Wikipedia talk:Neutral point of view/Archive 66

Instructing the reader

I feel pretty confident we ourselves as editors shouldn't be calling out attention to logical fallacies or incongruities, nor specifically instructing readers to notice any such things (e.g. literally saying in an article, "This cited claim by [subject of article] should be evaluated by the reader while considering this other cited claim from earlier in the article"). That's the editor inappropriately injecting themselves into the prose, right? We state the facts, cite the sources, and allow readers to determine the meaning behind such mismatches of cited facts. Does that track? — Fourthords | =Λ= | 01:04, 17 November 2023 (UTC)

That would be WP:SYNTH (synthesis of two claims) unless there is a source that makes the connection. Sennalen (talk) 02:10, 17 November 2023 (UTC)

SYNTH! That's what I was trying to think of, yes! Not a part of that policy I've had to reach for often. Thanks so much! — Fourthords | =Λ= | 02:29, 17 November 2023 (UTC)

Yes, we should avoid that, but it can be appropriate if there is something that is frequently misunderstood by non-English readers. -- Valjean (talk) (PING me) 19:21, 17 November 2023 (UTC)

Or even by English speakers. The List of common misconceptions is long. However, the example above goes well beyond what is reasonable. "He said X (quick definition of X) on Monday" is usually okay. "He said X on Monday and Y on Tuesday" might be okay. "Most cancer can be prevented through lifestyle changes, but you should decide whether that's true after you read the paragraph in the lead that says most cancer can't be prevented through lifestyle changes" is not okay. That's a sign that an editor doesn't know about the {{contradictory inline}} tag. WhatamIdoing (talk) 18:52, 22 November 2023 (UTC)

Giving "equal validity" can create a false balance

For Wikipedia:Neutral point of view#Giving "equal validity" can create a false balance

"Keep in mind that neutrality is not the average between bollocks and reality. In science, any compromise between a correct statement and an incorrect statement is an incorrect statement. We do not post erroneous statements here without contextualizing why they are wrong. Wikipedia becomes a source of disinformation if we fail to do that."

The first two sentences are from User:JzG.

Since this has been deleted by an editor who generally opposes everything I do here, I'd like to see what others think. -- Valjean (talk) (PING me) 04:34, 28 November 2023 (UTC)

As a core content policy page, that language is a bit too curt and off the cuff as official policy language. It immediately moves to dismiss anyone that comes to a discussion in good faith with a fringe viewpoint as unwanted. (granted the fraction of editors coming with this type of argument in good faith in comparison to those coming with bad faith is very small, but we still don't want to dismiss them like that). What's already in the false balance section is sufficient for policy. What's suggested is far better at a guideline or essay (say, perhaps WP:YESBIAS). Masem (t) 04:59, 28 November 2023 (UTC)

I don't think it's necessary to include this particular text in this policy.

Also, it's very science-specific, and most of our articles aren't about science. You can make the mistake of giving equal validity to unequal views in non-science subjects without producing any sort of disinformation. Consider "Most scholars believe the French play is about the recently concluded war, but some Soviet writers interpreted it as a metaphor for class struggle." You don't want to treat the views as being equal in popularity, because they're not, but if you did, the result wouldn't be incorrect statements or disinformation; the result would, at most, be the reader thinking the minority viewpoint was a bigger deal than it actually is. WhatamIdoing (talk) 05:17, 28 November 2023 (UTC)

Yes, we want to give due weight where it is due. I agree that not everything is cut-and-dried science right/wrong stuff, yet we do have many subjects outside of science where false/fact is an issue, so maybe tweak this? -- Valjean (talk) (PING me) 07:19, 28 November 2023 (UTC)

yet we do have many subjects outside of science where false/fact is an issue Agree

If language is a bit too curt and off the cuff as official policy language then should be change not removed. I think idea is good and false balance is a problem in lot of articles. YESBIAS page is good but so is short version because most editors dont read every page Softlem (talk) 11:45, 28 November 2023 (UTC)

Softlem is right: if people think it's too curt, then change the wording, don't remove the text, because right now a subset of people in the world appear to believe that they are entitled to their own facts. Guy (help! - typo?) 13:58, 28 November 2023 (UTC)

• I just see very little value in it compared to what is existing in the section. Yes it is poorly written and that could be improved but that does not mean it belongs in this policy page. As other and myself have suggested, such rhetoric is better off in an essay. It also promotes a misinformed view that life is binary and it simply is not. As an aside, statements like an editor who generally opposes everything I do here are not helpful, don't personalize content disputes. PackMecEng (talk) 14:27, 28 November 2023 (UTC)

It "promotes" no such thing. It's okay to address one side of an issue. Some things really are binary. We just disagree on things like whether Trump really lost the 2020 election and whether Trump and his campaign really did cooperate with Russian interference in the U.S. elections. There should be no doubt in any editor's mind about those things. The evidence is incontrovertible. With all the evidence we have, those are now, and nearly always have been, binary matters for those who had the evidence we have here as editors. -- Valjean (talk) (PING me) 17:20, 28 November 2023 (UTC)

The issue tends to be on topics where there is yet no confirmed "correct" answer but WP editors have decided that their view which may be shared by popular media must be the "correct" one, and thus write that as fact and push any other possible view out as fringe. There are things that do not fall into a clear binary distinction between correct and incorrect (the current debate around the situation in Gaza a prime example), but the suggested wording suggests that all such debates fall into this. I know that "in science..." predicates this sentence, implying this should only hold for scientific topics, but the implications are written to apply across the board.

If we want to say that we do not try to equate wacky claims of pseudoscience against established theories (such as "some say the earth is round while others say the earth is flat." Then absolutely let's focus on the issue of promoting scientific disinformation. But in other areas like politics, there a lot more grey area that we should be careful in trying state things as black and white. The suggested text pushes the black and white approach with considering nuances. Masem (t) 19:02, 28 November 2023 (UTC)

Actually, no, I think the worst examples are on subjects where there is a correct answer, and one, deeply entrenched, side, really doesn't like it. Look at the length and ferocity of disputes over climate change, which is objectively happening. Guy (help! - typo?) 16:24, 1 December 2023 (UTC)

Climate change is obviously an area that there is clear scientific agreement it is happening and attributable to human progress (all as fact), so when we have someone trying to promote a view that dismisses climate change and/or the human impact of that, we clearly should not give that equal validity. That's the same as the flat earth case. We should be treating any theory that dismisses climate change and its human origin as pseudoscience.

On the other hand, the origin of the COVID virus is yet well known, and while the prevailing scientific consensus is a natural origin, that is not yet treated as fact, so we should be careful in writing to treat the natural origin as "fact" and dismissing the lab leak theory as pseudoscience. Of course, there's a lot less support of the lab origin theory, and so treating both of these as "equal" is still wrong. Masem (t) 17:39, 1 December 2023 (UTC)

The possibility of a lab leak is a complex case, because there is a minority scientific view that a lab leak was the possibile origine of COVID and a much larger and more widely reported conspiracy thoery that the original of COVID was a lab leak. One should be given as a minority opinion in the scientific literature, and the other handled as any other conspiracy theory. The disconnect Wikipedia's articles can sometimes have is not showing the minority view, because the loudest voices calling for inclusion want to add conspiracy theory ideas as being equal to the majority scientific consensus of zoonotic origin. -- LCU ActivelyDisinterested «@» °∆t° 17:59, 1 December 2023 (UTC)

Thats not entirely correct. The more widely reported conspiracy theories are that it was an intentional release and/or the result of biological engineering/bioweapons programs, I think you're getting those confused with the lab leak theories (even if there is some overlap in the biological engineering/bioweapons program conspiracy theories). Horse Eye's Back (talk) 18:27, 1 December 2023 (UTC)

These lots of Soros fund the lab and China covered up a release (not saying they wouldn't, but that they have is baseless) nonsense going on out there. -- LCU ActivelyDisinterested «@» °∆t° 18:50, 1 December 2023 (UTC)

There's also a whole series of conspiracies propagated by the Chinese government that it was a release/leak from American military facilities and that the whole idea of a Chinese origin is an American coverup. So much baseless nonsense thats its almost impossible for any one person to keep track of it all. Horse Eye's Back (talk) 19:16, 1 December 2023 (UTC)

That was my point the conspiracy theories visibility far outweighs any actual minority view. -- LCU ActivelyDisinterested «@» °∆t° 20:33, 1 December 2023 (UTC)

Is that true though? Its true within the larger COVID space, but with the lab leak space its the scientific minority view which appears to have more coverage. Horse Eye's Back (talk) 20:42, 1 December 2023 (UTC)

It could be that we're each just describing our own observation bias, something we won't be able to clarify in this discussion. -- LCU ActivelyDisinterested «@» °∆t° 14:42, 2 December 2023 (UTC)

Very true, someone who spent most of their time reading the scientific literature would come away with one observation and someone who spent most of their time reading the fringe literature would come away with another. Perhaps this is why our Skeptic faction consistently over-emphasizes the fringe and de-emphasizes the scientific while most editors barely even notice the fringe. Horse Eye's Back (talk) 16:06, 2 December 2023 (UTC)

As someone who isn't a Skeptic I would refer that statement back to my comment about observation bias. -- LCU ActivelyDisinterested «@» °∆t° 16:14, 2 December 2023 (UTC)

• If anything it isn't strong enough, giving equal validity almost always results in a false balance because very very rarely do two sides actually have equal weight. Horse Eye's Back (talk) 17:27, 28 November 2023 (UTC)

  Agreed, rarely should two viewpoints be given equal weight, but that is already covered in the section. PackMecEng (talk) 17:46, 28 November 2023 (UTC)

As a sidebar, to flatly state that a statement is correct or incorrect means that it has a single explicit meaning and that that in that context it is clearly known to be (always) true or (always) false. Such statements are extremely rare. In reality, most statements have vagueness, variable meanings, spin, leave impressions of meaning etc.. For example, what exactly is the meaning of the above "Trump and his campaign really did cooperate with Russian interference in the U.S. elections." North8000 (talk) 18:34, 28 November 2023 (UTC)

Right. Who counts as "his campaign"? What counts as "cooperation", and how does one differentiate that from things like "benefited from"? What's the difference between "interference" and other related things, like "influence"?

But while acknowledging that sometimes statements are vague or could be interpreted in somewhat different ways, we also don't want to push editors into "if it's not plagiarism, it's original research". I saw an editor recently going through some sources and claiming that the sources didn't truly support the sentence because there weren't soundbite-sized quotations that support the exact words in a sentence, even though we all know that it's true, we all know that the (multiple) cited sources were actually saying/meaning what's written in the article, and we all know that we have an entire article about the subject of that sentence. But, hey: if you can't find a source containing a quotable sentence that says the same thing, in almost exactly the same words, then some of us can and will try to wikilawyer that information out of the articles. It wasn't really "his campaign", since most of the work wasn't campaign employees; they didn't really "cooperate" so much as "coordinated with"; it wasn't really "interference" so much as "exploration of social media"... WhatamIdoing (talk) 01:25, 29 November 2023 (UTC)

The answer to that is in the Mueller report. He welcomed offers of assistance, and gave internal polling data to the GRU. But it's not a particularly useful example.

Unfortunately, to quote Isa Blagden, "If a lie is only printed often enough, it becomes a quasi-truth, and if such a truth is repeated often enough, it becomes an article of belief, a dogma, and men will die for it".

We have the unenviable task of holding back the hordes of people who will die for lies ranging from the age of the Earth to who won the election in 2020. Virtually everything now has been turned into an article of faith, by cynical bastards who know that people don't understand climate change, but if you tell them it goes against God, they will pick up their AR-15 and march on whatever passes for Fort Sumter in the culture war. Probably CNN head office. Guy (help! - typo?) 16:22, 1 December 2023 (UTC)

So my point is that most statements are are imprecise and thus have multiple possible meanings. Calling it categorically "true" or "false" based on one's preferred selection of one of those meanings is itself flawed, and a common POV warrior tactic. North8000 (talk) 22:14, 1 December 2023 (UTC)

This is why in text attribution is so vital. When dealing with topics that are “conspiracy theory adjacent”, a proper balance can be best achieved by including who says what. Blueboar (talk) 14:59, 2 December 2023 (UTC)

What I've long argued for in most cases. Unless we are talking about a clearly established fact like the earth being round or the existence of global warming, there is rarely harm in adding attribution to a statement that can be taken as contentious. Masem (t) 15:02, 2 December 2023 (UTC)

Yes. But also to not place much value on somebody's evaluation of multiple meaning statement. Here's a real wiki-world example (with the name changed) I think that science says that (vaguely speaking) that masks reduce the spread of covid. Which means that sometimes they stop it and sometimes they don't. So now John Smith says "masks don't prevent the spread of covid." . This can have these meanings:

1. "Prevent" means "always stops it". Then, according to science, this John's statement would be true / accurate
2. "Don't prevent" means never prevents the transmission. This is a false statement and against science.

Now a WP:RS publication which is a political opponent of John Smith that wants to make him look bad says that John Smith has made a statement contrary to science. Their basis / defense for this was based on presuming the #2 meaning. North8000 (talk) 15:21, 2 December 2023 (UTC)

Research conflict of interest

Assume an editor has performed research that has been verified and published. Would it be considered a violation of neutrality to add the results of their research to a page? This is just a hypothetical, but I'm curious to know. Gloryreaper (talk) 17:15, 6 December 2023 (UTC)

WP:SELFCITE. Jclemens (talk) 04:24, 7 December 2023 (UTC)

Oops, sorry! I didn't see that. Thanks! Gloryreaper (talk) 13:14, 7 December 2023 (UTC)

Semi-protected edit request on 25 February 2024

Please change the text "non-negotiable" at front of the This policy is [...] to "non-negotiable". Making the bold text red draws the user's attention to the fact that this policy cannot be overridden. 95.141.97.245 (talk) 18:22, 25 February 2024 (UTC)

 Not done for now: please establish a consensus for this alteration before using the {{Edit semi-protected}} template. M.Bitton (talk) 18:33, 25 February 2024 (UTC)

I don't think that will help. Also, a common misunderstanding is that non-negotiable means that negotiation and compromise (i.e., normal talk page discussions) are inappropriate. The principle is non-negotiable, but it may take quite a lot of negotiation to figure out what neutrality actually means for a given article. WhatamIdoing (talk) 20:46, 7 March 2024 (UTC)

False dating

you cannot use CE or BCE because it is all based on a ANC model this Gregorian calendar, which was studied and scientifically made as accurate as possible is done by Jesuit priests and therefore you cannot remove that from the calendar. Therefore, you cannot reference it by creating a CE or a BCE . But in the hard work and do it yourself otherwise you millennials step off 2605:B100:323:19E6:D149:4DCF:55BE:CD41 (talk) 19:47, 6 March 2024 (UTC)

My changes were based on policy, per WP:MOS. Did you read the rules there? Oh, and I'm not a millennial. Even if I was, that seems an awful lot like an ad hominem argument. Please try not to do that, per our policies. Thanks. Professor Penguino (talk) 02:53, 8 March 2024 (UTC)

Listen, if you really think the dates should be BC and AD, go ahead. Just make it consistent throughout the article, and don't edit war. (Speaking of which, "BCE" and "CE" have been used since the early 18th century, so I'm not "creating" it. Professor Penguino (talk) 02:57, 8 March 2024 (UTC)

One last note: I found this comment by accident. You should ping users (@ them) if you want them to see your messages. I assume you're the person whom I reverted several days ago, in which case I would inform you that block evading is extremely discouraged. Professor Penguino (talk) 03:01, 8 March 2024 (UTC)

Discussion at Talk:Tiffany Henyard § Tag

 You are invited to join the discussion at Talk:Tiffany Henyard § Tag. The discussion has focused on due/undue (or reasonable/excessive) coverage within the article. RunningTiger123 (talk) 02:48, 18 March 2024 (UTC)

Question about WP:WEIGHT

This policy states "in determining proper weight, we consider a viewpoint's prevalence in reliable sources".

My question is this: when assessing a statement made in an article, should WEIGHT be assessed based on the sources cited in that article, or based on the sum total of all sources "out there somewhere", whether they are cited in the article or not? If it's the latter, what method can we use for assessing how the sum total of uncited sources would affect weight?

This question arises from a conversation here, but I'm not seeking dispute resolution on that topic - I'm interested in how to interpret the WEIGHT policy in a more general sense, specifically, how to deal with uncited sources when assessing weight. Thanks. Philomathes2357 (talk) 20:30, 25 May 2024 (UTC)

There's no requirement to cite every source that supports article content, in fact there's an essay and a template to discourage that behaviour. So I would have thought that implicitly you must weigh sources not in the article. In the weight discussion the sources need to be presented so they can be scrutinised, but I would think it a negative to have every single one in the article. -- LCU ActivelyDisinterested «@» °∆t° 21:45, 25 May 2024 (UTC)

Let me see if I understand - it is acceptable to invoke uncited sources "out there somewhere" in a debate about WP:WEIGHT, but the burden of proof is on the editor who invokes those sources to 1) demonstrate that they exist & are reliable 2) demonstrate that, if they were in the article, they would constitute sufficient weight to support the content in question.

