Wikipedia talk:Reliable sources

Reliable for opinions

This page's overview opens with:

Articles should be based on reliable, independent, published sources with a reputation for fact-checking and accuracy. This means that we publish only the analysis, views, and opinions of reliable authors, and not those of Wikipedians, who have read and interpreted primary source material for themselves.

I think there's some issues with "analysis, views, and opinions", and with the overview's focus:

• "Paris is the capital of France" is not really analysis, views, or opinions, yet facts such as these are the bulk of what the reliable sources' guideline pertains to. I understand that this asserting a flipside of the WP:OR policy, but it's a pretty niche side of the guideline.
• Views and opinions are not meaningfully distinct
• It's hard to understand what it means for an author to be a "reliable" for their opinion. They may be noted for it, e.g. film critics or food critics. But reliable? In at least one sense (WP:ABOUTSELF) every author is a "reliable author" for their opinions.
• Following the previous bulletpoint, I would venture that most explicit opinions/views in Wikipedia are contained within #Reception sections of media articles. As a source's reliability is a factor in determining if material is DUE (e.g. WP:BESTSOURCES), we see for instance the Daily Mail not being used as a source for film reviews, even though we include Rotten Tomatoes scores which use the Daily Mail in their calculations. Some Daily Mail film reviewers are notable. I understand that The Daily Mail may be used in rare cases in an about-self fashion, but editors seem to view it as banned for this purpose.

Would love to hear some thoughts. Rollinginhisgrave (talk | contributions) 03:08, 5 April 2025 (UTC)

There's a lot to talk about here. I'm going to pick out one from the middle:

Do you understand the difference between viewpoints, such as:

• Republicrats say the country needs more ___, and Demicans disagree.
• Consequentialists say ___ about whether it's okay to steal from a thief, and deontologists disagree.
• Russia says their military action in Ukraine was morally justified, and Ukraine says Russia is waging an unjustifiable war of aggression.

and opinions, such as:

• Coffee tastes good.
• Chocolate tastes better.
• That was a good movie.

WhatamIdoing (talk) 03:27, 5 April 2025 (UTC)

I don't think so. If I had said "In the opinion of Republicrats, the country needs more ____, while Demicans disagree" would you have corrected me, telling me these are viewpoints rather than opinions? Conversely, if I said "from my point of view, chocolates tastes better than coffee", would that be corrected? It seems like the distinction you're drawing is between opinions on quality, vs opinions on other things, which can't be summed up as viewpoints vs opinions. Rollinginhisgrave (talk | contributions) 03:40, 5 April 2025 (UTC)

I think maybe a distinction can be drawn as opinions being discrete, and viewpoints being patterns of opinions (attitudes), although I don't think your examples exemplify that. Rollinginhisgrave (talk | contributions) 03:42, 5 April 2025 (UTC)

The distinction we normally use is that an opinion is a personal (perhaps "individual") thing, and a viewpoint is something that depends on "the point from which the viewer is viewing". Millions of people think coffee tastes good (or bad), but that's still an opinion. Whether coffee tastes good cannot be proven true in general; it can only be proven that some people like it (and I don't).

On the viewpoint side, you could say, about (e.g.) a war that it's logical for (e.g.,) opposing belligerents to hold differing views, and that you could even predict some part of the views: The instigators will declare their actions justified, the invaded will deem their territory violated, the taxpayers will see their taxes go up, the sellers of war goods will see their sales increasing, the neighboring territories will see refugees streaming in, the soldiers will have a different view of the war than the senior officers, the victor will have a different idea of which decisions and actions were most important than the vanquished. In contrast to an opinion, you can prove that people looking at the war from "this point" will usually "view" it differently than people looking at it from "that point".

I agree that there is some overlap. One can have an individual opinion based on a viewpoint. Perhaps "From the viewpoint of technical photography, the lighting in that film was bad" would be an example of that. Perhaps even "From the viewpoint of fiscal conservatism, this candidate is better than that one".

Consider another pair of examples:

• It is good for Wikipedia to have a lot of information. (opinion; cannot be proven right or wrong)
• From the viewpoint of readers, it's great to have lots of new articles to read, but from the viewpoint of the NPPers and AFC folks, a sudden influx of articles is a lot of unexpected work. (viewpoint; whether it's good or bad from that viewpoint is more or less objective, but there is no universal answer that applies to everyone)

It is my opinion that the most practical way to address this problem is not by editing guidelines, but to improve the Wikipedia articles on Opinion and Point of view (philosophy). WhatamIdoing (talk) 04:25, 5 April 2025 (UTC)

I generally like these distinctions you've drawn. I think a link to point of view (philosophy) or changing "viewpoint" to attitude would give more clarity. Rollinginhisgrave (talk | contributions) 04:39, 5 April 2025 (UTC)

I've no objection to linking, but I don't think that "attitude" is the right idea. Teenagers have "attitudes". Different countries have different "viewpoints" about various situations. "Climate change is a big deal because it will swamp most of my island home and destroy all the drinking water" is not "an attitude". It's a viewpoint. Specifically, it's a viewpoint, because someone who is viewing it from the "point" of living on that soon-to-be-submerged island will have a different view of climate change than someone who does not live there. WhatamIdoing (talk) 21:15, 5 April 2025 (UTC)

Okay. Do you see it as at all problematic that all your examples of views are referring to those held by groups, while the overview is referring to individuals? (Publishing the views of reliable authors; can climate-change affected countries/Wikipedia readers/NPPers etc as a class be an author?) Rollinginhisgrave (talk | contributions) 21:30, 5 April 2025 (UTC)

Viewpoints from groups (including unofficial groups of authors, e.g., "all the moral philosophers" or "most of the biologists researching COVID-19 origins") are more likely to be WP:DUE than viewpoints from individuals.

