Wikipedia talk:Neutral point of view/Archive 63

What is a fact, and what is an opinion? It's time for a more rigorous delineation between these concepts.

NPOV makes two things very clear. Avoid stating opinions as facts, and avoid stating facts as opinions. Therefore, it is absolutely crucial that editors have at least a basic grasp of formal epistemology in order to apply NPOV correctly. I think that as a community, we need to have a clear, formal definition of the two terms that we can turn to in order to resolve content disputes. Otherwise - as I have seen time and time again - editors get sucked into a frustrating and unproductive back-and-forth about whether or not a certain contentious claim is a "fact" or an "opinion". I have seen some downright baffling and very dubious arguments suggesting that something is a "fact", when from an epistemic outlook, it is definitively and incontrovertibly an opinion. And vice versa. This is madness, and NPOV is literally pointless if there's no way to agree upon the difference between facts and opinions.

I'm sure this issue is salient in the hard sciences, social sciences, and humanities, but this is of particular concern to me as it relates to contentious political labels applied to controversial people and groups. Examples include "terrorist", "extremist", "radical", "fringe", "far left", "far right", "conspiracy theorist", and so on. Are these labels facts, or opinions? It is an objective, empirical fact that Al-Qaeda are terrorists, for instance? Or is it a widely held opinion? How about some socialist intellectual - is it an empirical fact that they are an "extremist" because some journalists said so, or is it an opinion held by those who disagree with the socialist intellectual? If I describe elephant poaching as "unfortunate" - is that a timeless, sacred truth, or my opinion about poaching, widely shared though it might be?

How about this - if I say "It is a fact that Joe Biden is the President of the United States". Am I merely repeating a fact, or expressing that it is my opinion that this is a fact?

This might sound like an exercise in intellectual wanking, but it isn't. It is vital that editors understand the difference, or else NPOV is a joke.

The answers to the above questions are very clear and self-evident to anyone who's, say, taken a graduate-level series of courses in epistemology. But we don't all have that opportunity, and lacking a rigorous formal philosophical training shouldn't be an impediment for sincere editors to contribute to articles about controversial topics. I've found that there is a general lack of epistemic literacy on Wikipedia. Some editors have suggested that "if an RS makes a statement with a declarative sentence, it's a fact". That simply won't do. RS are reliable because (we presume), their factual assertions are accurate, and their opinions are thoroughly-considered. What we cannot continue to do is elevate all statements in an RS to the level of "fact", regardless of the epistemic status of the statement. An RS might have a conservative contributor who writes an article (not an opinion piece) that describes Joe Biden as a "far-left radical". But we would clearly understand, despite the use of a declarative sentence in an RS, that this claim, regardless of its merit or lack thereof, is an opinion. Even if this became a trendy talking point in American media, and 5 or 10 RS used the same "far left radical" language, it would still be an opinion, rather than an empirical, factual description of an objective reality. Right? I hope you're all nodding in agreement.

My proposal is this - we need to have some sort of standard for what constitutes a fact, and what constitutes an opinion, that is more rigorous and epistemically sound than "well the RS used a declarative sentence, so it must be a factual statement" or "I personally contest that assertion, so it must just be an opinion". Also, whether or not a given claim is a "fact" should not be determined by a popularity contest, wherein widely held opinions gain the status of "facts" by community consensus, and widely-disliked facts get dismissed as mere opinion by consensus. Facts are facts, and opinions are opinions, regardless of consensus to the contrary.

We need to have either an informal guideline or a formal policy that helps sincere but epistemically uneducated editors to clarify this issue. I'd like to invite the community here at the NPOV page to have a spirited, rigorous, and serious discussion about epistemology and how to balance NPOVs guidance to "avoid stating opinions as facts, and avoid stating facts as opinions". Perhaps, together, we can formulate an outline for such a guideline or policy.

I look forward to reading the thoughts of other editors. Philomathes2357 (talk) 23:33, 12 August 2023 (UTC)

Honestly until epistemologists can agree on this I don't really see what the point of us hashing it out is... The line between fact and opinion is in fact fuzzy, anyone who can unfuzz it will be the most celebrated thinker of our generation. Horse Eye's Back (talk) 23:40, 12 August 2023 (UTC)

The point is that a proper application of NPOV demands that we be able to reach a common understanding of what is a fact, and what is an opinion - and "an RS said it, and they used a declarative sentence, so it must be a fact" or "if enough editors think it's a fact, it magically morphs from an opinion to a fact" (i.e. a popularity contest) is not a satisfactory way of doing so. If we just decide that there is no point in trying to distinguish between fact and opinion, there's no point to Wikipedia, and the platform might as well be shut down. Philomathes2357 (talk) 03:12, 13 August 2023 (UTC)

NPOV demands no such thing, it is specifically formulated to deal with the ambiguity of the real world in which we live. If that is what you really think of wikipedia I suggest you find a different hobby, you are unlikely to receive satisfaction. Horse Eye's Back (talk) 03:17, 13 August 2023 (UTC)

Wikipedia doesn't dabble in epistemology and truth. If something is really in doubt as to whether it is factual, it should be attributed. These are disputed facts, not opinions. Opinions are normative value judgments. Wikipedia includes opinions as well, but they have to be well-sourced. Some opinions, if consensus among experts and largely universally held (except for some fringe) will become Wikivoice facts, even if your Philosophy 101 class might call it an opinion. Andre🚐 23:56, 12 August 2023 (UTC)

Nice. Philomathes2357 (talk) 03:15, 13 August 2023 (UTC)

I agree with others that we aren’t going to solve any perennial intractable philosophical issues here, but one improvement I think we could make is to set a higher bar for the use of wikivoice in contentious topics. In those articles, where facts are harder to readily distinguish from opinions, we should err on the side of treating them as opinions. Where ideological lines are drawn around facts and opinions, we should expect to see greater use of attribution. Note that I am not proposing any revolutionary reinterpretation of what is considered factual. WP:WIKIVOICE already says this: Avoid stating seriously contested assertions as facts. I think what I would propose is that this existing wording is augmented with something like Take particular care with contentious topics. Note also I am not talking about entertaining raving loons and fringe pseudoscience, but about the contentious topics where intellectual honesty demands we recognise the existence of multiple competing perspectives within mainstream discourse. Barnards.tar.gz (talk) 21:34, 14 August 2023 (UTC)

I think a big problem is that what we consider reliable sources are at times used by national establishments as surreptitious outlets of propaganda or at least of controlled information. In Guatemala, José Rubén Zamora has been attacked at his home and has been jailed for, journalists say, criticizing the government or running news about government wrongdoing. The government hasn't touched the other newspapers or stations. Probably they are not as much as a problem for it.

Each country probably has its own issues regarding the true independence of reliable sources from undue influences from their own national establishments. In the US for example, the government pushed censorship of certain topics and pushed for only its own narratives to be published. It is unclear the extent of such effort. The government of France even threatened a Wikipedia administrator into deleting the article Pierre-sur-Haute military radio station. That makes me wonder, how independent are reliable sources in France? How much undue influence does governmemt has on them?

Another thing is the self-censorship professionals, scientists, journalists, researchers have on themselves, fearing for their reputation, grants, jobs. As a result, their work reflected in reliable sources is in an unknown number of times not truly independent or objective.

There is also the issue that many news outlets are owned to an extent by the same corporations, like BlackRock and The Vanguard Group. How independent are then from each other, specially when in topics like politics they espouse the same views? Regards, Thinker78 (talk) 22:09, 15 August 2023 (UTC)

Wrt to state-controlled media, I often find that it's important to consider what we're writing in the Wikipedia article. Sources can't be considered "reliable" in a vacuum (although that is, unfortunately, the oversimplified story that some editors will get from the existence of Wikipedia:Reliable sources/Perennial sources). Sources are only reliable when they are evaluated in the WP:RSCONTEXT of the exact sentence/paragraph that they are being used to support.

It still surprises and dismays me when editors remove sources to state-controlled media because "RSP says so" for content that we would accept a state-issued press release for. We don't need fewer citations for things like which person that state appointed to which office.

One would not want to rely solely on state-controlled media for some things ("There is no shortage of bread in our country", "People are not rioting in the streets", "The president has not been seen in public recently because he is too modest, and not because he was murdered in his office last week"), but WP:YESPOV requires us to describe their story, too: "State-controlled media said that there was no shortage of bread, but independent news and other governments report rioting in the streets over hunger.". WhatamIdoing (talk) 15:22, 16 August 2023 (UTC)

.

Good post and point. True fact (and conversely true declarations of "falseness") exists only in a tiny fraction of cases where it is claimed. But it does sometimes exist, contrary to claims otherwise. Taken from WP:Accuracy it is when "The metrics of a correct answer are agreed on, and the answer is overwhelmingly considered to be known. For example, "Did the US land a man on the moon?" All would agree what "US land a man on the moon" means, and at least 99% of people with good access to information would agree that the answer is known to be yes." North8000 (talk) 00:26, 13 August 2023 (UTC)

I am aware that you are just quoting an article, but 99% is too high. This is just people in the US, so it is not really representative. Good access to information does not help if one does not want to accept it.

Anyway, voting is obviously not a good way to determine whether something is a fact. --Hob Gadling (talk) 06:52, 14 August 2023 (UTC)

We absolutely need editors to be far more aware that just because an RS publishes a seemingly news-only story (not tagged as an opinion, analysis, or similar type of secondary source), that the use of accountability journalism today will allow the interjection of the author's opinion or view into the prose. To our end, that means we should be more wary and make sure that statements of fact versus statements of opinion are clearly separated, using WP:YESPOV appropriately if we are going to use the statements of opinion (eg with attribution). This may require more talk page discussion to determine how to break down critical sources, but that also is going to require editors to leave their personal biases and feelings aside when discussing these. Unfortunately from a policy stance, there's little we can change, we can only point to editor behavior and willingness to engage in these types of discussions when necessary. --Masem (t) 00:36, 13 August 2023 (UTC)

There is certainly a tendency among some editors to declare that anything in a news article they disagree with is just an injection of the author's opinion or view into the prose.

David Eppstein recently suggested "evaluative content" as an alternative way of talking about opinions, and I kind of like it, but I also find it confusing. What we want is:

• The economy would improve if interest rates were lowered: opinion.
  • But: Ev Economist said that the economy would improve if interest rates were lowered: fact (that Ev said it).
• Ruritania needs to modernize its infrastructure: opinion
  • But: At the end of the long hearing, in which eight experts presented a wide range of information about the human and economic results of poor infrastructure, Paul Politician concluded that Ruritania needs to modernize its infrastructure: fact (that Paul concluded it).
• 20% more pre-term babies would live if their mothers were given steroids less than one week before their birth: fact.
  • Also: Si Scientist evaluated 17 papers on the treatment of pre-term babies and concluded that 20% more pre-term babies would live if their mothers were given steroids less than one week before their birth: Also a fact.

Generally, in the "radical left-wing" category, we focus on what the words mean (e.g., by looking them up in a dictionary) and then trying to represent what the source really means in plain English (because we can and should Wikipedia:Use our own words). Sometimes this is a bit difficult, as multiple meanings may exist. To give an extreme example of contradictory meanings, one could imagine a troll claiming to "just follow the dictionary definition" and writing that Caitlyn Jenner is a man ("I'm just following the dictionary definition that says a man is any adult who was assigned male at birth") and a woman ("I'm just following the dictionary definition that says a woman is anyone with a feminine gender identity").

We tend to use this in the "negative" direction, i.e., not repeating words that sources have used incorrectly or in confusing ways. For example, if a source calls a transphobic US politician a TERF because he's trans-exclusionary, even though he's definitely not any kind of feminist (much less a radical one), then we would not repeat that word, because it's a false fact, rather than an opinion. Similarly, pseudoscientific gets used as a smear word, but you shouldn't find it in articles about subjects that claim to have nothing to do with science (e.g., religion), even if a source uses it. We're not "editorializing" when we try to understand and clearly communicate what all the sources, taken as a whole, actually mean, even if a given source wrote "radical" when there's no evidence of actual radical politics.

(Minor nitpicking: @Philomathes2357, an editorial is an opinion piece written by the Editor-in-chief of the publication. It's a contradiction in terms to talk about "an editorial (not an opinion piece)".) WhatamIdoing (talk) 02:01, 13 August 2023 (UTC)

What I have in mind is more statements like "So-and-so is an astrophysicist known for their research on black holes", which might appear in the lead sentence of an article. Although we might maybe accept a statement from the individual in question that they specialize in black holes, we could not say "known for" without a secondary source suggesting (or suggesting through its existence) that they are actually known for this research. On the other hand, this is not the sort of thing we would want to attribute to an individual opinion-holder, more or less because of MOS:DOUBT: we don't want to cast doubt on a BLP expert's expertise by stating their expertise in an unnecessarily constrained way. —David Eppstein (talk) 02:19, 13 August 2023 (UTC)

Misplaced attributions can also violate NPOV. The last bullet point in its WP:WIKIVOICE section addresses this. (That reminds me, I never did clean up the mess in one of the BLPs about an AIDS denialist. Someone seems to want his characterization as an AIDS denialist to be considered just one person's personal POV, rather than the view of every single reliable source published in the last 25 years.) WhatamIdoing (talk) 03:45, 13 August 2023 (UTC)

The thing on that parenthetical example is, what is being an AIDS denialist is still subjective, carrying the same problems that "Climate change skeptics vs denialists" has. Instead, a factual statement can be made by simple expansion of exactly why some consider the term applies, saying something like that "So-and-so has denied that AIDS exists and is a government conspiracy". A lot of this when it comes to labels is that it is far better to explain what has led to the use of the label in context, most which will be driven by factual statements, and avoid the use of the label or at least then have attribution about it. Continuing the example, you could "So-and-so has claimed AIDS is a conspiracy theory by the gov't, leading him to be considered an AIDS denialist." Masem (t) 13:09, 13 August 2023 (UTC)

@Masem, first of all, thank you for being kind, and not being rude. I don't expect anyone's sympathy, but I am autistic, and the rude & flippant comments are genuinely bothersome to me. Sometimes, they are so hurtful that I respond with snark and scorn, which is a weakness of character on my part, and one of the main reasons I endured a topic ban. Thank you for not being that way, even though I believe we may not be in 100% agreement. I want to make sure I understand you precisely. Is the following an accurate summary of your view?

