use common sense. Blueboar ( talk) 12:55, 18 November 2022 (UTC)

It jumps around a bit, doesn't it?

What would people think about moving WP:BALANCE up to be the parent of WP:BALASP and WP:FALSEBALANCE? And moving "Bias in sources" either up adjacent to WP:BESTSOURCES or down under "Attributing and specifying biased statements" in "Handling neutrality disputes"? Sennalen ( talk) 16:38, 16 December 2022 (UTC)

In the final paragraph of § Naming, it currently says:

an article titled "Criticisms of X" might be better renamed "Societal views on X".

The text's origin dates back to July 2007 and at that time, it had two defined (definitive? whatever) examples: renaming allegations that drugs are evil to societal views on drugs. It was edited one month later to include the "Criticisms of" vs. "Societal views on" examples and in 2010, drugs was replaced with X.

Since only two redirects and no articles use that "Societal views on..." format, I propose changing the text to something closer to the original version, so that it reads:

an article titled "Allegations that X is evil" might be better renamed "Criticism of X" (singular criticism). Or alternatively, "Allegations that X is bad".

Any thoughts? ~~ lol1 VNIO (I made a mistake?  talk to me) 15:41, 16 December 2022 (UTC)

I prefer the "X is bad" example. I think it might be easier for editors to understand it as a generic stand-in for any type of judgmental viewpoint. WhatamIdoing ( talk) 03:11, 20 December 2022 (UTC)

The context is that Neutral titles encourage multiple viewpoints and responsible article writing. I'd keep the current wording because it's better to have both sides of an argument. Even if a lot of articles ignore this advice, it's still good advice. Shooterwalker ( talk) 03:16, 20 December 2022 (UTC)

Hi. I'm currently in a disagreement at Reactions to the death of Elizabeth II, so I just have a question. I think it would be better to have partially recognised countries (Northern Cyprus, Somaliland and Western Sahara) under a "partially recognised states" section as it makes the articles neutral, whereas another editor thinks they should be amongst other normal/recognised countries' sections as the sources that cite them in the article consider them a country. I was just looking for some advice on this as per the NPOV policy? Thanks. - Therealscorp1an ( talk) 21:29, 15 March 2023 (UTC)

I have concerns with the recent addition of the shortcut to Fringe theories and pseudoscience subsection. The stated rationale for this addition was "since the shortcut name has caused confusion about the scope, provide another alt", but i believe that the confusion comes from the text of the paragraph. As a background, this addition was added as a result of a disagreement elsewhere see here here here here and especially here. To quote that last one the editor in question stated "All fringe subjects fall under WP:PSCI. This is explicit.'

I believe that all fringe subjects fall under WP:fringe and if WP:PSCI is meant to be a shorthand for the longer and not policy WP:fringe then we should rewrite the paragraph under WP:PSCI to match what is in WP:fringe, which most definitely does not say that all fringe topics are pseudoscience. Looking at the paragraphs under WP:PSCI, there are 10 sentences, 8 of which talk about pseudoscience explicitly and exclusively and we say the word 'pseudoscience' 9 times. The only mention of anything aside from pseudoscience is one sentence which states (in part) "This also applies to other fringe subjects, for instance, forms of historical revisionism that are considered by more reliable sources to either lack evidence or actively ignore evidence, .." (example removed)" Im not totally sure what the 'this' in that sentence is referring to.

Conversely, WP:Fringe states in its lede "... the term fringe theory is used in a very broad sense to describe an idea that departs significantly from the prevailing views or mainstream views in its particular field." That guide goes on to to specify different types of fringe theories (Pseudoscience, Questionable science, Alternative theoretical formulations) and in fact goes on to state that neither questionable science or alternative theoretical formulations should be described as pseudoscience.

So i propose the following, we rewire most of Fringe theories and pseudoscience to be more or less in line with Fringe. Perhaps:

"We use the term fringe theory in a very broad sense to describe an idea that departs significantly from the prevailing views or mainstream views in its particular field. A Wikipedia article should not make a fringe theory appear more notable or more widely accepted than it is. Thus, when talking about fringe topics, we should not describe these two opposing viewpoints as being equal to each other. While a fringe theory may in some cases be significant to an article, it should not obfuscate the description of the mainstream views of the scientific or academic community. Any inclusion of pseudoscientific or fringe views should not give them undue weight. Pseudoscientific views should be clearly described as such. An explanation of how scientists have reacted to pseudoscientific theories should be prominently included. This applies to pseudoscientific claims, forms of historical revisionism that are considered by more reliable sources to either lack evidence or actively ignore evidence, questionable science, alternative theoretical formulations, medical misinformation and conspiracy theories. (updated per feedback below)

Or something like that to make clear that this applies to all fringe topic, but that not all fringe topics are pseudoscience, or questionable science or history revisionism for that matter. Bonewah ( talk) 14:11, 7 April 2023 (UTC)

Well, no. Because it leaves the door open (as it seems you want) to giving undue weight to medical misinformation, as well as conspiracy theories, alien stories, Bigfoot, creationism, etc. etc. etc. The broad wording of NPOV currently is exactly right: all "fringe subjects" are subject to the wording in WP:FRINGESUBJECTS. The section name is, after all "Fringe theories and pseudoscience" - that is, deliberately broad. To omit (as proposed) this existing policy requirement would strike at the heart of Wikipedia's neutrality:

The pseudoscientific view should be clearly described as such. An explanation of how scientists have reacted to pseudoscientific theories should be prominently included.

Bon courage ( talk) 14:50, 7 April 2023 (UTC)

I think everyone's pretty clear that due and undue weight apply to all subjects. This argument seems more about wounded pride. Sennalen ( talk) 14:58, 7 April 2023 (UTC)

OP also seems to misunderstand: WP:FRINGE applies to both fringe subjects and minority views which are not fringe, pseudoscience and theories which are not pseudoscience, but pseudoscience-adjacent. The FRINGE guideline describes how to differentiate them and how to treat each of these. One cannot skirt FRINGE (or this section of NPOV) simply because one believes a theory is more plausible, or because one finds slightly more evidence in the literature in favor of one or another theory.

This proposed rewrite does not fix any problem, and no current discrepancy exists, as far as I can ascertain.

This also applies to other fringe subjects, for instance, forms of historical revisionism that are considered by more reliable sources to either lack evidence or actively ignore evidence, .." (example removed)" Im not totally sure what the 'this' in that sentence is referring to.