Is that a fair summary of what you've said? Philomathes2357 (talk) 23:58, 25 May 2024 (UTC)

1) Yes (if asked to). As with a lot of Wikipedia should is probably important. There has to some limit to it, as editors could disruptively demand excessive sources for every minor detail. Also proving reliability is a bit nebulous, there was a discussion on the reliability of RTÉ at RSN awhile ago. An editor was asking why it wasn't on WP:RSP, the answer is because no one has ever doubted it's reliability. So asking someone to prove the reliability of all their sources is a bit back to front, they should have to show the reliability of those sources if another editor has good faith reasons to doubt them.
2) Not quite. Demonstrate in that they would be arguing that these sources show weight, as part of finding some consensus among editors. Your phrasing makes it seem they would have to pass some test.
WhatamIdoing's comment below on the quality of sources showed up as my comment was half written, and covered half the points I was thinking of much better than I could have expressed them. -- LCU ActivelyDisinterested «@» °∆t° 00:37, 26 May 2024 (UTC)

Good point on #2. Thanks.

@ActivelyDisinterested and @WhatamIdoing, If you'll humor me, I have another question about how NPOV is commonly interpreted. I'd like to know your personal opinions, and also your general sense of the evolution of common practice over time.

Of course, NPOV says "avoid stating seriously contested assertions as facts. If different reliable sources make conflicting assertions about a matter, treat these assertions as opinions rather than facts, and do not present them as direct statements."

How is this commonly interpreted in terms of WP:WEIGHT? In other words, how much "weight" is regarded as sufficient to establish that an assertion is "seriously contested"?

More specifically, how is "contested" usually understood in terms of labels? Let me give you a hypothetical example to clarify:

Imagine we are working on an article about Israel. Let's say there are currently 80 sources cited in the article. 6 of them describe Israel as "terrorists" or a "terrorist state". The other 74 represent a variety of points of view on Israel - some are extremely critical, but do not use the label "terrorist", while others are more or less neutral, and others are, to some degree, sympathetic to Israel.

The sympathetic sources may say thinks like "Israel is justified in its actions", "Israel complies with international humanitarian law", or "Israel is a tolerant and democratic society". But none of the 80 sources explicitly say "Israel is NOT a terrorist state".

Would it be correct (or generally accepted by consensus as acceptable) to conclude "the statement that Israel is a terrorist state is an uncontested factual assertion, and, since no sources explicitly contest the claim, it can, should, or must be presented in Wikipedia's voice."

Or, would it be correct to say "although no source has explicitly negated the phraseology 'terrorist state', we can still assume, without violating WP:SYNTH, that the aforementioned quotes by sympathetic sources represent "conflicting assertions", so we should treat the description "terrorist state" as a seriously contested assertion."

Or, do neither of those views accurately represent the common understanding of NPOV? Philomathes2357 (talk) 01:00, 26 May 2024 (UTC)

For your example (or anything like it), we throw up our hands and point to WP:TERRORIST. We semi-sorta-kinda say that you should not directly call anyone or anything a terrorist. That is, we can have thousands of articles that say that the subject "has been called a terrorist by Alice, Bob, Chris, David, etc.", but you usually can't say that the subject "is a terrorist" (until the weight is so overwhelming that it's silly to attribute the view to a small set of people or groups. NB that when this happens, you will almost always have some editor says "But Wackos R Us and this one political influencer I like doesn't agree, so we still can't say this in WIKIVOICE!").

You may choose to think that this is mealy-mouthed of us, if you wish, but that's our usual practice.

Something that may help overall is that the goal is to have the Wikipedia article, both in the overall impression given by the article and in individual pieces, accurately reflect the mainstream (scholarly, if it's a scholarly subject) notion of the subject, even when that means being wrong or unfair to the minority POV. For something involving nation-states, that usually means scholarly sources. For example, if the mainstream scholarly opinion is that whether to call Israel a terrorist state depends on your definition, then the article should reflect that (e.g., "have been called terrorists by Bob, who uses the definition he got as a cereal box prize in 1962"). If the mainstream scholarly view is that the "deep state" is an implausible conspiracy theory believed primarily by blue-collar white men who feel, with some justification, like life gave them a raw deal, and believe that the only plausible reason for them not being socioeconomic winners is that something sneaky and disreputable is going on in the halls of power, then the article should reflect that.

I don't think the policy has changed much over the years. What has changed is the media around us. Using a lot of in-text attribution seems to be a way to signal that the writer is trying to be impartial. Consequently, we all (editors and readers alike) may have different ideas now about what a neutral tone sounds like. WhatamIdoing (talk) 02:09, 26 May 2024 (UTC)

Thanks for sharing your thoughts, I really appreciate it.

I think WP:TERRORIST is a great policy, and should probably be expanded to include a select few other terms that are often abused for POV-pushing, but that's a topic for another day.

I completely agree with you when you say "If the mainstream scholarly view is that the "deep state" is an implausible conspiracy theory...then the article should reflect that." - and to be clear, I posted here at NPOV because I wanted to gain a deeper understanding of how the policies have been interpreted in the past - I am not seeking input on the RFC, and don't want to discuss it here.

"Terrorist", on second thought, is not a great example, because I'm more interested in cases where WP:TERRORISM does not apply. For instance, if a small minority of sources calls a person, "corrupt" or "incompetent", and, while there are plenty reliable sources that are sympathetic to the person, no sources specifically say "he/she was NOT incompetent/NOT corrupt", would that be commonly understood as an "uncontested" or "contested" assertion? Let's assume they are not a living person, so BLP does not apply. Philomathes2357 (talk) 02:20, 26 May 2024 (UTC)

As it happens, I've spent this morning on a similar case, in which an editor wants to call something altmed. As it turns out, I can't find any MEDRS-style sources that explicitly says that it is or isn't. What I can find is that MEDRS-ideal sources do not use that language, and their overall presentation mostly leaves you with the opposite feeling. This differs importantly from your case in that good quality sources don't say anything either way, rather than a few saying "yes" and the rest being silent.

On the one hand, there are problems with relying solely on positive statements. You could end up with "Three sources said it is" and nothing to balance that – even if the implicit statements from all the other sources is the opposite.

For example, imagine that I have found 100 sources about chemotherapy. Three say that it's worthless. The other 97 are dealing with side effects. The other 97 implicitly suggest that there is value, because why would you deal with these side effects for a completely worthless treatment? If those 97 thought it was worthless, then side effect management would be short and simple: "Don't take this worthless stuff."

You don't want to say "Every source that explicitly comments on this subject says ____", when all the other sources imply that the opposite is their actual view. But you can't actually say "Three sources say it's worthless, and 97% of them imply that there's value", because although it's true, it's a NOR violation (because we require a source that Wikipedia:Directly supports the claim, not just 97 that imply it). Also, it's not appropriate to say that if three sources say it's worthless, seven imply that it's valuable, and 90 are, upon closer inspection, irrelevant to that particular point, that this proves that the point isn't worth mentioning at all.

Which takes me back to the main point: However we get there, the goal is to have an article that accurately and fairly represents the views of high-quality sources, including on facts so basic or obvious that they didn't explicitly and directly state them. WhatamIdoing (talk) 20:46, 26 May 2024 (UTC)

Thanks for your thought-provoking reply. Your chemotherapy example is very interesting. Based on what appears to be current consensus regarding NPOV's interpretation, the approach there would be to begin the chemotherapy article by saying, in Wikipedia's voice, "chemotherapy is a worthless treatment", since there is no source that explicitly/directly says "chemotherapy is NOT worthless".

But that would be completely absurd and untenable, would it not?

Maybe one could avoid the NOR issue by simply saying "the other 97 sources, irrespective of their value judgement about chemotherapy, simply do not lend weight to the claim that it is "worthless", so, therefore, there is very little weight behind the claim that it is "worthless", even though that claim has not been directly rebutted. That means that we can characterize the notion 'chemotherapy is worthless' as a minority-held view."

In your view, would an approach like that be, on its face, a violation of NOR, SYNTH, or some other policy or guideline?

The more I think about this, the more potential cases come to mind, and the more important it seems to get cases like this right and establish some sort of common understanding for how to deal with them. I appreciate you helping me dive deeper into this and providing very thoughtful replies. Philomathes2357 (talk) 22:11, 26 May 2024 (UTC)

@Philomathes2357, it has to be about sources that exist in the real world, because otherwise, I could remove all the sources I disagree with, and then re-write the article to say whatever I believe.

The ideal process is something like this:

• Do some research to find out what kinds of sources are out there. This could mean spending time with your favorite web search engine, with Wikipedia:The Wikipedia Library (which I am absolutely loving; De Gruyter's books might be useful for that article), or even in a library. Figure out what the overall lay of the land is, especially among the best sources.
• Write the article so that it provides a reasonable summary of what you learned during the hours/days/weeks/months of research you did.
• Cite whatever you need to, to verify the individual statements as you make them. If it's a pro/anti type of subject, this process can sometimes include citing "pro" sources to support "anti" content (and vice versa), which is another way that the balance of cited sources might not match the desired balance of the article itself.

If you are interested in an example, I think I have had as many NPOV arguments over Breast cancer awareness as for all the other articles combined. The problem is that the low-quality sources (e.g., puffy local news stories), were all rather glurge-y and irrelevant: "Look at the pretty woman who is soooo nice and strong and sweet that she's raising money to help other cancer victims!"

When I got into the scholarly literature, though, the story was quite different: "Look at the patriarchal assumptions that say sick women must be superheroes who never inconvenience anyone. Look at the unfair expectations that say sick women have to wear makeup and wigs so that the rest of us aren't reminded about their vulnerability or our own mortality. Look at how breast cancer was considered an obscene disease for so many decades. Look at the way society polices the things sick women say about their fears and experiences. Look at the way screening programs get promoted but prevention efforts gets downplayed. Look at those deceptive fundraisers, which imply unlimited donations but actually make paltry donations. Look at the billions of dollars we have spent without reducing the number of deaths materially. Look at the alcohol manufacturers putting pink ribbons on their products, instead of putting on labels that say '15% of breast cancer is caused by drinking alcohol. If you don't want breast cancer, then don't buy our product'."

The human problem we have is that when article content isn't what we expect, then we think it's wrong. So whatever your/my/anyone's own filter bubble says, that's what you automatically (whether you want to or not) think that's what the article should say. The only way I've found to get out of that "confirming my own pre-existing biases" mode is to do a deep dive into high-quality sources. WhatamIdoing (talk) 00:22, 26 May 2024 (UTC)

WhatamIdoing, as usual, you think deeply and show great wisdom. We may not always agree, at least initially, but I respect you.

You write: "when article content isn't what we expect, then we think it's wrong." Yes, too often one sees that is the initial reaction, especially by newbies and driveby POV pushers and vandals, and we often delete those comments on sight, with no explanation. "When article content isn't what we expect", the proper reaction is to:

1. AGF in fellow editors, IOW that they have tried to follow PAG;
2. assume the article is based on RS;
3. assume the article narrative is therefore correct;
4. assume that other editors and the sources they have found "know more" than I currently do;
5. assume that I am likely less informed and likely wrong;
6. assume this is a learning opportunity;
7. adopt a scientific attitude and follow the evidence, IOW, follow the sources;
8. bring my own POV into line with the sources, IOW, change my mind(!!!), no matter how painful;
9. before objecting to what I think is wrong, read the whole article, or at least the relevant parts, and also read the sources;
10. then, and only then, if I am still convinced the article, or part of it, is wrong, formulate a good case using RS, and open a thread to discuss a very specific issue, using exact quotes.

That's my method, and I have been forced to change my mind many times over the years. That's also what I love about editing here. I learn so much from other editors and sources. I may be stubborn at first, but good evidence will usually change my mind. -- Valjean (talk) (PING me) 17:25, 27 May 2024 (UTC)

We need a Barnstar of Publicly Changing Your Mind. It's one of the rarest behaviors. WhatamIdoing (talk) 05:34, 28 May 2024 (UTC)

This is a hard one. I don't think there's a general rule that always works. I go back to "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". So start there. Are all the significant and reliable views included in that article? After that, there's a debate to be had about how to present in-significant or un-reliable views (usually somewhere between leaving them out, or reporting them through the lens of more reliable sources). And then there's still a discussion to be had about each of the significant views. Are they all equal? Is one more significant or reliable than the other? The debate may be moot: our article may end up representing all of them, and letting the reader form their own opinion. Unfortunately, I think policy only gets us so far, and you need a lot of good faith editors doing a lot of quality research to settle each discussion, case-by-case. The only thing I'd add is if an editor digs their heels in and insists that some viewpoint is equally significant or more significant than what's currently in the article, there is eventually an onus on them to prove it, with actual WP:Verifiable reliable sources. Shooterwalker (talk) 19:06, 26 May 2024 (UTC)

Shooterwalker, that's a great analysis. Start with all the reliable sources you can find, and let them speak. Sources, not editorial beliefs and opinions, have the primacy. Then describe what unreliable sources say, but only using the lens thru which RS look at what those unreliable sources say (as contrarian, false, and inaccurate views). Unreliable sources alone have zero due weight and should not be cited directly, only indirectly by citing RS that mention them, with the POV of the RS. Use attribution. Avoid bothsidism and false balance. -- Valjean (talk) (PING me) 17:02, 27 May 2024 (UTC)

I find that a lot of novice editors come in because they have something to say, and then they find a source that supports it. It does risk pushing their POV (especially if they stray into primary research or other unreliable sources). More experienced editors approach it like a literature review. Shooterwalker (talk) 16:19, 28 May 2024 (UTC)

Should "of reliable sources" be restored?

WITHDRAWN BY OP. SEE AT BOTTOM.

THIS IS NOT A FORMAL RFC, BUT MAY LAY THE GROUNDWORK FOR A FUTURE RFC. A consensus here can be used to change content. An RfC is not necessary when there is a strong consensus.

WAS: "If a viewpoint is held by an extremely small minority, it does not belong on Wikipedia, regardless of whether it is true, or you can prove it, except perhaps in some ancillary article."

THEN: "If a viewpoint is held by an extremely small minority of reliable sources, it does not belong on Wikipedia, regardless of whether it is true, or you can prove it, except perhaps in some ancillary article."

NOW: "If a viewpoint is held by an extremely small minority, it does not belong on Wikipedia, regardless of whether it is true, or you can prove it, except perhaps in some ancillary article."

IMPLIED COROLLARY: "If a viewpoint is held by an extremely large majority, it does belong on Wikipedia, regardless of whether it is false, or you cannot prove it."

PROBLEM: "of reliable sources" was added and then removed.

THE ISSUE: "extremely small minority" refers to number of RS, not the number of people. That addition addresses a very real need here.

Let's use homeopathy as an example. "An extremely small minority of reliable sources" take it seriously. In fact, no really RS take it seriously. They universally criticize and debunk it. Yet, billions of people believe in homeopathy, primarily in India. That's not "an extremely small minority" (of people), so, per the implied corollary, it belongs on Wikipedia. But we don't include it because of the number of people who believe it. We do that because of the number of RS that document it as fringe, pseudoscientific, nonsense. RS are the reason we do things here.

Some fringe nonsense viewpoints "held by an extremely small minority of reliable sources" happen to be documented by myriad RS and are thus notable enough for documentation here, even with whole articles about fringe nonsense. It is the coverage in RS that give it enough weight for mention, not the number of people who believe the fringe nonsense. Fortunately, for the purposes of serving our readers, including the deluded masses, there is a tendency for widely held nonsense to be described in enough RS so we can document it. That keeps us inline with our mission here, to "document the sum of all human knowledge" as it is mentioned in RS.

At Wikipedia, we don't give a flying f##k how many people believe something. The masses can be deluded and deceived. They are not RS. Our only concern is what RS say (and unreliable sources have zero due weight). "extremely small minority" refers to number of RS, not the number of people, therefore that wording is important.

PROPOSAL: That "of reliable sources" be restored as it is an important modifier that keeps the wording in harmony with our PAG.

Yes or No? Let's discuss this. Provide your reasoning. (I have already done so above.) -- Valjean (talk) (PING me) 17:49, 2 June 2024 (UTC)

• Yes, as explained above. -- Valjean (talk) (PING me) 17:49, 2 June 2024 (UTC)

I think it's a good idea. There's obvious misunderstanding of this point, and what someone said twenty years ago doesn't dictate policy (especially as the point was always implicit). -- LCU ActivelyDisinterested «@» °∆t° 20:43, 2 June 2024 (UTC)

Please read #The key question to ask and a proposed answer, which bring out the very specific need answered by the notion of reliable sources and the fact that it is after 2003 that this need was felt. In fact, I am curious to know when the term "reliable sources" and its use for articles in medicine, etc. first appeared. Dominic Mayers (talk) 21:47, 2 June 2024 (UTC)

The dependence on RS has always existed, even if I can't remember the exact formulations. For discussion's sake, let's play with the idea that "reliable sources" was not an "original" concept. Let us also compare our PAG to the Constitution. The Constitution was quickly found to be lacking, hence the creation of Constitutional Amendments. At Wikipedia, our PAG grow all the time, and one could view those changes as amendments and improvements to the imperfect "original" ideas at the creation of Wikipedia. When we see a need, we fix it.

