Jump to content

Wikipedia talk:Reliable sources

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Jc3s5h (talk | contribs) at 16:28, 30 September 2022 (→"Generally acceptable" USERG: Reply). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

zakzak.co.jp

Moved to WP:Reliable sources/Noticeboard § zakzak.co.jp

Preventing Ad Hominem

Sock drawer. Generalrelative (talk) 21:47, 10 September 2022 (UTC)[reply]
The following discussion has been closed. Please do not modify it.

I was reading the definition of a source section and read that "any of the three can affect reliability" (The three being, author, publisher, piece of work itself). Might I suggest adding that the extent to which any of these three affects reliability is context dependent. For example, for a peer-reviewed scientific paper, the publisher is a lot more important than the author. That is what the peer-review process is about, being able to be confident that at least when it comes to factual matters, even though the author of a paper may be biased, said factual matters have been checked and double checked by disinterested reviewers. Ohcanada123 (talk) 07:44, 26 August 2022 (UTC)[reply]

"Bias" doesn't necessarily prevent a source being reliable. But subject to that, isn't the guideline already saying this eg WP:CONTEXTMATTERS? DeCausa (talk) 12:35, 26 August 2022 (UTC)[reply]
It's a fair point to suggest "context matters" isn't prominent enough. Why are we giving WP's definition of "published" right after the list of RS criteria, since it is used as a qualifying criteria of a source, not a qualifying criteria of reliability. Thus it should go in a separate section, or simply be limited to WP:V and WP:PUBLISHED. (Also this propagation of tangential policy cross-refs can be problematic, as I recently edited to show.) So I propose a cursory, if any mention of what disqualifies a source from being usable -- just wl to WP:V#Reliable sources == WP:SOURCE and separate the disqualifications there, as the point here is to talk about evaluating reliability. Then remove the "published" section (wl PUBLISHED and/or V), and presto, ContextMatters is now right after the list of three major things affecting reliability. SamuelRiv (talk) 16:16, 26 August 2022 (UTC)[reply]
Genuine CE here, but not on my post, to make this point: I made the mistake that WP:PUBLISHED above should actually be WP:PUBLISH, because the former redirects to the RS section, while the latter is its own information page, to which I suggest we wl instead. Is that potentially confusing? Needless duplication of identical material? Nah, that's just crazy-talk. SamuelRiv (talk) 16:20, 26 August 2022 (UTC)[reply]
I agree it's confusing. I feel like both should lead to Wikipedia:Reliable_sources#Definition_of_published, which provides a link to Wikipedia:Published where people can drill down. Valereee (talk) 18:05, 26 August 2022 (UTC)[reply]

Ok, I suggest we change the following:

"In general, the more people engaged in checking facts, analyzing legal issues, and scrutinizing the writing, the more reliable the publication. The relative importance of author, publisher and publication also depends on context. For example, a non-peer-reviewed scholarly book might be unreliable if the author is known to be biased. However, if a publication is peer reviewed in a reputable journal, or if the underlying data is verifiable in public data bases, then the author can be irrelevant." Ohcanada123 (talk) 22:09, 28 August 2022 (UTC)[reply]

This extrapolation is simply incorrect in general -- I don't know what to tell you. What's wrong with just noting the three criteria affecting reliability to consider, and context matters? Stuff about peer review, bias, etc. all belongs in subsections explanatory essays and policies specific to news and academic fields. Not sure what public databases have to do with anything -- they range in reliability from useless to gold-standard (depending on what you use them for and how). As a running theme on RSN, bias in general does not affect reliability. SamuelRiv (talk) 22:18, 28 August 2022 (UTC)[reply]
Scholarly books are never peer reviewed. All the good scholarly books are "biased" in some way (e.g., toward my idea of why they ousted the old ruler and away from your ideas about the same event). I wonder if we still aren't internalizing the idea that biased sources are okay. Wikipedia articles should not be biased, but reliable sources can be biased. WhatamIdoing (talk) 23:12, 29 August 2022 (UTC)[reply]
It's not correct to simply say that "scholarly books are never peer reviewed." Academic presses do employ peer reviews of their books, sometimes of portions of the books and sometimes of entire books. That is the primary thing that makes them academic presses (aside from their specialized topical foci, of course). ElKevbo (talk) 23:19, 29 August 2022 (UTC)[reply]

ElKevbo Interesting, I was not aware of that. In that case I cannot think of a situation (at least when it comes to scholarly topics) when the author of a paper or a book is really relevant to the reliability of the source, if their work is checked and approved by independent scholars in the field who work with reputable publishers. Does anyone know of an example when the author of a a piece of scholarly work that is published by a reputable academic publisher or journal can affect reliability?Ohcanada123 (talk) 03:29, 30 August 2022 (UTC)[reply]

Crude “reliability” isn’t usually the differentiator between academic authors if they’re published by reputable publishers. It’s usually about weight (WP:DUE etc). A world expert’s work is goingto carry more weight than a monograph based on a recent PhD thesis. But I come back to my point earlier in thus thread: context matters is already appropriately covered: nothing needs fixing, particularly asis demonstrated by this thread trying to “fix it” digs a hole full of cans of worms…DeCausa (talk) 03:47, 30 August 2022 (UTC)[reply]

I know crude reliability is not usually the differentiator between academic authors, but I was wondering about situations where it was. If these situations are exceedingly rare, may I suggest simply saying that the more the author's work or claim has been vetted by peer-review and other scholars in the field, the less relevant the author becomes in terms of reliability, but there still may be concerns about due weight. Ohcanada123 (talk) 05:37, 30 August 2022 (UTC)[reply]

Within academic sources, author identity usually matters at the extremes: the widely recognized world experts on the one end, and the crackpots and disgraced fraudsters on the other. WhatamIdoing (talk) 01:07, 31 August 2022 (UTC)[reply]

That is what I thought, but think of the following situation. Suppose Deepak Chopra ran a large scale double-blind, placebo-controlled, randomized controlled trial and found that homeopathy actually works. Suppose after intense peer-review and scrutiny The New England Journal of Medicine found that this study was of sufficient methodological quality to allow publication. Is it really relevant that the author is a crackpot or a disgraced fraudster? It seems to me the definition of ad hominem. Indeed, if a crackpot like him was able to jump through all the necessary hoops to get their methodology checked and their results published by a respected publisher, then in my opinion that is worth at least a mention. Ohcanada123 (talk) 03:30, 31 August 2022 (UTC)[reply]