You can have viewpoints from individuals, but they tend to be like "It hurt me when you stepped on my toe" and "It didn't hurt me when I stepped on your toe (but I recognize that if our positions had been reversed, I'd have been the one feeling the pain instead, and I'm sorry because I didn't mean to hurt you)". That tends not to be relevant for an article.

Often, especially for current events, we see one author being chosen to represent a group, e.g., "Chris Columnist wrote something scathing about the politician", but it's not usually important that Chris wrote it; it's usually DUE because a group (e.g., people who are affected by a policy, people holding a given set of values, people who subscribe to a particular political stance) holds that view, and Chris is just a handy example of that viewpoint. WhatamIdoing (talk) 01:28, 7 April 2025 (UTC)

Right. I'm thinking more for This means that we publish only the analysis, views, and opinions of reliable authors: "moral philosophers as a class" cannot be an author, a secondary source has to write what the view of moral philosophers is. In such a case, we are not only publishing the views of reliable authors, because the author writing about the view of moral philosophers may not share such a view.

Viewing the moral philosophers as the "reliable authors" is also problematic. For many political articles, the most relevant views we write about may be held by a populace, and determined by opinion polling. They are neither reliable (experts on a topic) nor published. They are reliable for their own opinion, as is (essentially) everyone. In which case "reliable author" is redundant. What we actually generally care about is the views of stakeholders, skewed towards experts.

I think this has come from the guideline's author wanting to emphasise "rely on reliable sources" as the opposite of "do original research". They contrast the above green text with and not those of Wikipedians, who have read and interpreted primary source material for themselves. This does not give an overview of what reliable sources are, it gives an overview of verifiability. Rollinginhisgrave (talk | contributions) 08:52, 7 April 2025 (UTC)

I don't think that the goal is "the views of stakeholders". Consider: Lake Woebegon and Smallville have a dispute. The neighboring towns (and their residents) are the stakeholders: either the border between the two is here or there. Are the views of the stakeholders the only ones that matter? Maybe the view of the non-stakeholding high court is actually the one that matters the most. WhatamIdoing (talk) 21:37, 7 April 2025 (UTC)

Perhaps that should say "views, etc., of the reliable sources" instead of "authors". WhatamIdoing (talk) 21:37, 7 April 2025 (UTC)

You're right. Stakeholders and authorities? I do really think substituting source for author doesn't solve the issue: we are looking for relevant views, skewed towards representing those of authorities. These can be those held by sources, or written about in sources. An issue here is that "reliable" in the paragraph refers to two things: 1) we want to skew coverage towards experts and 2) we want the views of all parties to be accurately rendered. Rollinginhisgrave (talk | contributions) 01:26, 8 April 2025 (UTC)

The view in any high-quality reliable source is relevant. The view of a "stakeholder" isn't necessarily relevant. Believers of conspiracy theories are stakeholders in those false ideas. Owners of companies who want to spam Wikipedia for their own benefit are stakeholders. We don't care about them. We do care about what an impartial, independent, non-stakeholder reliable source says. WhatamIdoing (talk) 19:28, 9 April 2025 (UTC)

I still find "the view in" confusingly ambiguous: The view of the author of the HQRS? The views described in HQRS? I might get clarification on this before I make further comments. It also might be helpful if I give some thought on how this would look and make an explicit diff proposal/bold edit. Rollinginhisgrave (talk | contributions) 02:20, 10 April 2025 (UTC)

We are looking for the views described in reliable sources. It does not matter if the author of the source personally holds the described view(s), though in some subject areas, it is more common to cite the directly expressed view. For example, one usually – but not through any policy, guideline, or other rule – cites the critic of an opera directly ("Miss Manners thought the climactic aria was sublime^{[Miss Manner's own review]} and Evelyn Waugh's grandson thought it was grand^{[Waugh's own review]}") instead of saying the same thing but citing it indirectly (e.g., citing The Big Book of Opera Reviews for both). Both are 'legal' but the former is more common. WhatamIdoing (talk) 04:50, 12 April 2025 (UTC)

Thanks. This is how I understood it. I have changed the guideline to reflect this. Rollinginhisgrave (talk | contributions) 05:08, 12 April 2025 (UTC)

Thanks Mathglot for reverting. I was considering self-reverting for a separate reason but you beat me to it. I don't think the wording necessarily implies that we can only publish views/opinions if they are described by secondary sources but I can see how it can be read that way.

If it wasn't clear the main issue I was trying to address is that we include views/opinions/analysis that are only described/expressed by the person who has the view/opinion/analysis (e.g. a movie review), and we also include views/opinions/analysis of a person/group when a secondary source describes them (e.g. a source saying this king was viewed by the people as wise). In what way does "we only publish the analysis, views, and opinions of reliable authors" account for the second? Rollinginhisgrave (talk | contributions) 09:17, 12 April 2025 (UTC)

How's this for a set of examples?