It sounds like you don't think it's a good idea to base all content on RS and are seeking to create an argument for "going back to the foundations" when there was no (as you imply) such requirement. Even if you are right that there was no such requirement, it's an improvement to have it, so casting doubt on it is an unwise idea. -- Valjean (talk) (PING me) 22:03, 2 June 2024 (UTC)

No, you misunderstand my goal. I have nothing against the new principles. However, repeating "use reliable sources" all over the place does not explain anything. It is just advertising criteria that will be explained elsewhere. If, to do this advertising, we emphasize due weight, which was only introduced after 2003, because it fits well with "reliable sources", then it creates a serious problem. For example, the notion of attribution in the principle "attribute opinions" has almost nothing to do with reliable sources. Yes, the opinion itself must be found in reliable sources, but once the opinion is sourced, the extra requirement for attribution has nothing to do with reliable sources. Similarly, the principle "do not take sides" per se, once all sides are properly sourced, has nothing to do with reliable sources. Therefore, the emphasis is on due weight, because it is directly connected with reliable sources. Yet, due weight is not the essence of NPOV. I do not want to cast doubt on the purpose of reliable sources, but just repeating "use reliable sources" all over the place is not explaining much and if in doing so we focalize on due weight and fails to also emphasize the original principles, then I think it is very bad. Dominic Mayers (talk) 23:06, 2 June 2024 (UTC)

Isn't attribution almost entirely about source reliability? What is it about otherwise? Horse Eye's Back (talk) 13:36, 3 June 2024 (UTC)

In the context of this policy, it is mostly a way to achieve the neutral point of view and, no, it is not mainly about reliability, in the same way that the way to Paris is not mainly about reliable roads. The latter is more basic, something that is better kept in the background when we explain the way to Paris. Really, it is strange that I need to explain this again and again. It is clear that we are far from having an attribution after we have only checked that the viewpoint is sourced. It should also be very clear that the purpose of attribution, which is to achieve the neutral point of view, goes way beyond the reliability of the sources. It makes no sense to suggest that the former can be reduced to the latter. Dominic Mayers (talk) 13:53, 3 June 2024 (UTC)

But isn't the decision about whether to attribute and if so what form that attribution should take almost entirely based on the reliability of the source/sources? I don't think your roads+Paris analogy works, I would drop it. Horse Eye's Back (talk) 15:42, 3 June 2024 (UTC)

You keep saying that reliable sources are the foundation, etc. and I don't disagree, but only in the same sense that roads are the foundation for the way to Paris. You do not accept my argument and my analogy with the roads to Paris. Having my arguments plainly rejected with a "please drop it", no further details provided and be left with nothing to build upon to further argue is not interesting. But fine, you win, I have no further arguments. Dominic Mayers (talk) 16:05, 3 June 2024 (UTC)

Policy isn't based on the exact phrasing that Wales used in 2003. I understand you point, I just don't agree with you. We're not going back in time to before October 2006, which is when the policy gained it's current form. If you want to use the form before that you need to get consensus for the change. -- LCU ActivelyDisinterested «@» °∆t° 22:56, 2 June 2024 (UTC)

The issue at the moment is that a small set of editors misinterpret the current wording, and how best to stop that from happening. The top wording of this section details that well. If you have a different way to avoid this misunderstanding I would be interested to hear it. -- LCU ActivelyDisinterested «@» °∆t° 23:25, 2 June 2024 (UTC)

I am not sure what is the misinterpretation. It may be that they have the correct interpretation of what it meant at the time and the problem is that some people do not like that reliable sources is not emphasized again. For example, it is clear from the context that the statement was not making a clear distinction between sources that hold a viewpoint and people that hold a viewpoint. Even today, this distinction is not always clear. That should not be an issue. I have a hard time to believe that the misinterpretation would be a confusion between people or media that are sources and wikipedia editors. It is almost impossible given the context to make this confusion. In any case, there is certainly no need to make an anachronism by mentioning reliable sources to remove that confusion. Dominic Mayers (talk) 00:05, 3 June 2024 (UTC)

They do not, it did in the preceding sentence of his statement, it is clear as per the next sentence in the current policy as mentioned below, whether you believe the issue or not it does happen, not an anachronism that would only be the case in NPOV policy was the 2003 statement unchanged which it isn't.
But Valjean makes a good point that this is already covered, so the solution would be to simply point out that next sentence. -- LCU ActivelyDisinterested «@» °∆t° 02:25, 3 June 2024 (UTC)

I am all for moving forward, but the notion of reliable sources does not need to be repeated all over the place, especially not when it creates an anachronism. It did not exist in 2003 and, here, we paraphrase a 2003 statement. The argument that it was implicit is wrong. If it was introduced in 2006, then it was something really new. Saying it was implicit is just playing with words. In a way, every thing potentially existed (or was implicit) at the time of the big bang ! It is also that modularity is important. There are principles that can be explained without reference to reliable sources. It is as if we were saying all the times take this reliable road and then this reliable road, etc to go to Paris. It creates confusion, because the fact that the roads are reliable is better kept implicit. It does not help to explain the way to Paris. And when I see that there is little interest here in better explaining the basic principles of NPOV and that we have that long discussion to emphasize "reliable sources" instead, some even say that I lack of focus because I complain, then I am sad. Dominic Mayers (talk) 23:41, 2 June 2024 (UTC)

I am withdrawing this proposal because of these words that immediately follow the text in question:

Keep in mind that, in determining proper weight, we consider a viewpoint's prevalence in reliable sources, not its prevalence among Wikipedia editors or the general public.

If you can prove a theory that few or none believe, Wikipedia is not the place to present such proof. Once it has been presented and discussed in sources that are reliable, it may be appropriately included. See "No original research" and "Verifiability".

That really covers my objective quite well. -- Valjean (talk) (PING me) 02:15, 3 June 2024 (UTC)

The "Keep in mind...." wording seems to come from a request. See Revision as of 01:37, 19 May 2008. I'm not sure it was a request I made, but it sounds exactly like something I might have added. I have added a number of things to this policy since 2003. -- Valjean (talk) (PING me) 02:30, 3 June 2024 (UTC)

I agree that this subsequent sentence covers the necessary territory. WhatamIdoing (talk) 02:57, 3 June 2024 (UTC)

This informal or pre Rfc is biased

It fails to mention the important point that it is about a paraphrase of Jimmy Wales 2003 statement. It also cherry picks what it considers relevant in that paraphrase. In particular, the following point in the 2003 statement is not mentioned:

If your viewpoint is held by a significant scientific minority, then it should be easy to name prominent adherents, and the article should certainly address the controversy without taking sides. (emphasis is mine)

It is important, because its reference to "prominent adherents" makes it clear that Jimmy Wales was thinking in very general terms. There is also a bias in making a pre RfC on reliable sources, which is already emphasized all over the place in the policy, while "without taking sides", which is a key point made in the nutshell at the top of the policy, is not even mentioned once after in the policy.

"Reliable sources" was not mentioned at the time of the 2003 statement, whereas the 2003 statement mentions "without taking sides". Since it is in the original 2003 statement, I think it is this that should appear in the paraphrase, not "reliable sources". It is also biased to refer to a "restoration". It is not that the policy was modified by removing "reliable sources" and now we want to restore it. No, it is the opposite. It has been proposed today to add it and this is contested. Dominic Mayers (talk) 18:47, 2 June 2024 (UTC)

It is not an RfC. It is for discussion to avoid any problems in the event an RfC is necessary. It is important to examine the issue from many angles so an RfC can be focused on a limited and specific issue. It is not possible to consider every single possibility in every discussion. This just leads to endless discussions like the ones you get involved in. Other editors finally just give up as the discussions are hopeless and goalposts keep getting moved. Try to stay focused. -- Valjean (talk) (PING me) 19:23, 2 June 2024 (UTC)

Please do not make judgements about my manner to proceed, but focus on the essential of my arguments instead. In this way, I will not have to reply about superficial issues like I am forced to do now and we will have the required focus. Dominic Mayers (talk) 19:28, 2 June 2024 (UTC)

The key question to ask and a proposed answer

Before we discuss the mention of "reliable sources" in Jimmy Wales's paraphrase, it is important to ask ourselves why, concretely, this emphasis on "reliable sources" is so important ? Why the criteria already there in 2003 ("commonly accepted reference texts" and "easy to name prominent adherents" as inclusion criteria and "extremely small minority" as an exclusion criterion) are not sufficient anymore ? By "concretely", I mean that we need to go beyond the obvious explanation. The obvious explanation fails, because there is nothing to be concerned about when it is something obvious that every one, even a wacko or a fanatic, accepts and understands. In other words, if there is a need for an emphasis, it is because it is something that must be explained to other people. I suspect that we will all agree that these people are anti-vax people or climate denial people, etc. In that context, principles such as "do not take sides", "do not engage in debates, but describe them" and "attribute opinions" do not seem sufficient. These principles were the key principles of NOPV, they are still very important and "use reliable sources" was not one of them.

Because these principles were not sufficient, in 2003, the NOR principle was added with something close to the reliable sources requirement, but it was not the reliable sources requirement. It was felt sufficient to require that the viewpoint was published in "commonly accepted reference texts". It was even considered sufficient that it is "easy to name prominent adherents". And for the rejection, it was sufficient that it was held by only an "extremely small minority". There was no mention of reliable sources. Now, again, the question is why nowadays we insist so much on the terminology "reliable sources". I propose that the explanation is that "easy to name prominent adherents" does not allow for a restrictive inclusion criteria. In contrast, with "reliable sources", we can insist that, in areas such as medicine, we must use only some special kind of meta-analyses. In other words, the reason is that "reliable sources" is more flexible when it comes to give the specific of the criterion in medicine, etc. There is nothing wrong with that.

Yet, please, please, let us not throw away the fundamental principles of NPOV, which is what this policy was about before 2003, before reliable sources, and should still be first about. I am concerned that this emphasis all over the place on "reliable sources" distract the readers from the essence of NPOV. Dominic Mayers (talk) 21:24, 2 June 2024 (UTC)

"use reliable sources" is fundamental to all content and PAG here. There is no content without RS, and all PAG exist with that background, whether it is said or not. All PAG exist in the service of using RS to create content. -- Valjean (talk) (PING me) 21:51, 2 June 2024 (UTC)

This is not in contradiction with what I wrote. The way to Paris cannot exist if there is no roads, but it does not mean that when we explain the way to Paris, we must emphasize the concept of road. On the contrary, it can be a distraction to discuss roads when we explain the way to Paris. So, because of what you said, this emphasis on reliable sources can be a distraction away from the main principles of NPOV. Dominic Mayers (talk) 22:03, 2 June 2024 (UTC)

NPOV is a huge and complicated topic with many aspects. We won't get anywhere if we must always mention every aspect in every discussion. Focus on one aspect and improve it. -- Valjean (talk) (PING me) 22:06, 2 June 2024 (UTC)

I think you are simply not understanding my point and, because of that, you simply considered it out of focus. But it is entirely on focus. In fact, it goes to the essence of the matter. You keep wanting mentioning "reliable sources" and the context is "We take side with reliable sources", etc. It is important to clarify for yourself why it is important for you. Just saying it is fundamental, etc. is not a concrete answer. It is so sad that you cannot step back and concretely answer the question why it is so important to emphasise it and then see the connection between this and the original principles of NPOV, which existed before the specific notion of reliable sources, which you want to emphasize. Dominic Mayers (talk) 22:21, 2 June 2024 (UTC)

I give up [on discussing with Dominic]. I'm not going to get caught up in another time sink with you. -- Valjean (talk) (PING me) 22:33, 2 June 2024 (UTC)

I was going to oppose, but if the OP has given up I guess I don't need to. Peter Gulutzan (talk) 23:30, 2 June 2024 (UTC)

@Peter Gulutzan:, I meant a discussion with Dominic (so have added that now). Such discussions drag on forever and yield little of worth. You can "oppose" above in the right section. -- Valjean (talk) (PING me) 02:09, 3 June 2024 (UTC)

Valjean The WP:TALK guideline re changing your own comment after there's been a reply includes Any inserted text should be marked with <ins>...</ins>, which renders in most browsers as underlined text, e.g., inserted. E.g. [on discussing with Dominic].Peter Gulutzan (talk) 14:09, 3 June 2024 (UTC)

Thanks for that clue. -- Valjean (talk) (PING me) 17:17, 3 June 2024 (UTC)

Curious what your objection is? The same paragraph says "Wikipedia aims to present competing views in proportion to their representation in reliable sources on the subject.", so adding "reliable sources" again seems unnecessary. This phrasing is also a bit more sophisticated, in that reliable sources might be telling us a view is held by a large majority, which we would believe in contrast to unreliable sources telling us a large number of people think the world is flat. Though I wouldn't necessarily object to adding "of reliable sources" either, since it doesn't seem like it would change anything. -- Beland (talk) 01:20, 3 June 2024 (UTC)

I have withdrawn the proposal because of words that immediately follow the text in question. See the section above about this matter. -- Valjean (talk) (PING me) 02:17, 3 June 2024 (UTC)

Beland, FYI I'd have objected again that when paraphrasing Mr Wales one should not insert words that bear no resemblance to his actual words, but I don't need to. Peter Gulutzan (talk) 14:18, 3 June 2024 (UTC)

Ah, gotcha. The page says the bulleted list is paraphrasing Wales, but the paragraph above makes no such representation; I would expect that to represent the modern consensus of the current community of editors. -- Beland (talk) 18:43, 3 June 2024 (UTC)

Dominic, I'm not sure that I understand what "without taking sides" means to you.

Imagine a world in which there was an online encyclopedia. We are all big fans and want to write an article about it. Imagine that every single independent source ever published about it consistently pans it: "All the content is bad, but fortunately, there's hardly any of it." "A survey of history articles indicates that the article creation priority is 'Every possible detail about the three historically unimportant minor battles that J. Henry Smith IV believes his great-great-great-grandfather was present during'." "When I find two grammatically correct sentences in a row, the surprise throws me off for the whole day." "We fact-checked 100 sentences and found 250 errors." "My aunt knows more than those goofs, and she can't even figure out how to turn off the flashlight on her smartphone."

What would "not taking sides" look like in such a case? Does that mean writing something like:

• "It has limited content, but covers some areas, such as certain minor battles during the War, in detail"? or
• "It has been harshly criticized for error-riddled content, low-quality writing and haphazard content"? or
• "It has been criticized some but is appreciated by others" [with Wikipedia editors being the unnamed 'others']?

WhatamIdoing (talk) 03:16, 3 June 2024 (UTC)

How to not take sides (reply to a question)

@WhatamIdoing:, I feel a bit embarrassed by your question, because this talk page is not the place to discuss the personal understanding of NPOV of anyone in particular. On the other hand, it is the place to ask if the principle "do not take sides", which is the very starting point in the nutshell and always have been central in the NPOV policy is misused. I think it is misused, by being not used enough. There are basic notions in life that we, human beings, share and that we should not try to define and the notion of not taking sides, being neutral, is one of them. We give examples to make sure we refer to the same notion, but examples are not definitions. If I take your fictive example, the way to not take sides is to write something like every single independent source ever published about it consistently pans it. This was an easy case, because you have given what is to be considered factual about existing viewpoints. When you are factual, you are not taking sides, but simply present the facts. Another example is to say "John said X" instead of directly "X". You are not taking sides with John when you say "John said X". As far as the readers are concerned, you might even disagree with John, but you simply give the fact that "John said X". In real practical cases, it is not so easy to find the facts, but in your fictive example, you made it very easy, because you wrote at the start what is to be considered factual. I am not saying that the definition of "do not take sides" is "be factual", of course not. These are only examples how to achieve it. We all share the notions of "not taking sides", "being neutral", but it is a basic notions that cannot be defined. It is impossible to define every thing, because every definition depends on other notions that also need to be defined and we will never reach the bottom of it. The most important notions, the one that we must use the most, are not definable. Dominic Mayers (talk) 12:08, 3 June 2024 (UTC)

And how do you determine what is a fact without determining what is a reliable source? Horse Eye's Back (talk) 13:46, 3 June 2024 (UTC)

This point has been covered again and again. I consider it a distraction from the essential when we try to explain the neutral point of view, just like mentioning all the times the importance of reliable roads can only bring confusion, if the goal is to explain the way to Paris. However, since you insist, I will expand on this. We learned here that reliability must be about the factual content of sources. If a source presents the point of view of a notorious wacko, the source is reliable as long as it is a fact that the wacko have this viewpoint. The source is not unreliable simply because it presents the viewpoint of a wacko. It is different, if the source presents the wacko as being a notorious scientific, because it is false that the wacko is a notorious scientific. This is why we need reliable sources that represent well every scientific domain. This is something that needs to be explained in the reliable sources guidelines. It is a distraction from the essential to emphasize this while we explain the neutral point of view. Dominic Mayers (talk) 14:08, 3 June 2024 (UTC)

Your reply reminds me of the FAQ at the top of WT:V, which says:

Are there sources that are "always reliable" or sources that are "always unreliable"?

No. The reliability of a source is entirely dependent on the context of the situation, and the statement it is being used to support. Some sources are generally better than others, but reliability is always contextual.

Another way of putting it is that all sources are reliable for something ("William Wacko said X", cited to him saying that) and that no source is reliable for everything (The best scholarly work of the previous decade is an impossibly bad source for last week's movie).