For myself, I'd probably believe that NEJM is trying to get some publicity. They're known for running controversial articles; getting people to cite an article for the purpose of disagreeing with it or using it to claim uncertainty in the field can do very good things for a journal's magic numbers. Also, you wouldn't be able to use it in any medical-related content on wiki, because it's a primary source instead of a review article. With the usual systems, there's a 5% chance of an erroneous outcome in a randomized trial.
But: If we leave aside all the details, and instead say that it's a famous woo-woo person, whom we'll call Cheeky Doppa, in a strictly non-medical hypothetical field of alternative babblology, who got an article published in one of the leading journals for the field, then I think the article would be considered reliable at least for some limited claims ("Cheeky Doppa claimed that..."), and the next step would be to figure out whether mentioning that was WP:DUE. Also, I'd suggest as much patience as you can manage, since complaints about bad work tend to surface fairly quickly, especially if it's made a big splash in the media. (Of course, that same level of media attention increases the demand for Wikipedia articles to say something, even if all they say amounts to "Yes, we know that he published a paper, okay?") WhatamIdoing (talk) 01:02, 1 September 2022 (UTC)[reply]

Objection to "Definition of published"

The "Definition of published" section contains the sentence "Like text, media must be produced by a reliable source and be properly cited." I object to this sentence because it is not actually part of the definition of published. There is an unfathomable amount of text and media that is neither produced by a reliable source nor cited in Wikipedia, but it is nevertheless published.

If we allow this style of writing, we will have to reproduce the entire guideline in every section. No, that's not good enough, we will have to reproduce the entire guideline in every word, which means the guideline would have to consist of a single word that says everything we want to say about reliable sources. Jc3s5h (talk) 20:29, 26 August 2022 (UTC)[reply]

z the leading sentence says: Published means, for Wikipedia's purposes, any source that was made available to the public in some form, which implies that any further discussion in that paragraph is specific to WP. Thus it is okay to require publishing to be from reliab!e sources. Masem (t) 20:38, 26 August 2022 (UTC)[reply]
It seems Jc3s5h is not considering the whole definition section and inadvertently taking this sentence out of context. The opening sentence agrees with what your are saying. The first sentence says, "Published means, for Wikipedia's purposes, any source that was made available to the public in some form." This fits with your definition of published and is as broad as yours. The sentence you have focused on pertains to being within the reliable sources guideline. The guideline page is the context for that particular sentence. Hope this helps. ----Steve Quinn (talk) 20:47, 26 August 2022 (UTC)[reply]
(edit conflict)Maybe the sentence is allowable, but it is unwise. The guideline should, as much as possible, separate different concepts to allow clear use and discussion of those concepts. "Published" is one concept. "Reliable" is a second concept. "Cited" is a third concept. "News organization" is a forth concept, and so on. If I want to say that a certain source, for example, https://www.capnhq.gov/ is unacceptable for use in Wikipedia because it is not published, I should be able to do so for the sole reason that it is not, and so far as I know, never has been, available to the public. I shouldn't get dragged into a discussion of whether it's cited or whether it's reliable when the sole issue is whether the public can get access to it. Jc3s5h (talk) 20:52, 26 August 2022 (UTC)[reply]
Your example (Civil Air Patrol? Anything more?) is on a website that I can theoretically access with an authorization that I can theoretically acquire (even if it's difficult to do so) -- I think the only way it could not be considered 'published' by WP's definition is if it's under some legal confidentiality that prevents disclosure (personal confidential info like med records, classified docs (including docs intrinsically classified), certain corporate docs, docs under court seal and/or general injunction, etc etc etc.). WMF and other higher-ups would be better suited to give the full run-down on that, but it should be more or less clear in the western world, and you should take uncertain cases (like releases of classified documents on video game forums, for example) to admins. SamuelRiv (talk) 22:29, 28 August 2022 (UTC)[reply]
The specific example I gave, the members-only website of the Civil Air Patrol, ordinarily would require one to join that that organization. A work that is only made available to members of an organization is clearly not made available to the public, and so is not published. Jc3s5h (talk) 22:37, 28 August 2022 (UTC)[reply]
That would still be published for all purposes. Publishing does not require to the public at large but to a subset of the public (here being the members of this group) Masem (t) 22:41, 28 August 2022 (UTC)[reply]
It is immoral to cite material that is only accessible to people who endorse particular views, such as the views of an organization, religion, or political party. But that is not the subject of this thread, it is about not intertwining definitions so that none of the words or phrases in this policy have a distinct meaning; everything means everything, hence nothing means anything. Jc3s5h (talk) 00:32, 29 August 2022 (UTC)[reply]
We didnt say it was okay to cite it... WP:PAYWALL discusses this further. Masem (t) 00:50, 29 August 2022 (UTC)[reply]
I agree with this point that, "published" is a concept, and "reliable" is another concept. We need statements to be verifiable, i.e. published reliable sources. Some sources are published, but unreliable. I believe that should be updated if it is not clear in the text. Andre🚐 00:52, 29 August 2022 (UTC)[reply]
Masem, our definition of "published" actually does require that the material be "made available to the public", and "available only to members" is definitely not the same as "available to the public". WhatamIdoing (talk) 23:15, 29 August 2022 (UTC)[reply]