"Reliable sources include both facts, and opinions. Sometimes, those opinions are expressed with declarative sentences, as if they were facts, but we still must treat them as opinions. Even if the same opinion has been repeated by multiple reliable sources, it should still be taken as an opinion, rather than a fact, and therefore, it should be attributed."

Please correct me if I have summarized your view improperly in any way. Also, I have a question for you: is there any circumstance where an opinion should be presented as a fact, in Wikivoice, simply because of the sheer number of RS that have expressed the same opinion?

Finally - your point that this issue cannot really be addressed at the level of policy is noted. I have a few other ideas for how this might be addressed constructively, besides a change in formal policy (which I concede would be a radical step to take).

1) The NPOV page, where it talks about "avoid stating opinions as facts" and vice versa could be expanded. The "genocide is an evil action" example is very helpful. Perhaps both of the sections could be expanded, giving multiple examples of the same idea formulated first in factual terms, and then as an opinion, for comparison's sake. Similar to the manner in which @WhatamIdoing has done above.

2) An informal guideline, rather than a formal policy, could be written, perhaps in the form of an essay. I've seen such informal stylistic guideline essays before. Such a guideline could provide, in 5-10 paragraphs, a basic overview of epistemology, with examples of "facts", "opinions", and some general guidance for how to distinguish between the two while reading reliable sources. I'd conceive of this as a companion to the Manual of Style.

3) The contentious labels section could be revisited and possibly expanded to include other commonly-used phrases that are markers of opinion. I'm not as excited about this option, and have been met with resistance when I suggested this in the past - but it's worth considering.

I have other ideas, too, but this post is long enough, so I'll leave it there for now and await a reaction. Philomathes2357 (talk) 18:59, 13 August 2023 (UTC)

I'm going to strongly oppose any change and especially the contentious labels bit. Far right wing fringe people are right wing fringe people, like conspiracy theorist Marjorie Taylor Greene or Breitbart News. These are Wikivoice facts. Andre🚐 19:57, 13 August 2023 (UTC)

I'm sorry, I see nothing in your comment but your own strong personal opinions about politics, Andre. While those might be interesting to discuss in a different setting, I don't see how they're relevant to NPOV - they're explicitly irrelevant, in fact, the same as my political opinions. And the term "Wikivoice facts"...I won't even go there.

Here's something substantive we could discuss - do you think the contentious labels section should be removed altogether, and that all the words currently included therein should be used freely in Wikivoice? Or do you support maintaining the contentious labels guidance, but oppose any changes or additions on principle? I'm genuinely curious. Philomathes2357 (talk) 20:02, 13 August 2023 (UTC)

Actually those aren't my opinions. Those are facts, on Wikipedia. Breitbart is a fringe far right publication. MTG is a conspiracy theorist. She theorizes about conspiracies. There's strong consensus and sourcing for this. Therefore it's a fact. It's the heart of your argument, so it's weird that you "won't go there." I recommend you drop this entire line of inquiry, actually. No change is needed or advisable to the current contentious label policy. Contentious labels may be used if they are widely used in sources and are accurate descriptions according to the predominance of prominent expert sources. That's why we only care about verifiability, not truth. It doesn't matter if you disagree. Only sources matter. Andre🚐 20:04, 13 August 2023 (UTC)

Thanks for sharing your opinions, Andre. Philomathes2357 (talk) 21:05, 13 August 2023 (UTC)

About "what is being an AIDS denialist is still subjective": No, not really. These are words ("labels", if you prefer) with sufficiently clear-cut, widely agreed-upon definitions that they are not subjective.

Consider:

• Definition of AIDS denialist: "Person who denies that HIV causes AIDS."
• Quote from BLP: "HIV has absolutely nothing to do with AIDS. I deny that there is any scientific evidence that HIV causes AIDS. HIV is perfectly harmless, and all those previously healthy young men who died from AIDS just had drug addictions."
• All reliable sources: "Yup, he's an AIDS denialist."
• Random editor: "Well, just because he says he denies that HIV causes AIDS, that's not a good enough reason for sources to call him an AIDS denialist. Besides, it's just their (i.e., the entire world's) opinion that the term AIDS denialist means someone who denies HIV causes AIDS. There's no objective reason why that term couldn't mean – oh, I don't know, maybe AIDS denialist should just be a synonym for cats. And he's not a cat, so that proves he's not an AIDS denialist, either."

The rhetorical trick in which you redefine key words to mean something else works in high school debate (at least, it works against those that are both poorly coached and unwary), and it turns up occasionally in serious sources, but pretending that most words (including "labels") don't have a pre-existing and well-understood meaning is really not a game we should play on Wikipedia. When the subject says that he denies that HIV is associated with AIDS, we should say he's an AIDS denialist and link to the article on HIV/AIDS denialism; we should not use circumlocutions to avoid using a "label". WhatamIdoing (talk) 05:59, 14 August 2023 (UTC)

Yes, of course you are right, WhatamIdoing. brain fart on my part. Thank you for the correction. So far, I'm glad to see that this has resulted in a spirited and (mostly) civil and constructive discourse. I'm going to read and absorb everything here and chew on it while other editors continue to chime in. Thanks to those who've taken the time to make very detailed, nuanced, thoughtful posts, and thanks to the 1 or 2 trolls for making me laugh. Philomathes2357 (talk) 03:05, 13 August 2023 (UTC)

This can an impossibly complex problem, but there is a simple partial solution which is to (unlike now) simply let the editor-related discussions / analysis on such topics occur. That doesn't actually happen because the practice is to follow metaphors of wp:VER and WP:Nor instead of the actual policies. For (an overly simplified) example, we should realize that one can search amongst WP:"Reliable"Sources to find anything that one wants, including conflicting "answers" to the above example questions plus to other questions that seem to be even more matters of "fact". The metaphor version of wp:Ver / WP:NOR pretends that it's illegitimate for editors to decide when to treat each of these as fact vs. opinion.

A second point is that in real life almost every statement is ambiguous / has multiple possible meanings, including all of the ones given as examples above, leaving any source that writes about them to choose amongst the many meanings and infer that it is the meaning. For example, let's say that John Smith said "masks don't prevent the spread of Covid"; here are two possible meanings of that:

1. Masks don't absolutely prevent the spread of Covid (a statement that is true)
2. Masks don't ever prevent the spread of covid, i.e. they don't reduce the number of incidences where it is spread (a statement that is false)

A "source" that doesn't like John Smith can be thinking about #2 and say that John spreads false info regarding Covid. A source that likes John Smith can be thinking about #1 and say that he spreads true info regarding Covid. Most "sources" will (and are allowed to) make either statement without clarification. And then at a contentious article, somebody is going to say that it's illegitimate to discuss the above because "we just write what sources say" which is the metaphor version of what WP:Ver & WP:NOR actually say. Sincerely, North8000 (talk) 11:43, 13 August 2023 (UTC)

This is an absolutely bizarre argument to make, because the overwhelming majority of scientific research and reliable news coverage will show that masks REDUCE the rate of COVID spread.

Someone trying to dishonestly frame something may use dishonest arguments, but the reliable coverage will address this.

I think it is a matter pf concern that multiple right-wing editors on Wikipedia have recently and in synchronous fashion adopted the same new talking point about how sources need to "define" terms such as far-right or else be removed from use. It's absolutely inconsistent with wikipedia publication policies.

Also, this entire post looks like Philomathes trying to take his fight to a new playing field after getting no traction trying to bludgeon a few talk pages with stuff like this [1].

Since Philomathes was previously indefinitely blocked and then topic banned for such behavior as well as making personal attacks and insinuations it's not a good look. Calling everyone that doesn't agree with their fringe / povpushing viewpoint "epistemically illiterate" [2] and posting walls of text is uncivil and it looks like this has been a return to the conduct by Philomathes that was a problem. Lois Lane of Earth-12 (talk) 17:33, 13 August 2023 (UTC)

This has been a very interesting conversation so far. Quite a few editors have noted that there is indeed a widespread confusion about determining the fact/opinion status of a claim. I'm not sure how attacking me personally, flirting with an AGF violation, and insinuating that this is a purely "right-wing" concern furthers the discussion.

And yes, I stand by my remark about widespread epistemic illiteracy on Wikipedia. It was not directed at any one editor, it was an observation of a systemic problem that I am attempting to constructively address. To insinuate that this is somehow malicious would be like calling someone an anti-black racist because they observe "there is widespread food insecurity in Africa". Clearly, if you read the above thread, I am not the only editor that has such concerns, so I guess there are a lot more "right-wingers" here than I thought. Philomathes2357 (talk) 18:42, 13 August 2023 (UTC)

Your comment makes no sense. I have never seen someone called an "anti-black racist" for observing that there is widespread food insecurity in Africa. Now if someone said that and added something like "because blacks cam't govern themselves" that would be an entirely different story.

Insulting every editor who does not 100% agree with the odd illogic in your wall-of-text polemics and calling them "epistemically illiterate" because they do not demand that every article include a multi-line definition of "far right" or "far left" or similar terms when placing a group or individual on the political spectrum, is absolutely a personal attack. You don't have to name individual people to clearly be incivil and demeaning. Lois Lane of Earth-12 (talk) 19:29, 13 August 2023 (UTC)

Where did I demand that every article include a multi-line definition of "far right" and "far left"? Nowhere, because I don't hold that view. That's a comical strawman of what I have said. What I've said is that value-laden terms, like "far leftist", "terrorist", and "extremist", are markers of opinion, not objective measurements, and should be attributed to the reliable sources that use them. If you dispute that "far left" is a value-laden or subjective term, I invite you to read the top of my personal page, where I have laid this out in some detail.

Clearly, you have not taken the time to read what I wrote and digest it - so why bother commenting? Notice that I don't weigh in on any of the debates here unless I feel that I have something genuinely constructive to offer. Also notice the contrast between your comment and the thoughtful, substance-filled comments above you. Please stop making these personal remarks about me here on the NPOV talk page. If you have an issue with my conduct that you think I need to address, feel free to drop me a line in my personal talk page. Thanks. Philomathes2357 (talk) 19:53, 13 August 2023 (UTC)

I have read what you've said. Which has taken far too long given that you could have said it in 1/4 or less the word count.

"Terrorist" not only has a dictionary definition, wikipedia defines when it should be used: when the WP:DUE weight of WP:RS use it in a verifiable and consistent way.

Thus far you have accused everyone who does not agree with your fringe views of being "epistemically illiterate" and above as "trolls," along with making other insinuations and insults [3][4]. You have made accusations of people being "rude and flippant" above while simultaneously accusing people of not reading your ridiculously longwinded polemics that boil down to how you dislike Wikipedia's policies and you think that everyone who disagrees with you is "illiterate."

I really and truly suggest that you stop doing that along with reducing your exhaustively long polemics, as it appears this is the exact reason you were blocked and topic banned previously. Lois Lane of Earth-12 (talk) 21:31, 13 August 2023 (UTC)

This is getting ridiculous. Please stop contacting me or I will request a two-way interaction ban. Thanks. Philomathes2357 (talk) 22:04, 13 August 2023 (UTC)

@Philomathes2357 Is this a reply to @Lois Lane of Earth-12? Or? Doug Weller talk 07:55, 14 August 2023 (UTC)

Unfortunately, yes. I've chosen to refrain from further replies so that I don't say something inappropriate. Philomathes2357 (talk) 14:45, 14 August 2023 (UTC)

After reviewing the apparently false accusation you made at Andrevan [5] and reviewing history I could not find any problem interaction from him towards you but it does appear that your starting of this thread is repeating your argument from your failed Wikipedia Talk:Manual of Style/Words to watch proposal that you were engaged in bludgeoning just before your block and topic ban. I am asking you politely again to recognize the problematic nature of this conduct.

Your last failed attempt at this can be viewed here [6] and it was overwhelmingly rejected. You appear to have returned to the same topic pages such as The Grayzone and repeating the prior mistakes.

Please for your own good, drop the WP:STICK. Lois Lane of Earth-12 (talk) 23:27, 13 August 2023 (UTC)

Hey @North8000 I think I understand your point, but can you clarify why you think WP:VER and WP:NOR are interpreted improperly (if that is, in fact, your view)? Philomathes2357 (talk) 22:15, 13 August 2023 (UTC)

What WP:Ver and wp:nor actually do is set requirements for the presence of content / exclude content that does not meet those requirements. I probably wouldn't choose the term "interpreted improperly" even though it is somewhat the case. What I was speaking of is that it common to derive overly broad/vague broad metaphors from them and then use the metaphor instead of the actual policy. For example, that sources answer every Wikipedia editorial question, or that if the source has the trappings to be a wp:RS that editors are not allowed to question, analyze or exclude something written by that source. North8000 (talk) 00:58, 14 August 2023 (UTC)

In simpler terms, it works like this:

1. WP:V and WP:NOR say that editors can't call Breitbart News "far-right" unless and until at least one reliable source says they're "far-right".
2. WP:V and WP:NOR say that editors are allowed to call them that if they find at least one (1) reliable source saying that.
3. But: Even if you have found one or more sources that say Breitbart News is "far-right", you are not automatically required to say that Breitbart News is "far-right".
4. The decision about whether to call Breitbart News "far right" (versus using other words that mean the same thing, versus not mentioning it all) is made through Wikipedia:Consensus-oriented editing and discussion.