It was very clear to me when re-reading it. "This" refers to the entire section. —  Shibbolethink ^{( ♔ ♕)} 14:54, 7 April 2023 (UTC)

I think there's broad agreement about what the policy says, but there's still some doublethink where editors will link these shortcuts while arguing to treat fringe as all-or-nothing. Sennalen ( talk) 15:01, 7 April 2023 (UTC)

If the concern here is that we might give undue weight to medical misinformation, as well as conspiracy theories, alien stories, Bigfoot, creationism, etc. etc. etc. as you say, then we should write the section in such a way so that it doesnt. Further i fully understand that WP:FRINGE applies to both fringe subjects and minority views, which is exactly why i want *this* section to reflect that. As it stands now, the section in question only mentions non-pseudoscience once while WP:FRINGE itself immediately specifies that it is not just about pseudoscience. If you really want to prevent editors from skirting FRINGE, then, in my opinion, the best solution is to rewrite this section of NPOV to make it clear that it applies to things that are not pseudoscience but are fringe. This is especially relevant as this page is actual policy whereas FRINGE is a guideline. And the notion that this is 'about wounded pride' is absurd as the stated rationale for the addition of the shortcut in question was to clear up confusion about the scope. I agree that there is confusion as to scope, this is an attempt to correct that. Bonewah ( talk) 15:44, 7 April 2023 (UTC)

The section is already written that way [my emphasis]:

The pseudoscientific view should be clearly described as such. An explanation of how scientists have reacted to pseudoscientific theories should be prominently included. This helps us to describe differing views fairly. This also applies to other fringe subjects, for instance, forms of historical revisionism that are considered by more reliable sources to either lack evidence or actively ignore evidence, such as claims that Pope John Paul I was murdered, or that the Apollo Moon landings were faked. ...

At this point I think you need to say what the intended effect of your change is, and give one or two concrete before-and-after examples of how an article would be different if you got your way. As I see it this opens the door to a load of WP:PROFRINGE-iness. Bon courage ( talk) 16:14, 7 April 2023 (UTC)

My intended effect is to clarify that WP:PSCI applies to all manor of fringe topics, not simply pseudoscience. I fully admit that i misunderstood you in the previous discussion and I agree that the shortcut PSEUDOSCIENCE is badly chosen shortcut name, as you stated here. As i stated earlier, the confusion stems from the fact that as written (prior to your edits) almost everything about this section references pseudoscience only, says pseudoscience repeatedly and was even shortcutted pseudoscience. I view myself as a fairly run of the mill editor and it stands to reason if I viewed this section as only applying to pseudoscience, then other editors most likely will as well. You must at some level agree as, again, your stated rationale for your change was to clear up confusion.

To provide a concrete example, if someone were to edit an article about economics, say Monetary policy, to include references to Modern Monetary Theory, you might (rightly) caution against providing false balance or undue weight, citing, (again rightly) WP:PSCI. The other theoretical editor might, also rightly, reply that MMT isnt pseudoscience, its more like an alternative theoretical formulation, and nothing in WP:PSCI says anything about that. Im sure you know how the rest of this exchange would go, you cite FRINGE, they point out that FRINGE isnt policy, etc etc. The solution? Make PSCI a paragraph or so summary of FRINGE, which mentions pseudoscience but also mentions all the other things that FRINGE goes on to clarify. Bonewah ( talk) 16:39, 7 April 2023 (UTC)

We can't make people read the policy, but in your example the 'solution' is to use the new shortcut and not PSCI to clue them in better. So it's unclear why you have 'concerns' about the new shortcut. Bon courage ( talk) 04:25, 8 April 2023 (UTC)

I agree here that "fringe" and "pseudoscience" are not synonyms, but my concern is mostly opposite of OP's. There are many fringe topics which do not claim to be scientific, such as astrology, and therefore using the word "pseudoscience" to refer to them is inaccurately narrow. But personally I would just remove the references to "fringe" from the section on pseudoscience, as it's perfectly reasonable to have a section on pseudoscience specifically. Loki ( talk) 21:45, 9 April 2023 (UTC)

That consequence of that would be to remove the policy requirement to identify non-pseudoscience fringe topics (moon landing hoax e.g.) as such prominently, and to contextualize those fringe views with mainstream expertise. Not a good idea and a WP:PROFRINGERS charter. Bon courage ( talk) 06:21, 10 April 2023 (UTC)

@ LokiTheLiar: You and I are saying the same thing, the word pseudoscience is inaccurately narrow and "fringe" and "pseudoscience" are not synonyms. As written WP:PSCI speaks almost exclusively about pseudoscience, we should expand it to cover all the things covered in wp:Fringe. I agree with Bon courage here, if we removed everything except the material covering pseudoscience, then we would, in effect, be saying, this policy does not cover non-pseudoscience material such as history revisionism or astrology or whatever. Bonewah ( talk) 15:36, 10 April 2023 (UTC)

@ Bon courage: Isn't that already in WP:FRINGE? If not it definitely should be, right? Loki ( talk) 16:28, 10 April 2023 (UTC)

FRINGE is only a guideline. We have enough trouble with WP:PROFRINGE editors lawyering about that as-is. I think nerfing core policy would require a policy RfC and it would send out a very strong message that Wikipedia was going soft on fringe topics, one of the very pillars of its reputation. Bon courage ( talk) 16:33, 10 April 2023 (UTC)

• Okay I've attempted a rewrite that hopefully doesn't change the meaning, but clarifies it. Dominic Mayers has changed "relevant experts" back to "scientists". The problem is that for certain fringe views (pseudohistory for example) it's not 'scientists' who represent the mainstream, but acknowledged experts in their non-science fields. How to square this circle? Bon courage ( talk) 16:58, 10 April 2023 (UTC)

  I did not realize what was the purpose. Why not just use "experts" by itself, without the "relevant" in front of it. Dominic Mayers ( talk) 17:14, 10 April 2023 (UTC)

  Because we often have a fringe problem because experts in one field support nonsense in another. Bon courage ( talk) 17:23, 10 April 2023 (UTC)

  I support this rewrite. Loki ( talk) 17:29, 10 April 2023 (UTC)

@ Bon courage: I dont see how this is a nerf as im proposing expanding the definition here to match what we currently have in wp:Fringe. As written, it looks as though WP:PSCI only applies to pseudoscience, when it should be a summary of what is in WP:fringe. Bonewah ( talk) 18:40, 10 April 2023 (UTC)

You removed the principal requirements: "The pseudoscientific view should be clearly described as such. An explanation of how scientists have reacted to pseudoscientific theories should be prominently included". Bon courage ( talk) 18:56, 10 April 2023 (UTC)

Ok, ive added that line as well. Bonewah ( talk) 19:07, 10 April 2023 (UTC)

You're all getting into the weeds and mixing subjects together that don't belong together. The solution, as proposed in the next section, is to deal with them separately. Pseudoscientific topics are fringe topics, but not the other way around. Instead of wasting a lot of time and effort to explain that, just deal with pseudoscience as a subcategory (in its own section) of fringe theories.

Now please read the next section before proceeding here. Maybe the problem can be solved in an easier way than continuing this discussion. -- Valjean ( talk) ( PING me) 19:58, 10 April 2023 (UTC)

The WP:V WP:NPOV preamble says:

"...representing fairly, proportionately, and, as far as possible, without editorial bias, all the significant views that have been published by reliable sources on a topic."

That means this policy is applicable only to the cases when more than one point of view exists on some concrete topic. However, the words "on a topic" are omitted in WP:DUE, which says:

"Neutrality requires that mainspace articles and pages fairly represent all significant viewpoints that have been published by reliable sources, in proportion to the prominence of each viewpoint in the published, reliable sources".