Masem, if you look at my edit more closely, I was showing that the proposal shows why archival sources are consensus not usable due to being unpublished. I was not changing the policy but changing the wording to reflect the policy. Andre🚐 01:12, 29 August 2022 (UTC)[reply]

that proposal you linked to us failed, meaning it doesn't have consensus. Masem (t) 01:20, 29 August 2022 (UTC)[reply]
Exactly, I was saying that the proposal shows why archival sources cannot be used. The proposal was a proposal to allow the usage of archival sources. Just, look at it again, please. Andre🚐 01:21, 29 August 2022 (UTC)[reply]
just because the proposed language failed makes the opposite a passing consensus, only that consensus for what was proposed did not happen. you cant make that type of assumption about that, in addition to only being the result of a straw poll rather than RFC. Masem (t) 01:25, 29 August 2022 (UTC)[reply]
We don't need to link to that from here, but that failed proposal contains a consensus (nevermind it being a "straw poll") that has as far as I know has never been overturned, that archives aren't valid because they aren't published. Andre🚐 01:27, 29 August 2022 (UTC)[reply]
Your reference for this policy is a 2006 RfC??? Let's use an example: a published-at-the-time obscure 18th century book or newspaper can only be accessed by request in library archives (say, only one library archive allows this option) as it is not digitized. This is not too abstract of a scenario, as in one Resource Requests I saw a 19th-century British paper which was only digitized for the latter half of its run, with the former half according to information online having few surviving copies and only available at a single library. Before you say this has become Primary at this point, you still need make an effort to look the original source up to do a proper secondary citation. So the question is, what is the essential difference between this scenario and what is being described in the 2006 RfC? What is it precisely the editors there had in mind when they were discussing archival materials? Was our current "WP definition" of 'published' even around back then (I back-traced the history on it a couple days ago, and it's only a couple years old, so the answer is "no"). SamuelRiv (talk) 01:26, 29 August 2022 (UTC)[reply]
Published "at the time" was published, the policy makes this clear, even if hard to access. Many archival sources were never published. [01:28, 29 August 2022 (UTC)] And you can see clearly from reading the 2006 RFC that the terminology of sources being "published" to be reliable existed at the time. It might have been written in a different way or in a different section. Oh yeah, and I participated in the 2006 discussion. Andre🚐 01:29, 29 August 2022 (UTC)[reply]

FWIW I started a thread and an edit on WP:V/WT:V about making it clear that archival sources, inasmuch as they are unpublished, are not verifiable, and that it is WP:OR besides to dig thru an unpublished archive for use on Wikipedia. Andre🚐 02:37, 29 August 2022 (UTC)[reply]

My overall impression of the WT:V thread is that people don't agree that archival sources are unpublished. The analysis seems to be something like this:
  • Is this allegedly unpublished archival source available to the public?
    • If yes, then it's actually a "published" archival source.
    • If no, then it's not published and cannot be used.
For example, if you could make an appointment at the archive, show up, and some nice archive librarian would bring you box 23-2494 and let you flip through the papers until you find a hand-written letter from the subject's mother, then that's "published" – as far as Wikipedia is concerned. If, on the other hand, you contacted the archive and they said "Oh, no, certainly not: only members of the Certified Careful Professor's Club are allowed to look at anything in the archives", then we would not consider that to be published for Wikipedia's purposes. WhatamIdoing (talk) 23:26, 29 August 2022 (UTC)[reply]
I think that's not clear yet from the discussion. The hand-written letter is at the least a primary, not a secondary source - and since it's not published, I'm not sure if it's reliable, I would hope that it would not be considered reliable for fact inasmuch as it is used to source statements from within the content of the letter. I have never previously heard that this kind of archival source was reliable, not-original-research, and verifiable. I would like to ensure we discuss this a bit more. We can discuss it here since this was the original thread and pertains to reliability, but I had started the discussion and edits there since I think this is really more about verifiability. Andre🚐 23:31, 29 August 2022 (UTC)[reply]
lets not worry about primary/secondary or reliability here. its about the ability of a reader to locate and obtain and read the source (within consideration of possible time and cost) which is the core verifyability aspect related to publication here. Masem (t) 23:36, 29 August 2022 (UTC)[reply]
@Andrevan, when you write things like and since it's not published, you're assuming that it's not published, even though I just told you that a hand-written letter can be "available to the public" and therefore what Wikipedia calls "published". Maybe it would help if you stopped thinking about real-world publication processes and starting thinking that the policy requires sources to be WikiJargonPublished, not real-world published. A hand-written letter that is posted in a store window is WikiJargonPublished (it is also LibelLawPublished, for that matter). A hand-written letter that is in an open-to-the-public archive is WikiJargonPublished. The policy does not require editors-and-printing-presses kinds of publication. It only requires available-to-the-public kinds of publication. WhatamIdoing (talk) 02:18, 31 August 2022 (UTC)[reply]
Part of reliability, at least used to involve, some oversight and reputability of the nature of publication. I'm not sure at what point the definition of published became so broad that anything that's available to any subset of the public is published, but I don't think it's a good idea for the reasons I stated, and I'm not convinced that really is the consensus view of the present policy, or that we shouldn't start some kind of discussion about tightening up that policy, provided that there are other editors who might agree (and I think 1 or 2 people seemed to agree, not pinging them out of respect for their time and attention). I like the notion of the broad distribution and wide availability, as a place to start. Regardless, a hand-written personal letter in an archive, in isolation and absent any other form of publication, would be the equivalent of a self-published blog post for reliability, and a primary source as well. So I can't see a lot of situations where this letter-in-an-archive thing is going to give anything useful. Do you know of any current articles that are using this kind of source? Andre🚐 02:27, 31 August 2022 (UTC)[reply]
The question of whether something is "reliable" is not the same as whether something is "published". This discussion is only about whether something is "published". Lots of "published" things are totally unreliable for any statement more complex than "Somebody posted ____ on the internet".
We have never had a definition of reliability. In practice, a reliable source is whatever source(s) experienced editors will accept as sufficient to support a particular claim. The acceptable source might – or might not – have some oversight. It might or might not have a positive reputation. It might or might not be one of the types of publication that we usually prefer. It might be a "good" source, or might be a WP:NOTGOODSOURCE.
If you're searching for sources that are likely to be reliable for most of the typical uses that an experienced editor would make of a good source, then we have always provided advice on qualities to look for, such as fact-checking, editorial oversight, non-self-published, etc. But you can't actually determine whether a source is actually reliable until you compare it against the statement it's meant to support. A corporate press release has none of the qualities we advise editors to seek out, and yet it's "reliable" for statements about what the corporation said once. A scholarly book about chemistry has all the qualities we advise editors to seek out, and yet it's "unreliable" for statements about which celebrity got arrested last night.
And, again, for this discussion, none of that has anything to do with whether the source is published/available to the public. WhatamIdoing (talk) 03:21, 31 August 2022 (UTC)[reply]
I wasn't conflating the definition of reliable and published. But part of reliability, at least in my heyday, has come from how and by whom the source was published. We do have a definition of verifiability: reliable, published sources. That's why I moved the thread, or tried to, to WT:V instead of here. Since this is WT:RS.
You didn't quite answer my question of a real-life example of an archival source being usable. I contend they should not be used. I can give my own example. A few years ago I went to the Rockefeller Archive Center which is open for visitors, who are mainly academics and historians, but you need to make an appointment, and presumably have a good reason as well and your area of research interest[1]. There is a bunch of unpublished material there that helped me research Cornelius P. Rhoads. At one point, I created an article about a borderline notable doctor and public health administrator called George C. Payne. Later it was deleted for insufficient references. I have scans and photos of my visit to the archive and I can cite the boxes. Should the archival docs count toward his notability and be usable for facts about him?
I would say, under my understanding of policy, they should not. Notability is determined by weight published, reliable sources. More reliability and prominence = more weight.
I agree with you that some self-published and unreliable sources are usable in context, but that is not about verifiability. If I started citing letters from an archive, it might be borderline unverifiable or very difficult to verify, if not impossible. That is a lot of surface for failed verification reference checks, in my view. Andre🚐 03:30, 31 August 2022 (UTC)[reply]
One of the factors that we encourage editors to consider, when assessing whether a source is reliable, is who/what published it. This is still the case. Things that are WikiJargonPublished in (only) an archive get basically zero points on that score. You would treat them the same way that you would treat a verified social media account.
If a source is usable in an article at all, then the source verifies the content. That is about verifiability. Any reliable source – even one that is "reliable only in context" – is a source that makes it possible for someone to check that editors didn't just make up the content, but instead took the content from a source that is reliable for that content.
But I'm not sure that archives would usually count towards notability, whose overall goal is to determine whether there is "sufficiently significant attention by the world at large". Most of the contents in a typical archive (e.g., a personal diary, a contract signed by the subject) also wouldn't be Wikipedia:Independent sources, and non-independent sources never count towards notability, even though they are frequently reliable sources. I would expect approximately none of the archives contents to be secondary sources, which the GNG theoretically also requires (though AFD routinely accepts WP:PRIMARYNEWS sources). WhatamIdoing (talk) 01:13, 1 September 2022 (UTC)[reply]
I definitely agree that things wikiJargonPublished only in an archive get 0 points since it's essentially, as you say, a verified social media account. But does everyone agree about that? And is that clear from policy or just our shared interpretation? Because archival sources have a lot of pomp and circumstance, so they seem really reliable and important if you aren't well-versed, but I agree they should be given no weight. I think it's worth figuring out.
You understand how it's confusing to say "who/what published it" is important but we actually just mean "who/what made it available," which isn't really quite the same non-wiki-jargon. Because there are reliable publications, and unreliable, and in some cases it's not a publication in the conventional sense. So it's also how, and where, and in what form it was published, and made available. But actually being real-world published in a journal confers notability, like in science that has a very specific meaning. So it makes it hard to understand and interpret the policy because we don't go into enough detail on what is the wiki definition, how it applies to reliability (this page), and different meanings that don't apply.
Archival sources are not always non-independent, and not always non-secondary. For example, the archive I referenced above included in their boxes, press clippings, internal reports, letters from various entities to and from each other, draft articles that could be published but weren't, etc. Some of those letters or other sources were from 1 independent person to another independent person about a 3rd person that was also independent from the archive.
Maybe, if you want, I'll upload a bunch of archival sources to an imgur so you can see what I'm talking about. Andre🚐 01:33, 1 September 2022 (UTC)[reply]