I think most people would find it easier to understand this with a different subject. Consider: Sources say that Vanilla ice cream is delicious. This is literally a matter of personal taste, and therefore an opinion. But WP:V and WP:NOR work the same way:

1. WP:V and WP:NOR say that editors can't say vanilla ice cream is delicious unless and until at least one reliable source says that vanilla ice cream is delicious.
2. WP:V and WP:NOR say that if you have a good source, you are Y permitted to write in the article that vanilla ice cream is delicious!
3. But: Even though we have the sources, you are N not required to write that in the article.
4. The decision about whether and how to describe these sources' is something editors decide through editing and discussion. They are allowed to decide to mention it, and they are equally allowed to decide not to mention it. If they decide to mention it, they are allowed to decide to mention it in a stronger or a weaker way. I can imagine editors deciding not to mention it, and I can imagine editors deciding to mention it in a weaker way, like "Many people believe that vanilla ice cream is delicious". (Personally, I would argue that it should be omitted in favor of "Vanilla ice cream is the most popular flavor".)

In the case of Breitbart (as well as other people and groups for whom "far-right" is a common description in news articles), editors pretty much always agree that this verifiable label should be included. But this is something that editors have discussed and agreed upon; neither inclusion nor exclusion is automatic. It is not enough to say "I feel like this word is about feelings and opinions, so we should exclude it"; it is also not enough to say "I have a reliable source that says this, so we should include it". Editors consider all the facts and circumstances, all the reliable sources, the nature of the article (e.g., is this a trivial detail about the subject or a key point?), and then use their best judgment. Using your best judgment is always a subjective process. We will not re-write the policies to eliminate the need for human judgment. WhatamIdoing (talk) 06:41, 14 August 2023 (UTC)

• It isn't complicated. We rely on the sources and how they frame things, including for the question of whether something is opinion or fact. It's not impossible for one source to be an outlier or to get things wrong, especially if it is low-quality, so for WP:EXCEPTIONAL or WP:BLP-sensitive things we might reasonably insist on multiple high-quality sources, or search for sources that contradict them to check and see if it's actually more controversial than the sources we have imply; and it is important to make a distinction between WP:RSOPINION and WP:RSes that can be used for statements of fact. If you feel a source fails to make that distinction internally, that is a good reason to argue that it is not a WP:RS at all. But presuming a source is a RS that is usable for facts, it is never appropriate to dismiss such sources simply by going "Yeah, well, that's just like, your opinion, man." That argument attempts to substitute an editor's personal opinions, feelings, and judgment for the judgment of the source, and will never be acceptable - people trying to make arguments in that style need to WP:DROPTHESTICK already. There are numerous options available if you think the sources aren't good enough (I listed several just now), but "we cannot include this ever regardless of the sources, because my gut tells me it's opinion rather than fact regardless of what the sources say" is not an argument compliant with our core policies. --Aquillion (talk) 15:08, 14 August 2023 (UTC)

  @Aquillion, I agree with you, but some editors might misunderstand the bit about If you feel a source fails to make that distinction internally, that is a good reason to argue that it is not a WP:RS at all. and end up going astray. Consider:

  1. I believe that classifying a subject in a certain way is (in my opinion) a statement of opinion.
  2. This source would normally be considered a reliable source for statements of facts on this subject (i.e., when compared against the usual list of criteria, it has most or all of those qualities).
  3. But this source classifies the subject in a way that (in my opinion) is making a statement of opinion.
  4. Therefore, the source isn't reliable after all!

  It doesn't matter what the source is saying. They could be claiming that a certain politician is right-wing, or that a certain cell type is cancer, or that light of 500 nm should be considered a somewhat green-ish blue, instead of a slightly blue-ish green. The problem with this is that an editor's personal opinion about whether something is "an opinion" doesn't invalidate the source's reliability. WhatamIdoing (talk) 01:59, 15 August 2023 (UTC)

I definitely don't think we can conclude that a source is opinion based on its conclusions, and I don't think we can conclude it's opinion based on using specific words or phrases, as some people have tried to argue above - since most political disputes do involve a dimension of people trying to invalidate language or conclusions they disagree with, and since a massive swath of many political movements seek to invalidate basically everything about their opposition as "emotional" and not "logical", I don't think arguments premised on that have any real weight. People on team X are always going to say that everything they perceive as supporting team Y as being emotional opinion and not fact - I think it's genuinely how they feel, because people actually believe what they believe and because large swaths of politics are designed to evoke that feeling, but taking that seriously would result in biased articles. That said, it is definitely possible to conclude that a source is publishing unlabeled opinion (or using label X to actually mean opinion, the way eg. The Atlantic uses the "Ideas" label) based on the overarching composition across multiple articles, especially when large numbers of people with different perspectives weigh in and agree - things written in the first person, things overtly intended to convey a plainly subjective interpretation, and so on. There's no one simple rule but I don't think I've ever seen RSN as a whole get it wrong - the difference between "yeah that's obvious an opinion piece" and "this starts from a premise I disagree with, therefore it's opinion" is pretty obvious once you have a bunch of people weighing in. --Aquillion (talk) 17:33, 15 August 2023 (UTC)

I think that much can come from realizing that, contrary to the common metaphors, wp:ver and wp:nor allow editors to analyze the words that they are writing, that much can be sorted by analysis of the words, and that editors are somewhat free to leave something out. Where this comes up is usually when somebody is trying to use the fact that something is sourced to force it into the article. Adding "Verifiability is a requirement for inclusion, not a reason for inclusion" in a prominent policy place would mostly fix that. This would not be a change, it would merely clarify actual policy vs. myth.

For "Vanilla ice cream is delicious", editors could decide that "delicious" is so subjective that it is uninformative and should be left out of the article, even if somebody searched and cherry picked 10 sources that said it. North8000 (talk) 15:47, 16 August 2023 (UTC)

I don't think it would be helpful to use the "subjective" language, because some editors struggle to understand what Objectivity and subjectivity are. And then it would probably be useful to remind them that a neutral encyclopedia article can (and often should) WP:ASSERT facts about opinions, namely "Many people think this is delicious" rather than "it is delicious".

The next lesson in the series would have to be that "it is spicy" is not so obviously a subjective opinion. A good editor might well write that "Ghost peppers are extremely spicy" as a simple summary that precedes the explanation about world records and the Scoville scale. WhatamIdoing (talk) 17:14, 16 August 2023 (UTC)

Thats true, but if we look at a less extreme example like banana peppers a reasonable editor may be of the opinion that they are extremely spicy and another may think them barely hotter than bells. Likewise when it comes to Ghost peppers there may be an disagreement between whether they are extremely spicy and inedible, it remains subjective even if the breadth of opinions is smaller. Horse Eye's Back (talk) 17:21, 16 August 2023 (UTC)

True, but we don't really care about the opinion of the Wikipedia editors. A quick search indicates that reliable sources don't describe banana peppers as being extremely spicy, and therefore we can't add that at all. WhatamIdoing (talk) 04:09, 17 August 2023 (UTC)

• Editors are free to leave things out, but when we're deciding what to include or exclude, it's important to encourage editors to focus on valid policy-based rationales, not their personal opinions; and it's important to make it clear that "I think that this entire broad topic, aspect of coverage, or field of academic study is opinion, and not fact" (or "I think the entire landscape of sources is biased in this particular way" or comparable arguments) is a bad rationale for sweeping efforts to omit entire categories of coverage from otherwise high-quality WP:RSes and can, intentionally or not, ultimately be a sign of WP:TENDENTIOUS editing. Many of the core disputes in the AP2 topic area in particular involve large swaths of talking heads basically dismissing entire areas of academic study with "yeah, well, that's just your opinion, man"; and the argument that the entire media landscape is biased (so you just need to go with your gut) is similarly a go-to argument for POV-pushers who don't like what the sources say. To avoid a situation where that POV ends up influencing article content, we have to decide what's fact and what's opinion based on the presentation in the highest-quality of sources, not based on editors' guts. We're not required to include everything every single individual source says, but we are required to determine what we include and exclude based on what the sources say overall rather than our personal opinions or perspectives; and our articles ultimately need to reflect the focus and weight of the sources, even if editors personally feel that their conclusions are biased, opinionated, or otherwise flawed. The argument that we could "correct" for what an editor sees as those biases or flaws is clearly a WP:RIGHTGREATWRONGS argument and needs to be staked in the heart - "the sources aren't good enough for this" is (almost) always a valid argument, but "we should ignore the sources because I think they're wrong" is almost never acceptable. --Aquillion (talk) 17:45, 16 August 2023 (UTC)

  I agree with you.

  I think that the way this policy talks about opinions vs facts is suboptimal. Consider this sentence: If different reliable sources make conflicting assertions about a matter, treat these assertions as opinions rather than facts.

  Disagreements between sources do not magically transform facts into opinions. Consider the case of cancer screening programs, like the PSA test or Breast self-examination. You can find reliable sources with differing views. But "this saves lives" (or doesn't, as the case may be) is never an opinion. The efficacy (or lack thereof) of a medical intervention is a hard fact. These particular facts have been available in "true" and "false" flavors during Wikipedia's existence, and at various points in time, we didn't necessarily know which ones were true and which ones were false, but they're not ever opinions.

  But we are telling editors to "treat these assertions as opinions", which in my opinion is a bad way of writing the policy, because when we know that some editors will read this as meaning "If reliable sources disagree with each other, then it's all just their opinions". I think we'd be better off saying something like "If different reliable sources make conflicting factual assertions about a matter, then present the claimed facts with the level of certainty they deserve, based on their relative levels of acceptance by relevant experts, as indicated in reliable sources". WhatamIdoing (talk) 05:07, 17 August 2023 (UTC)

I think that that would be a reasonable rewording. At the very least it needs to make it clear that the relative weight of sources needs to be taken into consideration - obviously WP:FRINGE or extremely marginal views cannot reduce things that are broadly-accepted among mainstream sources to opinion. The traditional examples are eg. climate change, the roundness of the earth, or various examples of historical denialism, but to show that this extends beyond science, a much more immediate and obvious example - one where it would be easy to find recent editorial disputes where the weight of sources clearly matters - would be the winner of the 2020 US presidential election. Marginal sources that dispute it exist but clearly cannot render it a matter of opinion when weighed against unanimous agreement among higher-quality sources. --Aquillion (talk) 06:18, 17 August 2023 (UTC)

A statement that is made by multiple quality reliable sources can still be a statement of opinion even in the absence of any sourcing (even low quality or non-reliable sources) that give a counterpoint. For example, if multiple sources came out to say "Einstein was the smartest person that ever lived.", a statement that you likely would not find many compelling sources that argue against that, that is still a subjective statement (there is no way to objectively measure "the smartest person"), and Wikipedia must present it in the lines of "Einstein is frequently considered the smartest person that ever lived." We should also be taking the same approach with controversial claims including labels, even when a multitude of sources assert the claim applies factually and there are absolutely no counter claims in any source.

A problem stems from when editors start from the insistence that "X is controversial" in their own factual voice when discussing the content on talk pages. For example, I'll take the example above that Breitbart is a fringe far right publication. Certainly numerous sources support the claim that it is fringe and that it is far right, and per DUE, to not mention those terms in conjunction with Breitbart would be a failure of policy. But both terms still are controversial and subjective - how do you measure "fringe"ness? How do you define "far right" objectively? You can't, so while we still are bound to present those terms, thinking that we must present those as "Breitbart is a fringe, far right publication" is a problem because we still pushing subjective claims as a factual statement. The only thing objectively we can say about Breitbart is that it is a publication, and separately that it is widely seen as fringe and supporting far right views - that is an objective statement of what the popular opinion is about the publication. And to that, that means editors should be able to discuss sources, those that appear to present opinions or anything that would require a subjective analysis, as whether they should be present as fact in Wikivoice or tempered with some type of in-prose attribution as per YESPOV. Given that we should always be taking the less risky ground, we should be defaulting to using in-prose attribution for these.