(I have more to say about this, but I'll add it below.) WhatamIdoing (talk) 17:28, 3 June 2024 (UTC)

I don't believe - and I don't think most editors believe - that the principle do not take sides applies in all instances. To use time-honored examples, Wikipedia articles do, in fact, "take sides" between geocentric and heliocentric models of planetary motion, and between young earth creationist and evolutionary accounts of life on this planet. We also take sides over the question, did Donald Trump win the 2020 presidential election? So the scope of topics to which "do not take sides" applies needs to be defined in practice, and I think WP:FRINGE represents a pretty good step to articulate many of the relevant considerations.

Put simply, the way I would articulate this is that sometimes WP:NPOV requires article text to take sides. Newimpartial (talk) 14:50, 3 June 2024 (UTC)

@Newimpartial:, I agree with you, but think it is better to say that Wikipedia "appears to take the side" of mainstream RS over unreliable sources or no sources by allowing RS to get the due weight they deserve. Wikipedia stays neutral by not interfering with what RS and by enabling RS to voice their views without the interference of editorial bias.

The very existence of WP:RS nails "Reliable Sources" ("Theses #96") fast to the PAG door as foundational to how we operate. Article content should reflect what RS say, and the bias found in RS should shine through, as we are not allowed to censor or neuter the bias and opinions of a RS. Editors must not get in the way. OTOH, Wikipedia does not take a side when there is a difference of opinion between RS. Then we "explain the sides, fairly and without editorial bias." -- Valjean (talk) (PING me) 17:13, 3 June 2024 (UTC)

I am not sure it is productive to discuss among us what "most editors believe". Instead, why not directly find among us about a good understanding of NPOV. I think you are saying that it is not always obvious how to apply "not taking sides". The problem is that, even if we try not to take sides, we might be taking sides. One example of this would be to attribute the viewpoint that "Donald Trump did not win the 2020 presidential election" to Biden. It seems that we are only being factual, because it is a fact that Biden has this viewpoint. Yet, it suggests that it is Biden vs Trump and this is not really factual. So, the problem is not with the basic concept of "not taking sides", but how to apply it correctly, Dominic Mayers (talk) 15:19, 3 June 2024 (UTC)

Well, the way I would put it is that our shared social reality sometimes regards one epistemic assertion as "correct", "factual" or "true" and alternatives to it as "incorrect", "non-factual" or "false". Under these circumstances, WP:NPOV requires us to state "facts" as facts, not as opinions, even though this would be seen by those taking the small minority view as "taking sides" against their worldview and with the dominant episteme. Newimpartial (talk) 15:26, 3 June 2024 (UTC)

The expression "dominant episteme" sounds philosophical. Philosophy is more practical that some might think. Despite of this practical value, there is a big danger that people think that, if it is philosophical, it is not practical and must not be part of the policy. We would get rid of concepts such as "not taking sides", "not engage in debates, but describe them", because discussing these practical principles do involve some philosophical points. We would only focus on "Reliable sources" as the key ingredient: we follow the reliable sources and all problems are gone. That would be a terrible mistake. I even stop here and remove every thing else I wrote before. I will summarize them in a single sentence: please, let us not reject the universal principles of neutrality, "not taking sides", "attribute opinions", "do not engage in debates, but describe them", only because they are not always easy to apply and might even raise philosophical questions. Dominic Mayers (talk) 16:35, 3 June 2024 (UTC)

I can assure you that no one here is proposing we "reject" any of that. I suspect that we are often "talking past each other" and therefore misunderstanding each other. -- Valjean (talk) (PING me) 18:23, 3 June 2024 (UTC)

Dominic, I think you and I have very similar ideas of what should end up in an article, and different ways of explaining it. IMO yours requires editors to have a much higher level of competence. I'm moving towards a least-common-denominator model.

For example:

• A: Donald Trump lost the election.
• B: No, he didn't!

Both of those are assertions of fact, using a definition along the lines of "a statement that can be proven to have a truth value". One of the statements is correct (a "true fact") and one of the statements is wrong (a "false fact").

These, however, are statements of opinion:

• C: Trump was a bad president.
• D: Trump deserved to win the election.

For matters of undisputed fact, we should WP:ASSERT the fact: "Biden won the election." For matters of opinion, we should assert facts about important/common opinions: "Many Republican politicians said that Trump was a good president and deserved to win". Sometimes we should even assert factual statements about the false facts: "Some people falsely claimed that Trump won the election".

One of the main reasons that I've been moving towards the least-common-denominator model is because we have editors who believe that "Trump won the election" is a true fact. (Another is because editors have so much trouble differentiating between opinion and fact.) So when if we say "When you are factual, you are not taking sides. Simply present the facts, and the article will be neutral", the response is "The simple fact is that Trump actually won the election. If you want a factual article, then say that Trump won. If you don't say that Trump won the election, then you are taking sides against me!"

Whereas when we say "Neutral means whatever the reliable sources say, and if it's not in a source that the community will accept as Reliable™ for that statement, then it can't be added", then that same person will yell about how we're taking sides but accept that our rules require us to present the mainstream media POV as factual.

In other words, we solve more disputes, faster, by emphasizing reliable sources than we do by emphasizing a subject that POV pushers do not understand. WhatamIdoing (talk) 18:38, 3 June 2024 (UTC)

My feeling is that most examples that we consider create false dilemmas, by restricting the problem to a choice between two statements such as "Won the election" or "Lost the election". There is always an almost infinite number of possibilities, most of them adding extra information. The neutral point of view must take advantage of this. Take, for example, the earth is spherical vs the earth is flat. The article Earth says "Earth is rounded into an ellipsoid with a circumference of about 40,000 km." It is not only "The earth is round" (in opposition to "the earth is flat"). It does not affirm the earth is spherical, but provides encyclopedic information that turns out to say that it is rounded (but they could and perhaps should have used "shaped" instead). Also, one could conclude way before that statement that the earth is not flat, by the picture, the mention of its core, etc. It does not engage in a debate. In this case, it does not even need to describe a debate. A similar attitude should apply to the last USA election. In my view, there is not even a debate here also. But, the exact claim made by Trump could be pertinent in some context. In that case, we should simply be informative about it. If it is done well, the readers will not be mislead (just as it would also be the case if we mentioned beliefs that the earth is flat). Again, the way to achieve that is by being informative. Of course, if we simply reduce it to "Trump won" vs "Trump lost", then we have a serious problem that even attribution would not solve. This is why the neutral point of view should not be reduced to "attribution of opinions". This limited view of the neutral point of view associated with false dilemmas is so wrong. Dominic Mayers (talk) 20:26, 3 June 2024 (UTC)

I don't think that it's a bad example. Depending on the needs of the article, "Trump lost" might be all that is warranted. Elections are binary; you either win or you don't.

There are some subjects that purely factual but aren't strictly binary (e.g., the proper dose of a drug, the interactions between tree roots and fungi), but the US presidential election is strictly binary: one person wins, and everyone else loses. WhatamIdoing (talk) 00:48, 4 June 2024 (UTC)

The important editorial process and the confusion between due weight and taking sides

Not taking sides is the most fundamental principle of NPOV. It is mentioned right at the start. Nevertheless, many people say that we must take sides. It is not that bad, because they actually mean that we must give due weight to the dominant point of view. Fine, but why that confusion between due weight and not taking sides ? I propose that the explanation is that not taking sides is only about the editorial process per se and does not refer to the content in the sources that finds its way in the article. So, "due weight" refers to the content in sources that finds its way in the articles, but not taking sides only refers to the editorial process per se. Therefore, if you naively think that the content in the article is essentially a transcription of the content in sources, that is, if you consider that the editorial process can essentially be ignored when we define "not taking sides", then you will necessarily confuse "not taking sides" with "due weight". You will consider that not taking sides means that the article reflects the bias in sources: it will not be about the editorial process, including attribution, etc. Therefore, I propose that the real confusion, the real problem, is a misunderstanding of the important rôle of the editorial process. Dominic Mayers (talk) 04:01, 4 June 2024 (UTC)

I've been trying to follow your logic for the last several threads, and at this point, you've lost any chance of anyone following this.
First, do not create a new section title for a continuation of a discussion. That makes it difficult to have others follow what you mean. The volume of what you're posting is also approaching WP:TE / WP:DEADHORSE levels of involvement, particularly given the struggles to get a rational debate out of this.
Second, you need more concrete examples, and more than this Paris and roads analogy. I have no idea where your initial concern rose from, but a good example that we can all judge in the present tense is something like how Trump is handled in sources. You don't have to use that, but its far easier to put things in concrete terms than these abstracts you're dancing around.
Third, if I think the point you are trying to make is what it is, I have long ago believed that our neutral POV should try to reflect the nature of a matter that we know exists in the real world but which might not be represented by reliable sources (particularly given the left-lean we have collalesed around due to the quality of these sources and the general trend of lack of quality on more right-leaning ones). That is, I once though we should strive to capture the state of views based on what the overall population thought, and not that told to us by reliable sources. But I've been through enough editing on my contributions and in debates in other places (eg Gamergate) that I recognize why we reflect what the state of views are on a topic as told by what RSes give us, and not what we cannot source properly. WP:V is more fundamental than NPOV in terms of content policies, so if it cannot be verified in a reliable source, it goes nowhere for us. We certainly should try to make sure that minority viewpoints that are captured in RSes are not outright ignored (that is, despite our RSes not including many right-leaning sources, we often get the right wing views from left-leaning sources that cover that), but we should not force a false balance or any type of artificial weighting to force minority viewpoints otherwise not covered by RSes. I will still argue that there is a need to apply careful application of attribution and language to make sure we aren't rushing to state majority views in Wikivoice, but that's a wholly different matter than acknowledging that NPOV depends on what's provided through the sources that meet WP:V.
If that is not what you are arguing, then I think you've completely lost what point you are trying to make here, and it may be better to take a short break, collect your thoughts, and come back with a more reasoned argument. — Masem (t) 04:40, 4 June 2024 (UTC)

@Masem:, I couldn't have said it better. I share your concerns about the bludgeoning and huge amount of material. I also agree with your comments about WP:V and editing. We obviously don't just take what RS say and plop it down and think that's all that needs to be done. We do exercise common sense and editorial judgment in how we frame our content. My concern is that RS, not editors, should be the only source of bias, hence my belief that we should let RS speak without editorial interference. By "interference", I am not referring to normal editorial processes, but to what happens when editors censor or manipulate content to fit their own beliefs and opinions. We want to hear the opinions of the sources, not the opinions of editors. -- Valjean (talk) (PING me) 16:32, 4 June 2024 (UTC)

Note added: I find it annoying that people insert comments without respecting the chronological order. Now, I need to indicate clearly that I am replying to the 04:40, 4 June 2024 (UTC) comment, not to mention that the extra indent needed for the inserted comment is violating manual of style for accessibility by jumping more than one level. Dominic Mayers (talk) 20:24, 4 June 2024 (UTC)

Your reply suggests that you believe that I am endorsing original research. I am speaking about a legitimate and required editorial process, not about some illegitimate original research. You have given the most trivial case of a valid editorial process: attribution. So, you do understand the distinction between a valid editorial process and illegitimate original research. That simple case, should be enough to understand the general idea. The idea, when seen in the light of that simple case, is that when you attribute and write "Wacko Smith said the earth is flat" instead of "The earth is flat" you are doing the editorial work that is needed for not taking sides. This is the key point: not taking sides happens through the important editorial process. As pointed out by Beland, just because one side gets more weight doesn't mean Wikipedia has taken that side, or even appears to be doing so. If people, despite this example, still do not fully realize the very important role of the editorial process, say because they confuse it with original research, then they will confuse "due weight" with "taking sides". They will not understand that the most important in the editorial process is to "avoid taking sides" while giving "due weight". Really, it is clear, that many people seem to do that mistake when they insist that we must take sides. With a proper understanding that "not taking sides" is the essential ingredient of a good editorial process, they will never say that. Dominic Mayers (talk) 12:16, 4 June 2024 (UTC)

Yes, tone, and word choice, and structure do have something to do with the neutrality we are seeking, but that's just it, it is a part and you still have to have communal standards shared measurements (so probably something objective to look outside ourselves) on how to get to that and the rest. Alanscottwalker (talk) 12:36, 4 June 2024 (UTC)

It is not just tone, phrasing or organization (the structure), it is also about the actual content. This is said in the second part of the nutshell: This applies to both what you say and how you say it. (emphasis is mine). Dominic Mayers (talk) 12:42, 4 June 2024 (UTC)

Well, I just said that, and the point remains. Alanscottwalker (talk) 12:47, 4 June 2024 (UTC)

I did not understand your point, but if you say that it says what I said and is in the nutshell, I will not dig further. OK, I just digged further and perhaps you simply meant that we still need to avoid doing original research. Of course, we all agree with that. Dominic Mayers (talk) 12:53, 4 June 2024 (UTC)

I thought the point clear, this and every policy is functional and its function is to get a bunch of people to pull together, so there are a bunch of agreements outside ourselves we have to make for measurement, standards, etc., and it is decidedly not an ontological exercise. - Alanscottwalker (talk) 13:59, 4 June 2024 (UTC)

I believe you are simply warning that getting an agreement on policy is not obvious. Sure, I agree. In fact, I am thinking that all the times. Dominic Mayers (talk) 15:27, 4 June 2024 (UTC)

Extra comment by Dominic Mayers (talk) on the process itself.
The process to agree on policy depends a lot on the kind of communication used in the community. In all cases, it depends much more on feelings, social capital, etc. than on intellectual discourse. Perhaps this is what Alanscottwalker is trying to warn me about. Dominic Mayers (talk) 20:06, 4 June 2024 (UTC)

Does Wikipedia "take sides"?

Added later: "The short answer is NO! It only appears to do so, and I explain why it appears to do so." -- Valjean (talk) (PING me) 16:02, 4 June 2024 (UTC)

Proposed new section:

Does Wikipedia "take sides"?

The nutshell of NPOV says: "Articles must not take sides, but should explain the sides, fairly and without editorial bias."

But what about when it appears that Wikipedia "takes a side" and is "biased" toward that side? How can that even happen? It all depends on whether or not there is any significant disagreement between reliable sources. In either case, Wikipedia remains neutral and lets reliable sources speak.

When there is no significant disagreement between reliable sources, Wikipedia appears to take the side of mainstream reliable sources (over the views in unreliable sources) because it allows reliable sources to get the full due weight they deserve (unreliable sources have zero due weight). Wikipedia stays neutral by not interfering with what reliable sources say and by enabling them to voice their views without the interference and distortion of editorial bias.

By contrast, when there is a significant disagreement between reliable sources, Wikipedia does "not take sides, but should explain the sides, fairly and without editorial bias".

The NPOV policy requires articles to fairly and proportionately represent the views published in reliable sources. It does not permit editors to "correct" or remove biases they see in sources, or to allow their own beliefs and opinions to "get between" the sources and the article content. Editors should put their own opinions aside and "stay out of the way" by neutrally documenting what a source says, including its opinions and biases. That means that when editors edit neutrally, Wikipedia content will reflect the biases found in reliable sources, and that form of bias is okay. It is "editorial bias" that is wrong.

Let's brainstorm this. -- Valjean (talk) (PING me) 18:27, 3 June 2024 (UTC)

I think the nutshell could profitably be re-written from the ground up, but I think it should be done after the policy itself is restructured, and I propose that the best method for doing this involves time travel. WhatamIdoing (talk) 18:41, 3 June 2024 (UTC)

Oh crap! Now you've got me crying when I think of Sarah "SV" (SlimVirgin). She is missed. -- Valjean (talk) (PING me) 18:49, 3 June 2024 (UTC)

This proposed addition seems redundant to the "Explanation" and "Due and undue weight" sections. The claim that reflecting the biases found in reliable sources is OK sounds bad, and is bound to be extremely controversial and generate a lot of outrage and possibly bad press coverage. I'm happy with the NPOV policy as it is. -- Beland (talk) 18:49, 3 June 2024 (UTC)

It may or may not be redundant, but some parts are not mentioned at all here. It's usually only controversial with the fringe who are unhappy with what RS say. They already "generate a lot of outrage" when we allow RS to speak and refuse to create a false balance to soothe their fringe feelings. It's also consistent with current practice and interpretation of multiple policies. It's largely a "no false balance" entry that also makes it clear that bias from sources, unlike editorial bias, is allowed. -- Valjean (talk) (PING me) 20:02, 3 June 2024 (UTC)

I have no problem at people pushing fringe POVs being outraged at the "due weight" rule. That's a sensible rule that I can stand behind. Saying things like "Wikipedia takes sides when reliable sources don't disagree" and "Wikipedia reflects the bias of its sources" will produce reasonable rebuttals like "Wikipedia shouldn't take sides, it should be neutral" and "Wikipedia should correct for the bias of its sources". Those rebuttals sound completely reasonable to me, and will to many people who will not understand the philosophical nuance of your argument. I do not see any circumstance where adding those sentiments to the NPOV policy would change the existing rules in any way or affect the outcome of any discussion, do you? If not, they are really best left unsaid. -- Beland (talk) 21:47, 3 June 2024 (UTC)

The idea of "taking sides" in the "no significant disagreement" case arguably doesn't make sense. Wikipedia doesn't appear to or actually take a side when there are no sides to take. It merely reports consensus reality. That is, in fact, nearly all of what the encyclopedia does. -- Beland (talk) 21:50, 3 June 2024 (UTC)

Something that WAID brought up as example would be something to consider in talking about the "no significant disagreement" section. While today the media sentiment around Trump is "Trump is the worst president in US history", it would seem far too early to treat that as a fact, dispute all other factors of "no significant disagreement" given above being met. It's not a stance that WP should take until some years have gone by, and we have more academic/less news media evaluation of Trump's presidency. It's not that we can't talk of this sentiment in attributed form, just not in factual. Basically we should not be trying to decide when there is "no significant disagreement" in the short term.