The Epistemic Issue With "Reliable Sources"

The article on “Appeal to Authority” highlights a major epistemic issue which itself reflects poorly on wikipedia’s knowledge base.

There are two claims outlined in the article — one, that “appeals to authority” are fallacious, and two, that they are not.

The first claim is not only sourced, but various arguments are presented from those sources in support of the claim. “Appeals to authority are fallacious because experts are not always correct, the human psyche has a cognitive bias in favor of authority, and not all authorities have the right knowledge on the matter.”

The second claim is only supported by the sourcing itself. “Appeals to authority are not fallacious because various authorities have claimed that they are not.” The issue with this claim should be apparent. And if you read these sources, they do have sentences like “the appeal to authority is not always fallacious,” but within context they are all outlining situations where epistemic certainty is not guaranteed.

I think I can see why this happened — editors are asked not to provide the justification that the authors give for their claim, but only the claim itself. Let’s say the claim is that person X was at street Y at Z time. In order for wikipedia to reliably publish this claim, the editor should be forced to also source the justification for it, i.e. “there is video evidence” or “multiple first hand accounts place them there.”

Obviously this article is pertinent to the situation at hand, as wikipedia’s entire knowledge base is essentially an appeal to authority. I get why this is, but as appeals to authorities do not and cannot guarantee epistemic certainty, far greater care must be taken than is currently being taken. I would suggest forcing editors to quote the justification for the knowledge claim being given, and if there is no justification other than “this source is subjectively reliable,” they should not be able to publish the claim. 2600:4040:A23F:B200:14DF:3CE1:5244:E5F0 (talk) 15:37, 6 September 2022 (UTC)[reply]