1. Paris is the capital of France is a fact, able to be demonstrated as true based on its formulation only from comparatively well-defined and well-accepted terms. It is also clear how changing each of the words would result in a false proposition.
2. France was the greatest military power in Europe during the reign of Louis XIV is a viewpoint. Most people would understand this as also being either true or false, but there is significant complexity in how one defines terms and the general parameters of the discussion, such that it simply requires expertise to discern one way or another. Most statements of value on Wikipedia live here?
3. Paris is (considered) a beautiful city – this is an opinion. I think most editors would make a serious editorial distinction depending on whether considered is part of it (i.e. whether it's an attributed opinion) but I think that matters less than some think—whether we ask for one's opinion relies either on how reliable a source is with facts and viewpoints much of the time (an expert opinion), or how prominently other RSes care to reproduce their opinion.

Remsense ‥ 论 21:47, 5 April 2025 (UTC)

I don't love this distinction, and I don't think it is a good thing that several editors are each coming up with their own ways of determining what opinion means vs view. I don't think it's worth a protracted discussion, I was just trying to see whether people thought much is added by writing "analysis, views and opinions" vs "analysis and opinions". Rollinginhisgrave (talk | contributions) 02:45, 6 April 2025 (UTC)

Sorry I wasn't helpful. Remsense ‥ 论 02:51, 6 April 2025 (UTC)

Remsense Very sorry, I don't intend this to be mean, I'm glad you commented. I said "it's not a good thing that several editors are each coming up with their own ways of determining" as I think it comment on the underlying ambiguity, not on the value of your input. Rollinginhisgrave (talk | contributions) 02:58, 6 April 2025 (UTC)

I'm not sure that "the greatest military power" is a viewpoint. It's not demonstrably true or false (in most cases). "France had the most soldiers" would be a fact, but "France was the greatest because that's determined by having the most soldiers, and France had the most" might be an opinion.

Perhaps "The Treaty of Versailles required the Germans to pay the French a reasonably fair amount of money in reparations after WWI" is a viewpoint. The French people seemed to think the payments were fair recompense for the destruction the Germans caused, even as they understood that it was difficult for Germany to make the payments. The German people seemed to think it was a economy-crushing level of debt, especially after the Great Depression hit, even though they understood that paying for damage caused is fair. Both viewpoints were correct (and we ended up with Hitler and WWII). WhatamIdoing (talk) 01:38, 7 April 2025 (UTC)

Nitpick… “The Capitol of France is Paris” is an accurate statement of fact… however “Paris is the Capitol of France” might not be (because there is more than one “Paris”). Blueboar (talk) 20:52, 9 April 2025 (UTC)

Re the OP's comment "I understand that The Daily Mail may be used in rare cases in an about-self fashion, but editors seem to view it as banned for this purpose." That's quoting from the essay-class WP:RSP page which is not a result of the WP:DAILYMAIL1 RfC and even if it had been it would violate WP:NOTCENSORED which unlike RS RfCs is a policy. On prior occasions where this claim was made I've asked the closers of the RfC, e.g. here, and the response has always been: opinions are allowed period, not just aboutself opinions. One can claim undueness which usually works but is not an RS consideration and is itself just editor opinion. So WP:RSP is misleading. However, that some editors seem to view it as banned may be true, I won't dispute that. Peter Gulutzan (talk) 19:24, 14 April 2025 (UTC)

Peter Gulutzan I think we're in agreement. I'll give an example of an AfD I'm looking at to see if I've got this right: An editor said a restaurant review in The Herald Sun shouldn't count towards establishing notability as "The Herald Sun is dubious as a reliable source". What do you make of this? Rollinginhisgrave (talk | contributions) 09:32, 16 April 2025 (UTC)

My take is that this review is reliable but notability is a higher bar. Peter Gulutzan (talk) 15:12, 21 April 2025 (UTC)

Discogs - It IS a Reliable Source

Where is all this new fuss about Discogs coming from?

When researching and writing about music, Discogs is absolutely the most reliable readily-available source we have, particularly for old music. Unlike other user-generated sites, Discogs has an onerous contribution process--people can't just type stuff into it. Users are also required to upload photos and/or scans of all records/albums, album covers, liners etc. So if there's any dispute about the facts, there's a visual record of everything. Not only is it utterly reliable, it is much more closely monitored than any other comparable site.

People seem to be confusing it with AllMusic, which used to be deprecated because it really is unreliable. People can enter anything they like--it's as easy as RateMyMusic, which I believe is still deprecated. Up until recently, editors were not even allowed to mention AllMusic; all of a sudden, it's being used all over Wikipedia as a reference. Its only real value lies in professional music reviews; jazz historians and the like post reviews on it. (The same goes for IMDB, which also used to be deprecated and apparently no longer is.)