Note that this still means that UNDUE applies as well as no false balances in the short term. There are not a lot of views from RS that present Trump's president as highly rated, so we'd still mostly have commentary from RSes that would easily summarize the presidency as one of the worst, just that we'd use appropriate language and attribution to keep the sentiment out of wiki voice. Masem (t) 19:01, 3 June 2024 (UTC)

I agree, as long as we interpret "should take" in your "It's not a stance that WP should take" as "documenting a stance that RS take". Wikipedia just reflects the RS, not some permanent "stance". That means that something like this example, can change over time. That's why our articles are "never finished". -- Valjean (talk) (PING me) 20:02, 3 June 2024 (UTC)

Yes, I mean, in my context, not a stance WP should take in wiki voice, but absolutely something we should document with necessary language and attribution in the short term, as long as we are otherwise following UNDUE. Masem (t) 20:42, 3 June 2024 (UTC)

Absolutely. We are not talking about wikivoice. That would be a case-by-case matter determined by local consensus. When in doubt, use attribution, by all means.

Wikipedia doesn't really "take a stance", it just sides with RS by giving them due weight over unreliable sources. To the fringe crowd, that will always seem like Wikipedia is "taking sides", when in fact that is only a "perceived" bias from Wikipedia's editors. In reality, it's the bias found in RS, and editors are neutrally letting RS speak. That's our job, and we must keep our opinions and biases out of the editing process. -- Valjean (talk) (PING me) 20:58, 3 June 2024 (UTC)

Yes, agree with all that; but as I've discussed elsewhere many times before, our reporting of the dominate POV on a recent topic or event in the short term should not be written as if the dominate POV was fact. (that is we should report, factually and in wiki voice, that a dominant viewpoint on a topic is X, but not to a point where we are saying X is factual in wiki voice) With time and more sources that are independent, secondary, and looking back in time, we may end up treating the dominant view as a fact in wiki voice, but that's on the order of decades. I see a lot of cases were editors want to rush to convert a dominant viewpoint into a fact too close to events or even while the event is ongoing, on the basis that if there is only one dominant viewpoint and no contestable positions, that viewpoint must be fact and the NPOV rules on "don't report facts as opinions" apply and state the domininant viewpoint as wiki voice fact.

A lot of this is the fact that we as a whole obsess on current event articles, wanting to include opinions from every posdible RS that reports on these. This is not encyclopedic writing where we are to summary of nature of opinions about a topic. That's a larger problem beyond NPOV but NPOV is affected by it. Masem (t) 21:12, 3 June 2024 (UTC)

I think due weight and point of view are orthogonal. Just because one side gets more weight doesn't mean Wikipedia has taken that side, or even appears to be doing so. Consider an article like Horoscope, where the majority of the text explains ideas which evidence shows are completely wrong. Just because that part of text is longer doesn't mean Wikipedia "appears" to be taking a pro-horoscope stance, given how thoroughly it dismisses the idea that horoscopes could possibly be or ever actually are more accurate than chance. Sometimes it's interesting to learn about wrong ideas in great detail so one can see how incorrect conclusions can be arrived at, or why people are motivated by them to do certain things.

I also think this debate is mostly philosophical and has no bearing on how NPOV is actually implemented. If the goal is to communicate a subtly of NPOV to people who you think just aren't getting it, perhaps this explanation would be better as a supplementary personal or small-group essay, rather than (almost certainly unsuccessfully) trying to become a policy modification. -- Beland (talk) 06:16, 4 June 2024 (UTC)

I think this is a valuable conversation. But maybe there's a better venue, like here or here? Philomathes2357 (talk) 06:36, 4 June 2024 (UTC)

Define 'significant disagreement'. Only in death does duty end (talk) 20:06, 3 June 2024 (UTC)

Let's use a couple examples. I would guess it means quite a bit more than the less than 3% of climate scientists who doubt global warming. There are real disagreements in science, and there are manufactured/fake disagreements that are exploited. The climate skeptics and anti-vaxxers claim there are serious disagreements in science and medicine, when there really aren't. There is a lot of fame and money to be made from pushing conspiracy theories and denying common knowledge and reality. -- Valjean (talk) (PING me) 20:51, 3 June 2024 (UTC)

The most common form of bias in Wikipedia isn't about presenting the two sides of the issue, (or Wikipedia explicitly "taking a side") it's about which information is covered/ overcovered and which information is isn't covered / is undercovered. And in certain areas (e.g. US politics) it certainly does have a systemic bias. At the core of that is the "unbiased means echoing what the wiki-selected US media sources say/cover" thought process. I've stopped worrying about the higher bar of being unbiased, and just get concerned when it gets so bad that it affects the informativeness of articles, which does happen. North8000 (talk) 20:17, 3 June 2024 (UTC)

In a way, "not taking sides" includes not being attach to the specific content that originally define each side. I explain what I mean here in my comment above. An example of extra information is the arguments used by each side or anything that makes Wikipedia more informative/descriptive than engaged. From this angle, I agree that the information not covered can often be the issue, but I am not sure what exactly is the point in your case. Dominic Mayers (talk) 21:04, 3 June 2024 (UTC)

From my perspective we pick the side of NPOV... That means inherently we pick a number of sides... It means that we're anti-authoritarian, anti-aristocratic, anti-mystic, and anti-fringe. One of the earliest complaints about encyclopedias is that they didn't respect the "natural order" of the world because they listed all the worlds things together... Meaning commoner could come before noble and sacred could be sandwiched by profane... So the simple act of arranging an encyclopedia has never been a neutral act in the larger sense, it can't be... But it can be done from a neutral point of view. Horse Eye's Back (talk) 20:52, 3 June 2024 (UTC)

Because some people here are not getting my point above, I feel compelled to provide an answer to the original question: "No, Wikipedia does not take a side on content. It only takes a side on sourcing. As proven by the WP:RS policy, it sides with RS." Any bias found in articles should not come from editors taking sides. It should only come from the sources. Therefore, any evidence of bias should not be perceived as Wikipedia taking sides. That's a misunderstanding of how Wikipedia functions. Wikipedia does not take sides. That is still unchanged. -- Valjean (talk) (PING me) 22:20, 3 June 2024 (UTC)

Because this is not addressd in the NPOV policy, I feel it should be addressed. That's what this thread is about. -- Valjean (talk) (PING me) 22:25, 3 June 2024 (UTC)

It all depends on how you formulate it... If NPOV is a side we take a side... If NPOV is the lack of a side we do not. Perhaps I am biased because of my political science background, in political science neutrality is a position not the lack of a position (for example Swiss neutrality) in the same way that being non-aligned is actually being aligned... There are I guess two questions here, a practical one in which I find myself in almost complete agreement and a philosophical side where there becomes a chicken and an egg problem if not taking a side is taking a side. Some see neutrality as being against everyone, some see it as being against no one, and some see it as both... That seems to me more than anything in the actual wording of NPOV to alter how different editors perceive the policy. How to address that problem? I really don't know, I wish I did. Horse Eye's Back (talk) 23:14, 3 June 2024 (UTC)

You do not understand that the neutral point of view is not obtained by following what reliable sources say. It involves a lot of editorial work. In the french version, they even refer to this editorial work as "personal". The trivial case of this editorial work is attribution. Attribution is not at all a trivial work. So, only with this case, we can see that the neutral point of view is NOT following what reliable sources say. Another example is this part of the policy that says try not to quote directly from participants engaged in a heated dispute; instead, summarize and present the arguments in an impartial, formal tone. It's way more than that. The very concept of a synthesis is a lot of editorial work that must be done in the neutral point of view. Because of all these points, the neutral point of view goes way beyond simply following what the reliable sources say. Dominic Mayers (talk) 23:29, 3 June 2024 (UTC)

"neutral point of view is not obtained by following what reliable sources say"?? Of course not. It is not "obtained". NPOV is an editing attitude. The expression "neutral point of view" is misleading because the "N" in NPOV refers to an editorial attitude and mindset; it is not a true "point of view". Editors have their inclinations and biases, but when they are editing they must put on their "editor's hat". That "hat" is a neutral attitude and mindset, since NPOV is not a true "point of view" which can be included in an article. Like a referee, they are responsible for presiding over the article with a neutral and objective attitude. As long as their biases do not cause them to violate policy, there should be no problem. While editing, editors must remain apathetic, disinterested, and even-handed towards the subject, regardless of their personal POV. -- Valjean (talk) (PING me) 23:50, 3 June 2024 (UTC)

For a long time we were clearly referring to the neutral point of view as the actual point of view adopted by Wikipedia and I believe we still do now, but less often. In any case, it's a perfectly fine way of using the expression "the neutral point of view". In fact, this way of using the expression helps a lot to explain that the neutral point of view is NOT the point of view in reliable sources. Dominic Mayers (talk) 00:02, 4 June 2024 (UTC)

I disagree. I find Wikipedia's articles on contentious topics are better when there are editors with multiple biases, who check each others' work and in the end all come to agreement that the resulting content treats their own point of view fairly, either by not unnecessarily saying something they object to, attributing opinions they disagree with, or doing a good job explaining their POV without endorsing it. Editors do need to refrain from changing content to reflect their own point of view by treating other POVs unfairly, but that's not the same as adopting a judge-like neutrality mindset.

I also don't think this is worth arguing over; what problem is it trying to solve? -- Beland (talk) 00:04, 4 June 2024 (UTC)

It's fine when multiple editors with different biases all come together, all wearing their editor hats, to set aside biases and write a comprehensive picture of the viewpoints about a topic. But that is the exception. Most of the time we have editors demanding POVs that align with their biases be included but typically failing to provide any RSes or showing how that view is more that fringe. But we also have editors with biases supporting the dominint view that also tend to gatekeep articles even when valid sources of alterive viewpoints are presented. That's all behavioral issues somewhat beyond the scope of NPOV but NPOV should be clear why both sets of behavior are unacceptable. Masem (t) 13:40, 4 June 2024 (UTC)

A box from my talk page. -- Valjean (talk) (PING me) 17:41, 4 June 2024 (UTC)

Talk page negotiation table

"The best content is developed through civil collaboration between editors who hold opposing points of view."
by Valjean. From WP:NEUTRALEDIT

"The quality of Wikipedia articles rises with the number of editors per article as well as a greater diversity among them."[1]

When all else fails, AGF and remember that

We Just Disagree
So let's leave it alone, 'cause we can't see eye to eye.
There ain't no good guy, there ain't no bad guy.
There's only you and me, and we just disagree.
by Dave Mason (Listen)

Note added: The previous comment does not respect the chronological order. Besides, I find that such opinion about good behaviour is much better expressed by simply doing it concretely when we interact with others, not by showing off at the wrong time. I decided to not participate further in discussions, but I do check that we do not disrupt the flow of existing threads. Dominic Mayers (talk) 17:27, 5 June 2024 (UTC)

You worry that editors' biases can result in undue weight. Undue and due weight is indeed an aspect of NPOV. It was added to it in 2005 by making use of the 2003 NOR statement of Jimmy Wales. Later, I believe in 2006, it was expressed in terms of "reliable sources". Nowadays, some people even feel that the essence of NPOV is to follow reliable sources while respecting due weight. I know that you have difficulties following my logic when I explain that it is NOT the essence of NPOV. The essence of NPOV deals with aspects of the editorial process, such as not taking sides, that go way beyond following reliable sources and respecting due weight. My concern with an emphasis on due weight is actually a concern that the notions of not taking sides, attribution of opinions, to not engage, but describe, etc. are not enough emphasized and even misunderstood. There is no issue per se in mentioning reliable sources, but if you understood the importance of these other aspects, you would understand my analogy with an emphasis on reliable roads when trying to explain the way to Paris. Dominic Mayers (talk) 14:04, 4 June 2024 (UTC)

Those concerns don't seem to be solved by the proposed text, and I'm not sure any text changes would address them. There will always be people trying to push a POV into the project, and new contributors hardly ever read WP:NPOV before doing that. WP:NPOV and WP:V and WP:OWN already explain why those behaviors are unacceptable. We simply need other editors to engage in civil conversation, point people at those policies, point out exactly how they're not being followed, make edits to enforce them, promote a culture of following those policies, and report persistent and willful violations to admins. -- Beland (talk) 17:06, 4 June 2024 (UTC)

This is pointing to the wrong place. First: it is wrong to say "WP:RS policy", it is a guideline and it does not "prove" anything. Second, the first sentence of WP:NPOV points not to WP:RS but to WP:V -- at one point in 2010 it was specifically pointing to WP:SOURCES within that (see "as defined by the WP:SOURCES sourcing policy", I haven't traced to where somebody made it vaguer but that's the relevant part of WP:V). Near the top of WP:RS are the words "for Wikipedia policy on reliable sources" and a pointer to Wikipedia:Verifiability#Reliable sources -- because the WP:RS guideline tail does not wag the WP:V policy dog. Sprinkling the vague linkless words "reliable sources" elsewhere in WP:NPOV is a mistake which causes confusion like this idea that WP:NPOV needs WP:RS, but not enough confusion to overthrow 14 years of pointing to WP:V. It is addressed in the WP:NPOV policy by pointing to WP:SOURCES. Peter Gulutzan (talk) 23:34, 3 June 2024 (UTC)

"this idea that WP:NPOV needs WP:RS,"?? No, it doesn't "need" RS. RS are just the basis for content, the dough and ingredients. NPOV is how we prepare and handle that dough. -- Valjean (talk) (PING me) 23:53, 3 June 2024 (UTC)

"In either case, Wikipedia remains neutral and lets reliable sources speak." That a source is reliable (by Wikipedia standards) does not mean much when a) it covers topics outside its area of expertise, b) it is significantly outdated, and c.) it expressed views held by a minority of professionals in a relevant field. Wikipedia is supposed to summarise the majority views in topics such as archaeology, and to be reasonably up to date. In relevant discussions, we have had to distinguish between sources covering the mainstream views of (for example) the 1920s, the 1950s, and the 1990s. Dimadick (talk) 00:01, 4 June 2024 (UTC)

This is an immensely important topic but has so many very different wide ranging scenarios and wiki-universes bundled into it that it really need to be split up if there is any hope of progress. North8000 (talk) 00:12, 4 June 2024 (UTC)

@North8000:, I agree. There has been such a huge volume of content and discussion of tangential, or even unrelated issues, that it's impossible for me to figure out anymore. I may have included stuff that wasn't essential, and that may have contributed to what's happened. I am not always good at explaining things. Maybe you can help to cut to the chase. Maybe in a new thread? -- Valjean (talk) (PING me) 16:02, 4 June 2024 (UTC)

@Valjean: Cool. Maybe: Are changes needed in wp:weight? North8000 (talk) 17:44, 4 June 2024 (UTC)

It's the wrong question

The question is Does Wikipedia take sides? We will have a lot of answers depending on the interpretation of "taking sides". Some, will say that Wikipedia must in many cases take sides, but this is because of what they associate with "taking sides". It's the wrong approach. The correct approach is to respect the idea that has been there and has defined Wikipedia for more than 20 years and try to agree on how to practically achieve it. The point is that, not taking sides, in the sense of being neutral, is generally accepted as positive for an encyclopedia and we should start with this as a premise and build around it. The way I understand it, "not taking sides" can be achieved by being more descriptive/informative, more factual, less engaged in superficial ill informed debates. Even those who originally say that Wikipedia must often take sides, without fundamentally changing their mind, might agree this is a good thing. Dominic Mayers (talk) 21:51, 3 June 2024 (UTC)

"more factual" You can't be factual while not taking sides. One of the reasons several Wikipedia articles tend to read like public relations campaigns is that the encyclopedia's house style tends to downplay the controversial aspects of several topics. That is obscurantism, not factual reporting. Dimadick (talk) 00:08, 4 June 2024 (UTC)

@Dimadick:, see my explanation in the next section. Dominic Mayers (talk) 02:14, 4 June 2024 (UTC)

Application of NPOV to the last USA election (and a bit to earth is flat)

I wanted to dig a bit more into an example given by Dominic, because I am not convinced that we are all on the same page. We say "Wacko Smith said the earth is flat" instead of "The earth is flat", and in doing so we follow NPOV. But we do not say "Mainstream Scientist says the earth is round" - instead, we refer to the earth as round in wikivoice (usually by referring to characteristics, such as the Earth's diameter or circumference, that follow from it being an oblate spheroid). In my view, it would be an NPOV violation were we to restrain our text to attributed statements about the actual, consensus reality, majority view/common episteme understanding of the shape of the earth.

Also, on a different but related topic (and since Masem is in this recent thread): I have seen him offer the following caution more than once:

...our reporting of the dominate POV on a recent topic or event in the short term should not be written as if the dominate POV was fact. (that is we should report, factually and in wiki voice, that a dominant viewpoint on a topic is X, but not to a point where we are saying X is factual in wiki voice) With time and more sources that are independent, secondary, and looking back in time, we may end up treating the dominant view as a fact in wiki voice, but that's on the order of decades.