Appeal to authority means you can't pull rank to win arguments - it does NOT mean that experts aren't experts or that Wikipedia shouldn't report what reliable sources say. Andre🚐 15:45, 6 September 2022 (UTC)[reply]
I never claimed that “experts aren’t experts” nor that “Wikipedia shouldn’t report what reliable sources say.” I’m not entirely sure how you derived that from the text provided. "Appeal to authority" means that just because a reliable source claims something, that doens't make that thing true.
I am suggesting that Wikipedia’s articles should quote the justification that the source gives for the claim, rather than just the source’s claim. The form should be “claim X, because Y,” rather than just “claim X.” If there is no “because Y,” the source should be deemed unreliable. 2600:4040:A23F:B200:74B7:1F05:89FF:8344 (talk) 18:15, 6 September 2022 (UTC)[reply]
I disagree. Verifiability, not truth, is the test. If sources all say something and they are reliable according to our policy and consensus, we do not need to cite their proof and reasoning. Andre🚐 18:25, 6 September 2022 (UTC)[reply]
If no justification is provided for the source, first, how can the information possibly be verifiable, and second, how is the source qualified as "reliable" aside from the subjective judgement of the editors?
Wouldn't this be the precise opposite of "verifiability, not truth," where the verifiability of the information is eschewed in favor of a person merely believing the claim is true?
I was hoping this issue wouldn't be one fundamental to wikipedia's practice and that it was just an oversight. Is the general opinion of other editors also that appeal to some authority ultimately justifies wikipedia's claims, rather than another form of justification? 2600:4040:A23F:B200:8451:2E22:12D1:FEDF (talk) 19:19, 6 September 2022 (UTC)[reply]
Since we have no process available to determine what the truth is (neither on Wikipedia or in the world for many topics), verifiability is the best we can do. MrOllie (talk) 19:22, 6 September 2022 (UTC)[reply]
Ignoring the (very postmodern) tautological contradiction in your claim -- how is verifiability possible if the source's justification for the claim is not given?
"Verifiability, not truth" means that the verification should be achievable and mere belief that a thing is true isn't enough to make a claim. Not that a source merely believing a thing is true is enough for wikipedia to state it as such without giving verifiability. That would be the exact opposite of what that means. 2600:4040:A23F:B200:8451:2E22:12D1:FEDF (talk) 20:17, 6 September 2022 (UTC)[reply]
We don't care what the source's justification is. We don't require sources to show how they got to their conclusions. MrOllie (talk) 20:22, 6 September 2022 (UTC)[reply]
Again, how is a source then verified as reliable?
It's not "truth, not verifiability." But that seems to be what is happening if you don't require the sources to verify their claims. 2600:4040:A23F:B200:C0F9:54E5:12FD:43A (talk) 20:49, 6 September 2022 (UTC)[reply]
Wikipedia is a encyclopedia. That means it's a tertiary summary source. We aren't arbiters of the truth of facts. We are simply assembling a summary of the treatment in reliable sources. Verification that the fact is sourced reliably, is what is required, not verification that the fact is true. It's not tautological - even if we read a claim in a source that we couldn't personally verify as true, and we thought it was false and wanted to verify its truth, verification might not be feasible. For example, claims made by the consensus of scientists based on experiments. We might not have the right equipment or skills or ingredients or permission, like centrifuges or medicines or whatever, to personally verify physics or biology results. We don't just blindly trust everything, but we are going to take the word of reliable sources, or balance them if there is a weight of disagreement on something, or attribute in cases where necessary. Andre🚐 20:23, 6 September 2022 (UTC)[reply]
An encyclopedia is a collection of knowledge, and knowledge is justified true belief. If wikipedia provides no justification for the knowledge it presents other than a bare appeal to authority, it is a very poor encyclopedia.
And again, how is a source deemed reliable? Is reliability is decided ultimately on an appeal to authority or is it decided on non-fallacious grounds, i.e. by providing the justification for the source's claims? I am not entirely sure how to make this more clear.
Science is a good example as it always provides justifications for its claims, or ways to verify its claims, other than an authority, as is noted in the article I am referencing. If "verification" is as you describe and is ignorant of the source's own verification, then shouldn't the rule be called "journalistic quotation with no verification, not truth?" 2600:4040:A23F:B200:C0F9:54E5:12FD:43A (talk) 21:08, 6 September 2022 (UTC)[reply]
I genuinely applaud the OP, but regarding the specific question in the OP, what you ask for is not feasible, and I think that Mr. Ollie summarized it well. But regarding other items which underlie this thread or what it touched on... "Verifiability not truth" is gone, and we got rid of it for good reasons. One is that one of the TWO common meanings for "truth" is accuracy, and "verifiability, not accuracy" is certainly not our objective. Regarding areas where objective accuracy exists, we strive for it, and verifiability is a means to that end. Also, it would be good to embed reviewing the expertise and objectivity of sources more deeply in policy. RSN actually does this, but outside of RSN too often sources get considered "reliable" not "not reliable" simply by whether or not they have certain wikipedia-defined trappings. North8000 (talk) 21:20, 6 September 2022 (UTC)[reply]
I'm not sure why it would not be feasible. Perhaps you could explain?
In order for the source to be qualified as reliable in the first place, it surely must have a justification for the knowledge other than "I'm an authority" if that knowledge is not to be fallaciously derived. That justification should be the thing that wikipedia sources, not just the bare claim. Otherwise a "reliable source" could make a false claim with no justification and wikipedia would have no safeguards other than mainstream opinion.
How does wikipedia verify its claims aside from subjective appeal to whoever the editors happen to find reliable? I'm not sure what the "verifiability, not truth" rule was replaced with (it says on the page that the core of the rule remains the same), but doesn't this process eschew actual verification in favor of whichever popular authorities merely claim certain things are true? 2600:4040:A23F:B200:C0F9:54E5:12FD:43A (talk) 21:54, 6 September 2022 (UTC)[reply]
Yes, "I'm an authority" is what we require, and is pretty much all we require. We follow whatever is written in (for example) peer-reviewed scientific journals from publishers with good reputations. If a reliable source makes a false claim Wikipedia will repeat it until a better source comes along to correct the issue. A hypothetical version of Wikipedia from 1800 would have stated that diseases are caused by Galen's miasmas, because that was the belief of doctors of the day, even though it turned out to be wrong (and some people were saying it was wrong even then). That is how the Wikipedia project is designed, because the capability to reproduce every observation and experiment from first principles is not within our means. MrOllie (talk) 22:05, 6 September 2022 (UTC)[reply]