Wikipedia editors, myself included, have been using Discogs for years. It is a written and visual encyclopedia of music and there's no logical basis for objection to its use.23:41, 9 April 2025 (UTC) LJA123 (talk) 23:41, 9 April 2025 (UTC)

This is probably the wrong page for this. However, if you look at WP:DISCOGS, there's a link to a big discussion at Wikipedia:Reliable sources/Noticeboard/Archive 456#RfC: using photos of record labels from Discogs? from five or six months ago. Might that be relevant? WhatamIdoing (talk) 02:17, 10 April 2025 (UTC)

AGEMATTERS

The section currently says use recent sources because new information has been brought to light, new theories proposed, or vocabulary changed, which seems mostly suited to hard sciences. Could something be added that warns against using old sources for topics related to colonialism? Or stresses the importance of using recent sources for those contexts, or any contexts where older sources are likely to have racist biases? Too often I see sources from the late 19th or early 20th century (literally from the time of the Scramble for Africa) used in articles on African history. At Talk:Copts an editor is trying to use sources from 1894 etc. and there isn't really any guideline that addresses this context. Relatedly, most of our articles on African colonies (mainly the British ones) are written with sources from pre 1970!!! (East Africa Protectorate, Nyasaland, Colony of Natal and Orange River Colony 1911!, Anglo-Egyptian Sudan, Southern Rhodesia, Transvaal Colony etc.). Kowal2701 (talk) 20:51, 14 April 2025 (UTC)

I've seen editors suggest that sources for history should be no older than 50 years, or no older than the 1950s. "Vocabulary changed" certain seems likely. WhatamIdoing (talk) 05:25, 16 April 2025 (UTC)

How about adding outdated biases, or replacing vocabulary changed with that? "Vocabulary changed" seems too superficial and doesn’t really capture the main concern with these sources Kowal2701 (talk) 15:31, 16 April 2025 (UTC)

"Vocabulary changed" is a significant factor in medicine. One should not use outdated and offensive language such as habitual aborter or mentally retarded in a Wikipedia article, even though older sources do. WhatamIdoing (talk) 16:27, 16 April 2025 (UTC)

Good point. How about or have outdated language or biases? So it sort of keeps the rule of three Kowal2701 (talk) 16:55, 16 April 2025 (UTC)

Maybe? Or maybe what we need is a good explanatory essay. WhatamIdoing (talk) 17:34, 16 April 2025 (UTC)

Agreed, I've thought about it before but don't think I'm best placed to write on history as a whole. Do you know of anyone who might be able to write it? I'd put something at Wikipedia talk: WikiProject History but it's pretty inactive Kowal2701 (talk) 20:17, 16 April 2025 (UTC)

You could see if the authors of Wikipedia:Identifying reliable sources (history) (a failed proposal) are still active. WhatamIdoing (talk) 01:29, 17 April 2025 (UTC)

Proposal to tighten policy on dissertations

I was very surprised that the discussion on dissertations does not mention the quality of the university. I am not going to suggest a specific wording and will leave that open for suggestion.

In my view any thesis at a university that scores below 1,000 on the Times Higher Education World University Rankings would be very questionable. Scores above 500 are good and above 200 very good. QS World University Rankings may also be used, of course. The dissertation by Kenneth Arrow was reliable the day after it was published, of course, but 5 years later it was not even necessary to refer to it.

And the fluidity of the field matters. From my old school days many years ago, I remember another student who did his PhD on an emerging topic, passed and got a university job and then tenure all in that field. But ten years later his thesis was totally outdated because the field was changing so fast. In emerging fields I would suggest a 5-10 year limit.

Please make suggestions. Thanks. Yesterday, all my dreams... (talk) 11:25, 16 April 2025 (UTC)

I think it's more complicated than that, and that editors should use the best sources they can, as much as they can. WhatamIdoing (talk) 16:28, 16 April 2025 (UTC)

I am generally quite suspicious about third-party indices from non-academic sources such as newspapers. As such I would not want Wikipedia to be endorsing this specific index as being a definitive statement regarding university quality. I would prefer simply to require dissertations be WP:USEDBYOTHERS before assuming reliability. Simonm223 (talk) 16:30, 16 April 2025 (UTC)

I agree. Also, a university's standing may change over time.

And importantly, the key question isn't "Is it reliable?" but "Is it reliable for this claim?" The same dissertation might be a reliable primary source for "Publications have used this terminology since at least [date of dissertation]" or "This has been discussed in sources as varied as PhD dissertations^{[dissertation]} and science comics^[xkcd.com]", while not being reliable at all for "This is absolutely the final truth forever". WhatamIdoing (talk) 16:34, 16 April 2025 (UTC)

Well, we must have studied logic at different universities. My point was not that the best universities can be blanket trusted but that the lowest scoring ones can be trusted less. Please consider modal logic and fuzzy logic. And if all else fails, fuzzy modal logic [1]. I will leave it there and move on. Yesterday, all my dreams... (talk) 22:15, 16 April 2025 (UTC)

I think that what's more important is that a university can be middling overall, and yet have a really important program in one specific area. WhatamIdoing (talk) 22:18, 16 April 2025 (UTC)

Agree with WhatamIdoing and Simonm223. For example, UC Irvine is generally seen as mediocre (90 on the THE rankings), but is ranked no. 2 in the nation in criminology (according to U.S. News and World Report). --Coolcaesar (talk) 18:58, 17 April 2025 (UTC)

I think the used by others needs to be substantial use and in a way that really validates it vs just references it. In my area a thesis commonly wouldn't contain much that wasn't already published in journal articles that could be cited instead. Absent some sort of clear evidence that a thesis/dissertation was getting quality references (more than just a statement that the work was done), I would suggest avoiding using them. Springee (talk) 04:04, 22 April 2025 (UTC)