Now, with time and further analysis detached from the current culture wars, the only prevailing stance on Breitbart (ideally from academics in 10+ years time) may be that it was a fringe, far right publication, at which point, it would be reasonably to write those terms without the "widely considered as"-type language. Its why we can identify the Manson Family objectively as a cult now, but we can't do the same for QAnon (though how it has been described as a cult is 100% fair game for inclusion per DUE). I won't get into this in detail, but I think that NPOV, particularly with the last decade+ of culture wars and how news media works, needs to make sure editors are thinking about NPOV from the long-term view and make sure that short term views on a topic are not artificially inflated as much as they are now. Masem (t) 23:17, 17 August 2023 (UTC)

Here's the problem: Breitbart was created in 2007 and was basically a fringe far-right publication from inception. And this type of thinking dangerously opens the door to attributing blue-sky statements because they're too controversial and too recent. Andre🚐 23:27, 17 August 2023 (UTC)

See, you're exemplifying my point, you want to walk in the door treating Breitbart as fringe and far-right. That does not help any discussions on neutrality aspects. And we should be attributing any "blue sky" statements that involve subjective determination that are in the recent, no matter how many sources back up that claim. The only factor that the number of sources make is the DUEness for inclusion of the claim. Masem (t) 23:35, 17 August 2023 (UTC)

No. Breitbart came out in 2007. That's quite a bit more than 10 years ago. And when something is so unequivocal in the source material and entirely unrebutted, it does not need any further justification. Andre🚐 00:45, 18 August 2023 (UTC)

However, you cannot prove a negative (that is, that there are zero usable sources that rebut the popular position), and just because you can't prove the opposite doesn't mean we take the current popular opinion as fact. That's certainly not how scientific theory works, and should be applied similarly when we are seeking to write something subjective in WP voice. That's why the time factor is important - as with scientific theory, if a stance stands the test of time, then theory becomes treated as a fact. Similarly, if a current popular opinion remains the way the topic is described far distanced in time, then it is fair to say that becomes fact. Editors need to put aside personal opinions on these topics when trying to determine how to write neutrally around them. Masem (t) 00:53, 18 August 2023 (UTC)

That is a significant departure from the standard interpretation of policy, and it echoes some of the more problematic rumblings of drive-by editors. It does not matter if we can prove a negative. We regurgitate what the sources say. Breitbart is fringe right-wing. Period. End. Full stop. Your opinion is your opinion. Standard, authoritative opinions are facts. Andre🚐 01:22, 18 August 2023 (UTC)

I would actually argue that the policy has been poorly applied, because of the current culture war outside Wikipedia, and that many Wikipedians, typically in thought agreement with mainstream sources (which is not a problem in itself), will become blinded by that agreement to not give any question to these sources (it is really easy to pile on everything one can that is negative around someone like Trump just because the news is similarly fasinated with that as well, but this is not how we should be writing an encyclopedia). Nothing in NPOV conflicts with what I've said, because its actually enshrined in YESPOV - don't state opinion as fact. When every source reports something that is still subjective at its core, we can absolutely include it, but make sure it is clear is the popular opinion or view of the topic.

What is probably at odds is back to the question originally asked: how to tell the difference between what is objective and what is subjective. Because YESPOV has both sides of this: just as we don't state opinion as fact, we shouldn't mask fact as opinion. But as I've said, something subjective does not suddenly become fact just because nearly all sources repeat it and/or no sources present a contrary view. And this is the issue when editors come in with a chip on their shoulders towards a given topic and unwilling to accept a more neutral, dispassionate approach to summarizing sources. Masem (t) 01:31, 18 August 2023 (UTC)

Actually, what you're arguing for is to give cover to fringe editors who want to claim that standard, authoritative, expert opinions on matters such as whether Marjorie Taylor-Greene is a conspiracy theorist or whether Breitbart is a far-right publication should be litigated. That it's somehow betraying Wikipedia's left-wing bias. There is not, and has never been, a culture war, unless you mean maybe 1968. What there is is a particularly loud, angry, and insurrectionist strain of anti-democratic crusaders in the United States, who are engaged in an online propaganda campaign to right great wrongs. We know this for a fact, just like we know for a fact that Marjorie Taylor-Greene is a conspiracy theorist because she is on record spouting conspiracies, and reliable sources call her that. Pop philosophizing and epistemology 101 has no place here or anywhere, least of all from a Wikipedia admin on the NPOV talk page. That is a very dangerous and wrong thing. Andre🚐 01:36, 18 August 2023 (UTC)

Just to offer a WP:3O, truth is outside of the scope of Wikipedia. (See WP:NOTTRUTH.) Arguing about truth usually makes disputes harder to resolve. I've seen editors borderline on harassment by swearing up and down that something is true, without producing a single reliable source to support their statement. I won't wax philosophical about "truth" except to remind people how Wikipedia articles are written, as a summary of what reliable third-party sources have said about a topic.

Whether you believe that "objective truth" is important or not, we only have our best practices to produce encyclopedia articles. If something is widely considered to be "true", then most reliable sources will say so. Vice versa, a statement widely understood as "false" will be directly contradicted by reliable sources, or otherwise not reported at all.

Sometimes reliable sources point in multiple directions, and we have a multitude of policies that help us resolve that:

• WP:V and WP:EXCEPTIONAL says that statements likely to be challenged will need to be supported in higher quality sources.
• WP:RS and WP:RSN tells us that some sources are better than others. We ignore WP:UNRELIABLE sources entirely.
• WP:FRINGE and WP:NPOV tells us that we treat minority viewpoints with caution, mentioning them in a smaller proportion than the mainstream view, if at all.
• There is also WP:OLDSOURCES that reminds us that sources may diverge because of newer, better information, correcting mistakes in earlier reports.

Those policies give us the tools to resolve contradictions between the sources, along with the sources themselves. Disputes tend to go off track when people try to make declarations, instead of using our best tools/policies/sources. Shooterwalker (talk) 15:26, 18 August 2023 (UTC)

Masem, you've made this claim about needing sources with a 10-year perspective for Breitbart and other right-of-center publications and people many times. I'm getting frustrated with it, because it seems like the goalposts are ever shifting.

So you say that we need to think about what sources will say after 10 years, because WP:RECENTISM is bad, and the WP:10YEARTEST feels appealing (even though that section says nothing at all related to sources needing to have a ten-year perspective), and you also say that we need to consider scholarly sources, but when you are given actual sources that meet your stated conditions – I provide here a quick list of scholarly sources with the number of years' of Breitbart's corpus they are responding to:

• "Breitbart News and other far-right publications" doi:10.17645/mac.v7i4.2280 – 12 years
• "the popular far-right Breitbart News Network" [7] – 13 years
• "Breitbart established itself as a media outlet between the extreme far-right and mainstream right" (i.e., "far right" but not "extreme far right") doi:10.1080/21670811.2019.1682629 – 12 years
• "Other highly influential sites include the far-right news site Breitbart" doi:10.1080/1369118X.2022.2118546 – 15 years
• "through the lens of far right website Breitbart.com" doi:10.1080/22041451.2018.1558790 – 12 years

– I expect you to say that these aren't the "correct" 10 years, because we can't write anything that makes you feel bad now unless we somehow have sources from 10 years in the future. A decade from now, I expect you to be saying that sources from 2033 aren't good enough, because we need sources from 2043.

This isn't a good approach to writing encyclopedia articles. It feels like a case of stonewalling, or a rather extreme case of Wikipedia:Bring me a rock, in which the only possible way to provide the Correct™ source actually requires time travel(!). I think you are wrong, and I would like to see you stop advocating for Wikipedia to be mealy-mouthed about negative "labels". WhatamIdoing (talk) 17:31, 18 August 2023 (UTC)

Respectfully, I don't think Masem's statements here are in anyway incompatible with WP:NOTTRUTH. Masem is not making and statement about the truth-value of these labels, like "fringe far-right". They are simply pointing out the reality that these labels, true or not, are subjective labels of opinion, and, barring exceptional cases, they should be treated as opinions - which doesn't mean "don't include them" (that would be whitewashing), it just means "include them, yes, but as attributed opinions, not factual statements".

Going back to the "vanilla ice cream is delicious" example - even if one could find 13 sources that say "vanilla ice cream is delicious", and no sources exist that definitively state "vanilla ice cream is not delicious", we would still obviously have to regard "delicious" as a marker of opinion, not fact, and should be attributed with "many authors, including X,Y,and Z, believe that vanilla ice cream is delicious". Any other course of action would lead to the utterly foolish outcome of an encyclopedia affirmatively taking a side on the question "what flavor is the most delicious?" The same is true of the Breitbart example. One can find all the examples you'd like of a news outlet describing them as "far-right", but that doesn't change the self-evident fact that "far-right" is, definitionally speaking, a subjective opinion, no less so than "delicious". "Truth" need not enter the conversation here - one can make a fact/opinion assessment without making a truth/falsehood assessment - they aren't the same.

If a claim is "verifiable", that means it might merit inclusion, but that does not mean that it is "true". I've seen several users echo the sentiment that "verifiability is truth", which is one of the most dangerous and wrongheaded ways of thinking about this issue that I could possibly imagine. Verifiability equals verifiability. Once we've found a verifiable source that makes a claim - hooray, we can add that claim to Wikipedia! But we can't weasel our way out of determining whether or not that "verifiable" claim is a fact or an opinion. Simply "trusting the source" (i.e. if the source utters something in a declarative sentence, that makes it a statement of fact) might work in the scientific domain, but if this guideline is pursued in the domain of persuasive political writing (and almost all journalism and political writing has become persuasive writing), it will lead to broken results. We have to treat obvious markers of subjectivity ("delicious", "extreme", "far", "fringe", "wonderful") for what they are. It might help confused editors to have a more thorough treatment of "what is an opinion?" on the NPOV page.

Option 1: Right way of thinking about verifiability - I've verified that 10 news outlets describe Breitbart as "far-right". Therefore, this is a notable and verifiable opinion that conclusively merits inclusion (attribution) in the article.

Option 2: Common, but wrong way of thinking about verifiability - I've verified that 10 news outlets describe Breitbart as "far-right", therefore, because the statement is verifiable, it is factual, and should be presented as a fact in Wikivoice.

I think that's part of the misunderstanding editors are having here. I see some editors trying to apply the same standards of fact/opinion analysis to persuasive, culture-war infused political writing that they would apply to a scientific journal article. Perhaps option 2 above is a good way to approach declarative sentences made in scientific journals, but we are going to consistently have these NPOV issues as long as we keep applying option 2 to the domain of persuasive political writing - and remember, almost all political writing is subjective and persuasive, even if it's not explicitly labeled as such. There needs to be more awareness of this amongst the editor community. Philomathes2357 (talk) 19:04, 18 August 2023 (UTC)

That would be a change to the current policy. One that I, and many others, would strongly oppose. Andre🚐 19:15, 18 August 2023 (UTC)

Specifically, requiring attribution for extremely widely held opinions could violate the "Indicate the relative prominence of opposing views" part of the NPOV policy by making it seem like this opinion is not held by a significant minority of sources. If you find yourself writing, "According to many sources, including X, Y, and Z, breathing in humans is preferable to not breathing", then you have screwed up, even though that is, technically, an opinion.

And, of course, for actual facts – such as how a publication is classified on a particular political spectrum – when the mainstream viewpoint is overwhelmingly held to be so by independent reliable sources, then we would not .

It might help to think about other spectrums. Autism spectrum disorders, for example, are classified these days in levels 1, 2, and 3. The difference between "low 2" and "high 3" is fuzzy, but that doesn't mean that when someone says "This student has ASD level 2", that we should say "Oh, that's just someone's opinion, and you could only write that 'According to sources X, Y, and Z, this student has ASD level 2'." Autism and its levels are facts, not opinions. According to the policy, you should just write (if it's relevant to the article) that the person has ASD level 2 and leave out the doubt. Adding "according to X, Y, and Z" creates doubt in the mind of the reader. Some of them see a list of examples and think "Oh, it says many sources such as', but in reality so few sources think this that they can actually name all three of them in the sentence". WhatamIdoing (talk) 00:47, 19 August 2023 (UTC)

If we said "Many journalists consider Breitbart a right-wing news outlet.", that does not imply that there are counterviews of any significance that deny that. Instead, that simply implies, at best, some journalists have not commented on Breitbart's position. Even when a position is so overwhelmingly held by the media and the only agency arguing against that is the topic at hand (eg Breitbart here), you still have the case that there are likely other major sources that simply haven't commented on that, so it would be wrong to say "Most" or "All journalists", and turn that statement of public opinion into fact.

For us as an encyclopedia, it is necessary to be impartial and dispassionate and separate our writing from that of the public opinion. It does not hurt our presentation in any way, and improves our neutral tone to make sure we don't take subjective claims as fact until far far down the road. Masem (t) 12:25, 21 August 2023 (UTC)

I don't agree.

• "Many journalists, such as Alice and Bob" = "Alice and Bob and maybe some others, but probably not a lot"
• "Many journalists" = "Okay, the journalists think that, but probably nobody else agrees with them, and not even all of the journalists think this".

We wouldn't write "Many scientists, such as Alice and Bob, believe that the common cold is caused by a virus" precisely because it would create doubt where none is warranted, even though "most" scientists have never researched the subject and therefore have no business commenting on it (just like one would expect that "most" journalists haven't commented on Breitbart, since "most" journalists write about sports, business, and local news instead of English-language political sites). Why should we be creating doubt about whether Breitbart is right-wing, especially given that we have many sources saying "right" and none saying "left" or "center"? WhatamIdoing (talk) 23:03, 21 August 2023 (UTC)

Many of the answers can be derived for a logical analysis of the words and recognition of the realities regarding them. Editors need to be allowed to do this. I'm going to start with an obvious example "John Smith is a bad person". Since there are infinite different criteria for "bad", that is an inherently ambiguous sentence. And it is completely subjective, because anyone can decide any criteria for "bad" It can't be shown to be true or false, and in common communication it denigrates John Smith. Editors should and should be allowed to analyze and decide this. Including that in this case that the statement provides no information and should be left out. To decide that the process of looking for sources that say "John Smith is a bad person" is an effort that should not be taken. WP:VER and WP:NOR allow this, many of the the commonly accepted metaphors of WP:VER and WP:NOR do not. The vague and dysfunctional wp:weight confuses it but does not preclude it.