These words are interpreted by some users as a criterion for inclusion of some topic into articles: the topic (not "the viewpoint") covered by many sources should be included, but the topics covered in an smaller number of sources should not. IMO, that is incorrect, and I propose to add the words "on a topic" to WP:DUE as folows:

"Neutrality requires that mainspace articles and pages fairly represent all significant viewpoints on a topic that have been published by reliable sources, in proportion to the prominence of each viewpoint".

I also think " in the published, reliable sources" is awkward and redundant, because the meaning of "published" is discussed in WP:V. -- Paul Siebert ( talk) 00:59, 24 May 2023 (UTC)

@ Paul Siebert: I assume you meant WP:NPOV not WP:V ......recommend correcting. North8000 ( talk) 17:31, 25 May 2023 (UTC)

Yes, fixed.-- Paul Siebert ( talk) 17:51, 25 May 2023 (UTC)

I think that adding "on a topic" is fine. I wonder if it should be "on the subject of the article"?

(I wonder whether we should ever address "viewpoints". A viewpoint is content about the subject that can differ based on how you look at the subject: When I look at ("view") the subject from this place ("point"), I think this is a problem; when I look at the subject from this other place, I think this is okay. However, there aren't any extant viewpoints about some rather boring subjects, e.g., Entomocorus benjamini. A neutral encyclopedia article in such a case is one that provides the expected facts and no viewpoints.) WhatamIdoing ( talk) 05:26, 24 May 2023 (UTC)

Consider this example. 50 sources say oranges contains a lot of vitamin C, 2 sources say otherwise. These sources are in an obvious conflict with each other. The "topic" is "if oranges are a good source of the vitamin C", and per WP:DUE we should present the view that oranges have a lot of vitamin C as a majority view (and, probably, not to mention another view at all).

Another example: 50 sources say oranges contain a lot of vitamin C, and 2 sources say apples are a good source of vitamin C. Unlike the first example, there is no conflict between the sources: the first group of sources say nothing about apples, the second group says nothing about oranges. Therefore, WP:DUE is not applicable: the topics are different.

However, I noticed that some users refer to WP:DUE and argue (I will continue using this artificial example) that, since oranges are mentioned in 50 sources and apples only in 2, then (ostensibly per WP:DUE) we should tell about oranges, but not about apples. Therefore, to avoid this type misunderstanding, I propose to add this clarification. Paul Siebert ( talk) 05:48, 24 May 2023 (UTC)

@ Paul Siebert Shoot I just noticed your last line after I made a bold edit. I agree that "the published, reliable sources" is a bit redundant—though I think that's the case because the sentence previously says "published by reliable sources" (and the use of the definite article "the" makes me think that it's a specific callback). I think a relative pronoun is especially appropriate there. As such, I changed the sentence to: Neutrality requires that mainspace articles and pages fairly represent all significant viewpoints that have been published by reliable sources, in proportion to the prominence of each viewpoint in those sources. diff.-- Jerome Frank Disciple 18:21, 24 May 2023 (UTC)

IMO, "on each topic", or "on a topic", is more important, because some users interpret WP:DUE as WP:NOTABILITY (a threshold for inclusion). We must clearly explain that WP:DUE is not about a coverage per se, but about a relative weight of at least two conflicting points of view. Paul Siebert ( talk) 18:39, 24 May 2023 (UTC)

...and "on a topic" applies to both the "significant viewpoint" and the "reliable source"? or is that beyond the intent? If the article topic is the chemical composition of fruits then it shouldn't be using a source which might as an aside say something about vitamin C, but the source topic and where the author is qualified is economics of orchard agriculture. fiveby( zero) 20:20, 24 May 2023 (UTC)

Of course, to viewpoints. The core idea of NPOV is: if more than one points of view on some topic exist in reliable sources, each of them (except fringe ones) should be presented in a balanced way. Therefore, when WP:DUE says about " all significant viewpoints", it is implied that these are the viewpoints of the same topic. These words are currently absent, which creates a false impression that WP:DUE discusses a threshold for inclusion (in terms of prominence of some viewpoint taken separately from others). Going back to my example,

• If some sources say oranges are a good source of vitamins, but other sources say otherwise, WP:DUE is applicable, and we should define which view is a majority view.
• If some sources discuss oranges (and not apples), whereas other sources discuss apples, these sources are not in a conflict with each other, but complement each other, so WP:DUE is not applicable. In that case, each source about apples and each source about oranges is evaluated on its own merit, not in each other's context. We cannot say "50 sources say about oranges, and only 2 sources say about apples, so we exclude apples per WP:DUE".

And the words "on a topic" make that clear. Paul Siebert ( talk) 02:42, 25 May 2023 (UTC)

I'm trying to think of ways your argument can be turned around and used to wiki-lawyer in inappropriate content. Recalling a certain disturbing detail on the famine article concerning Harry Lang and a "reliable source", Tapay. We have some very qualified sources (oranges) telling use what is appropriate to relay to the reader concerning the famine and the regime's responsibility. There is one source, Tapay (apples) discussing a poster. The argument for inclusion was that since no one disputes the poster's existence and it's in a "reliable source" it can't be excluded. My argument is that we shouldn't have enough confidence in the sources to say anything at all about apples. Also of the big names, those well-qualified to tell us what's important for the reader, none have included it in their works. Maybe i am reading too much into your proposal and "on topic"? Not that don't agree in general, if you were to include some apple content in and article there's probably a good reason behind it. But more often the problem seems to be editors trying to sneak in apples content in order to imply something about oranges. fiveby( zero) 16:20, 25 May 2023 (UTC)

Yes, what you say is correct. With one reservation. WP:DUE is not the only policy. Another relevant policy is WP:ONUS: if some statement or viewpoint is not in a conflict with other views, it still may be removed, for example, because its relevance to the article's subject is questionable, or because the users decided to replace it with some other piece of text. As you probably remember, the words about that posters were removed, but a description of cannibalism stayed: we just decided to add other facts, which were better documented.

Therefore, we removed Tapay per WP:ONUS, not per WP:DUE. Paul Siebert ( talk) 17:03, 25 May 2023 (UTC)

I have some hesitation about this. I don't want to sidetrack this conversation by detailing too many specifics, but I'm going to borrow from one content dispute I've seen recently, though I'm going to alter the facts to make it a more apt illustration of my concerns. My point isn't so much that one answer is clearly right or wrong, but that the ambiguity in the policy is good.

By your argument, DUE only applies to sources in direct conflict with eachother. But, sometimes, a fringe viewpoint isn't in conflict with any established viewpoint. For example, there's recently been some dispute over whether the Columbine massacre was a terrorist attack. Now, without debating whether it was or wasn't, let's imagine/pretend that only one single source exists describing the massacre as a terrorist attack. (Don't fact check that—just go with the hypo.) On the basis of that source, an editor adds a terrorist-attack category to the article and adds the phrase terrorist attack to the body—"The terrorist attack began at ...."