What I want to know is, from this perspective, when would be the right time to say in wikivoice that Biden won the 2020 presidential election, rather than rattling off a long list of authorities saying that he did? Division on this question is more substantial, and more consequential, than the flat-earth POV; for people on the "attribute more" side of the spectrum, and especially Masem's "history requires a baseline of decades" presupposition, when would we be allowed to state the election result as fact? Newimpartial (talk) 16:31, 4 June 2024 (UTC)

I just want to say that a policy is always made of abstract principles which we do not illustrate by polemical examples (or examples that can be seen as polemical), even though they are the practical cases for which we need the policy. It is more pedagogical to take extreme examples such as the earth is flat, which hopefully avoid polemical discussions, to illustrate the principles. The goal is only to convey the basic ideas, which can then be further discussed in practice when we meet polemical cases. In this section the previous section, the goal was only to convey the basic idea that the editorial process is important, way more than a simple transcription process, and that it is at this level that the notion of not taking sides applies as something always needed and completely compatible with giving due weight. This is the only idea that this section the previous section wanted to convey. The purpose was not to explain all the possible approaches to achieve the neutral point of view and not taking sides and how these approaches can be used in practical polemical cases. Dominic Mayers (talk) 17:17, 4 June 2024 (UTC)

When reliable sources say "won" rather than "appears to have won" or "claims to have won" or "is projected to win". Generally this is when official preliminary vote counts are released and when the number of votes yet to be counted is smaller than the margin of victory, in enough states to win the Electoral College. They would refrain from doing so if there were credible allegations of widespread vote fraud or rigging, which for this election there were not. We would follow those sources in noting any irregularities such as voter suppression, jailing of opponents or people attending rallies, exclusion of one party from the media, etc. -- Beland (talk) 17:17, 4 June 2024 (UTC)

It makes sense. I imagine myself describing the result of the last USA election and I think I would keep it simple and gives the results that are recorded. I don't see where is the problem. I feel this is going too much into details. There are people here that say "Wikipedia must take sides". I believe this is a much more serious issue, because it could eventually result in a modification of the nutshell. That is why it is very important to remove the confusion between "giving due weight" and "taking sides" and that is the purpose of that section. It may seem a very abstract purpose in relation with practical concerns regarding the last USA election, but it is actually very concrete, because we are talking about the nutshell of the policy. Dominic Mayers (talk) 17:48, 4 June 2024 (UTC)

@Dominic Mayers:, you write: "There are people here that say "Wikipedia must take sides"." Where is that being said and by whom? Please provide exact quotes. -- Valjean (talk) (PING me) 17:58, 4 June 2024 (UTC)

I have to search them just like you could. I just made a small search and obtained

Wikipedia articles do, in fact, "take sides" between geocenteic and heliocentric models of planetary motion, and between young earth creationist and evolutionary accounts of life on this planet.

The idea that Wikipedia must take sides, must even be biased, etc. has been mentioned in other discussions in the past, like weeks ago. It must be a recurrent phenomena. And this is often done in clear opposition to not taking sides as stated by Jimmy Wales and still today in the nutshell. Otherwise, I would not be concerned. I don't feel the need to show all the diffs. Anyone that doubts it can do it for himself. Dominic Mayers (talk) 18:10, 4 June 2024 (UTC)

It is a question of definition. Taking the referee simile from the last section, there is a time when the referee has to decide "yes, that was a foul" or "no, that was no foul". After that, lots of fans from the side that had a disadvantage from the decision will say the referee is biased and is taking sides. Using their definition of "taking sides", yes, we are taking sides. Using their definition of "bias", yes, we are biased. (That is the attitude of the tongue-in-cheek WP:YWAB essay.) By the usual definition, we are not. (That is the attitude of the opposite essay WP:NOTBIASED). It is pointless to argue which is right. They are both right but the second one does not understand that. --Hob Gadling (talk) 13:37, 5 June 2024 (UTC)

I said I will not reply, but will be brief. The problem is also that we assume that "not taking sides" is a simple process such as doing a transcription of a POV without interfering. Reducing "not taking sides" to "attribution" is also not the solution: it results in excessive attributions or to the wrong question "should we take sides". It is a complex active process that may include giving the arguments, adopting a more neutral tone, etc. It must be done as much as reasonably possible–this should never be put in question. It naturally results in giving due weight (some might say without interfering here), so it is not at all opposed to due weight, but it is far from being only that. Dominic Mayers (talk) 13:58, 5 June 2024 (UTC)

In terms of American politics, Wikipedia is like a referee who's wearing one of the team jerseys while he works.

Of course Wikipedia is biased, by the "usual" definition. That's a widely acknowledged fact. Through Olympian-level mental gymnastics, words like "truth", "bias", and "neutrality" have been manipulated, to disguise this fact. All of those words have common definitions, and separate Wikipedia definitions. If the "public" knew this, I don't think they'd be very happy about it. Philomathes2357 (talk) 23:30, 7 June 2024 (UTC)

Whether Biden won the election or not is not dependent on any subjective measure. It's an objective yes or no answer once the vote tally and electoral college processes are completed. There is no viewpoint aspects here. The few that argue about stolen election have been shown to be operating on conspiry theories, which falls into Fringe.
When the collusion is based on subjective assessments, like Trump being the worst president, that's when we should wait to determine how perspective settles down on the topic to treat it as fact, even if the short term set of viewpoints would suggest this was the only view of the topic. — Masem (t) 17:28, 4 June 2024 (UTC)

And to add, usually when we are writing about the results of the election between voting day and inauguration, we typically use the same language in sources like "president elect" or "presumptive winner" that states the results are still yet to be 100% validated, but that's still and objective stance, not subjective — Masem (t) 17:34, 4 June 2024 (UTC)

I appreciate that answer, but to take the next step, what about the claims of widespread voter fraud that fuelled Trump's "stolen election" narrative? The article Efforts to overturn the 2020 presidential election currently describes these among efforts by Trump and his allies ... to promote numerous false claims and conspiracy theories. Is it still TOOSOON for this to be pronounced in wikivoice? At what point was it - or will it be - appropriate to make this kind of statement, in terms of NPOV? From the latest polling I have seen, at least 30% of Americans currently believe at least some of this "stolen election" narrative, so it is not a settled issue in terms of public opinion. But for RS, it is. Newimpartial (talk) 17:49, 4 June 2024 (UTC)

My comment here is the same as my previous 17:48, 4 June 2024 (UTC) comment. The answer to the specific question is not as simple, but the basic point of my comment still applies. Dominic Mayers (talk) 17:53, 4 June 2024 (UTC)

Again, the starting point is that who won the election is an objective measure. It has also been an objective measure by courts and independent parties that no significant fraud has occurred in comparison to other elections.

That a significant fraction of people want to ignore these facts does not play into how we report the election and lack of fraud as fact, and treat those claims otherwise as conspiracy theories. Thus, there's no time issue here in that. NPOV does not cave to cases where there is this denial of objective reality. Masem (t) 18:18, 4 June 2024 (UTC)

I think you missed the question. The question was not referring to the result of the election, but to "Efforts to overturn the 2020 presidential election". It is a bit more complicated. I don't know what Trump said, but his lawyers could argue that it was not the purpose. Well, maybe now, after the recent jury decision, there is something to say, but it will go into appeal. It seems more complicated. Dominic Mayers (talk) 18:31, 4 June 2024 (UTC)

There is no issue from NPOV to document the efforts by these groups to overturn the election, as long as we express their reasons in attributed voice and state that the bulk of these have been shown by expert sources (courts and independent studies) that these beliefs are wrong and treated as conspiracy theories. The objective facts are settled by RSes. — Masem (t) 19:06, 4 June 2024 (UTC)

It is not too soon to consider the matter settled; reliable news sources consider it settled and simply say that the claims of fraud are false. The claims have been adjudicated by many different courts and certified by officials in both major parties. There's a difference between factual claims where there is plausibility on both sides, like "gun control will reduce the murder rate" and factual claims which have been proven false beyond reasonable doubt. -- Beland (talk) 19:16, 4 June 2024 (UTC)

I must be missing the point, but asking whether the claim that the election is invalid is itself valid or not seems the same as asking directly whether the election was valid. In other words, it is the same question as before. In fact, the arguments for deciding if it must be attributed or not are also the same as before. I thought it was a different question, this time about the intention of people when they make that claim. In any case, I maintain that it is not what this section the previous section is about and I don't think this diversion from the original purpose of that section could result in any improvement of the policy, because it goes way too much into details. Dominic Mayers (talk) 19:32, 4 June 2024 (UTC)

@Newimpartial, Beland, Masem, and Valjean: I am going to follow what was perhaps a hidden advice from Alanscottwalker and not participate in discussions in this talk page anymore, but eventually we will need to make sure the wording of the policy removes any confusion between taking sides and giving due weight. It is just not a good time now. I will appreciate that we do not notify me, refer to me or to the content of my previous comments in an explicit manner unless it is necessary, because, of course, I would then feel oblige to respond. Dominic Mayers (talk) 20:39, 4 June 2024 (UTC)

There is a difference between saying who won the election and statements that it was stolen (by whatever their criteria is for "stolen" or even mindlessly saying/thinking "stolen"......for some it might mean allowing too much proxy voting vial mail-in ballot) Many people who say it was stolen would also agree that Biden won. So wording should be accordingly. Now, if you want to see a more common example of a wikipedia NPOV problem it is lumping the 30% making a vague claim of "stolen" in with the tin hat few % who explicitly say that Trump won the election. North8000 (talk) 14:38, 5 June 2024 (UTC)

So I am unsure what you are referring to as lumping ... in or what kind of NPOV problem you intend to point to. It seems to me that, from an objective standpoint, (i) Biden won the 2020 election, (ii) no significant irregularities took place that could cast doubt on the result, and therefore (iii) the 2020 election was not "stolen". I also believe - and this is a related but not an identical point - that the consensus of WP:RS, especially WP:HQRS, supports all of these as being objectively valid statements. I therefore conclude that all three of these should be stated in wikivoice (with the typical language for the third point being a reference to something like "false claims that the 2020 election was stolen"). And in fact we do currently state all three things in wikivoice in various articles.

What I am less certain about, is whether all three of these things are clearly "facts" rather than "opinions". Or rather, I'm fairly certain that the enwiki community doesn't entirely agree on whether they are all facts rather than opinions - I suspect there is less agreement about the inherent nature of the statements than there is agreement that they should all be stated in wikivoice. And I think this is where the "follow the RS" line of argument comes in - statements that involve some degree of judgement are assessed as valid or invalid all the time, the RS and HQRS we use do so, and when there is near-universal consensus among sources that a statment is valid, we should treat it as a "fact" for purposes of "don't give attribution for agreed-upon facts". I believe we already usually apply this principle (correctly) even when the "facts" involve such a degree of judgement that, if RS were to disagree about them to a significant extent, we would consider them to be opinions.

The way I see it is that whether a specific statement represents a "fact" or an "opinion" depends as much on the episteme as does whether or not the statement is valid. Most of the time editors can agree whether a statement is fact or opinion for the same reason they can usually agree whether it is valid - because they are able to decide based on a shared episteme. But in liminal cases editors can disagree both about whether a statement is valid and about whether it is fact or opinion. And my view of this situation is that the key question for article text should not be whether the statement is fact or opinion, but whether it is universally (or near-univerally) regarded by the relevant RS as valid.

To return to the shape of the earth for a second, it pretty clearly represents a fact, not an opinion, in most possible episteme. But if we were writing in an episteme where it was an unsettled factual question, as the existence of Dark matter is for us now, we should make statements about it using attribution even if we regard the question itself as factual. And likewise, while the question whether the 2020 US election was "stolen" involves a degree of judgement and therefore - in an important sense - opinion, it is a claim sources of any reasonable quality are unanimous in indicating as unfounded and should therefore be stated as such in wikivoice. To bring this back to the topic of this page, in my view WP:NPOV requires us to adhere to what HQRS univerally regard as valid, without attribution. Meanwhile, statements that HQRS regard as values, rather than valid or invalid statements, we will attribute even when our HQRS share the same values. I'm not saying "take the POV of the best sources in wikivoice all the time"; I'm saying, "take the POV of the best sources in wikivoice for statements those sources regard as valid (or invalid) and where they all agree as to the statement's validity". Which is strongly correlated with, but not quite the same as, "present facts as facts" - as I hope this long digression has illustrated. Newimpartial (talk) 16:11, 5 June 2024 (UTC)

A fact is something, at the ultimate time scale, can objectively measured as true, likely in a rproducable way via the scientific method or similar. An opinion relies on some subjective assessment by one or more experts at that ultimate scale, and thus may not produce the same result if you asked some other expert the same question.
This, the winner of the election is a fact, just that for a short period between election day and certification we denote it as pending results, but there is no subjective aspects here.
As a different example, the origin of the COVID is effectively a fact if we had perfect investigative capabilities, but one that hasn't yet been completely affirmed due to the fact we are piecing together evidence. We say now that the leading theory of its origins is the zootropiv route, but we have to acknowledge that some non trivial voices still consider the lab leak. In time scientific methods will affirm which ever is true or otherwise consider it accepted knowledge (much like the dark matter example)
On the other hand, a statement like "Trump was the worst present" absolutely relies on a subjective call. One could point to objective evidence like the deficit or inflation, but whether and what degree those contribute to determining if Trump was the worst is subjective. Over time, just as with the COVID origin, this stance may be accepted by pol Sci and historians as a true statement at which point we can treat it as fact, but that's likely decades out from now
The reason to avoid presuming that some subjective measure, held by a majority of RSes (mostly mainstream news rather than academics) in the short term is that they are writing about the topic "now" without the impact of time and how that could reshape opinions. There is a reason we consider news reports on current events as primary sources as they are documenting as media eyewitnesses with a focus on the near term even if they engage in commentary and analysis. But broad stances and public opinion can change over time. Racism was "accepted" in the early 20th century, but now it's seen as a negative. Or for example the attitude around the Iraq War has significant changed over time from being necessary to being one of those wars America "lost". Public opinion may not change at all with time, but to try to predict that on current sourcing near an event falls into CRYSTAL. Thus before transforming a majority viewpoint, even one without any record able opposing voices, we should wait until we are reasonable certain that viewpoint has stabilized as the only viewpoint held by experts well distant from the original events (eg those independent of those events). Hence the need to consider time factors in NPOV considerations.
None of this means we cannot report the current public opinion, just that it should be taken out of wiki voice. — Masem (t) 17:39, 5 June 2024 (UTC)

One other thing which is required to be absolute fact is widespread agreement on the exact meaning of the question and of any answer. For "did Trump win the election?" one exact meaning of that question is "Was Trump deemed the winner by the USA's official process for determining the winner?" North8000 (talk) 18:41, 5 June 2024 (UTC)

I'm going to say something that is hopelessly vague but still very useful. If it's an area where the knowledge is mature and 90%+ agree, we state it as fact in the voice of Wikipedia. So we don't need to say "according to many sources, Hitler was a bad person". But there are lots of people who want to word things as if their 50% view is the 90% view. And wikilawyer to accomplish that. "According to "reliable sources"... " capitalizing on the fact that "wp:reliable source" doesn't necessarily mean actual reliability. Actual reliability would be expertise and objectivity with respect to the text which cited it. North8000 (talk) 12:40, 5 June 2024 (UTC)

This subsection has not much to do with "not taking sides" as used in the policy and it does not fit in that section. Clearly, in the policy and it has been like that for more than 20 years, "not taking sides" do not mean to attribute all statements. It means to attribute opinions, not simple facts. I understand that the questions what to attribute and what is a fact are very difficult questions, but they have nothing to do with the requirement not taking sides. I did not want to participate in that discussion, especially not about "what is a fact" or "what to attribute", because this has been discussed again and again and it will not change anything to the general policy. They are just difficulties in applying the policy, nothing more. However, when we reorganize the sections and subsections in a way that makes a very confusing point, especially this one, then I feel oblige to express my concern. Dominic Mayers (talk) 20:30, 5 June 2024 (UTC)

Dominic, please stop adding new subsection headings, whether to move the comments of others (as you did here) or to highlight new comments of your own as you just did a moment ago. - MrOllie (talk) 20:39, 5 June 2024 (UTC)

Dominic, I disagree that This subsection has not much to do with "not taking sides" as used in the policy. I get that editors mean different things when they refer to "taking sides", but one of the main things they are likely to mean (in my experience) is taking "one side" of a debate and entrenching that perspective in wikivoice. I therefore believe that various aspects of what counts as fact and what as opinion - what counts as valid within our rules governing valid statements and what counts as a value statement that can't be called "valid" except when a value commitment is presupposed - are all relevant to this discussion.