The same concepts can apply for less obvious cases. Sincerely, North8000 (talk) 20:06, 18 August 2023 (UTC)

No, because "John is a bad person" is a normative value judgment. "Breitbart is a far-right publication" is a factual statement. "Marjorie Taylor-Greene is a conspiracy theorist" is a factual statement. Editors want to claim that political spectra are subjective. But they aren't. There is a factual basis that an academic consensus of experts in politics believe Breitbart is far-right. Just like experts in science and medicine, get to decide what a virus is and what a frog is, political experts get to decide the special political facts of the day. And we just regurgitate that information according to weight, prominence, reliability, verifiability, NPOV, etc etc. But not put a finger on the scales of NPOV to keep our noses clean of anything that has a whiff of offending the delicate sensibilities of those who have argued that "far-right" is a pejorative or offensive description. It's not, in fact it is grossly anatomical. You have simply deprived the politics subject area of its expert accuracy in sourcing and in determining what is factual. It's just a fact that Bernie Sanders is on the far left of the American political spectrum, though not as far left on the global scale. That is not a pejorative. We didn't say Bernie Sanders was a bad person or a radical or a troublemaker. We made a factual observational analysis of political information. Actually, we didn't do it. The sources do. Andre🚐 20:43, 18 August 2023 (UTC)

Let's take your Breitbart example. I actually agree with using the "far right" label in an article, but not with your argument which arrived at it. Since the definitions/meanings of "far right" are variable, one can cherry pick which meaning they are using which means that the application of that label is also subjective. However, you picked an example which is at the extreme which makes the selection of a term more of a slam dunk. Probably 98% of the definitions of "far right" would classify Breitbart as far right which moves it into the realm of fact. If it was 51% it would be a statement of opinion, and if it was 20% it would be a misleading/deceptive statement. North8000 (talk) 13:33, 20 August 2023 (UTC)

It sounds like the definition of "fact" you are using is "something most people agree on", and that "opinion" is "something reasonable people could disagree about". I'm not sure that those are really the best definitions for Wikipedia's purposes. WhatamIdoing (talk) 03:24, 21 August 2023 (UTC)

See my post below. North8000 (talk) 12:58, 21 August 2023 (UTC)

20% is basically the definition of a minority opinion... Not a misleading/deceptive statement. Horse Eye's Back (talk) 06:44, 21 August 2023 (UTC)

Due to various reasons (especially there being many different definitions of the terms used), most statements are ambiguous but leave a certain impression based on the common meanings of a term. You might call that "opinion" but for me that is too vague of a term and only a secondary use for the term "opinion" For example if I say "The Watergate break-in is a myth", the common meaning of "myth" includes "false" but the technical and uncommon (20%) meaning of "myth" can include a re-told story which is not false and so by that definition my statement is technically true. But I would call my statement "misleading", something a POV warrior is likely to insert, and something that the editors should feel free to leave out (or attribute or explain) despite any argument that "it's sourced". North8000 (talk) 12:58, 21 August 2023 (UTC)

But in this context it would be 100% just one meaning, that the event had been mythologized. Thats not how being true works unless you're making some extremely esoteric philosophical point. Horse Eye's Back (talk) 22:49, 21 August 2023 (UTC)

Ah, but it's not just one meaning. I'd read "The Watergate break-in is a myth" and not think about the classical myths (The new Twelve Labours of Hercules: killing the Nemean lion, killing the hydra, capturing the hind, discovering political corruption...), but instead assume it was written by someone who thinks Richard Nixon was forced from office by a shadowy, world-controlling conspiracy. I wouldn't call that idea "a myth", though; I'd call it "an error". WhatamIdoing (talk) 23:13, 21 August 2023 (UTC)

Unless you're a subject matter expert who chooses to publish on the topic how you read it has zero bearing on NPOV. Horse Eye's Back (talk) 17:52, 26 August 2023 (UTC)

I think that the information we are trying to communicate, and how readers would likely (mis)understand a given phrasing of that information in an article, is the core job of Wikipedia editors. If we write "myth" and some people hear Urban myth and some people hear National myth and some people hear Political myth and still others hear Greek myth, then we've failed at our most basic task. Wikipedia editors need to write clearly, and that means understanding how people could (mis)read what we have written. If you read it one way and I read it another way, then that suggests we should be picking other words.

Perhaps the new words would be "Alice Expert called Watergate a "myth", meaning that it is a story that is handed down between generations to explain that even the US president cannot commit crimes with impunity". We could write this because presumably Alice didn't just say "It's a myth", but actually explained this label in some detail, which we can then summarize without violating NOR. (If she didn't provide anything except those three words, then her three-word comment is probably UNDUE anyway.)

Perhaps instead the new words would be "Chris Conspiracist called Watergate a myth and claims it never happened."

But when editors read a given word choice in divergent ways, then how they read it has quite a lot to do with NPOV, because negligently letting people think that Watergate was a Greek myth is a violation of all the core policies. WhatamIdoing (talk) 20:09, 27 August 2023 (UTC)

• While I agree with Philmathes2357's comment that "This has been a very interesting conversation so far", especially considering my [ex-]membership in a local philosophy book club, I question the appropriateness of this discussion here, and wonder if anything has been accomplished in any of this ginormous wall of text so far. What I see, is a very interesting conversation that should've been held perhaps at WT:PHILOSOPHY but hasn't resolved anything here, nor, imho, has it helped to improve WP:NPOV, which is the only raison d'être of this page. In this, I agree with User:Horse Eye's Back's comment in the very first response to the OP at the top of the section. To be blunt: Philmathes2357, do you have a specific proposal in mind about how to improve NPOV? By "specific proposal" I mean, not your OP or other comments, which I've read and found very interesting, but rather a before–after text presentation indicating exactly what words of the current project page you wish to alter or add to? If you could please make your proposal below this, using the {{TextDiff}} template, it would make it easier for me to understand what it is you are proposing. If that isn't possible, then I would say this whole conversation should be hatted and closed, or moved to another venue. I'm sorry if that seems harsh, and by the way did I mention that I really am enjoying this very interesting conversation? Mathglot (talk) 08:11, 21 August 2023 (UTC)

IMO there are some mostly unrecognized underlying structural elements involved here and those need to brought into daylight and sorted out before a specific proposal can be made or have any hope of passing. IMO this discussion is working on that. North8000 (talk) 13:02, 21 August 2023 (UTC)

I agree with North8000 on the benefits of this discussion. Not every talk page discussion needs to include a proposal for specific language to amend the policy. Sometimes we need to have preliminary discussions that may (or may not) result in more formal proposals. Those preliminary discussions can be lengthy and become very philosophical. That’s actually a good thing. It helps us move beyond just wikilawyering about policy language. It means we are examining the intent behind that language. Blueboar (talk) 13:50, 21 August 2023 (UTC)

I'd like to echo what @North8000 and @Blueboar said. I agree with their replies 100%. Perhaps they have some thoughts on how this could be brought into the realm of concrete, specific, actionable ideas. Because @Mathglot's critique is wise and well-considered. Allow me to restate my position more concretely in terms of what I'd like to see, and how it would affect Wikipedia, with less general philosophical commentary.

Specific idea

Here's one direct, concrete way to benefit NPOV: we have an example of a fact/opinion dispute on the NPOV page already - the "genocide is an evil action" example. I think we should expand the "avoid stating opinions as facts" and "avoid stating facts as opinions" sections to include multiple examples of each. That could help to address the underlying problem I'm trying to probe here, which is that a concerning number of editors simply don't know the difference, or distinguish between the two along some frankly bizarre criteria. I'd also support some language that encourages users to err on the side of attribution when the empirical, factual nature of a claim is not absolutely clear and incontrovertible. We obviously don't want to enter the realm of clownish absurdity by attributing "according to many geographers, France is in Europe", but short of the absurdly clear examples of "facts", it almost never hurts to err on the side of attribution.

The core of the problem - common yet problematic ways to attempt a fact/opinion distinction

Editors will say "stop whitewashing - it's just a fact that such-and-such is X", but curiously, all the "facts" seem to align with their personal political opinions, and the "facts" that don't align with their personal POV are all somehow just "reliably sourced opinions". That logic, applied community-wide, creates an epistemic wild west where things that are definitionally opinions can become Wikivoice facts by winning a popularity contest, and vice versa. That won't do - Wikipedia should aspire to be more than a mirror that reflects and magnifies the norms and opinions of the editor community. Another popular approach comes down to "look at how the RS frames things - if they call it an opinion, it's an opinion, but if the RS feigns factual objectivity by using declarative sentences to express their opinions, we should uncritically treat those declarative sentences as facts, even if we know that they're clearly opinions". That won't do either. Reliable sources sneak their opinions into supposedly objective political commentary constantly and incessantly, especially in the post-2015 era. There have to be standards that are grounded in the definitions of the words "fact" and "opinion" that aren't subject to the whims of RS sentence structure or editor discretion.

My vision for how to improve NPOV

In more general terms - it would take a fair amount of further discussion to formulate a specific policy here - I think NPOV would benefit from counseling editors to almost always err on the side of attribution. Especially in the domain of political labels. I see very significant upsides to NPOV and overall article quality by heavily leaning towards an attribution model. I personally think that political labels, outside of the bare-bones basics of "left (of center)" or "right (of center)", should always be attributed, even when the "reliable sources" that we can find often throw around a plethora of other labels.

I'll share one example - terrorist. We used to call Al-Qaeda "terrorists" and "extremists" in Wikivoice, but we no longer do - even though reliable sources do. That is because we know that "one man's terrorist is another man's freedom fighter", and describing a group as "terrorists" clearly indicates that the author has an unfavorable view of the group. The many hundreds of thousands of people who have supported Al-Qaeda's activities certainly don't think of themselves as "sponsors of terrorism". It doesn't take a reliable source, or OR, to recognize that "terrorist" is an opinionated marker, it just takes basic political literacy and a knowledge of what "fact" and "opinion" mean. Even though we may not have RS in English that describe Al-Qaeda as "not terrorists" or "freedom fighters", we can recognize "terrorist" or "extremist" as a negative, value-laden term of opinion, and thereby make the wise choice to attribute the term, even if 5, 10, or 20 RS have used the term in a declarative sentence. We don't remove it - just attribute it to the reliable sources that have expressed it.

Basically:

• Al-Qaeda is a terrorist group - opinion
• Al-Qaeda has been described by many western scholars and media publications as a "terrorist" group, including sources W, X, Y, and Z. - fact

Avoid stating opinions as facts.

I have a list of labels similar to "terrorist" that I think we'd be wise to always attribute and never state in Wikivoice, regardless of the sheer quantity of opinions expressed by reliable sources on the matter. I do not share the concerns of other editors that we will run afoul of NPOV by violating "avoid stating facts as opinions". I don't see how Wikipedia benefits in any way whatsoever by affirmatively jumping into the fray and stating widely-expressed political opinions as Wikivoice facts, but I do see significant downsides. On the other hand, I see significant benefits to NPOV and article quality from erring on the side of attribution.

Please note that in this model, absolutely zero reliable sources would be removed or discarded - I would strongly oppose that - but a significant number of them would be attributed, rather than used to justify & enable stating opinions in Wikivoice. Philomathes2357 (talk) 17:00, 21 August 2023 (UTC)

What do you think of the wording I suggested above? I think it’s broadly aligned to what you are saying. Barnards.tar.gz (talk) 17:25, 21 August 2023 (UTC)

I think that your analysis and ideas are good and a good start. I think that there's another pretty good concept to use is to lean towards avoiding value laden words. (strengthen our obscure advice regarding this) By their nature they are subjective, based on the values/definitions of the persons who said it. What makes this easier to decide is that we are here to provide information, and as such those statements tend to be either un-infomative or not very informative. To make an extreme example, I'll go Godwin and take a case where there is such universal agreement that it can be considered as fact. It is not useful/informative to a reader to put "Hitler was a bad person" in the Hitler article. Putting in particulars regarding that would be informative. North8000 (talk) 17:48, 21 August 2023 (UTC)

But that's a strawman argument. "Hitler was a bad person" is in no way analogous to correctly indicating that a website or political party is "far-right" on the Left–right political spectrum. 76.142.90.140 (talk) 01:06, 22 August 2023 (UTC)

This statement: short of the absurdly clear examples of "facts", it almost never hurts to err on the side of attribution is not true. A true statement probably sounds like "I personally do not perceive the harms that unnecessary attribution cause, and the issues that I would have to admit are true – such as making articles more verbose, which does not align with the value of concision in encyclopedic writing styles – are not issues that I personally value". WhatamIdoing (talk) 23:20, 21 August 2023 (UTC)

Another thing that strikes me: The argument that we must not say that ____ is a terrorist group is based on the opinion that "one man's terrorist is another man's freedom fighter".

This may be true in some cases, but in those cases, you should be able to say "Alice says he's a terrorist but Bob says he's a freedom fighter". When nobody actually makes that latter claim, then we can (and do, and should) say that it's a terrorist group. We thus would not say that the Irish Republican Army is a terrorist group, because viewpoints legitimately differ. The IRA used terrorist tactics to advance their freedom-oriented goals. OTOH, we should and do say that Boko Haram is a terrorist group, because they are not fighting for anyone's freedom, not even in theory.

As for it being "subjective", I refer you to Jonah Goldberg, who wrote in The Tyranny of Cliches that "It is simply absurd to contend that because people may argue over who is or is not a terrorist that it is therefore impossible to make meaningful distinctions between terrorists and freedom fighters." These are actually objective labels; most definitions involve objective criteria like "non-state entity" (as opposed to government military organizations) and "for political reasons" (as opposed to common criminals).