By your argument, WP:DUE wouldn't be a concern unless other sources directly contradict a source. So, in this scenario, you'd essentially have to have sources explicitly taking the negative position—that Columbine was not a terrorist attack. But I think that has to be wrong ... if, of all the articles and sources discussing Columbine, only one article used the phrase terrorist attack, then I think calling the massacre a terrorist attack would present a DUE concern. And why would any sources go out of their way to dispute a descriptor used in only one source? (You might say that it's a WP:FRINGE concern and not a WP:DUE concern, but bear in mind that FRINGE is, by its own text, clearly related to WP:DUE ("Because Wikipedia aims to summarize significant opinions with representation in proportion to their prominence, a Wikipedia article should not make a fringe theory appear more notable or more widely accepted than it is.")-- Jerome Frank Disciple 16:33, 25 May 2023 (UTC)

^ I tried to craft that hypo to specifically illustrate the type of issue I'm concerned about. Yes, sometimes sources contradict eachother without directly addressing eachother—for example, a source that calls oranges a fruit directly contradicts a source that calls oranges a vegetable. But sometimes that's not the case—there's no "not terrorist attack" descriptor that a source would naturally use.-- Jerome Frank Disciple 16:43, 25 May 2023 (UTC)

Yes, that situation is quite common. However, in your example, these two sources discuss oranges, i.e. the topic is the same. We cannot refer to WP:DUE to reject the view that apples are a vegetable because a majority sources say oranges are a fruit. Paul Siebert ( talk) 17:24, 25 May 2023 (UTC)

I think I have some confusion about what "on a topic" is supposed to mean here. Can you refer to specific content disputes to show where you think the additional language would be appropriate?-- Jerome Frank Disciple 17:27, 25 May 2023 (UTC)

I would prefer to avoid bringing that dispute to another forum. I hope my post below explains my point. Paul Siebert ( talk) 17:34, 25 May 2023 (UTC)

WRT "By your argument, DUE only applies to sources in direct conflict with eachother." No, DUE applies to the viewpoints that are in a conflict with each other. Thus, if 50 sources say oranges are a fruit and 2 sources say oranges are a vegetable, there is a conflict between them even if they do not cite each other and do not mention each other. Therefore, DUE is applicable here.

However, if 50 sources say oranges are a fruit, and 2 sources say apples are a vegetable, there is no conflict, and we cannot claim the view that apples are a vegetable should be removed per WP:DUE. Actually, if NO sources claim apples are a a fruit, the view that apples are a vegetable is a majority view. Paul Siebert ( talk) 17:32, 25 May 2023 (UTC)

What about the example I used?-- Jerome Frank Disciple 17:35, 25 May 2023 (UTC)

Do you mean Columbine?

Speaking formally, I envision the following description of this event:

1. Columbine was a terrorist attack;

2. Columbine was not a terrorist attack;

3. The perpetrator's motives were not clear.

4. something else

No matter if the sources from groups ##1-4 mention or cite each other, the views ##1-4 are, to some degree, in a conflict with each other, AND they relates to the same topic (Columbine). Therefore, DUE is totally applicable. Paul Siebert ( talk) 17:46, 25 May 2023 (UTC)

I think that's fighting the hypo a bit, no? As I said, "2" is unlikely, because it's a negative that would almost certainly only be brought up in response to the positive claim, but, as I stipulated, the positive claim is only in one source. "3", as I see it, is only, at best, loosely related to the question—the motivations behind a terrorist attack can be unclear. And "4" is just a catchall? I'm not sure what "something else" is supposed to indicate.

Again, as I said, the central problem is that there isn't likely to be a "and it WASN'T a terrorist attack" viewpoint that gets published, even though, as I stipulated, the terrorist-attack descriptor is only embraced by one source. This was just one example, but I'm sure we can brainstorm many—if a single source characterizes an event in an unusual way, it's not always true that other sources (or the viewpoints expressed therein) will directly contradict that characterization.-- Jerome Frank Disciple 17:51, 25 May 2023 (UTC)

When a single source characterizes the event in an unusual way, and there is a conflict with other sources, then DUE applies. It doesn't matter if this conflict is acknowledged in any of those sources.

When there is no conflicting views, or when there is just one view, WP:DUE is not applicable. We cannot say "this information should not be included per WP:DUE because only two sources say about that." Paul Siebert ( talk) 18:10, 25 May 2023 (UTC)

Ah, I appreciate the clarification. Given that I think that such a policy opens a fairly large hole, as per above, I have to oppose this proposed change.-- Jerome Frank Disciple 18:12, 25 May 2023 (UTC)

Which hole? Paul Siebert ( talk) 18:13, 25 May 2023 (UTC)

It seems that hole is covered by WP:ONUS. Paul Siebert ( talk) 18:15, 25 May 2023 (UTC)

You might be conflating the requirement to "fairly represent" viewpoints with how to go about doing that. " EMDR is an empirically supported treatment" and "EMDR is pseudoscience" are both well-supported valid views. You might see those viewpoints as conflicting or not and this might seem an intractable problem. But really isn't within the body of the article, and "on a topic" helps to do that and organize the content. If you focus on how to "fairly represent" those viewpoints with an MOS:LABEL, category, or entry in an infobox, etc. it's a harder problem and one i don't think the proposal is addressing? fiveby( zero) 18:28, 25 May 2023 (UTC)

WRT "You might see those viewpoints as conflicting or not and this might seem an intractable problem." I see no problem at all. From a point of view of a formal logic, two viewpoints may either contradict to each other or complement each other. Tertium non datur.

In the former case, WP:DUE is applicable, in the latter case it is not. Paul Siebert ( talk) 19:10, 25 May 2023 (UTC)

I think that "on a topic" is a good thing. Much POV happens by inclusions of things that are more removed in WP:Relevance just for effect, and more emphasis that it is sort of a balancing provision on a articular topic/question would be a good thing. For example, if I'm, an orange hater, I can't use the previous wording of that provision to muscle into the above "Oranges were Adolf Hitler's favorite fruit". North8000 ( talk) 12:22, 25 May 2023 (UTC)

• Not sure that “topic” is the right word. This could cause confusion with “article topic”. Sometimes the disputed viewpoints concern relatively minor details that don’t necessarily directly relate to the article topic, ie details that flesh out the article. For example: say we are writing an article about a politician, and we wanted to cover that politician’s stance on something controversial… say gun control. Within this article, we still need to be neutral in how we present any coverage of gun control, even though the “topic” is the politician and the sub-topic is his/her stance. Blueboar ( talk) 18:19, 25 May 2023 (UTC)

  Understood. Maybe, "each subtopic"? Paul Siebert ( talk) 18:27, 25 May 2023 (UTC)

  Well, one of the reasons for this discussion is to make it consistent with the lead, and the lead says "topic". Also, this whole section is not that explicit and so we probably needn't struggle for unrealistic precision on "topic". North8000 ( talk) 19:06, 25 May 2023 (UTC)

The words "on a topic" were added to the preamble on 15 June 2013 by Nikkimaria with edit summary = "ce". I believe a better way to get consistency is not to repeat them elsewhere but to revert that bold change. Peter Gulutzan ( talk) 14:23, 26 May 2023 (UTC)

Support revert. The "on a topic" for the scope of the article is redundant and reiterates published sources that are directly related to the topic of the article in WP:OR.

I share Paul's and Blueboar's concern about use of UNDUE in disputes about subtopics, but my reservations with the original proposal would be:

1. To what extent is it "abuse" of WP:UNDUE as opposed to editors mistakenly referring to WP:UNDUE where they mean WP:MINORASPECT. Changing the wording of WP:DUE as originally proposed would have no bearing on it.