Often dissertations will go into more detail and background than journals will allow, and are targeted to a broader audience. Because of this they can be useful for sourcing things that are omitted from the journal papers because they're considered obvious to specialists. I agree with you to an extent: if the claim to be sourced can be found both in a journal version and a dissertation version of some work, then the journal version is preferable. But I think that doctoral dissertations are generally reliable enough to use without special care. Master's theses need some care and bachelor's theses are usually better avoided. —David Eppstein (talk) 05:25, 22 April 2025 (UTC)

Several doctoral dissertations in philosophy have subsequently been packaged as books by those philosophers and are often quite significant books within their bibliographies at that. Look at Discourse, Figure as a perfect example. Doctoral dissertations *can* be quite useful. But they're also, pretty much definitionally, the first useful piece of work an academic is likely to produce. As such, yeah, not every dissertation is going to be the best source for whatever it is citing. But the main criterion should be the reception of the work, not the reputation of the school. Simonm223 (talk) 13:15, 22 April 2025 (UTC)

Reliable yet profringe (again)

About two years ago, the WP user Sovkhozniki made a failed attempt to modify the WP:SCHOLARSHIP guideline. The guideline normally states, Material such as an article, book, monograph, or research paper that has been vetted by the scholarly community is regarded as reliable, where the material has been published in reputable peer-reviewed sources or by well-regarded academic presses. Sovkhozniki added the qualifier, except in cases where it promotes fringe theories. The Wikipedia community rejected this change, and Sovkhozniki later was blocked as a parody account. This account was part of a trolling project which originated in 2018 at RationalWiki, which has tried to hoodwink other users into supporting violations of content policies, and then publicized the violations on social media and external sites.

Although the community rejected this proposal, in several topics recently there's been a widespread trend to classify such sources as non-RS, as Sovkhozniki had argued to do. This has happened most commonly on articles covered by the race and intelligence arbitration case, although it isn't confined to that topic. It is argued that the consistent consensus of editors has been that such sources are not RS. More recently, classifying such sources as non-RS was described as a longstanding, topic-wide practice.

Some of the sources removed, such as New York Times articles about recent human evolution, [2] [3] [4] [5] do not directly present any fringe ideas, and are classified as non-RS because of views expressed in unrelated publications from the publisher or author. This basis for classifying sources as non-RS also was explained with respect to the journal Personality and Individual Differences. Aside from journals and newspapers, this practice also applies to certain academic books (that comment is referring to three books from Cambridge University Press: [6] [7] [8]). If necessary I can provide further diffs of sources removed for these reasons, but for now it's best to stay focused on the overall principle instead of individual examples.

The argument being made is not only that these sources are fringe; it's that they fail the requirements defined by Wikipedia:RS, and thus can't be included in evaluations of what balance of views is required by NPOV policy. The assertion that such sources are inherently non-RS has been used to reject proposals to review the source literature in depth for such an evaluation. (When all sources taking a certain position are non-RS by definition, then the question of what's NPOV is answered a priori.)

There is a contradiction between the community's rejection of Sovkhozniki's change to this guideline, and the local consensus in some topics that when Wikipedia editors decide a particular view is fringe, reputably published sources supporting it become non-RS. In the near future, I plan to open a request for comment to resolve the contradiction. I've discussed this proposal with a member of Arbcom, and they suggested that before opening the RFC, I should initiate a discussion about how it should be formulated. I have two questions:

1. The assumption that such sources are innately non-RS is being applied across several dozen articles, so any such RFC would have to occur in a centralized place, not on an individual article's talk page. Would this talk page be an acceptable location, or is there a more appropriate venue?
2. Three options that I suggest should be voted on in the RFC are as follows: to change the guideline in the way Sovkhozniki proposed; to keep the guideline in its current form (and make it clear this community consensus supersedes any local consensus to use Sovkhozniki's version of it); or to modify the guideline to say that reputably published academic sources are RSes in most topics, but that some topic areas are exceptions. Are there any other choices that should be included?

Ferahgo the Assassin (talk) 22:42, 20 April 2025 (UTC)

1. Yes, if you want to make a WP:PROPOSAL to change this guideline, then the discussion(s) should happen on this page. Please read about the process for a formal proposal.
2. It sounds like you are trying to narrow the space in which editors can use their judgment. Is that what you really want to accomplish?

The situation you describe above doesn't surprise me. Personally, I usually see this in Wikipedia:Contentious topics. An author (or organization) will write something that an editor morally disagrees with (e.g., J. K. Rowling opposes something about trans rights, so she is anathema with one group of editors; the Freedom From Religion Foundation supports something about trans rights, so they are anathema with another group of editors), and now it is urgent to de-platform them completely, even for content that is unrelated to whatever the editor is disturbed by.

But:

• There's a long tradition in academia of people being excellent at one thing and bad at another. Nobel disease has a good list if you'd like to read some stories, or think about how widely scientists accepted and supported eugenics from the beginning of modern science until the horrors of the Holocaust became apparent. The world is not divisible into people who are always perfect and people who are always wrong. You need to use a holistic evaluation. You need to let humans be complicated.
• Similarly, even "the best" academic publishers and journals will publish bad content. Sometimes it seems to just be a mistake; sometimes it seems to be intentional, as part of a plan to provoke discussion in a field or to provide a balanced set of sources (e.g., one from a conservative, one from a progressive, one from a libertarian, etc.). But publishing a small percentage of bad things doesn't make everything they publish bad.