Forgive me for any frustration but it's very hard to get you guys to respond to the the things that I am saying. I am not suggesting that wikipedia reproduce every observation and experiment from first principles. I have no idea how you derived that from the text provided. I am merely suggesting that editors quote the justification that the source itself gives. In this way, only sources that themselves provide justification will be trusted.
This cuts to the heart of the question I've been asking -- how is "authority" defined, and "reliable sources" decided upon? Why have the doublespeak of "verification" if all that is meant is "appeal to whoever the editors subjectively deem an authority without any need for verification?" 2600:4040:A23F:B200:3D20:D009:EEF:1C62 (talk) 14:06, 7 September 2022 (UTC)[reply]
Well, that's just not how it works, here, and I haven't heard a good reason why it should be. Reliable sources are defined by WP:RS and there is some explanatory info on WP:RSP that explains the consensus from past decisions on why some sources are considered WP:GENREL. Andre🚐 14:08, 7 September 2022 (UTC)[reply]
I can see that they are vaguely defined, but it ultimately seems to appeal to group opinion and does not explain how this appeal works.
Is "verification" just a subjective vote with no appeal to actual verification done by the source? As in, if a source says "it's raining cats and dogs," that source needs to provide absolutely no verification for that information in order to be included on wikipedia?
Why exactly would necessitating the source provide justification be a bad idea? 2600:4040:A23F:B200:3D20:D009:EEF:1C62 (talk) 14:20, 7 September 2022 (UTC)[reply]
As in, if a source says "it's raining cats and dogs," that source needs to provide absolutely no verification for that information in order to be included on wikipedia? Yes, that is correct. Why exactly would necessitating the source provide justification be a bad idea? Why would it be a good idea? At some level we just have to trust that the author of the citation did the work. MrOllie (talk) 14:30, 7 September 2022 (UTC)[reply]
I'm a bit flabbergasted. You are honestly asking why it would be a good idea to prevent false unjustified claims from appearing on wikipedia? If you aren't being purposefully obtuse, one good reason is that the literal only thing preventing wikipedia from publishing any kind of misinformation is whether or not the editors happen to subjectively distrust the source. So the only thing preventing wikipedia from publishing right wing misinformation is that the editors happen to subjectively not trust right wing news sources, rather than that those sources do not provide justification for their claims. I am surprised that suggesting a fix to this problem is met with this response.
And no, "we" do not just need at some level to trust that the author of the citation did the work. "We" can very, very, incredibly easily show that they did the work. Otherwise "we" will easily fall prey to whichever misinformation happens to be popular. 2600:4040:A23F:B200:3D20:D009:EEF:1C62 (talk) 15:00, 7 September 2022 (UTC)[reply]
No, that's not what I said. I don't believe requiring some poorly defined 'justification' will actually keep false claims from appearing on Wikipedia. Anyone untrustworthy enough to publish a false statement will probably publish a false justification as well. Despite your claims, many statements in reliable sources are based on detailed statistical analysis, large data sets, and the like. There is no 'incredibly easy' way to double check the work, that is why we rely on editors or peer-reviewers (another 'authority') to do it. MrOllie (talk) 15:20, 7 September 2022 (UTC)[reply]
“That’s not what I said, but yes I will reiterate that Wikipedia shouldn’t prevent unjustified claims because that is exactly what I said.”
And again, I'm not sure why you aren't grasping this point, but I am not suggesting that editors necessarily double check the information themselves, merely that if a source is reliable, then it necessarily will have some justification for a claim other than that it is an authority, and that justification should be cited by the editor.
-"Anyone untrustworthy enough to publish a false statement will probably publish a false justification as well."
And anyone that publishes no justification is by definition untrustworthy. Requiring no justification just makes it that much easier for anyone happening to propagate misinformation. This is incredibly easy.
I also don't understand why you're dodging the question of how exactly "reliable sources" are decided upon. If the base for the claim is not the thing being debated by the editors, then what exactly are the editors verifying if not baseless claims? 2600:4040:A23F:B200:389A:C155:155D:F9AD (talk) 16:27, 7 September 2022 (UTC)[reply]
I think I'm done responding here, have fun arguing with that straw man. - MrOllie (talk) 16:34, 7 September 2022 (UTC)[reply]
You asked why it would be a good idea to cite the source's justification and suggested that appeals to authority were the only possible base of all knowledge claims. In what way is my wording a straw man? 2600:4040:A23F:B200:EC55:CED4:84D8:6E8E (talk) 18:04, 7 September 2022 (UTC)[reply]
Reliable sources, for news for example, are decided based on how a WP:NEWSORG meets our standards for accuracy, fact checking, and the other hallmarks of a reliable news source. For academic sources and books, we look at the reputation and credentials of the institutions and publishers etc., whether it is a predatory journal or self-published or a pre-print, or affiliated with the standard academic consensus process for research on the other hand. For different kinds of statements in articles, and for subject areas, WP:CONTEXTMATTERS on what is reliable. I'll give a few examples. For WP:MEDRS and science there is a higher standard. For WP:CVG, there are a lot of sites used for video game articles, used that are probably not going to pass muster for something like politics and general news, where the sources on WP:RSN that are labelled as such, may have caveats for controversial or contentious information.
Anyway, the point is that what is reliable depends on a bunch of factors, but in the end, if it's reliable for general facts, we are going to accept those facts at face value when not contradicted by other sources. The logical justification does not enter into it. Sure, if you had something that seemed obviously wrong and obviously illogical, you could start a discussion and obtain a WP:CONSENSUS to leave out an otherwise apparently reliable source, or to consider an otherwise borderline/questionable source reliable when attributed appropriately, etc. Andre🚐 16:43, 7 September 2022 (UTC)[reply]
Let's say one editor finds a source reliable and another finds it unreliable. In what way is it decided that the source is reliable? Is it a bare appeal to authority, where the only thing looked at is credentials and reputation? Is it just a subjective vote by editors?
Do the standards of accuracy and fact checking include a process whereby the accuracy of the authority's claim is actually checked, or is this process also deferred to an appeal to authority? 2600:4040:A23F:B200:EC55:CED4:84D8:6E8E (talk) 18:12, 7 September 2022 (UTC)[reply]
It's based on a consensus of editors and valid arguments considering the source's track record and reputation in light of policy. Editors may offer evidence such as failed fact checks or reliable studies of the accuracy. See WP:RSP WP:RSN. Examples of sources making statements that turn out to be false may be used in that context. Andre🚐 18:15, 7 September 2022 (UTC)[reply]