To take an example from Masem's reply above, if evaluating the claim that the 2020 election was stolen relies on some subjective assessment by one or more experts at that ultimate scale - and by any realistic assessment it probably does - then his position is that we shouldn't describe the claim as "false" until that viewpoint has stabilized as the only viewpoint held by experts well distant from the original events. Now I would say that stabilization has actually I already happened, but Masem prefers a time scale of decades or longer before accepting the consensus of HQRS. I personally don't think that is how an encyclopaedia should work - I think a good encyclpaedia published since 1967 should present plate tectonics as objectively true, not as one perspective about continents - but I do think these are all brass tacks examples of what "taking sides" (or not) actually means in practice to most editors. Newimpartial (talk) 20:55, 5 June 2024 (UTC)

In the case of the stolen election, or let's call it a question of if sufficient fraud existed that the election results fell within the margin of error created by that fraud, that was a question that did take time to be answered but after a year or so, we had judicial and measured evidence to clearly affirm that the results were outside the margin if error and the amount of discover fraud was comparible to other elections, giving the objective result fraud did not happen
At which point, we do not have to give credibility to claims it was (just as we don't give credibility that the earth is flat). Now, there is the factor that in that year long period before that fact was established, that most RSes gave little weight to considering fraud (if anything their focus was on voter suppression) because in many many decades of voting, there never was a significant level of fraud. So the default position in RSes was no fraud, and in our case, we'd likely handle it the same way, and the position there was fraud the equivalent of a conspiracy theory, with that position (claiming fraud being a conspiracy theory) strengthened as reports and court de visions affirmed nothing out if the ordinary happened. So no, here's a case that involves an objective stance that ultimately can be measured but there was no significant doubt it would be proven that way by RSes, so we treat it as fact.
As a different example, let's take trickle down economics, which at the time in the 80s was claimed to bring up everyone in the economy. There is no way to objectively prove this, since the results are heavily dependent on specific society behaviors and lack of outside influence. You can only observe and make educated conclusions if there is a connection. At its time in the 80s, it seemed right but today the theory of it working has been heavily criticized. So despite the amount of time involved, the benefits or criticisms of trickle diwn theory should still be presented outside of wiki voice. Maybe far in the future there will be enough agreement among economists that trickle down would never work to then state that factually, but we are not at that point yet. — Masem (t) 21:20, 5 June 2024 (UTC)

To show how complex these things are, "trickle down" is a pejorative term applied to supply side economics, which starts with a (straw man) assertion that supply-side's claim on how it helps/works is to give money to rich people and then be dependent on them giving it to poorer people. So, regarding the types of study that you just described, there needs to be agreement on what exactly is the question? North8000 (talk) 21:38, 5 June 2024 (UTC)

The claim - made up and down the Reagan era - would be that by giving things like tax breaks to the wealthy, that that would boost the wealth of lower classes, as to justify why tax breaks for the wealthy and corporates were passed. In the US, the expected results of this breaks did not result in improvements for lower classes, but that neither invalidates or validates the idea of this because its one example with a huge number of variable aspects that must be considered. Even arguing in the specific that the Reagen-era push for this principle was flawed still is uncertain despite the growing number of reports asserting it likely was. So from an NPOV stance, we'll never have a objective answer to this question, but that's because its impossible to measure something that has this much uncertainity. Contrast that to the more narrowly focused law of supply and demand, which while still is based on uncertain behavior, has been shown repeatedly in economic science to work over and over again, and thus generally accepted as true, hence why we present that at supply and demand as generally factual. — Masem (t) 04:44, 6 June 2024 (UTC)

@Masem: In the big picture and for the conversation here, I agree with you. My post was a bit of a tangent. It was pointing out that some putative subjects are mis-descriptions of them through the lens of a pejorative term. And so even using the pejorative name or pejorative description of the subject as the "subject" it is itself POV.North8000 (talk) 18:58, 7 June 2024 (UTC)

@Newimpartial:, yes, it is often the case that an expression has different meanings in different contexts. Not surprisingly, it is like that for "not taking sides" as well. It is not a big deal if someone uses "not taking sides" with a different meaning than the one it has in Jimmy Wales 2003 statement, as long as this is made clear in the context. I insist, you are perfectly right that "taking sides" can have the meaning "take the statement as a fact and do not attribute it", just affirm it, "take this side". In that case, "never taking sides" means "never take statements as facts and always attribute them." But clearly, given the title of the section, "The 2003 statement of Jimmy Wales", I am not referring to that other meaning. Yes, some times, we must accept a statement as a fact and not attribute it (i.e., take side in the other meaning of it), but the request to not taking sides, as in the policy, is not put in question because of that, not at all. In fact, it is part of the policy and thus of not taking sides that doubting facts as if they were opinions is not the neutral point of view.

By the way, it is not the attribution per se that is the problem. The problem is the part where we are not neutral and take position by doubting. This last point is very important, because it might be possible to attribute a fact in a way that expresses no doubt, but instead is informative, pertinent, interesting, encyclopedic, etc. Certainly, there is nothing wrong with that kind of attribution of facts. There might exists other ways than an attribution to present a fact in a neutral manner without taking sides, by adding a context. So, attributing or not is a false dilemma. Not taking sides might require that we think out of the box so that we are more informative, more neutral than simply affirming the fact without any context. It is very wrong to reduce it to attributing or not. This discussion tries to deal with concrete cases, but it cannot do well outside the real articles.

Anyway, I did not intend to participate in that discussion. After I saw that this section is now a subsection, I just wanted to explain that it is off topic with respect to the original subject of the section. It is not that what is a fact is not related to NPOV. It is a part of it, but NPOV is way more than that. It often requires that we think out of the box, outside false dilemmas such as attributing or not, fact or not. Sometimes, we must provide the arguments, etc. Now, I am out of here. Dominic Mayers (talk) 22:29, 5 June 2024 (UTC)

Notes

Proposal of a RFC about a better unification of include-info and reject-info principles

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

This RFC should happen before we consider any decision about the expression "not taking sides", because if we address this first issue, the superficial terminological debate about "taking sides" will also be taking care of. Dominic Mayers (talk) 15:02, 9 June 2024 (UTC)

Before actually starting one, perhaps it would be better to discuss first. I'd start with what you mean by "taking sides" (or perhaps better said, what you object to or what you believe others mean by "taking sides" that you disagree with), and what you mean by include-info and reject-info principles, and why that should happen first and why it would resolve the other issue, however you define it. The connection between the two may seem obvious to you as you know what you mean, but there is too much vagueness here for me to easily see it. Maybe once clarified, everybody will agree with you, and an Rfc won't be needed. Mathglot (talk) 16:19, 9 June 2024 (UTC)

"Not taking sides" is an expression similar to "the neutral point of view". Both were never "defined" in any formal sense. Instead, they were explained through examples: describe debates instead of engage them, provide the arguments, attribute opinions, etc. These are important include-info principles and these examples answer your question what you mean by include-info. It will resolve the current question about "not taking sides", because the actual concern is that it is seen as opposed to reject-info principles. It's clear that the sentence When those pushing fringe views complain about articles obviously omitting or contextualizing those views ... is a concern that "not taking sides" can prevent us from rejecting fringe theories. So, clearly, the deep issue is a reject-info vs include-info debate. I don't think we should superficially debate if we remove or not the expression "not taking sides", before the policies clearly explain the different inclusion and exclusion principles themselves, through examples irrespective of terminology, and how they work together toward a same goal. Dominic Mayers (talk) 17:02, 9 June 2024 (UTC)

It seems to me that for the principles mentioned here, such as "describe debates", "provide arguments" and "attribute opinions", the key aspect to articulate is the range of application of each. We do not "describe debate" about the round earth or the heliocentric solar system, except when contextualizing FRINGE theory or relating history of science. We do not "attribute" the "opinion" that conspiracy theories about George Soros are false: we note false claims in wikivoice. What would help clarify our policy the most (and reduce debate) is if we had some kind of consensus about when NPOV and encyclopaedic writing benefits from following these principles, and in which contexts each of them does not in fact apply. Newimpartial (talk) 18:38, 9 June 2024 (UTC)

Yes, but the first step is to describe the include-info principles and their usefulness through examples. Once they have been explained in a positive manner, it makes sense to illustrate through examples how they are well complemented by reject-info principles, including the fact that the info must be relevant. BTW, the requirement for reliable sources is often the same as a requirement that the info is relevant. For example, we require sources that represent well a scientific domain, but this is the same as saying that religious views or even views in an unrelated scientific domain are not relevant. It's not always as easy, but this challenge is not a reason for not explaining well the include-info principles. This is what I fear. I fear that because we had difficulties to reject info in the past, we started to remove otherwise useful include-info principles. Dominic Mayers (talk) 20:15, 9 June 2024 (UTC)

I don't think this policy has ever had substantial examples (WP:NOR does), but as part of this WP:RFCBEFORE discussion, perhaps you could suggest something that you think would be helpful. WhatamIdoing (talk) 21:33, 9 June 2024 (UTC)

Maybe you could name some of these principles? I can't find any reference to "include-info and reject-info principles" in previous discussions or the text (or used by any other editor anywhere else on wikipedia), did you just make them up or are they real things? Horse Eye's Back (talk) 16:30, 15 June 2024 (UTC)

I will not answer the question, because that would be repeating myself. Dominic Mayers (talk) 16:54, 15 June 2024 (UTC)

You have talked around the ideas of include-info and reject-info ideas but you have never defined them. It makes it flvery hard to understand what you are asking for. Masem (t) 17:14, 15 June 2024 (UTC)

Yeah, I'm intrigued by what Dominic has to say, and it feels like there might be something of value here, but it needs to be more concrete and specific before I can wrap my mind around it. @Dominic Mayers, maybe you could give some specific examples of what these" include-info and reject-info principles" would look like, and why they are necessary? Perhaps you could share some specific examples of content that you feel was inappropriately included/excluded on an article, and explain how putting clearer principles in place would address those specific examples? Pecopteris (talk) 19:11, 15 June 2024 (UTC)

Then at least link the diff where you say it, I've been through this entire conversation and you never define those concepts and nobody else in the wold seems to use them so there isn't anyone else in the world to ask besides you. Horse Eye's Back (talk) 17:17, 15 June 2024 (UTC)

Please air your Rfc question first

Before starting an RfC, could you please see WP:RFCBEFORE? I am happy to take part in an Rfc, and though I have participated in many and started several, I must admit I had trouble understanding your proposal both in this section and the previous one. Since an Rfc is very costly in terms of editor resources, it's important to write an Rfc question that is brief, neutral, easily understood, and which contains a small number of options to choose from. I fear that if the Rfc question is worded similarly to the way your proposals have been, I will not understand it, and consequently won't be able to vote on it. Perhaps others will find themselves in the same boat. That could cause the Rfc to become bogged down and fizzle out in endless discussion of what it means, rather than come to a community decision based on something everyone understands and can discuss and vote on.

To avoid such a costly waste of Wikipedia's most priceless resource, may I ask you to air your proposed Rfc question here first, not in order to discuss the merits in favor or opposed, but solely to get feedback from other editors about whether they understand what it is you are asking. I think this would help promote a successful Rfc in the end, should you decide to start one. (edit conflict) Thanks, Mathglot (talk) 17:34, 15 June 2024 (UTC)

When people ask to define "adding information" and "rejecting information", it is a sign that the discussion has stalled and we need external input. In that situation, I don't think I should have to try discuss further without first inviting more people to join the discussion. As it happened before with other explanatory statements in the policy, some people here are concerned that "not taking sides" could be used to say that we cannot reject contents such as "Wacko Smith said Trump won the last USA election". In that particular case, the argument is that Trump partisans could argue that rejecting this statement is taking sides. Therefore, they say that Wikipedia do and must take side, but the real fundamental issue is the same old one: should the policy be mostly a tool to reject fringe theories, etc. or a tool that requests that we add all relevant information to achieve the neutral point of view (and explains through examples what this means). I am only concerned that the policy is superficially modified without the true agenda being discussed. Dominic Mayers (talk) 18:51, 15 June 2024 (UTC)

Not asking you to define anything. If I understand correctly, you wish to start an Rfc. It is a basic feature of an Rfc, that it begin with a brief, neutrally worded question. Without defining anything, would you please quote in this subsection the question you intend to use at the top your Rfc? Thanks, Mathglot (talk) 19:12, 15 June 2024 (UTC)

You might have missed the part This RFC should happen before we consider any decision about the expression "not taking sides". As long as there is no serious intention to remove "not taking sides", I certainly have no intention of doing a RfC. Meanwhile, I might even change my mind regarding the best strategy to involve more people. Dominic Mayers (talk) 19:17, 15 June 2024 (UTC)

That's fine. If there is no intention of starting an Rfc, it seems this section, "Proposal of a RFC"... is now moot. Do you have any objection if I mark it closed? It would save further investment of time to no end. Mathglot (talk) 19:44, 15 June 2024 (UTC)

Did something special was implicitly open simply because there is a the word "RfC" in the title ? If that is the case, perhaps a renaming of the title that brings out its uncertainty is enough. Or close it, if you want, but this is like saying that there is a retraction, when there is nothing to retract. Only the uncertain aspect needs to be clarified. Dominic Mayers (talk) 19:55, 15 June 2024 (UTC)

I'm afraid I really can't quite follow. I certainly don't mean to imply that a closure involves a retraction, far from it. The only point of closure here, would be to save editor time and nothing more. This section appears to be a discussion about whether or not to start an Rfc, and I thought I understood you to say you were no longer planning to do so, in which case closure simply saves other editors the time of reading through a now-long discussion, only to discover in the end that there is nothing further to add as an Rfc is not in the offing.

However, it appears that I may be wrong about your intent; I really cannot tell. Given that whether or not you intend to start an Rfc is basically a yes or no question and I haven't managed to work out the answer, I think I'm not the best person to contribute here further, and other editors who better understand your points will have better responses than I. So at this point, I will bow out, and leave it to others to carry on. Given some level of communication problems thus far, at least in my level of understanding if nothing else, I would just reiterate my plea for you to please quote your intended Rfc question in this section before actually starting an Rfc, should you decide to start one, so that you can get feedback on the wording of the question itself first. Thanks, and good luck. Mathglot (talk) 20:44, 15 June 2024 (UTC)

@Mathglot: if it does not mean a retraction, just a clarification that there is no actual proposal for the time being, please close it. It is also that I do not know what is a closing of a section. The closing statement, if one can be added, could be "No actual proposal of a RfC for the time being", the "for the time being" being important. Dominic Mayers (talk) 20:55, 15 June 2024 (UTC)

Thanks, done. Mathglot (talk) 21:08, 15 June 2024 (UTC)

should the policy be mostly a tool to reject fringe theories, etc. or a tool that requests that we add all relevant information to achieve the neutral point of view (and explains through examples what this means). this sounds like you want us to reconsider FALSEBALANCE. And speaking from my own experience in trying to create viewpoint neutral articles and running into problems with sourcing, I am not seeing a major issue with how we use FALSEBALANCE. I know that this part of policy means that on some topics we will appear biased because a contentious topic shows more weight to one side in contrastt to the apparent balance In the real world due to how it is covered In RSes, but it is far more an uphill road to actually try to write to reflect the real world POV rather than the POV given by RSes. — Masem (t) 20:19, 15 June 2024 (UTC)

I agree, @Masem. I'm not even sure how we would determine what the "real world POV" is, other than consulting published sources, so any attempt to make Wikipedia reflect "the real world POV" by changing the FALSEBALANCE policy is likely doomed to failure.

One thing we can do to avoid FALSEBALANCE and create viewpoint-neutral articles without the overt appearance of bias is to avoid sneaking loaded language into Wikivoice, especially when the loaded language is not supported by an overwhelming weight of RS. While our sources are not required to be neutral, we are.

Another way that I think our FALSEBALANCE policy can be misused is by using it as a cudgel to completely exclude significant minority views, especially in contentious areas like modern politics. I'm not sure how to solve that problem yet. Philomathes2357 (talk) 20:37, 15 June 2024 (UTC)

Those who want to remove "not taking sides", because of a concern that it could be used to a create false balance toward inclusion of fringe theories, are the ones who worry the most with the explanation of false balance in the policy, not me. I only react to that and say, if the goal is to reject fringe theories, let us discuss the issue, while considering its two sides, inclusion and rejection, especially inclusion which was clearly neglected and could be misunderstood. Dominic Mayers (talk) 20:42, 15 June 2024 (UTC)

For that specific problem I think the ongoing internationalization of enwiki will largely provide the solution without us having to do anything... Its really only US politics which are contentious and as the percentage of US editors declines that contention will pose less of a stumbling block. Horse Eye's Back (talk) 20:43, 15 June 2024 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Why does Wikipedia appear to take sides?

Let's try this angle.

The NPOV 'nutshell' says: "Articles must not take sides, but should explain the sides, fairly and without editorial bias."

So why do articles sometimes appear to "take sides"? Here are a few possibilities:

1. Choice of reliable vs unreliable sources
2. Bias in sources
3. Due weight
4. Mainstream (fact) bias
5. Anti-fringe bias
6. Systemic bias
7. Language bias
8. Editors who violate PAG

Let's develop this. -- Valjean (talk) (PING me) 17:49, 8 June 2024 (UTC)

Great and important topic. I wish I wasn't going to be gone until the 17th.

• The overall problem is systemic bias on US political issues, and being systemically too easily game-able in one direction on such.
• Most of the time it is due to what is included and excluded. Wp:weight is supposed to help on that but instead the way that it is written it is one of the causes of the problem. Even on day 1 it was not objectively implementable (leaving it to be just gaming based) and now it has also become obsolete on two fronts in the post-journalism era and the current media-are-now-partisan-advocates era.