WhatamIdoing (talk) 00:38, 21 April 2025 (UTC)

Tagging @Generalrelative to this discussion, as their input seems relevant here given their previous involvement in some of the cited incidents. Harryhenry1 (talk) 01:10, 21 April 2025 (UTC)

I’m not sure I have anything to add that hasn't been stated before, except that I'm surprised to see Ferahgo reappear to flog this dead horse once again. The last time she weighed in on the race & intelligence topic area, an area from which she'd previously been t-banned, she immediately self-reverted, stating "it was a mistake for me to get involved in this issue again." I was impressed, at the time, by the personal growth that showed. Yes consensus can change, but this exact same issue has been relitigated again and again and again. At some point you're just wasting the community's time. Generalrelative (talk) 04:23, 21 April 2025 (UTC)

I have suggested before that this guideline would benefit from a definition of reliable source. My current working definition is that a reliable source is any source that experienced editors will accept as being sufficient to support a given piece of content.

It has not been popular (surely the definition needs to say something about independence and secondary sources and peer review?!), but it appears to be true (e.g., {{cite tweet}} is none of those things, and is used in 42K articles). And I wonder if it would solve this sort of problem, because instead of a tis/tisn't argument over whether a source is "really" fringe, we could focus on what really matters, which is consensus. Either we agree that it's acceptable for the given use, or we don't. We will hopefully have good reasons for our acceptance/non-acceptance, but in the end: it's reliable if we accept it, and unreliable if we reject it. WhatamIdoing (talk) 20:13, 21 April 2025 (UTC)

@WhatamIdoing: I think the amount of leeway editors currently have to declare sources non-RS, based only on the positions those sources take - even when the publisher is a clearly reputable one such as Cambridge University Press or The New York Times - is obstructing the ability to follow NPOV policy. WP:SOURCEGOODFAITH directly mentions this as a practice to be avoided: The danger here is in judging the reliability of sources by how well they support the desired viewpoint. I don't view the prospective RFC as trying to narrow the space in which editors can use their judgment, so much as trying to uphold both the letter and the spirit of the RS guideline.

I also think it's a problem for WP:SCHOLARSHIP to say one thing about whether reputably published academic sources are RS, while local consensus in certain topics says something else entirely. There ought to be consistency between what the guideline says and how articles are edited, even if it might potentially mean turning Sovkhozniki's change into an actual part of the guideline. (I hope that won't be the outcome, but it's still an option the community should vote on.)

The WP:PROPOSAL page that you linked to suggests workshopping a proposal before starting an actual RFC, so here's an idea about the options to be voted on.

• Option 1: Modify WP:SCHOLARSHIP the way Sovkhozniki proposed, to say that books and papers "published in reputable peer-reviewed sources or by well-regarded academic presses" are not reliable sources if they present fringe views.
• Option 2: Modify WP:SCHOLARSHIP to say this part of the guideline has exceptions for certain topic areas. One exception would be articles covered by the race and intelligence arbitration case.
• Option 3: Don't modify the guideline, and decide that its current form is supported by community consensus, which supersedes any local consensus to use Sovkhozniki's version of it.

Lest I be misunderstood: if the community decides to uphold what WP:SCHOLARSHIP currently says about the reliability of reputably published academic sources, it would not mean that any given such source must necessarily be cited. All of the standard sourcing restrictions such as WP:PRIMARY would still apply. Suggestions about wording are welcome. Ferahgo the Assassin (talk) 20:44, 21 April 2025 (UTC)

I am not convinced that the problem you're seeking a fix for actually exists. Fringe views (which in this case would seem to be racist pseudoscience) are omitted for a number of reasons, and sometimes because the sources are unreliable. And that isn't purely down to the publisher, we're supposed to consider the author as well, among other factors. One of these 'clearly reputable' Cambridge University Press books you mention was written by a person who is also notable for attendance at white supremacist conferences and writes for The Unz Review. Any change to the sourcing policy that would require editors to ignore factors like that would not be for the good of the encyclopedia. MrOllie (talk) 20:55, 21 April 2025 (UTC)

If you think I'm seeking to change the guideline, you've misunderstood my argument. Currently, the part of the guideline we're discussing states: Material such as an article, book, monograph, or research paper that has been vetted by the scholarly community is regarded as reliable, where the material has been published in reputable peer-reviewed sources or by well-regarded academic presses. My preferred outcome of the RFC would be for the community to support this guideline in its current form. You seem to be suggesting sources by authors such as Rindermann should be an exception to this part of the guideline. In its current form it doesn't include any such exception, but if you think that exception should exist, modifying the guideline to include such an exception is the first of the three options I've listed above. Ferahgo the Assassin (talk) 22:22, 21 April 2025 (UTC)

I am not suggesting that, no. I am suggesting that you are misreading the current guideline. The omissions of these sources and views are entirely supported by the present policy. My preferred outcome would be that you stop wasting community time on this periodically (we have had several RFCs already) and simply accept that consensus is against your views. I imagine if you keep this up at some point somebody will file the required wikipaperwork to get your topic ban reinstated if you don't stop voluntarily. MrOllie (talk) 22:31, 21 April 2025 (UTC)