Additionally, consider this viewpoint: a person can be a person who uses the means of terrorism – that makes him a terrorist – to fight for the end goal of freedom – which makes him a freedom fighter. These are not mutually exclusive categories; one person can be both at the same time.

I don't think we should adopt an opinion (or "pithy hogwash", in Goldberg's words) to tell us that we should not follow the sources when they say that a person or an organization meets the definition for terrorist. We should consider whether it's DUE; we should consider whether it's a passing comment or a quotation from an affected person, but when we have serious sources stating that this group is a terrorist organization, and no significant viewpoints opposing this, then we should just state that fact.

(Try this blog post, which makes a similar point, if you can't get Goldberg's book.) WhatamIdoing (talk) 00:49, 22 August 2023 (UTC)

I agree with the statement made by Mathglot. For every article, the important thing is that the precise wording was worked out on all of those articles by consensus. Editors didn't want to write "Hitler is a bad person" but "genocidal dictator" covers it. Al-Qaeda is described as a paramilitary jihadist organization or whatever precise term editors, by consensus, determined to describe them as. "Jihadist paramilitary group" isn't exactly a complimentary term. It's quite laden. But it is nonetheless descriptive. It's a very slippery slope, since every term has some amount of value laden in it. Guidelines are guidelines and Wikipedia intentionally does not provide very bright lines and specific guidance that is incontrovertible. As far as terms such as "far right wing" that is a term that editors by consensus have determined to include. I can't imagine there will be consensus for a formal change to bar such wording. Do you really think that a consensus of editors will agree to that change? Andre🚐 19:05, 21 August2023 (UTC)

I can't imagine and wouldn't want categorically banning such wording. I'd image a framework where editors would discuss including these considerations:

1. Is it informative vs. just value laden? Reasonably informative but somewhat value laden.
2. Is it pejorative or deprecating? In many context it is, in many it isn't
3. How clearly is it actually the case? Does the classification comply with the common meaning or majority of the definitions of the term? Review sources to help on this (preferably neutral ones) ....not cherry picking searches.

A typical set of neutral editors would probably decide that if the answer to #3 was a strong "yes", yes, if not not. If they decided yes, then get it sourced to comply with wp:VER and WP:NOR. Sincerely, North8000 (talk) 19:28, 21 August 2023 (UTC)

When you mention consensus, that immediately brings up the issue it'd by Philomathe above under "The core of the problem - common yet problematic ways to attempt a fact/opinion distinction", in that in these type of discussions you get groups of editors that share the same view as the sources and thus they do not consider that taking that POV as fact being problematic. I've seen enough of those discussions to know that self-feeding loop happens too frequently. While I don't expect policy to deal with behavior issues, we should at least remind editors that when dealing with NPOV, their personal opinions about a topic should be left at the door to avoid impacting content discussions or at least making editors blond to other options. --Masem (t) 20:25, 21 August 2023 (UTC)

I don't agree that editors are making editorial decisions based on whether they share the same view as the sources. Editors try hard to separate their own views from the views of the consensus of academic and expert authors. But yeah, editors and sources are likely to agree many things, especially when sources are stating uncontroversially shared basis points of reality, such as the statement "Breitbart is far-right." Nobody serious disputes this, and when they do, it's automatically FRINGE. Andre🚐 20:38, 21 August 2023 (UTC)

Automatically dismissing any option for content that doesn't agree with "We must state Bretbart is far right as fact" as fringe is exactly the problem I describe. Key to this is the mislogic that if there is no published dispute for a statement of opinion, then we must be treated as fact.

Note that when editors come around and demand insertion of a true fringe view eg "Breitbart is a liberal work!", without RSes to support that, then absolutely we should dismiss those. Masem (t) 20:51, 21 August 2023 (UTC)

That isn't what WP:FRINGE tells us. I won't quote it at you, but it tells us quite clearly that fringe views are any views that depart from the mainstream significantly. WP:RGW also tells us in-depth how to view and with editors who do not drop the WP:STICK but insist that a fringe view ("Breitbart isn't a far-right publication") has some kind of currency. Without evidence, or any supporting facts, or any reliable consideration whatsoever that I've seen, just tautological, syllogistic argumentation about epistemology, which is really a form of sophistry and belaboring or bludgeoning a dead, dead equine. It doesn't matter what the truth value of the statement is - it's a view. NPOV is clearly talking about views which, in another synonym, are opinions. You and I don't experience the reality tunnel the same way (obviously) and most things you think are "Facts" are actually suspended disbelief through lossy faulty measurement substrata. Not to get all metaphysical on ya, but it's hardly for Wikipedia to solve the hard problems of logical positivism, consciousness, relativism, quantum reality blah blah blah... we decided long ago not to get into all that. What we have decided is that when different POVs exist, the vast majority or near unanimity of respectable positions say X, and some very whackadoodle dingdong jimmies think that 5G gives you mind control - well, that doesn't go in. But that fringe extends to everything, even commonly held insane cult ideas like Qanon. We are an encyclopedia of science and rationalism in our way of constructing the NPOV and RS policies around verifiability. On Wikipedia, verifiability IS truth. That's because we don't open the door to trolls, cultists, and brigadiers who are attempting to create more of a market for a specific idea than demand would organically do for. In the name of WP:FALSEBALANCE and false-NPOV. NPOV never said do not state any views in Wikivoice. It actually is very clear that you should do so. WP:YESPOV does not say that we need to open the door for "controversy" like that there isn't anthropogenic climate change or that vaccines aren't effective or cause autism. Wikipedia doesn't start making these things subjective just because there's a cult out there believing it, a loud minority of wrong people, does not make it open to interpretation. That fringe gets put in the right box. Andre🚐 21:26, 21 August 2023 (UTC)

However, my point has nothing to do with FRINGE, but we do need to make sure that how we deal with FRINGE is still handled correctly.

My point is to understand if and when as WP editors can take a subjective statement and turn that into a statement of fact. (Which ties to understanding when something is a subjective vs objective statement). To use the Breitbart example, I would hope that it is clear that "far right" is a subjective term due to the vagueness of the definition of "far right, so a claim made by one RS that a publication is "far right" should be taken as subjective for purposes of NPOV.

If only one or two RSes out of many talking about the publication make that statement,, then inclusion may be UNDUE, When a fair of RS start to say the same thing, but as per YESPOV, we should use attribution here.

Now, let me jump to the other side of the issue, where many sources, far distant in time and independence from the topic, still call a publication far right, then we likely have a good reason to treat that statement as fact despite it being subjective. But we need a long distance in time (at least 10 years but likely more based on the current culture war) and the sources that are independence of thought (like academics) to make that call. But I absolutely agree that we can eventually make subjective claims with only wide agreement in sources as fact without need of attribution.

That leaves the area between when a subjective statement is DUE to include, and when we can eventually treat it as fact, which is where we are for Breitbart. I believe we should be very cautious and impartial here and keep such statements as attributed claims, as to avoid getting caught out in the prevailing view of popular opinion. When we can leave the point of being an attributed claim included via DUE, and switch to a statement fact as a huge grey area, and that is a major point of contention in these arguments, but we should be waiting far far longer, and particularly for most topics where this is an issue, while we are not in the midst of a culture war. Masem (t) 23:39, 21 August 2023 (UTC)

What exactly is this "argument from vagueness"? Who said there was anything "vague" about "far-right"? We know exactly what it means. It's not vague. If you're into the stuff Breitbart are into, those are far-right on the political spectrum. Nobody really disputes this. I don't know why you say it's so clear that it is vague and subjective. No, it's a political descriptor. They are relative, not exact, and the lines are somewhat blurry, but they are not subjective. You can't just read Breitbart and be like "oh, yeah, it smells/tastes more centrist to me." "Breitbart is good, bad, happy, sad, wrong, right" - those are subjective. "Right-wing" is factual. Your understanding of it is contrary to that of political experts.

There is no current culture war that gives you the ability to impose or extend a recency deadline. The culture war's existence is subjective, according to you. and it's meaningless. Eventual objectivity of subjective claims is also contradictory. You just go on arguing that some things are inherently a certain way. You're not Plato. Andre🚐 23:47, 21 August 2023 (UTC)

Some people personally feel like far right is too vague for them to personally figure out whether something falls in that category. Therefore, instead of accepting that reliable sources are reliable, they try to discount the sources ("they can't mean that, because it's too vague").

The same claim can be made about cancerous vs benign tumors, only most editors don't know that. The difference between "pre-cancer" and "very early stage cancer" is sometimes a marketing decision rather than a strict scientific one. And yet we don't see editors claiming that the sources aren't reliable when they say that a particular morphology should be called "cervical dysplasia" or "cervical intraepithelial neoplasia" or "carcinoma in situ of the cervix"; we accept that reliable sources are reliable for labeling this vague, fuzzy spectrum.

It's hard not to assume that there is a certain amount of political POV pushing going on behind the insistence that vague, fuzzy, spectrum-related labels must be rejected for matters involving politics, but are swallowed whole without complaint for hard sciences. Perhaps that's a talking point on some political website – to say "You can't label us as being extreme far-right folks, because we entirely reject the idea that anybody could figure out how to classify anyone, even if we say ourselves that we're more to the right of 99.9% of this country's adults. Being in the 0.1% isn't 'extreme' or 'far' or 'right'." WhatamIdoing (talk) 03:06, 22 August 2023 (UTC)

IMHO, we accept fuzzy labels in science where they are not contentious. Is there anyone arguing the other side of the tumor example? It’s the contentiousness that should cause us to raise our standards for the use of wikivoice.

Ten years ago, I’m not sure there was a lot of contention over what “far-right” meant. Today, we see it being deployed to cover an ever wider range of positions. Here’s a reliable source describing Jordan Peterson as “far-right”: [8]. It’s mutating from being a relatively well-grounded political science term, into being an attack term that just means whatever the author doesn’t like. In that respect, it’s becoming no better than “woke”, used on the other side for the same reasons. Barnards.tar.gz (talk) 07:29, 22 August 2023 (UTC)

Let's unpack that. Why would a reliable source place Jordan Peterson in the Alt-right segment of the Left–right political spectrum?

Casual misogyny? Check.

Casual racism? Check.

Weird stuff about "alpha males" and pseudoscientific stuff about social hierarchies? Check.

Anti-LGBTQ attacks and bigotry? Check.

Rants about "marxism"? Check.

Conspiracy theories? Checkitty check check.

Perhaps analysis by some of his contemporaries[9] will lend us an idea of why a reliable source might describe him as "alt-right" legitimately?

He is also adored by figures on the so-called alt-light (basically the “alt-right” without the sieg heils and the white ethnostate), including Mike Cernovich, Gavin McInnes and Paul Joseph Watson. His earnings from crowdfunding drives on Patreon and YouTube hits (his lectures and debates have been viewed almost 40m times), now dwarf his academic salary.

Not everybody is persuaded that Peterson is a thinker of substance, however. Last November, fellow University of Toronto professor Ira Wells called him “the professor of piffle” – a YouTube star rather than a credible intellectual. Tabatha Southey, a columnist for the Canadian magazine Macleans, designated him “the stupid man’s smart person”.

“Peterson’s secret sauce is to provide an academic veneer to a lot of old-school rightwing cant, including the notion that most academia is corrupt and evil, and banal self-help patter,” says Southey. “He’s very much a cult thing, in every regard. I think he’s a goof, which does not mean he’s not dangerous.”

Oh there's more?

You can read this stuff any day of the week and perhaps agree with some of it. However, Peterson goes further, into its most paranoid territory. His bete noire is what he calls “postmodern neo-Marxism” or “cultural Marxism”. In a nutshell: having failed to win the economic argument, Marxists decided to infiltrate the education system and undermine western values with “vicious, untenable and anti-human ideas”, such as identity politics, that will pave the road to totalitarianism.

Peterson studied political science and psychology, but he weaves several more disciplines – evolutionary biology, anthropology, sociology, history, literature, religious studies – into his grand theory. Rather than promoting blatant bigotry, like the far right, he claims that concepts fundamental to social-justice movements, such as the existence of patriarchy and other forms of structural oppression, are treacherous illusions, and that he can prove this with science. Hence: “The idea that women were oppressed throughout history is an appalling theory.” Islamophobia is “a word created by fascists and used by cowards to manipulate morons”. White privilege is “a Marxist lie”. Believing that gender identity is subjective is “as bad as claiming that the world is flat”. Unsurprisingly, he was an early supporter of James Damore, the engineer fired by Google for his memo Google’s Ideological Echo Chamber.

Aww gee Wally, I can't imagine why anyone would ever think this guy is "alt-right" on the political spectrum. I mean sure, his associates are primarily alt-right, his supporters are primarily alt-right, and everything he's said and done for the past decade waves a giant signal flag with "alt right" spraypainted on it. A major part of his bete noire is to openly use an antisemitic conspiracy theory. But still, none of that could possibly mean he falls into the alt-right segment of the political spectrum, could it? 76.142.90.140 (talk) 12:55, 22 August 2023 (UTC)

Barnards, the experts don't seem to have big discussions about how horrible it is that there is a fuzzy line between "barely cancer" and "almost cancer", because they accept that it is inevitable, especially when initial classifications are made with incomplete information. It probably also helps that they know that benign tumors can kill you much faster than than some cancerous tumors, so while the distinction is important for some purposes (e.g., grant funding), it's not necessarily the most important thing in terms of prognosis.