2. Is the proposed change an attempt to solve a conduct issue (editors wikilawyering to exclude content to shift balance in articles) with a content policy change.

3. Can specific issues in specific articles be solved by applying other policies? Anecdotally, when I encountered controversial 50 vs 2 sources oranges/apples examples, it'd often be an editor using 50 WP:PRIMARY sources to argue that 2 WP:SECONDARY sources should be excluded. This is already covered in WP:BALANCE and WP:PST. PaulT2022 ( talk) 16:16, 26 May 2023 (UTC)

Actually you are right about MIONORASPESTS. However, I am not sure the reference to "published sources that are directly related to the topic of the article" resolves the issue. The question remains if NPOV is applied to the article's topic as whole, or to each aspect taken separately. In my opinion, DUE is dealing with the situations when different viewpoints exist on the same subtopic/aspect (and these viewpoints contradict to each other), whereas MINORASPESTS explains what we should do when the sources complement each other, although one of the aspects may be minor.

In other words, the situation when 50 sources say oranges are rich in vitamin C, and 2 sources say they are not may fall under DUE; the situation when 50 sources say oranges are rich in vitamin C, and 2 sources say apples are rich in vitamin C may fall under MINORASPECTS.

Therefore, "on a topic" (or its equivalent) may be useful in DUE. Paul Siebert ( talk) 17:25, 26 May 2023 (UTC)

That whole section is only usable as vague general guidance. It's not operatively usable (except in infinite ways by wikilawyers.) E.G., using the example, it's practically undo-able to set the criteria for inclusion of a source with the opinion on oranges, search the world, and find that there are exactly 52 of them, then interpret their responses in the context of the wiki-question and then tally them up. IMO someday that section will need to get replaced with something better. So I don't think that the discussed word change will have much impact either way. The reason I lean a bit towards including "on a topic" is that this section is more usable (and probably originally intended)) only as a "balancing" measure (= on a specific topic or question) rather that as an inclusion mandate along the lines of the point/concern introduced and explained by Jerome Frank Disciple. North8000 ( talk) 18:02, 26 May 2023 (UTC)

Agree with this and describing that DUE is dealing with the situations when different viewpoints exist on the same subtopic/aspect (and these viewpoints contradict to each other), whereas MINORASPESTS explains what we should do when the sources complement each other, although one of the aspects may be minor., but I wish it'd be communicated with a different wording than the proposed. "on a topic" is way too vague and I think it conflicts with use of "a topic" in other parts of the policy. For example, "a topic" refers to the topic of an article in WP:POVNAMING. Subtopic is better; WP:NPOVVIEW refers to is as a "facet", WP:YESPOV calls it a "matter". A footnote might be a solution as well. Otherwise, the change won't solve what it intends to solve, with editors still being confused about what "a topic" means in the context of DUE. PaulT2022 ( talk) 19:11, 26 May 2023 (UTC)

Maybe, another solution could be to add the footnotes to DUE and MINORASPECTS that explain it? A tentative wording may be the sentence quoted by PaulT2022. Paul Siebert ( talk) 19:52, 26 May 2023 (UTC)

One of the problems I have found in topics where most sources are news coverage is that the media tend to cover exceptional rather than routine events. For example, man bites dog is more likely to be covered than dog bites man, because the first happens so rarely. Establishing weight in this case using media coverage, articles would imply that the first happens more commonly.

This becomes a problem with subjects that are marginally notable. A city councillor for example was arrested for possession of child pornography, which later turned out to be unfounded. So most of his article was about the charges rather than his political and professional career.

Is there anything in policy that addresses this?

TFD ( talk) 14:19, 19 December 2022 (UTC)

• Somewhat addressed at WP:RECENTISM, and hinted at in WP:NOTNEWS. Blueboar ( talk) 14:30, 19 December 2022 (UTC)

  Taking an example of the last few weeks of Twitter and how fast and as many twists that has taken, I feel we need to revisit our policies around news to caution editors that some stories should just wait until better resolution is known rather than jumping on a burst of coverage. Masem ( t) 14:36, 19 December 2022 (UTC)

• This problem arises by deciding WP:DUE based on "mentions of the subject" rather than based on "top sources about the subject". Levivich ( talk) 14:36, 19 December 2022 (UTC)

  DUE desperately needs a statement that nearness to event must be taken in account when figuring weight. Masem ( t) 14:38, 19 December 2022 (UTC)

• No problem, no need to address any problem. This is coming out of The Daily Caller article where editors have decided to clean it up to protect that outlet's reputation from association with white supremacists in the past. It has nothing to do with recentism or notnews in this case. Andre 🚐 16:12, 19 December 2022 (UTC)

• The fact that it was the Daily Caller article that inspired the question does not mean the question is invalid. I see this in many bio articles and articles on current events (especially politics). I think it is worth discussing further. Blueboar ( talk) 17:00, 19 December 2022 (UTC)

That comment was not cool, Andrevan. Levivich ( talk) 17:33, 19 December 2022 (UTC)

Well, I'm sorry if you took offense, but I believe that is basically the discussion that's going on Andre 🚐 17:47, 19 December 2022 (UTC)

It's not just recentism. A deeper fix is needed. That would be to also factor in the degree of enclyclopecicness of the content in inclusion decisions. If somebody fell down on stage while discussing a complex economic policy, the majority of the media coverage will be about them falling down. And if POV warriors don't like him, using wp:due, that means that the Wikipedia article will cover them falling down but not the economics content.North8000 ( talk) 17:22, 19 December 2022 (UTC)

Not even just POV warriors, but editors with good intentions that seem to take "lots of news coverage" equals ripe for inclusion. Understanding when massive news coverage equates to encyclopedic content that meets all content policies is a skill of art and does require more scrutiny of one's own edits to make sure we aren't capturing the tone of the media as well. Masem ( t) 17:54, 19 December 2022 (UTC)

There are several good points in this discussion. What you're pointing to is a subtle but important point that's at risk of getting lost among the rest: A source's interest may be in information that's not what is most appropriate to an encyclopedia. A lot of news outlets love scandal that drives clicks. That choice of emphasis may say more about the source than it says about the topic. Sennalen ( talk) 14:15, 21 December 2022 (UTC)

I found Andre's comments about differences between primary and secondary coverage in news media in the #News sources discussion above very helpful (TL;DR: not every press article is secondary). I think it'd cover a lot of issues if primary vs secondary would be seen more critically by editors (i.e. 10 reports about someone falling on a stage in the example above would be primary). PaulT2022 ( talk) 18:00, 19 December 2022 (UTC)

Good point, and thanks. Andre 🚐 18:02, 19 December 2022 (UTC)

Another "fix" is to take wp:due in what is probably its intended context. Which is (only) how to cover the "sides" of particular topic within an article. Rather than how to totally exclude coverage of an encyclopedic topic within an article.North8000 ( talk) 18:26, 19 December 2022 (UTC)

We might need (in this policy) that says something like "You have to provide a basic summary of the subject, even if the sources really really really dive into this totally niche detail."