Sincerely, North8000 (talk) 18:44, 8 June 2024 (UTC)

I think that's right, @North8000. You make a good observation about WP:WEIGHT, and you're spot-on in regards to the "post-journalism" era/media-are-now-partisan-advocates era. That phenomenon is well documented, and Wikipedia editors absolutely need to be more aware of that. What, specifically, is it about the way WEIGHT is written that helps "cause" this problem? Philomathes2357 (talk) 19:45, 8 June 2024 (UTC)

Obvious question: Is it about how that section is written, or about how it is applied? Most editors never read the written section. We teach people the rules by citing WP:UPPERCASE shortcuts, and new editors believe us even when we're wrong. WhatamIdoing (talk) 19:48, 8 June 2024 (UTC)

IMO the cause of the problem is how it is written.North8000 (talk) 13:24, 17 June 2024 (UTC)

If you could re-write WEIGHT, how would you re-write it to address the problem you've identified? Philomathes2357 (talk) 22:31, 19 June 2024 (UTC)

Not whom you are posing the question to, but given the question, I would absolutely add the need to value sources far separated in both time and relationship to the topic at hand than those that are very near the topic in both ways, which is tied to factors like WP:NOT#NEWS. Far too much of our articles on US politics are running to jump in to use material from journalists making comments on the same day the news breaks. A good amount of these, the balance of POVs at the time of the event doesn't change in the long term, but we should still be careful to make presumptions based on individual media reactions and instead seek sources that give a rounded view of the event so that we can see, from either a long-term or a higher-up position of what the range of viewpoints are on a topic, and thus where DUE, WEIGHT, and FRINGE apply. Masem (t) 22:48, 19 June 2024 (UTC)

This discussion is interesting, but it does not seem to suggest in the immediate a change in the text of the policy. Perhaps it should occur somewhere else, such as the village pump. I really mean it when I say that it is interesting. I will try to continue this thread in the village pump. Dominic Mayers (talk) 08:57, 20 June 2024 (UTC)

To avoid scattering the discussion amongst many different places, here is a link to the right place to continue the discussion. Dominic Mayers (talk) 13:19, 20 June 2024 (UTC)

Reader bias, e.g., Confirmation bias. (In this sense, editors are also readers.) How you determine whether you can trust a website depends heavily on it telling you the things you expect to hear. If you genuinely hold extreme beliefs (e.g., Biden lost the election), then any website that is neutral/mainstream will appear biased to you. WhatamIdoing (talk) 19:46, 8 June 2024 (UTC)

Note also that WP:WEIGHT says:

Generally, the views of tiny minorities should not be included at all...

—Bagumba (talk) 07:06, 9 June 2024 (UTC)

I assume the purpose of this section is to improve the policy so that the original question about "taking sides" is answered or disappear. A unified understanding of include-info principles and reject-info principles is the needed improvement, because the include-info principles (describing instead of engaging, etc.) explain "not taking sides", the neutral point of view. Dominic Mayers (talk) 12:58, 9 June 2024 (UTC)

I would support changing the nutshell so it avoid language about "taking sides". Obviously, we do and should take sides. When those pushing fringe views complain about articles obviously omitting or contextualizing those views, we shouldn't be in the position of having to say "exceptions apply", or "you're interpreting that plain language wrong", or "your side isn't really a side". I'd favor nutshell language along the lines of "Wikipedia should present views and details in proportion to their coverage in reliable sources." Firefangledfeathers (talk / contribs) 13:39, 9 June 2024 (UTC)

At the end, if it is only a question of language, there is no big deal. My concern is that "not taking sides", when properly understood, is way better than "present details" to bring out include-info principles such as "describing instead of engaging debates", "providing the arguments", which say way more than "present views and details (emphasis is mine)". The main point is let us stop seeing oppositions between these inclusions principles and rejection principles. Instead, let us see that they all converge toward a same goal. BTW, I am not saying that "not taking sides" is an include-info principle. It also requires reject-info principles and more (neutral tone, etc.), but it is a way better unifying concept than mere due weight regarding the include-info side. Dominic Mayers (talk) 14:40, 9 June 2024 (UTC)

Do we think it's fair to consider neutrality as having these three components?

• What to do if there is disagreement/diversity of viewpoint ("weight"): For example, economists say X, and politicians say Y – but not scientists say X, and my neighbor's uncle says Y.
• What to include in general ("balancing aspects"): Articles should provide a basic encyclopedic summary of ordinary facts, without going into extraordinary detail about some aspect that is generally considered minor. For example, we should say that Ronald Reagan was born in 1911 (because encyclopedic context requires placing people in both time and place), but we should not have even one paragraph about the astrological significance of his birthday.
• How to say stuff ("impartial tone"): Neutral language, which is both direct and formal is appropriate for an encyclopedia. For example, "Lotteries disproportionately harm low-income people", rather than "Lotteries are a tax on people who are bad at math, especially if they are credulous or impulsive".

Have I missed any general categories that are important for this policy? WhatamIdoing (talk) 15:53, 9 June 2024 (UTC)

I think those are fairly comprehensive broad categories. If you focus the third more on the "How to say stuff" and less on the "impartial tone", I think it covers the important parts of attribution vs. wikivoice present in multiple parts of the policy. Firefangledfeathers (talk / contribs) 16:26, 9 June 2024 (UTC)

"Describing debates, e.g., providing arguments, instead of engaging them" is not covered in these categories. The third category should be also be about "What to say", not only about "How to say stuff", so that it covers this include-info principle. This is a part of the nutshell: [Not taking sides] applies to both what you say and how you say it. Not taking sides is certainly not about impartial tone and due weight only, unless we can expand on due weight so that it includes "describing debates, e.g., providing arguments, instead of engaging them". The key point is that the include-info principles should be better explained with examples. Dominic Mayers (talk) 17:19, 9 June 2024 (UTC)

I think that both of the first two are "what to say", and that "describing debates instead of engaging them" is "how to say stuff".

Would you like to share an example that you think would be helpful? WhatamIdoing (talk) 21:29, 9 June 2024 (UTC)

That's my analysis as well. Firefangledfeathers (talk / contribs) 01:12, 10 June 2024 (UTC)

For you how is due weight distinct from not taking sides? Horse Eye's Back (talk) 21:01, 10 June 2024 (UTC)

The problem related to the meaning of "not taking sides" will disappear once we will have better explained the include-info principles and the reject-info principles. The reason is that the motivation to say that "Wikipedia must often take sides" is closely related to the need to reject fringe theories. The motivation is not at all a respect of the ordinary meaning of "taking sides". On the contrary, the use of "not taking sides" in the policy, as something mandatory, is perfectly in line with its most common meaning. Dominic Mayers (talk) 01:39, 11 June 2024 (UTC)

I appreciate that, but it doesn't answer the question I thought I was asking. Let me rephrase... What is the difference between the meaning of "don't take sides" and the meaning of "include all verifiable points of view which have sufficient due weight"? Horse Eye's Back (talk) 02:39, 11 June 2024 (UTC)

Let me be more explicit then. Because of what you appreciate, I will not take the time to answer the question. It would make us enter into superficial terminological issues that lead nowhere, whereas explaining the principles is more useful. Dominic Mayers (talk) 02:45, 11 June 2024 (UTC)

Beyond the hyper specific does it suffice to say that the differences are so small as to be negligible? Horse Eye's Back (talk) 02:52, 11 June 2024 (UTC)

If you believe that, let's keep "Not taking sides", because it resonates well with people in general and says the essential. I am not answering your question, just making a comment on the affirmation included in your rhetorical question. Dominic Mayers (talk) 03:02, 11 June 2024 (UTC)

Its not a rhetorical question, you don't have to answer it but I think it would be helpful if you did (your argument doesn't really make sense otherwise, you just have to look around to see that your argument appears to be confusing, illogical, and internally inconistent to your fellow editors) Horse Eye's Back (talk) 16:27, 15 June 2024 (UTC)

In your case, it is not even needed to look around to see the fallacy in your argument. Dominic Mayers (talk) 16:49, 15 June 2024 (UTC)

Yes... Please look at my massive fallacy... Anything of substance to add or are you done contributing in a constructive manner? Horse Eye's Back (talk) 17:20, 15 June 2024 (UTC)

I just reacted to you just have to look around to see that your argument appears to be confusing, illogical, and internally inconistent to your fellow editors. I consider this a fallacy, because it refers to opinions of others instead of addressing my actual arguments. You see, in my case, I actually point to the exact part that is the fallacy. I do not say to you that others don't understand you, find you illogical, etc. I expect the same from you. Dominic Mayers (talk) 20:15, 15 June 2024 (UTC)

If it were true you could say it... I'm not going to lie to you, and saying that people seem to understand what you're saying would be lying to you. Since I noted this additional editors have as well, thats been the predominant reaction to the section you opened below. Horse Eye's Back (talk) 20:40, 15 June 2024 (UTC)

It is still a fallacy when seen as an argument. The number of persons who agree or understand is not a valid argument. If you do not intend to argue about the policy, please do that elsewhere. Dominic Mayers (talk) 21:26, 15 June 2024 (UTC)

But it's not an argument it's a comment. Failure to express an argument clearly doesn't make it wrong, but maybe it does need to be stated in a way that is easier for others to understand. -- LCU ActivelyDisinterested «@» °∆t° 13:24, 16 June 2024 (UTC)

That's playing with words. A comment on arguments that does the same as a fallacy is a fallacy. Dominic Mayers (talk) 15:08, 16 June 2024 (UTC)

Not playing with words, just pointing out that your argument needs clarification. -- LCU ActivelyDisinterested «@» °∆t° 15:39, 16 June 2024 (UTC)

We are trying to help you understand that no one has a good idea of explicit changes or discussion you want to have about policy as you are introducing terminology without definition no provide clear examples (real or hypothetical) to encourage discussion. Instead you are doubling down on saying you've made your argument andeveryone else is missing it. That's not helping to get to policy talk faster. — Masem (t) 15:47, 16 June 2024 (UTC)

If you want to change the policy, the two sides, including info and rejecting info, should be considered. It will be a good thing that people have this discussion regarding the two sides anyway, but I will not start it. On the contrary, I wish not to discuss further here. Dominic Mayers (talk) 16:06, 16 June 2024 (UTC)

┌────────────────────────────────────────────────────────────┘
You seem to be the only editor here that has this idea to change policy, but you are not articulating exactly the change to be considered (even if it is a high level change that you cannot immediately express in terms of a diff revision to existing text) and then as soon as we ask you to explain more, you say you will no longer participate. Either you are trolling us (which I doubt) or you likely know you change would happen when you are questioned about it, so you remove yourself from the discussion. Either way, that's no way to proactively and efficiently engage in discussions on policu changes. You may want to take a break, think exactly what you think needs change and then present a case using establish terminology and clear examples to back up your suggestion. But right now, this overall approach is not a path forward. — Masem (t) 16:41, 16 June 2024 (UTC)

A global discussion to change the policy is a good idea, but more people should be involved and we should consider both sides : the addition of info and the rejection of info. Please do not take it personally, but I am not interested in pursuing that specific discussion. It seems that this last point is what you blame me for. Dominic Mayers (talk) 17:11, 16 June 2024 (UTC)

Can you explain why those are seperate sides/proscesses? From my perspective its a single side with a single critieria, you can't choose what to include without also choosing what not to include and vice versa. Horse Eye's Back (talk) 18:41, 16 June 2024 (UTC)

That's not true. The rules to reject info, that is NOR. V and RS, are not sufficient. For example, attributing opinions is not a rule to reject info, but to add info, in this case, the source in an explicit manner in the text. The challenge is that, except maybe for attributions of opinions, it is not obvious what info should be added to achieve the neutral point of view. Even how to best attribute in a neutral manner, without taking sides, is not obvious. People ask me to illustrate by examples how including info is useful to achieve the neutral point of view, which is a valid request, but a difficult one and I am passing the hot potato to others. Dominic Mayers (talk) 19:32, 16 June 2024 (UTC)

You can't choose when to attribute without choosing when not to attribute. Choosing not to attribute is rejecting info. Horse Eye's Back (talk) 21:12, 16 June 2024 (UTC)

You won, definitively not because this trivial logical fact proves me wrong—it does not, but because you exhausted me. I insist you won. I will not argue. Dominic Mayers (talk) 23:23, 16 June 2024 (UTC)

Dude come on there is no winning and loosing, I'm trying to understand what you mean not prove you wrong. Horse Eye's Back (talk) 00:37, 17 June 2024 (UTC)

We are trying to find a way to get more involved via an RFC but you are not helping by not giving a clear statement to work from to form an RFC, and claim that you don't want to participate in a discussion you stated. That's creating frustration here, and we are asking for more specifity based on establish terminology. Masem (t) 18:46, 16 June 2024 (UTC)

There is no proposal for a RfC. Do you want to do a RfC about removing "not taking sides" from the policy ? Dominic Mayers (talk) 20:01, 16 June 2024 (UTC)

I recommend that we drop this series of discussions. Over the past few weeks, this talk page has seen multiple threads vaguely about "taking sides", but I am not seeing evidence that there is a problem. After a certain point, it becomes unfair to expect editors to keep explaining why the policy is the way it is. I want to remind editors of WP:IDHT.

It's time for the main participants in this discussion to get back to working on articles, and accept the last stable version of the NPOV policy. If there are any POV issues on Wikipedia, they can be addressed at the article level. And if one of those articles reveals an intractable POV dispute, maybe it will need a clarification of NPOV, or maybe it won't. But editing articles would be a better use of everyone's time. Shooterwalker (talk) 20:23, 16 June 2024 (UTC)

I agree with @Shooterwalker. @Dominic Mayers, I have a gut feeling that, if I could figure out what you are talking about, I would agree with you, but I honestly cannot decipher what you are saying. I see the words, and the sentences, but I cannot figure out what they mean, even though I am a native speaker of English. My gut tells me that you have some pretty big and deep ideas that you are trying to get across, but perhaps they are so big and so deep that you're struggling to present them in an easily-digestible way.

I think it would be a good idea for you to do some off-Wiki writing, to figure out specifically what you mean, and how to say it in a way that is easier to comprehend. If you want help honing your thoughts into something more precise, you can ping me at my talk page or email me, and I'd be more than happy to talk some more. Philomathes2357 (talk) 21:40, 16 June 2024 (UTC)

I don't think I am saying anything very deep. There are issues about the neutral point of view that are intrinsically complex. I am struggling to understand them, so I never tried to explain them. Perhaps, you meant that these complex issues surface a bit in my simple points, but it just means that I do not try to hide them. Dominic Mayers (talk) 22:04, 16 June 2024 (UTC)

@Dominic Mayers: you need to listen to @Philomathes2357:. They have a point. They write "if I could figure out what you are talking about". @Horse Eye's Back: wrote "I'm trying to understand what you mean". I have the same issue with what you write, and many others have said the same thing. It's so bad a problem, that when you started to comment in this thread, I was considering shutting it down immediately as I knew you would drown everyone out with walls of text that few would understand. That's exactly what has happened. You kill threads everywhere you comment. Please be more concise and speak the language of commoners. Not everyone has a Ph.D. here. We're trying to understand you, but it's really difficult. At this point, the resolution is not for you to try to explain, but to be silent. -- Valjean (talk) (PING me) 17:57, 17 June 2024 (UTC)

I knew that even though I said that I do not wish to pursue this discussion, people will continue to present me as someone who kills threads, who nobody understand, etc. It is even predictable. It is because I do not want to endure this, that I do not wish to continue this discussion, but of course I have some dignity and I respond when we present me in this manner. My response is that I do not kill threads. My points are simple. In fact, I can summarize them in two sentences:

• Let us not see opposition between rules for including information (describing debates instead of engaging them, providing the arguments, etc.) and the rules for rejecting information (NOR, V, RS), but see them instead as working together toward a same goal.
• This unified understanding of all the rules should happen before we discuss "not taking sides", because the true concern here is that people worry that "not taking sides" means we cannot reject fringe theories.

Weirdly, people ask what do you mean by including information ? HEB even suggests that we only need rules to reject information, because including information is only the opposite. I will not go into this. As I said to HEB, you won. I have no further argument. Dominic Mayers (talk) 19:00, 17 June 2024 (UTC)

Thats not what I suggest... I'm actually suggesting that we already have a "unified understanding of all the rules" rather than seperate rules for rejecting and including information... There is no opposition because its the same thing... Rejecting is including and including is rejecting. One can't be done without the other, its a philosophical impossibility. Horse Eye's Back (talk) 17:26, 19 June 2024 (UTC)

Thanks for clarifying: "Rejecting is including and including is rejecting". I have to think about the deep meaning of this in the overall picture of what is going on in Wikipedia. Dominic Mayers (talk) 17:48, 19 June 2024 (UTC)

Think about it this way... A group that includes only you and I excludes everyone else on the planet... It is not possible to include without excluding unless you are including literally everyone. Horse Eye's Back (talk) 18:51, 19 June 2024 (UTC)

+1 to an RFC to remove "does not take sides" from the nutshell and body of NPOV, because Wikipedia does take sides: it takes the sides of RS. Levivich (talk) 16:24, 17 June 2024 (UTC)

Probably the most common problem with this is not when talking about including the "sides" of an issue, it is about deleting material when not talking about sides. I'd guess that 60% of all of the content of Wikipedia violates a strict reading of this wp:weight, and so a POV warrior can selectively use it to knock out any of that 60% that they choose. North8000 (talk) 13:33, 17 June 2024 (UTC)