• Something to remember… per WP:VNOT… Verifiability does not guarantee inclusion. Even if the source is 100% reliable, if there is a consensus to not mention something, we don’t mention it. Blueboar (talk) 21:19, 21 April 2025 (UTC)

• The guideline as it stands now looks good to me. The problem here seems to be local instances where the guideline was not adhered to. That the passage "...human populations living on different parts of the globe have been evolving on divergent trajectories reflects the different conditions of their habitats" cited to a 2013 NY Times science article was removed under the argument that the source was not reliable, is rather egregious, although I'm not surprised that this happened. Jweiss11 (talk) 23:58, 21 April 2025 (UTC)

  Jweiss11, please clear this up for me: Are you saying that if the New York Times publishes something, we should ignore the fact that the author was shortly afterward the subject of a truly massive rebuke by subject-matter experts –– perhaps the most resounding refutation in the history of science journalism –– on precisely the same topic? Or were you just assuming that the removal was "egregious" without looking into the matter? Generalrelative (talk) 00:57, 22 April 2025 (UTC)

  The passage I cited here is clearly true. It's a non-controversial statement about the basics of biological evolution. Is human evolution not subject to environmental conditions? I think it clearly is, e.g. High-altitude adaptation in humans, [9], etc. If Nicholas Wade shot the president the next day after this article was published, it wouldn't make it false. Jweiss11 (talk) 01:16, 22 April 2025 (UTC)

  If a claim is non-controversial, we should be able to find a non-controversial source to back it up. If you read that talk page discussion, you'll see that I made this point in my first comment. You are free to go back and re-add the claim with a solid reference right now. Unless you think the passage is so "clearly true" that WP:BLUE applies, in which case no reference would be needed. But Nicholas Wade didn't get in trouble for shooting anyone; he got in trouble for flagrantly misrepresenting the scope of recent human evolution –– precisely the topic for which we decided he is unreliable. Ferahgo appears to believe that editors shouldn't be trusted to make this sort of determination through consensus, but she is wrong. We do it all the time and it's one of the foundational reasons this project works. Generalrelative (talk) 01:42, 22 April 2025 (UTC)

  It's my editorial judgement that Nicholas Wade making this straightforward claim about evolution in the NY Times is indeed reliable irrespective of the other controversy with Wade. I also have doubts that Wade "flagrantly misrepresented" anything. Surely many people in academia thought this was so, but I suspect this was case in large part because of politically-motivated reasoning. This project works very well for many topics, but in the case of sensitive political ones, like this one, it tends to import massive political bias both from scholars in academia and the interpreters of sources here on Wikipedia. This episode right here with Wade looks a textbook case of such a failure. Jweiss11 (talk) 02:55, 22 April 2025 (UTC)

  This isn't just 'many people in academia', it is many of the people whose work Wade cited. Surely if anyone is qualified to state that research is misrepresented, it is the people who did the research in the first place. MrOllie (talk) 03:08, 22 April 2025 (UTC)

  +1 to this. That said, Jweiss11 appears to understand that they're in the minority here, and that ultimately editorial judgement, through the consensus process, needs to prevail. To that extent, we all agree. If they feel that one journalist is somehow more reliable (because less politically motivated?) than 139 senior faculty members in population genetics and evolutionary biology –– when it comes to matters of population genetics and evolutionary biology no less –– they are fully entitled to their opinion. But they will fail to establish consensus for such a view among any reasonable group of editors, which is as it should be. Again, this is not a bug, it's a feature. Generalrelative (talk) 03:46, 22 April 2025 (UTC)

Now that the various entrenched sides in this dispute have explained their positions, could uninvolved editors such as WhatamIdoing please weigh in?

The main thing I'd like to know from uninvolved users is whether they feel the current wording of the guideline, Material such as an article, book, monograph, or research paper that has been vetted by the scholarly community is regarded as reliable, where the material has been published in reputable peer-reviewed sources or by well-regarded academic presses is consistent with current editing practices and with the local consensus that Generalrelative explained above, which classifies such sources as non-RS in some cases. And if the current wording of that part of the guideline is not consistent with this local consensus, what do uninvolved editors think should be done to address the inconsistency? Ferahgo the Assassin (talk) 20:25, 22 April 2025 (UTC)

Ferahgo: I understand that you'd like to hear less from me but you're blatantly misrepresenting the guideline. Directly above the line you keep quoting it states:

However, some scholarly material may be outdated, in competition with alternative theories, controversial within the relevant field, or largely ignored by the mainstream academic discourse because of lack of citations. Try to cite current scholarly consensus when available, recognizing that this is often absent.

In the case of race & intelligence, there is a strong scholarly consensus, which is why we try to cite that consensus wherever appropriate, and consider those who try to build a career in opposition to that consensus generally unreliable on the topic –– just as we would consider the work of someone who builds a career arguing that vaccines cause autism unreliable when it comes to the topic of vaccine safety. This attempt to revive the stunt of a long-term abuser is beyond misguided. The appropriate response to such abuse is to deny recognition. Generalrelative (talk) 22:22, 22 April 2025 (UTC)

The part of the guideline that you're quoting, which includes the instruction Try to cite current scholarly consensus when available, is discussing how editors should decide what sources to cite in articles, or how to choose which sources take priority. There are many reasons sources that satisfy the criteria of WP:RS sometimes don't get cited, and that's one. But the issue I'm trying to discuss here is about what sources satisfy WP:RS at all, which is a separate question.