In the example of a falsely accused public figure, we'd normally yank all of the unfounded information out, so that the article goes from "He was elected and accused of crimes" to "He was elected". But sometimes you'd want to acknowledge it, so that the article goes from "He was elected and accused of crimes" to "He was elected and became the victim of false accusations".

In other cases, you have sources going on at great length about things that never need to be mentioned in an encyclopedic summary. Wedding dress of Lady Diana Spencer is definitely a notable subject, so I'm not saying that all clothing worn on a particular day isn't worth mentioning somewhere in Wikipedia, but you can find sources for who wore what to the Met Gala every single year, and what we need in an encyclopedia is more like "It's a fundraiser" than "Chris Celebrity once wore flip flops to the Met Gala". A mathematically precise replication of what was written about the subject, even if restricted to likely sources (e.g., magazines rather than social media), is not going to produce an encyclopedia article. WhatamIdoing ( talk) 03:29, 20 December 2022 (UTC)

I pretty much agree with WAID and N8000. Thanks to TFD for bringing this up. Huggums537 ( talk) 16:51, 6 January 2023 (UTC)

Prominence is not frequency

Perhaps DUE could be worded to say that weight of viewpoints shouldn't be evaluated merely by counting sources? Just some general guidance that "prominence" should not be confused with "frequency". Frequency could establish prominence, but it's dangerous to automatically assume it always does, as evident from the examples above. PaulT2022 ( talk) 03:21, 21 December 2022 (UTC)

I agree, prominence needs to take into account reliability and weight. Andre 🚐 04:09, 21 December 2022 (UTC)

I agree there's a problem here. I want to be careful that any fix doesn't create more side effects, so it would be safest to start with the low hanging fruit. For example, one source trying to squeeze twenty articles from the same story isn't the same as a story covered by twenty sources. Shooterwalker ( talk) 15:33, 23 December 2022 (UTC)

That's already covered by policy I believe. Andre 🚐 18:21, 23 December 2022 (UTC)

I'm pretty sure that @ Colin has talked about this problem in the past.

Do we need something that says "You have to write an encyclopedia article"?

Not only do we have to write an encyclopedia article, we are normally doing this from non-encyclopedic sources. Even if all of the sources are scandal sheets and trashy tabloids, the resulting Wikipedia article must be WP:NOTGOSSIP and WP:NOTSCANDAL. Even if all the sources are promotional, the resulting Wikipedia article must be WP:NOTPROMOTION. Even if all the sources are instruction manuals and textbooks, the resulting Wikipedia article must be WP:NOTHOWTO. Even if all the sources are up-to-the-second breaking news, the resulting Wikipedia article must be WP:NOTDIARY and WP:NOTNEWS. This means that some facts that are in reliable sources will not be suitable for inclusion in the resulting Wikipedia article, or will need to be given much less emphasis, prominence, and detail than they were given in the non-encyclopedic sources. WhatamIdoing ( talk) 20:08, 5 January 2023 (UTC)

I'm not totally sure which issue WAID is referring to but I agree with some of the comments in this overall section that heavy use of news media coverage seems to not be working well with DUE. Editors are saying "I've got a handful of reliable sources covering this factoid so therefore we must insert it". All those "reliable sources" have an agenda to push on some political issue, either at the level of newspaper editorial policy, or the journalist's own record. More neutral media outlets or those with an opposite agenda simply didn't cover it at all. But when you Google for "X was accused of Y", say, Google doesn't list all the equivalent sources that chose not to cover that story. So it is hard to determine what the weight of reliable sources is saying.

We weigh using reliable sources, but reliability doesn't mean neutral. If your Google search of reliable sources only turns up sources with one agenda, then it may seem that agenda is important.

FWIW, I've seen this recently with some of the articles on transgender, where we have otherwise "reliable sources" like The Telegraph publishing trans hostile articles daily, sometimes several times a day. It is clearly out of all proportion to the tiny number of people who are trans or their impact on most people's lives, but is a very hot topic politically. How then do we stop DUE making Wikipedia just as trans hostile as The Telegraph when it comes to weight? (If your politics swing the other way, I'm sure you can think of alternative examples). I am wondering whether in addition to assessing if a source is reliable, we take account its bias (at any level, publisher, editor, journalist, etc) and consider that too when assessing weight. -- Colin° ^Talk 21:16, 5 January 2023 (UTC)

I don't agree that most reliable sources have an agenda. Most reliable sources want to sell papers, get clicks, and monetize eyeballs. Thus clickbait headlines. Beyond that most journalists have a commitment to accuracy and fact-checking and getting the news out. We have stuff like WP:NEWSORG. WP:RSP has carve-outs for opinionated sources that might be reliable but should be attributed for WP:RSOPINION. I do not think we should consider bias when assessing weight other than requiring certain biased sources to be opinionated by consensus, which naturally eliminates most of their weight. It opens a significant can of worms. Andre 🚐 21:56, 5 January 2023 (UTC)

I think Colin has it right. But it's a very hard problem to address in the WP:PAGs. Sometimes I wonder if news sourcing should not count for establishing WP:WEIGHT. It would make Wikipedia less contemporary in what it covered, but that would be a good thing IMO. Too much drama on the English Wikipedia is generated by politically POV editors engaged in warfare over current political topics, using news sources as their cudgels. If instead these editors instead went down the library and starting reading books and journals to think about what content could be made from the knowledge therein, it would be a win all round. Bon courage ( talk) 07:30, 6 January 2023 (UTC)

Andre's opening sentence conflates "reliable sources" with "newspapers that we regard as reliable sources", which is the focus here. I didn't say all newspaper sources have an agenda, I said that all the ones that POV-pushing editors will find with their google search have an agenda (at some level, journalist, editor, publisher), because it is the agenda that makes that news item newsworthy, and nothing else. As soon as you have enough such sources, the pushers claim they have weight. We do have advice in policy about not including every little news story that may not have lasting significance, etc, etc, but it is difficult to get them to be accepted. I see this attitude on all sides of controversial topics, not just one, and some articles become a list of things folk found in the news.

In the UK at least, the press is strongly political and journalists are expected to biased. Of course there are topics where the journalist do a fairly neutral job but there are plenty articles where the whole point of a story is to continue the agenda of that journalist or newspaper. If one's only sources are such articles, then Wikipedia ends up becoming the mouthpiece of activists and political lobby groups. In the UK, the only mainstream news source that is required to be neutral is the BBC (and ITV/C4 news), though for political stories that tends towards the false-balance trap of always giving the UK government the last word.

I think this is a difficult problem if we rely on an algorithmic mechanism like WEIGHT because there is no way to put on the other side of the balance all the potential sources who chose not to cover that story. We aren't weighing a variety of comprehensive textbooks on topic X to see how they all cover subtopic Y. With strictly biomedical matters, we have MEDRS, which declares newspapers unreliable, so we are lucky that also eliminates them for WEIGHT purposes when they end up giving platforms to cranks and fraudsters and celebrity-no-nothings. If a neutral and subject-knowlegable editor sat down, as WAID suggests, and considered how to "write an encyclopedia article" they would naturally dismiss many of the press stories as distracting agenda promoting nonsense. But if you are an editor who hates the subject, then distracting agenda promoting nonsense is exactly what you want included.