So "reliability" has nothing whatsoever to do with whether or not the source actually justifies their argument validly, from the perspective of the editors vetting those sources? If you would give a straight answer you hopefully might begin to see the problem. 2600:4040:A23F:B200:EC55:CED4:84D8:6E8E (talk) 18:42, 7 September 2022 (UTC)[reply]
I believe I have answered that yes, reliability is not based on a singular individual perspective of an editor determining whether individual sources make valid arguments, though it is possible for a consensus of editors on an article or in a subject area to determine that a certain source isn't reliable in context, perhaps based on other RS, or they may as a whole determine that certain sources are or are not generally reliable across the board - which doesn't preclude exceptions if they are merited. I don't see it as a problem. We have a core idea of No original research that addresses this. Maybe if you have a more specific, concrete example of where you've seen a problem, we can dig in that way. Andre🚐 18:57, 7 September 2022 (UTC)[reply]
And the thing the editors are debating has nothing whatsoever to do with if or how a source justifies their argument? How do you not see a problem?
I already showed a specific, concrete example as my first comment. Maybe if you finally address that we can dig in that way.
And I think for the third time, I am not in anyway suggesting that editors produce original research. I am merely suggesting that the sources cited provide justification other than an appeal to authority, otherwise they should not be considered a reliable source, and that that justification should always be cited along with the knowledge claim.
The editors should at the very least be able to cite "X because Y." 2600:4040:A23F:B200:EDA6:D850:D04F:2FA5 (talk) 19:40, 7 September 2022 (UTC)[reply]
Not every statement needs a "because." Sources aren't making arguments, they're reporting facts, and we just reflect statements that are encyclopedic and neutral according to their weight. When it's a contentious argument it may need to be treated carefully. Then we offer a reference so users can see where they came from. I don't see the specific concrete example. By specific and concrete I mean you should give me a real article, a real reference, and the source, and why you think the source needed additional justification in the way that it's being used in the article. Andre🚐 19:46, 7 September 2022 (UTC)[reply]
When a source is claiming a fact, if they are reliable, they will provide their justification for reporting it as a fact. I am really having difficultly understanding why this is hard to grasp.
And I've sourced the article. It's real, the things I referenced are real, and I've explained why the additional justification was needed. Maybe if you can address what I said in my first comment we can move this conversation forward? 2600:4040:A23F:B200:B984:7852:F72:8224 (talk) 20:43, 7 September 2022 (UTC)[reply]
So in the "appeal to authority" article, what is the text you are saying is problematic, and what source specifically are you saying is problematic? Because I am not seeing what you wrote in the article. Maybe the article text has changed. I do see this text, "Some consider that it is used in a cogent form if all sides of a discussion agree on the reliability of the authority in the given context." Which is fine for me - is that the issue? Andre🚐 21:09, 7 September 2022 (UTC)[reply]
Reading the article, do you get the impression that some sources argued that appeals to authority might not be fallacious within logical arguments? 2600:4040:A23F:B200:9C91:98D0:ECB7:E2DF (talk) 23:04, 7 September 2022 (UTC)[reply]
Are you referring to the reference to "VI. Appeal to Authority This is the argumentum ad verecundiam. Such an appeal is not necessarily fallacious." from "Logic and the Common Law Trial"? Andre🚐 23:11, 7 September 2022 (UTC)[reply]
That's a good example. Can you see in the source how it justifies that claim? Can you tell what type of argument the source is talking about in what circumstances? 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 00:09, 8 September 2022 (UTC)[reply]
It mentions witness testimony. It talks about "legitimate expert witnesses," who presumably aren't being used fallaciously. It goes on to talk about "the scientist, like the movie star, regularly lends his or her name to a claim or cause without regard to the relevance of his or her special knowledge-argumentum ad verecundiam-allin the service of the "agenda." So what else is new? "Each of them believed himself to be extremely wise in matters of great importance, because he was skillful in his own art: and this mistake of theirs threw their real wisdom into the shade." All of this convinces me that arguments from authority are reasonable in some contexts. Andre🚐 00:23, 8 September 2022 (UTC)[reply]
Those were justifications for why arguments from authority are fallacies from weak induction but can be persuasive. That text was after "When listed as a fallacy, the appeal to authority is said to be one of weak induction-the premises may lend some support to the conclusion. The question is, how much?"
The only justification is the source quoting the claim, which is an appeal to authority. I don't have access to the text it is quoting so I have no idea what justification that source gives for the claim, if they are talking about logical arguments or legal arguments or rhetorical arguments or dialectic arguments. But this cuts to the heart of the issue -- this could be amended by requiring sources to provide their justification in their claim, so that editors at the very least have to argue about whether the justification supports the claim rather than whether the claim was spoken by an authority.
Elsewhere in the text there is this quote:
"In the case of a strong inductive argument, if the premises are true then the conclusion is only probably (not necessarily) true. In the standard treatments, the fallacies of weak induction include appeal to authority, appeal to ignorance, false cause, slippery slope, and weak analogy."
Which quote should be believed? That appeals to authority are standardly fallacies, or that they are equally historically nonfallacious? Can wikipedia at least require the editors to check if the text itself is at the very least logically sound and consistent? 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 00:49, 8 September 2022 (UTC)[reply]
It's clearly talking about different contexts. Appeal to authority is not always a fallacy. For example when you're qualifying an expert witness. Anyway, from a practical standpoint, nobody else is going to endorse your suggestion to require justification, so may as well drop it. The text IMO is perfectly logically sound. It describes things in different scenarios and contexts, which is OK. A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines.-Emerson Andre🚐 00:53, 8 September 2022 (UTC)[reply]
"But, gentlemen of the jury, the good craftsmen seemed to me to have the same fault as the poets: each of them, because of his success at his craft, thought himself very wise in other most important pursuits, and this error of theirs overshadowed the wisdom they had" -- Plato, The Apology, part of the thing you quoted as convincing you that appeals to authority are justified 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 01:36, 8 September 2022 (UTC)[reply]
Open-source intelligence (OSINT) adds transparency to journalism, and Wikipedia articles do cite sources that make use of that technique, such as Bellingcat (RSP entry). However, since the vast majority of reliable sources do not make use of OSINT, it is simply not feasible for Wikipedia to make the use of OSINT a requirement for a source to be considered reliable. If you would like to start a wiki that exclusively cites OSINT-based sources, you are free to do so. MediaWiki, the software that Wikipedia runs on, is a free and open-source solution that anyone can use.