There are, however, periodic debates about the classification of individual conditions. The most common form of what was then called cervical cancer was renamed to "not a cancer" status a few decades ago. The result was much more proportionate responses to the condition (in terms of both medical treatment and fear). There's an effort now to rename Ductal carcinoma in situ (from considering it "early-stage breast cancer" to classifying it as "pre-cancer of the breast"). The general news media really only cares about these when the condition is common (e.g., cervical cancer or breast cancer). WhatamIdoing (talk) 15:40, 22 August 2023 (UTC)

Brand new editor 76, I really couldn’t have asked for a better illustration of the problem. I cited a source misusing “far-right” and you respond with a defence of the use of “alt-right” - a notably different term which doesn’t appear in that source. Worse, your reasoning is a tenuous chain of guilt-by-association. He is admired by people who are alt-light -> alt-light is adjacent to alt-right -> alt-right has overlap with far-right -> therefore he is far-right. This is exactly the kind of Overton-window-shifting semantic creep that is rendering the term useless. Barnards.tar.gz (talk) 09:32, 23 August 2023 (UTC)

"The alt-right (abbreviated from alternative right) is a far-right, white nationalist, white supremacist, anti-LGBT, anti-immigration and antifeminist movement." If someone is alt-right, and he clearly is by reliable source coverage, he is also far-right. The reporting of the last decade since he began his anti-lgbtq crusade and stopped being an unnotable pop psychologist almost universally uses either far-right or alt-right to describe him. 76.142.90.140 (talk) 13:12, 23 August 2023 (UTC)

If you really think this mess of equivocation and guilt-by-association results in the label of "far-right", the place to argue it is on the Jordan Peterson article, where "far-right" has failed to achieve consensus multiple times.

As far as this thread goes, I maintain that you are doing sterling work in demonstrating how the term gets redefined by culture warriors (including those writing for ostensibly reliable sources) for their own purposes, and has drifted far from its baseline meaning. Barnards.tar.gz (talk) 14:30, 23 August 2023 (UTC)

Peterson has his own little army of fans who pop up to defend him when anyone points out his promotions of bigotry, so I'm not surprised some of those fans have been ardent in trying to whitewash the article on wikipedia. You make just one good suggestion here despite your misrepresentation of sources, there's a full decade of reliable sources describing him as far-right and alt-right so it's time for a review of that on the article page. 76.142.90.140 (talk) 14:48, 23 August 2023 (UTC)

Also, the source you listed is not "misusing" the term far-right. They describe him as far-right and justify this in text citing

the plethora of examples detailing Peterson’s vehement anti-LGBTQ+ sentiments – including several dozen tweets and audio snippets of him comparing the trans community to “borderline schizophrenic” individuals

his refusal to correctly gender actor Elliot Page in a set of tweets in July, prompting his several-month-long ban from Twitter before being reinstated after Elon Musk’s acquisition of the site.

Since being reinstated, Peterson’s rhetoric against the transgender and LGBTQ+ community has continued among his repeated tweets of lockdown scepticism and far-right ideological rhetoric.

Your claim that the source is "misusing" the term is itself misrepresentation of a source. Please don't do that. 76.142.90.140 (talk) 13:16, 23 August 2023 (UTC)

Not sure I understand the assertion that political science is more subjective than the other social sciences, if these are opinions then is there anything beyond such treatment in the social sciences? Horse Eye's Back (talk) 23:50, 21 August 2023 (UTC)

I agree with North's list of questions. Saying that we must always/never use a given term without attribution is an example of all-or-nothing thinking. There are times when we should "label" (e.g.,) a terrorist organization, and also times when a different description would be more informative. We have to take multiple factors into account. This requires nuanced thinking, not the all-or-nothing thinking of "Oops, it's on the bad word list! You can't say that unless you name the people who espouse that view!" WhatamIdoing (talk) 00:54, 22 August 2023 (UTC)

I think I tried to explain that I agree that this is a point that a subjective claim including labels goes from "must be presented with attribution" to "can be stated as fact in wilivoice". I also agree that part if this transition is based on the relative portion of RSes that use the subjective claim in covering the topic. What I am arguing for is that we also need to have some type of time and independence-of-though aspects as well before making that transition. We want the subjective claim to be more enduring and not an overnight burst of usage, and we want to watch for cases of subjective claims that can arise due to broader events, such as near elections, or in the past decade+, this culture war we are in. Eg we should wait until we have academic sources looking back at these periods and reuse these claims themselves. In this way, the approach is very similar to scientific theory...the theory of gravity is still a theory, but in absence of any proofs against it or alternate theories in the centuries since proposed, it is scientifically accepted as fact. Or as a more recent example, we can't state factually the origin of COVID but it would be expected in years to come that the prevailing theory of zoological origins will be considered as fact...until then we still are clear its a leading theory but not the factual origin. Masem (t) 13:18, 22 August 2023 (UTC)

@Masem, I've personally given you multiple peer-reviewed academic journals that were based on 12 to 15 years of Breitbart's existence. They call it a far-right website. I'm asking you (personally): Are you going to keep complaining about Breitbart getting labeled as far-right, or will I never hear that particular complaint from you again or see you give that as an example of a website that we should be reticent to label that way? Because if you say that we need to have "academic sources" looking at the website over "some type of time" that is "enduring", and I've given you exactly the sources that you asked for, but you keep whingeing about it being unfair to follow the sources that you identified as the ideal ones, then the labels that are going to be on my mind are labels like hypocrite and two-faced. To be absolutely clear: This is one of those points at which you basically get to decide what I think about you. I really do hope you'll be choosing the option that makes me think you are a thoughtful, consistent, trustworthy member of our community. WhatamIdoing (talk) 15:51, 22 August 2023 (UTC)

Was that really you that wrote this? :-) North8000 (talk) 17:47, 22 August 2023 (UTC)

I expect the above is hard to hear. If you're not in a place to deal with it today, it's okay with me if you set this aside for another week. One of the problems with online communication is that you can't figure out whether the other person's already having a bad day, and thus time unpleasant feedback for more suitable moments. I am in no hurry to get a response. Anyway, the response that will matter most to me is the one that happens in the coming months and years, not any comment posted on this page. WhatamIdoing (talk) 15:54, 22 August 2023 (UTC)

Yes, there are those academic sources to support that we can call Breitbert as far right as fact. And to wit, these seem to be non passionate pieces so they are trying to look at Breitbart from a disinterested angle, so that type of sourcing fits the bill. But thats also what we should be looking for in consider any subjective claim, both the time and the thought independence. Instead more often than not we are using mainstream media at the current time of events to assert a subjec/I've claim as a label. We need to more like this Breitbart example. Masem (t) 16:00, 22 August 2023 (UTC)

With the caveat/context that it refers to where they sit in the spectrum of medias, I think that 90% + of the definitions of "far right" would place them in that category. Which means that 90% + of objective people would call them "far right" in this context. IMO this puts it in the zone of fact. And also informative, albiet brief....e.g. for the lead of an article. North8000 (talk) 17:54, 22 August 2023 (UTC)

Related to that last point, whenever we are reasonable assured we can state something as fact, like Breitbart being far right, we should always try to provide context for that as close as possible to the statement. In the body, I would think that we can write something to the effect of "Breitbart is a far right publication because it supports..." followed by sentences providing examples. In a lede this would be more difficult but one can still likely include a sentence to summarize that body. What comes off as a problem me often that the subjective claim-as-fact is just thrown in there in eikivoice without context, which is not helpful to explain to the ready why it is treated that way. Masem (t) 18:02, 22 August 2023 (UTC)

It's incredibly strange to watch people argue that the literal "platform of the alt-right" as run by Steve Bannon and declared to be such in his own words [10][11][12] is somehow not a far-right website. It's almost impossible to believe that such arguments are offered in good faith. 76.142.90.140 (talk) 20:22, 22 August 2023 (UTC)

We typically do not rely on dependent sources to self-describe themselves for fact, even if in this case it aligns with how they are called by others. Masem (t) 22:03, 22 August 2023 (UTC)

There's so much to say here that I could write a book. I've been keeping my participation here down to 1 comment per day, and I'll stick to that. Here's my takeaways from the way the conversation has evolved

Nobody here should care if Breitbart is far right or not - if you care, it's evidence that your POV is influencing the discussion

I have never taken a position on whether or not Breitbart is "far right" or not. Neither has Masem, who's been on the receiving end of AGF violations here that would normally be met with a very stern warning if not for the unpopularity of his POV. In fact, this mysterious new IP editor and one other editor are the only two who have opined about whether or not Breitbart is actually "far right" or not. I haven't opined about it because I don't care, as its immaterial to the topic at hand. You could substitute Breitbart for any outlet, even one I've never read or heard of, and my commentary would be unchanged.

A more relevant question is whether or not ANY statement following the format "person/news outlet X is 'far Y'" should be best understood as a statement of empirical fact, or an expression of opinion, regardless of who says it or how many times it's been said. Breitbart is one random example among many.

This is why the thread was created

The fact that some participants in this thread have entertained a debate about whether or not Breitbart is "actually far right" is troubling. It's an absurd conversation with no possibility of factual resolution without recourse to one's personal opinions of what constitutes "far". It's also a perfect illustration of why this thread was created. A concerning number of good-faith editors simply don't bring the the necessary tools to the table that would enable them to distinguish between facts and opinions in persuasive political writing, or in their own thinking. And they don't have the necessary tools to notice when their own POV is coming into play. That's not a comment on the intelligence of editors - it's a sad commentary on Western education systems, I suppose, and a good argument for Wikipedia putting more clear guidance into place - but most saliently, its a serious impediment to Wikipedia's reputation and flourishing. Reminding editors to "leave their POV at the door" accomplishes nothing when an editor sincerely believes that they don't have a POV, since their worldview consists of nothing but "the facts" and any suggestion to the contrary is "right-wing thinking".

This is a really big problem, and although there may not be a consensus to do anything about it yet, I hope more editors will start to notice the magnitude of the problem and reflect upon its potential long-term consequences. I believe that Wikipedia's treatment of modern politics is going to descend into complete absurdity in the coming years if we do not give editors some very clear guidance about what "facts" are, how they are different from "opinions", and how, specifically, to go about "leaving your POV at the door". Just saying "don't state opinions as facts, and leave your POV at the door" means nothing to somebody who thinks that all of their opinions are facts and that they therefore don't have a POV.

Towards a new model of attribution for articles about modern politics

@North8000 I've really appreciated your contributions here, which is why I have to be upfront and say that I'm concerned by the idea of reaching a point where "we are reasonably assured we can state something as fact". If something is not an inherently factual statement about politics when uttered by only one person, it does not ever, under any circumstances whatsoever, become "more factual" by virtue of being widely repeated. What it becomes is "more verifiable", and thus a relatively more notable opinion. Facts can be verifiable, and opinions can be verifiable. Verifiable facts should be given due weight in Wikivoice, verifiable opinions should be given due weight in the form of attributed statements. But opinions can't "become" facts by being repeated loudly enough and persistently enough by the mainstream press.

I know that's roughly how it works, conditionally, within the confines of the scientific method and scientific debate - the most-repeated scientific opinions tend to be supported by more empirical evidence, and therefore they become "conditional facts" - but I think that applying that logic to the domain of persuasive writings about modern politics is fraught with danger. In the same way that medical sourcing is held to a different standard than other Wikipedia article topics, it's possible that articles about modern politics need to have a separate standard for verifiability and attribution, too - because we simply cannot draw an analogy between scientific experts debating counter-hypotheses & different bodies of evidence in the annals of a scientific journal and political writers (who are not necessarily 'experts' in anything they write about) opining about politics in the popular press. The "consensus" that Breitbart is far right, for example, is reached by an entirely different process, with different standards of analysis and evidence, than the consensus amongst astrophysicists about the Big Bang is the best model for understanding the early universe. I'm comfortable stating the Big Bang is a "fact", with the understanding that this fact-status is conditional. I'm not comfortable stating widely-repeated opinions in the media as facts in the same way - I think we should avoid doing it.

Hopefully that makes sense. This is not an easy issue to parse. I'm glad there are at least a few of us who think it's worth parsing. I believe it continues to be a fruitful conversation and will be a seed that produces more nuanced, specific, and actionable conversations in the ensuing months and years. Philomathes2357 (talk) 23:14, 22 August 2023 (UTC)

It's an absurd conversation with no possibility of factual resolution without recourse to one's personal opinions of what constitutes "far". The constitution of "far" is verifiability. That's a fundamental tenet of Wikipedia. As Masem and North8000 have also admitted, but you do not seem to. It is not personal editor views but source views. it's possible that articles about modern politics need to have a separate standard for verifiability and attribution It would only ever be a higher standard that would still demand we adhere to the sources' opinions on many topics that you would consider subjective. It would not be a standard that holds that verifiability should be subject to different tests. I'm comfortable stating the Big Bang is a "fact", with the understanding that this fact-status is conditional. Everything on Wikipedia is conditional only on reliable sources. I believe it continues to be a fruitful conversation and will be a seed that produces more nuanced, specific, and actionable conversations Here's the thing - Wikipedia decides things based on consensus. This conversation does not create any precedent as no action was take and no consensus obtained. You are not a brand-new user, but you have several mistaken views on mainly WP:V not NPOV specifically. Andre🚐 23:52, 22 August 2023 (UTC)

I appreciate the spirit of the suggestion that Wikipedia should "not state opinions as facts". But I think this is easily misapplied advice, and ends up begging the question of what's an opinion and what's a fact. If you follow it to its most cautious conclusion, we might rewrite our articles in the following voice:

Water is widely considered by scientists to be an inorganic compound with the chemical formula H2O. According to most human experience using the five senses, it is a transparent, tasteless, odorless, and nearly colorless chemical substance, and it has been observed to be the main constituent of Earth's hydrosphere and the fluids of all known living organisms (in which it acts as a solvent). According to mainstream theories, it is vital for all known forms of life, despite not providing food energy, or organic micronutrients. Using the popular nomenclature in mainstream chemistry, dihydrogen monoxide, indicates that each of its molecules contains one oxygen and two hydrogen atoms, connected by covalent bonds.