I don't think we can disregard newspapers entirely for news, because many subjects are only covered at present by such stories. I do wonder whether for certain political hot topics, we might require editors to demonstrate weight using a variety of sources of differing bias and not just those clearly biased against (or for) an agenda. -- Colin° ^Talk 10:13, 6 January 2023 (UTC)

Agree with Colin & Bon courage here. It's hard to think of a realistic policy fix, but Masem's DUE desperately needs a statement that nearness to event must be taken in account when figuring weight and Bon courage's I wonder if news sourcing should not count for establishing WP:WEIGHT seem most promising. Have these been offered for discussion at WP:VPI yet? DFlhb ( talk) 17:29, 10 April 2023 (UTC)

I think Colin and Andre are both right.

Most mainstream media probably don't have an agenda beyond pandering to the views of their audience - which they do extensively. Not by inventing facts, but by shifting the amount of coverage and the angle (as in who gets the last word in the article). In the UK, there's a whole spectrum of reliable news publications to match every worldview from Guardian to Spectator. Anecdotally, I have made a few successful investments by using Bloomberg to understand what's priced into the market and what isn't - because its coverage tends to bias so strongly towards what its audience believed in yesterday (rather than presenting facts neutrally), that it can actually be used as a signal to measure this.

Letting editors to decide biases of sources and account for them wouldn't improve anything - if the problems are exacerbated by editors' own biases in the first place.

Deciding DUE/WEIGHT from how often a certain news or a POV was reported in the news media leads to articles shifted in the non-encyclopedic direction, both POV-wise (per Colin), and style-wise (per WAID).

Personally, I would support any change that decreases mechanistic influence of amount of coverage in news sources on weight. (I don't think there's a realistic prospect of achieving consensus to exclude them entirely for all topic areas though.) PaulT2022 ( talk) 18:23, 6 January 2023 (UTC)

Maybe we just need some kind of hint written into the policy and leave folks to argue how strongly to take that hint. It isn't easy though and we'd have to watch it didn't end up excluding or diminishing minority groups simply for being minority groups standing up for their minority.

PaulT2022, you forget that Wikipedia exclude "most mainstream media" in the UK: Daily Mail, Sun, Mirror,... all well known for "inventing facts". Goodness, we even had a prime minister who was well known for "inventing facts" which were duly repeated as factual by sympathetic source we think of as "reliable". And know it is quite possible to deliberately indicate things that aren't so without actually lying. -- Colin° ^Talk 19:07, 6 January 2023 (UTC)

I think that most reliable sources have "an agenda", if you understand that concept broadly. For example, the agenda for most scientific papers is "I hope publishing this will advance my career and promote my ideas within my field". The agenda for most magazines is "We will stay in business by publishing what our subscribers want to read, even if that's not what's most important". These agendas do not necessarily harm Wikipedia articles. WhatamIdoing ( talk) 00:27, 12 January 2023 (UTC)

How does the latter not "harm Wikipedia" if that newspaper's subscribers are unrepresentative and we count that newspaper's weight the same as anyone else? It isn't like there are sufficient newspapers covering all possible sets of subscribers that the whole thing balances out. In the UK, the press is overwhelmingly right-wing. Anyway, I think this concept of a newspaper operating like a listed company tasked with doing whatever is necessary to make most profit through selling the most papers isn't valid, certainly not in the UK or perhaps anywhere that a newspaper is privately owned by an individual. It is well known in the UK that most newspapers push the political agenda of their owners and titles are not always profitable. They don't follow the views of their readers; the job of many newspapers is to instil views in those readers. Perhaps at the level of core values the paper cannot shift readers much, but at the level of individual issues, the reader is very much told what they should care about through relentless coverage of things that poll very very low in most people's concerns.

I think where it becomes most unstuck is when a newspaper, an editor or a journalist has a very specific agenda about some political hot topic. For example, whether the journalist focuses on the disruption and potential harm (to health or businesses) that strikes cause or focuses on the dire state of these businesses or services and how they treat and pay their employees. Whether the they focus on the concerns of trans people, or on the concerns of a minority sect of feminists. Whether they focus on refugees from war and oppression and hatred, or talk only of "asylum seekers" taking "illegal routes" to the UK. And so on. The weight given to one aspect or another is entirely political and agenda based in UK newspapers, with no attempt or desire whatsoever to find any kind of balance or neutrality. I think that has consequences if we are naïve about weight.

Plus I still don't think we have a solution to the problem that newspapers who don't consider a story worthy to report don't get counted when discussing weight. We are not comparing how comprehensive source A deals with topic X vs how comprehensive source B deals with topic X, which is where weight works best. We are dealing with biased cherry picking sources A, B and C highlighting topic X and biased cherry picking sources D, E and F ignoring topic X. We have no way of knowing if there are also G, H, I, J and K also ignoring it, and only A, B and C turn up on any google search. -- Colin° ^Talk 10:31, 12 January 2023 (UTC)

The first duty of a newspaper is to stay in business. If you don't do that, nothing else matters. You can't break the important news of the day if the newspaper's closed. This might be less true for those with billionaires using them as propaganda machines, but it's still very true for smaller and locally owned papers, as well as for many magazines.

Of course newspaper readers are unrepresentative of society as a whole. For example, they tend not to represent the ~15% of US adults who can't read well enough to bother with subscribing to a newspaper. They also tend not to cater to the interests of people who are too poor to subscribe. But the goal isn't really to fairly and proportionately represent the views of humans. If we did that, then we'd be writing that ghosts are real, that birds will die if they eat dried rice, and that it's the mother's fault if she has daughters instead of sons. We are trying to fairly and proportionately represent the knowledge contained in sources, rather than to fairly and proportionately represent the things people believe.

When it comes to questions of individual words and phrases, I think we are often best using formal, encyclopedic language, which tends to be non-judgmental but is not afraid of "labels" (only, we might call it "classification", because that's that formal term for that). We should be wary of following the advice of parenting experts in Wikipedia articles ("Oh, dear, you mustn't hurt his poor feelings by saying he's a bad boy! Just say that he made a bad choice instead"), and instead cheerfully say that some people are right-wing and some are left-wing and some are best classified in other ways, or whose classification is unclear; that there exist things that are scientific and things that are non-scientific, and also things that are pseudoscientific, and it's reasonable to know which is which, when that's clear (and not just because certain sources like to use pseudoscience as a smear word that means 'thing I disagree with'); and that some people would prefer that their actual, avowed beliefs and actions not be described in the standard terms for that concept ("I deny being an AIDS denialist! I'm an extremely brave AIDS dissident, standing up for my right to contribute to the spread of disease and death by promoting misinformation!"). WhatamIdoing ( talk) 04:46, 15 January 2023 (UTC)

Sorry for joining this discussion late. I am wondering if the footnote "c" ( The relative prominence of each viewpoint among Wikipedia editors or the general public is irrelevant and should not be considered.") explains that. I think coverage of some topic in mass-media also fits "prominence among general public" criteria. In that sense, if one viewpoint has been expressed in 5 reliable scholarly publications and was extensively covered in various mass-media, and another viewpoint was described in 4 scholarly publications but not in mass-media, these two viewpoints have the same weight and should be treated as such.