This isn't just a nit pick or a semantic point. Sources that fail the requirements of WP:RS not only can't be cited; they also can't be used in examinations of the balance of views that exists in the source literature. The local consensus that all hereditarian sources are inherently non-RS, regardless of where they're published, was one of the main reasons for rejecting Sesquivalent's proposal to review the source literature in depth for such an examination. Defining what's a reliable source in this way also means that if the scholarly consensus about a topic ever shifts so that a view is no longer fringe, Wikipedia would nonetheless be required to continue treating it as a fringe view, because evaluations on Wikipedia of what is or isn't fringe can only be based on views that exist in reliable sources, and sources that support the view are regarded as non-RS by definition.

You probably still think I'm misunderstanding the guideline, but we aren't accomplishing anything by continuing to repeat the same argument that I (and Jweiss11, Stonkaments, etc.) have been having with you for the past four years. Can we please wait to get feedback from uninvolved editors? Ferahgo the Assassin (talk) 23:53, 22 April 2025 (UTC)

That's nonsense. If the scholarly consensus truly changes, you will have no trouble convincing a noticeboard that the scholarly consensus has changed. You will show folks the new sources and they will be persuaded by the evidence. Then you can add sources consistent with the new consensus to article space. There is nothing I could do to stop this from happening, if the science was on your side. You and that handful of other editors doggedly pushing racial hereditarianism (most of them indeffed by now) have failed to persuade the community because your arguments have been unpersuasive. And now that I've said my piece, yes I'll be happy to leave it to uninvolved editors to weigh in. Generalrelative (talk) 00:46, 23 April 2025 (UTC)

Ferahgo, I don't think you'll be happy with my answer, but here it is:

• To echo what Blueboar said above, it is possible for a source to be reliable but unusable for reasons such as WP:UNDUE or WP:PRIMARY.
• To echo what a couple of editors said above, if an idea is actually accepted as a majority or minority POV, rather than a fringe idea, then you will normally be able to find unrelated reputable sources saying the same thing. For example: If Prof. I.M. Portant says that he achieved cold fusion, then we want people to say things like "I replicated his results, and it worked for me, too". We don't want sources that just believe him, or say that they've always believed it was possible, or that his assertion proves my idea is correct, too. We want more than WP:LINKSINACHAIN.
• If Prof. Portant publishes something – even in a gold-plated "ideal" scholarly source – and a large section of the field smacks him for getting it wrong, then Prof. Portant's source stops being reliable "for" whatever he claims and becomes only reliable "for" narrow claims that he said it. This is WP:RSCONTEXT, which is equally a part of this guideline.

People often say "it's unreliable" when they actually mean "it's unreliable for the specific extraordinary statement (but not all possible statements)" or "this source, even if it has various qualities that we associate with reliability, isn't strong enough to demonstrate that this content is DUE".

Which brings me back to the basic definition: A reliable source is one that editors accept for a particular use. Consensus is ultimately required. A "bad" source can be reliable, and a "good" source can be unreliable. You cannot look at a source and say "It's WP:SCHOLARSHIP, so it's absolutely guaranteed to be reliable no matter what". I think you will find that deciding whether to use a given source can be much more complicated than that. WhatamIdoing (talk) 00:52, 23 April 2025 (UTC)

I see no reason to change the current policy. If a reliable source suports a fringe view then what we should really be doing if we really think there is something wrong with it is find other sources. The suggestion sounded to me more like applying ones own feelings to try an bias Wikipedia rather than using the sources. NadVolum (talk) 23:09, 22 April 2025 (UTC)

Article for awareness on journal publications

Nature article on how reputable journals get bought and changed to poor quality works [10]

Nothing we need to act on, but should be aware that a reputable journal from the past may change over time. Masem (t) 16:23, 22 April 2025 (UTC)

RfC regarding names in sources' titles and URLs

 You are invited to join the discussion at Wikipedia:Village pump (policy) § RfC: Exclusion of a person's name following consensus. Some1 (talk) 00:39, 29 April 2025 (UTC)

Question about Definition of published

On Definition of published I noticed it said this

Published means, for Wikipedia's purposes, any source that was made available to the public in some form.

I must ask about this because I have stumbled upon sources at many libraries. These sources aren't accessible anywhere online, you can't buy them anywhere, and you can't check these books out. However, anyone can enter these libraries and anyone can freely read these books. I was even allowed to scan these books by hand for free.

Would these sources be considered published by wikipedia standards?CycoMa2 (talk) 18:01, 4 May 2025 (UTC)

Yes, very safely so. Remsense ‥ 论 18:02, 4 May 2025 (UTC)

Okay thank you. CycoMa2 (talk) 18:03, 4 May 2025 (UTC)

To me, the key elements are the present guarantee of verifiability by the public, and additionally the institution (a library, generally having the imperfect expectation of operation on compatible terms through the coming years and decades) that suggests future verifiability also. Whatever other hurdles there are toward reliability are a separate matter. Remsense ‥ 论 18:08, 4 May 2025 (UTC)

Yes, anything that is accessible to the general public is 'published'. As an example of a limit, a source that is only available to the employees of a business, or to members of a religion or a club, is not 'published' for Wikipedia's purposes, even if several Wikipedia editors happen to have access. WhatamIdoing (talk) 20:15, 4 May 2025 (UTC)