... it just clutters the message. And while there may be new scientific observations that could come along and make our level of understanding obsolete, we should confidently what current scholarship says.

There will never be a set of Wikipedia policies that will eliminate the need for intelligent discussion. We'll always need editors who can provide a neutral reading of the sources, and engage in consensus building within the spirit of our policies. I can't say I agree with every detail of every article on Wikipedia, but I think we generally do a good job, and the policies are basically working. What we need are more good editors, and fewer POV pushers. Shooterwalker (talk) 00:24, 23 August 2023 (UTC)

Except that everything you are ascribing with attributes are objectively-proven scientific facts. That is purposely obfuscation that "Avoid stating facts as opinions." says not to do. A better example is what is currently said about the COVID origin, that "most scientists believe it is zoologic in origin", because we cannot objectively prove its origin.

Now, in flipping to when we are talking subjective claims which presently lack the weight and time of sources as in the case of Breitbart, there are ways to include attribution that does not flood the prose purposely like that as well. Once a claim has been attributed in a manner once in prose, then we don't have to keep writing to that attribution. Masem (t) 02:13, 23 August 2023 (UTC)

Philomathes, I think you might be dealing with an Etymological fallacy. You're talking about any type of "person/news outlet X is 'far Y'", but Far-right politics is not fully defined as "being really far to the right of center". "This person posts far-right political views" is not the same thing as "This person posts very right-wing political views". (Compare Chronic fatigue syndrome, which is not just "being tired all the time", even if they are tired all the time).

Far-right politics does fall on the one end of the left–right political spectrum, and that's what it's named for, but that's not a complete definition. Far-right politics is typically authoritarian and based on fear that my group is losing out in some way.

Consider these three "right wing" personas:

• Wow, my grandfather is the most conservative person I know. Not only does he object to higher taxes or more government services, he even wears the same style of pajamas that his father wore a hundred years ago! → Conservatism, even taken to an extreme, is not far-right.
• He's very right-wing. He absolutely insists that people and businesses get to do whatever they want, with no government regulation at all. He says pollution is best solved by expensive lawsuits for damages instead of regulations to prevent it, and that if you're the victim of a crime, you should either get revenge yourself or forget about it, instead of expecting a government to provide a police force to help you. → Libertarianism, even when taken to an extreme, is not far-right.
• My neighbor is far-right. He believes in the Great Replacement conspiracy theory, and he once told me he wished the president was strong enough to prevent any more immigrants and make all vaccines optional, and he says white men are somewhat discriminated against in every area of life, but most of the time he'd rather talk about football. → The actual far right is far right.

All of them are on the political right, and the first two hold their right-wing views to an extreme amount, but only the last one is far-right. WhatamIdoing (talk) 19:29, 23 August 2023 (UTC)

Above, you are basically outlining why there is no single, objective measure of "far right", which is why when present in RS it should be taken as a subjective claim. Which means to us, that we have to look at the number and proportion of sources and the endurance of time to determine how to speak to that in WP (that in some cases like Breitbart above we can treat it as fact).

This gets back to the main point of this dialog, in that we absolutely need to know what are subjective claims made in RS in contrast with objective ones, and if we are going to include a subjective claim that is DUE for inclusion, the decision speaking to that in a factual Wikivoice comes to the proportion and enduring use of the claim over time. So it should be fair game that if an editor sees a subjective claim being spoken in Wikivoice, to question that on the talk page, where then there should be a review of sources to show (as was done here for Breitbart) that we are confident that the claim has wide and enduring use in sources and without any other non fringe counterviews (such as the case if the the Covid origin theory). This type of questioning attitude should not be dismissed by other editors unless the matter was previously discussed in this manner.

To that the only change I think NPOV needs to that is something in the sense of NOPOV, saying that subjective claims with a significant proportion of use in sources and endurance in time and into independent works (academic journals) should be treated as facts without attribution. Right now we have nothing like this so editors often just make claims as facts in Wikivoice just based on a handful of sources, but the bar should be higher so we are right most cautiously and neutrally. Masem (t) 20:14, 23 August 2023 (UTC)

WAID just outlined some examples that are illustrative (though she didn't cite her sources but I'm sure she could find some) of how right-wing politics (normally, conservatism, as we think about it in the USA) differ from far-right wing politics, which is almost a euphemism for proto-fascism. These descriptions are not subjective. Any given label is relative, and may be capricious, or arbitrary, if controversial. But there is nothing subjective about the fact that far-right politics has a range of definitions which are relative. Relative means there may not be an exact way to measure these things. But that in no way indicates that they differ from person to person (subjectivity). You are NOT entitled to a different definition of "what you consider to be" far-right politics as pertaining to definitions or labeling. Any given label or definition comes from the sources and we should only seek to understand and accurately summarize how the sources define, and label, such things. They are terms of art no different from dosing antibiotics in medicine. Different doctors might dose them slightly differently. But nobody says "clarithomycin isn't a macrolide antibiotic to me, that's more subjective." Far-right politics and far right politicians are a category, and everyone is in it, so long as a number of reliable sources (for a sustained, sufficient amount of coverage to meet our policy and the consensus of editors) consider that to be the case. Now, in a borderline case, where you think maybe the coverage isn't sufficient or sustained enough, you can make that argument. But instead some people are making the argument that nothing' can be far anything. That's a ridiculous argument. Andre🚐 20:26, 23 August 2023 (UTC)

And I am not saying at all that we can't call anything or anyone far right. Just that, as you have written, "far right" is still a relative term with grey lines to delineate it from simply being a conservative "right", so the term us subjective, period. I am not saying that that immediately demends attribution, but that our mindset in writing that way is to assume attribution is needed unless it is shown that the proportion and endurance of sources are there to support us saying it in wikivoice. Right now. I think editors are too fast in jumping to treating such claims in wikivoice without attribution (vilification is of human nature) and we need editors to have more restraint, but we should not be stopping such when sources are beyond clear that we can state it as fact. Masem (t) 22:25, 23 August 2023 (UTC)

I don't think that we should treat relative and subjective as being interchangeable. WhatamIdoing (talk) 06:42, 24 August 2023 (UTC)

Something that is relative means that how one person views it will differ from how another person views it. That's fundamentally what subjective is. I'd certainly agree that if I drew Venn diagrams of what is commonly treated as far right in reliable sources, there would be a large overlap (agreement) but there still would be segments without any overlap, and that's the subjectivity that comes into play. That's typical of most labels. Masem (t) 12:35, 24 August 2023 (UTC)

Something that is relative means that it is more or less than some other thing, as in:

• Your ice cream cone is bigger than mine.
• There are more bot edits being made today than there were 20 years ago.
• This city is more accepting of LGBT people than the rest of the state.

Subjective means that different people have (legitimately) different views, as in:

• Wim Hof says a long, cold swim improves his health, but it's dangerous for babies.
• She loves hot, sunny days, but they make me feel sick.
• The student thinks Brexit was bad because it restricts his opportunities for overseas education, and his mother thinks Brexit was good because she wanted him to live at home while attending university.

WhatamIdoing (talk) 01:10, 25 August 2023 (UTC)

But we are still not talking about an objective measure (like sizes of ice cream cones or number of bot edits which anyone can count and measure and get the same answer) but some subjective aspect of where one's views sit on a political spectrum. And since you can't quantify that (such as the inability to quantify what is "more accepting"), "relative" is still subjective in this context. Masem (t) 01:15, 25 August 2023 (UTC)

It is possible for something to be both relative and subjective, but the words are not interchangeable.

The relative statement above that "This city is more accepting of LGBT people than the rest of the state" is also neither an easily quantifiable nor a universal statement. It's relative because it's making a comparison. It could equally say "Relative to the rest of the state, this city is more accepting of LGBT people".

The relative part is the part that compares the city to the rest of the state. The subjective part is that different people could have different ideas about what it means to be accepting. WhatamIdoing (talk) 01:36, 25 August 2023 (UTC)

This may seem ethereal/gnomish, but it often comes down to analyizng words and realizing that they are usually ambiguous (= have multiple meanings). It's only in the rare cases where every meaning is defined and the answer to that specific statement is widely accepted amongst knowlegable people that you can have an absolute fact.WP:Accuracy But quite commonly we get pretty close on all of those things and that is generally accepted as fact. The farther away you get from all of that (including choices of definitions of terms used) the more you get into spin/opinions and even further than that is misleading / deception. I think that at least 95% of the tussles over "fact" in Wikipedia are NOT about where some information is unknown, they are about which words are used to describe, characterize or summarize KNOWN information. To risk being overly simplistic, if the definitions of words used are 90% the same as those of the listeners, and 90% of knowlegable objective people agree that the statement is correct, that clearly passes as fact in normal writing as in an enclyclopedia article. North8000 (talk) 13:58, 23 August 2023 (UTC)

For example, on close analysis, the statement "George Washington was one of the founders of the USA" is technically ambiguous because there are many definitions of "founder". But at least 90% of definitions of "founder" would include him. So this ambiguous statement passes as fact. North8000 (talk) 14:10, 23 August 2023 (UTC)

For better or for worse, Wikipedia is based on a standard of intersubjectivity. We are an aggregate of what reliable sources say about a subject. The best we can do is decide which sources are unreliable, and which sources have a reputation for accuracy (or at least fact-checking and corrections). Shooterwalker (talk) 14:05, 25 August 2023 (UTC)

Indeed, I agree with North and Shooterwalker. Ambiguity is a part of all language and all life. 80% should be the bar. If 80% of the most prominent and reliable material agrees about something, that is de facto factual. Andre🚐 17:48, 25 August 2023 (UTC)

Then that absolutely needs to be demonstrated on relevant talk pages with a source survey. Even if that means surveying 100s or 1000s of articles to verify the numbers. I can tell you most of where we write subjective language as facts in Wikivoice, this has not been done at all.

Instead, as I've agreed with in the case of Breitbart above, it is far better to show that a good proportion of sources support that (here, even something like 25% of all reliable sources covering a topic would be fair), that there is no non-fringe counter view also reported via reliable sources, and the view is supported through enduring coverage far beyond the point of origin and ideally into academic sources. That still means a source survey needs to be done, so that's on the onus of those that want to write in factual terms. Masem (t) 19:24, 25 August 2023 (UTC)

I don't think I'd set a particular number, although I agree that 80% is better than 99.9999%.

Masem, I don't think there is any point in documenting a thorough source survey if nobody is credibly challenging the accuracy of the statements. If someone sees something in an article, and says "here's a non-fringe source that has a different POV" or "yeah, but I can only find that one source", then the situation is different, but for the most part, I think we're getting things about right. WhatamIdoing (talk) 04:00, 26 August 2023 (UTC)

And to clarify, I'm also not in favor of enshrining any number, but I think as a rule of thumb, even 2/3 of sources might be enough if the remaining 1/3 was sufficiently fringe-y. Though I'd be hard-pressed to substantiate a specific case of that so please don't call me on that. Like other policies on Wikipedia, there isn't a specific quorum or a mathematical threshold, because those are prone to gaming and obscure the point of things like a rough consensus mechanism or a system of evaluation. Though of course, a source survey is often a good idea, and I believe it generally has been done - for example, to label Marjorie-Taylor Greene a conspiracist, there were extensive sources provided. Since, for example, everyone agrees that Breitbart does have sufficient sources, if there's an example where we think that something is, contra-NPOV or contra-RS/OR, being branded unfairly, but where the sources are a bit more split or there's a non-fringey minority that is closer to 56-43 territory and somehow is losing out to the tyranny of the majority? 'Cause I don't think that exists - if anything, on Wikipedia, I think there's a bit more frequency of a situation where the vocal minority of 20% opinion holders have litigated their POV into a lot of articles at a bit heigher weight than it might truly be merited. Again, I don't have a good example in mind of that, but it's just as likely as the opposite (ie, the claim that some articles are unfairly POV by excluding or labelling minority fringe views that happen to align with groups that have a persecution complex onwiki and allege that wikipedia is biased against their group) Andre🚐 23:12, 26 August 2023 (UTC)

One of the problems we encounter in this is that editors who don't want a particular POV to be mentioned want a survey of "all" the sources instead of the "relevant" sources. To give a non-inflammatory example, sources consistently identify Dodi Fayed in one of three ways:

• a film producer,
• the boyfriend of Diana when they died, and/or
• the son of a billionaire.

Probably 80% of sources identify him as being one (or more) of these three things, but I don't think that you could find any given item from the list in even 50% of sources. That doesn't mean that the sources disagree about, e.g., who his father was; it just means that a source writing about (e.g.) a film he produced is more likely to care about the fact that he was a film producer instead of caring about who his father was. There is a tendency to argue that any source not mentioning a given fact is silently disagreeing with it or indicating that it is unimportant. We should not accept such arguments. WhatamIdoing (talk) 20:27, 27 August 2023 (UTC)