However, if the footnote "c" cannot be interpreted in that way, maybe, we should make it more explicit?

In addition, do I understand it correctly that the footnote "c" explicitly prohibits us to resolve neutrality disputes by majority of Wikipedians' "votes"? Paul Siebert ( talk) 02:07, 27 May 2023 (UTC)

Support making it more explicit. I think it'd have to be done in fairly general terms as there's no clear agreement in this thread what would impact prominence, but there seems to be a consensus that it shouldn't be done by counting alone (which is echoed in the existing footnote). Different factors affecting relative prominence were listed (nearness of the coverage to the event, amount of coverage in news etc), and I suspect they need to be applied depending on common sense and topic area. PaulT2022 ( talk) 03:30, 27 May 2023 (UTC)

Prominence and coverage

I apologise in advance if that has already been discussed, but I have a feeling that the above discussion hadn't led to any adjustment of the policy, and some users still do not discriminate between "prominence" and "coverage". I am frequently facing a situation when one fact found in a single source is being repeated and massively reproduced by a large number of secondary sources, which either directly cite this source or cite each other. The most blatant example was the "NKVD gas van" story: the claim that the Gas van was ostensibly invented by Soviet NKVD. This fact was taken from a single publication from one Russian tabloid, but it was reproduced by more than ten secondary sources, including the book by a Nobel prize winner Alexandr Solzhenitsyn. For a long time, all those sources were used in that article as independent confirmations of that story.

I think, to prevent this type situations in future, it is necessary to explain in the policy that, to determine relative weight, we have to take into account only independent sources (derivative sources should not be considered as independent sources): thus, if some number was published in the article by a scholar X, the publication in a newspaper Y or in an encyclopedia Z (both of them just take that figure from X) should not be taken into account when a relative weight is being determined. -- Paul Siebert ( talk) 02:58, 27 May 2023 (UTC)

I found Andre's comments in Wikipedia_talk:Neutral_point_of_view/Archive_60#News_sources very insightful. A TL;DR version is in the WP:OR: A secondary source... contains analysis, evaluation, interpretation, or synthesis. A simple reproduction doesn't make a source secondary. WP:BALANCE already requires to use secondary and tertiary sources. PaulT2022 ( talk) 03:16, 27 May 2023 (UTC)

Please add links to the articles you can think of that are most often associated with contention as related to NPOV.

Controversial subjects

1. Fringe theories
   1. Fringe theories,
2. Economics/Money
   1. Economics, Capitalism, Communism, Socialism, Social democracy, Trickle-down economics, Supply-side economics
3. Fact vs Opinion
   1. Wikipedia:Facts precede opinions (essay),
4. History
   1. History, Historical revisionism, Pseudohistory
5. Information
   1. Information warfare, Propaganda, Conspiracy theories, Disinformation, Firehose of falsehood
6. Legal disputes
   1. Law,
7. Politics
   1. Politics,
8. Pseudoscience
   1. Pseudoscience,
9. Religion
   1. Religion, Atheism, Agnosticism, Religious violence, Freedom of religion, Separation of church and state
10. Sexuality
    1. LGBT, Transgender, Gender dysphoria, Legal status of transgender people, Transgender discrimination
11. Social justice
    1. Civil disobedience, Revolution

Valjean ( talk) ( PING me) 22:31, 10 April 2023 (UTC)

I suspect that you want to look at difficult cases in different categories and try to build a specific interpretation of NPOV for each category. The first issue is that the interpretation of NPOV within a category would still not fix the decision for each case in the category. It would still have to be interpreted in each individual case. Therefore, the goal in itself is far from being a full solution to the application of NPOV. The second issue that any solution for a particular case will only reflect the perspective that prevails among the wikipedians that worked to resolve this contentious NPOV case. Trying to extract an interpretation of NPOV from a few cases like that within a category will be yet another challenge. After all this work, the result would still be relative to the perspective of the wikipedians that worked on this. It would not be an objective interpretation. In fact, I worry that, if the interpretation is accepted as a permanent principle, it would only have crystallized a bias that exists among a few wikipedians. You might have in mind that a very large group representative of the global community would work on each case and each category : that would explain why we pick a few contentious cases only in each category. That is a lot to ask and seems unrealistic, but even then it would represent the bias of the community at the time and could not be accepted as permanent : only the basic principles are non negotiable. It would be an ongoing work. To conclude, this approach requires an unrealistic participation of the community only for a partial solution that would be an ongoing work. It is an interesting idea, but there is a simpler approach. Let us just focus on a single ingredient that is required in your approach: have the community learn to apply the basic principle of NPOV in each individual case, without worry about categories and intermediary interpretations for these categories. The point is that the basic principle is clear enough in itself and there is no need to write down intermediary interpretations for different categories. Dominic Mayers ( talk) 00:45, 11 April 2023 (UTC)

I agree with your last point in the previous section, so am not going to go much further with this. I'm interested in finding the controversial articles so I can search their archives and related dramaboard decisions. I don't think that can apply to other topics or that we can even begin to do any type of full justice to the matter. At this point, I'd just like to do some research. -- Valjean ( talk) ( PING me) 01:05, 11 April 2023 (UTC)

WP:ACDST for the contentious topics on Wikipedia currently. Bon courage ( talk) 05:33, 11 April 2023 (UTC)

Thanks! The work is already done for me. -- Valjean ( talk) ( PING me) 05:43, 11 April 2023 (UTC)

I think you're missing a large section of articles by going for the obvious... IMO the areas with the most pervasive NPOV issues are transportation, sports, and celebrities (particularly musicians) because in those areas more than others articles are created by fans. Horse Eye's Back ( talk) 16:11, 23 May 2023 (UTC)

• Another take -- Not sure if looking for NPOV or controversy here, and whether this is wanting areas or specific articles. I like specifying these areas and I can seek "most often" by looking for what frequently comes up WP:Contentious and WP:Contentious topics in the WP:NPOV link to NPOV noticeboard, or the WP:LCI list and links Wikipedia:Controversial articles and Category:Wikipedia controversial topics. In addition to "most often" re that 'most often associated with contention as related to NPOV', I might also suggest looking at the causes. It seems to me I often seem issues from subjective values (computer game TALK was the rudest I've seen), wanting only one POV (just the fact and no POV, or only the 'best' side), or wanting a particular non-neutral wording. Cheers Markbassett ( talk) 19:21, 5 June 2023 (UTC)

An article should not give undue weight to minor aspects of its subject but should strive to treat each aspect with a weight proportional to its treatment in the body of reliable, published material on the subject.

This term is defined nowhere on the page, and I've seen users misinterpreting this policy and trying to apply it to articles where I believe neutrality is not a concern.

As far as I understand this whole policy, it, and everything written in it, is applicable only if there are alternative viewpoints on a topic present in different, reliable sources. And that would apply to the term "aspect" as well, whatever the hell it's supposed to mean. Can someone clarify on this point with examples if possible? Aberration ( talk) 10:09, 13 June 2023 (UTC)

• You can find the meaning of the word “aspect” in any standard English language dictionary. Blueboar ( talk) 11:27, 13 June 2023 (UTC)