Also, Wikipedia has an entire article on epistemic certainty. It states, "The philosophical question of whether one can ever be truly certain about anything has been widely debated for centuries. Many proponents of philosophical skepticism deny that certainty is possible, or claim that it is only possible in a priori domains such as logic or mathematics. [...] It is generally accepted today that most of our beliefs are compatible with their falsity and are therefore fallible, although the status of being certain is still often ascribed to a limited range of beliefs (such as "I exist"). The apparent fallibility of our beliefs has led many contemporary philosophers to deny that knowledge requires certainty." Wikipedia relies on verifiability, not truth, and the threshold for verifiability is not set at epistemic certainty. — Newslinger talk 23:37, 7 September 2022 (UTC)[reply]
It is unclear to me why the common response has been that quoting a source's justification is "unfeasible." Why is that so? 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 00:24, 8 September 2022 (UTC)[reply]
Most reliable sources are not based on raw data that is made publicly available. If Wikipedia articles were only allowed to cite sources with accompanying raw data, almost all content on Wikipedia would be eliminated. The resulting drop in topic coverage would make Wikipedia significantly less useful for our readers, which is why the suggestion is not feasible. Wikipedia welcomes reliable sources that use open-source intelligence when these sources are available, with the understanding that these sources only cover a fraction of the topics that Wikipedia covers. — Newslinger talk 01:15, 8 September 2022 (UTC)[reply]
I'm not entirely sure that OSINT is exactly what I'm talking about, but I think I see what you mean. However just because certain sources might not themselves have openly available sources doesn't mean that all sources should be treated as such. For articles like the one I referenced, about logical fallacies, it should theoretically be more than possible to cite the justification that an authority provides for their claims.
Is there any rule anywhere in wikipedia that addresses something of what I am talking about, where certain sources about certain topics that are able to are required to provide the verification for their claims? Or is the reliability of every source ultimately if the claim is spoken by a perceived authority? 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 02:09, 8 September 2022 (UTC)[reply]
Neither the verifiability policy nor the reliable sources guideline requires that. If a secondary source provides the primary sources that were used in its reporting, that adds to its credibility but is not a necessary condition for it to be considered reliable on Wikipedia. The following measures act as safeguards against including inaccurate information:
  • The reliable sources guideline ensures that only sources with a "reputation for fact-checking and accuracy" are cited.
  • The notability guideline requires each article to cite multiple reliable sources that are independent of each other, which helps mitigate the inaccuracy risk associated with any single source.
  • In-text attribution is used in some cases, such as when the available sourcing is inadequate to repeat a claim in Wikipedia's voice. The use of phrases such as "according to", paired with the corresponding citations, increases the certainty of the claim.
These measures do not achieve epistemic certainty on Wikipedia, but epistemic certainty was never a goal of Wikipedia in the first place. Wikipedia is an encyclopedia, a tertiary source based on information from other sources, mainly reliable secondary sources.
When a reliable source does provide primary sources, the primary sources can be cited alongside the secondary source. The use of primary sources on Wikipedia is governed by the WP:PRIMARY policy. — Newslinger talk 02:47, 8 September 2022 (UTC)[reply]
Gotcha. So there is literally no policy in place that requires any editors at any time to ever check if any source has actually verified or supported any claim outside an appeal to authority, even in cases where that verification is readily available within the text provided, and even in the policy known as the "verifiability policy."
And I'm not saying that epistemic certainty will be guaranteed with my suggestion, merely that wikipedia does not theoretically need to limit itself to solely justifying things by an appeal to authority, and that some policy change should be entirely feasible and probably epistemically beneficial. 2600:4040:A23F:B200:F483:370D:2B54:A744 (talk) 03:31, 8 September 2022 (UTC)[reply]
Let's look at a recent featured article: Benedict Joseph Fenwick. For the sources cited in this article, how exactly would you "check if any source has actually verified or supported any claim outside an appeal to authority" to achieve your goal? — Newslinger talk 03:44, 8 September 2022 (UTC)[reply]
Take the claim "Born in Maryland." How does the specific source cited justify that claim? If it is from a diary written by a family member during the time Benedict was alive, that would make it a primary source, and the justification would be that the person was there. If it is an author writing a biography and looking at historical documents, the justification would be secondary. If the source itself merely quotes another source that the editor has not cited, the justification would be an appeal to authority. Right now the editor is not required to cite which of these it is, so I won't know unless I have access to the book. If you get to articles on non-empirical information providing justification becomes a bit different but easier, and many articles do provide the literal logical or mathematical justifications for their claims. This is more the situation I was talking about, where on the page for "Arguments from Authority" claims like "appeals to authority are not necessarily fallacious" should at least be justified outside an appeal to authority, but it can definitely apply to the situation you outlined too. 2600:4040:A23F:B200:FCB3:BBE4:78D2:9AC7 (talk) 04:45, 8 September 2022 (UTC)[reply]

The text "Born in Maryland" is in the lead section of the Benedict Joseph Fenwick article. Although lead sections do not require citations in many cases, the featured article criteria require the corresponding content in the article body to be supported by inline citations where appropriate. For this article, the corresponding text in the article body is "Benedict Joseph Fenwick was born on September 3, 1782, at Beaverdam Manor in Leonardtown, Maryland", which cites "The Organizer of the Church in New England: Bishop Benedict Joseph Fenwick (1782–1846)", a July 1936 publication from the The Catholic Historical Review. This is a secondary source that does not list primary sources to back up most of its claims.

Thanks to The Wikipedia Library, I have access to this source through JSTOR, and I am able to reproduce the part of the source that covers the claim and check that the article text satisfies Wikipedia's verifiability policy:

Benedict Joseph Fenwick was born September 3, 1782, at his father's plantation on Beaver Dam Manor, near Leonardtown, St. Mary's County, Maryland. His father, George Fenwick, a planter and surveyor, who took a notable part in laying out the District of Columbia and the City of Washington, remains a somewhat dim figure. All the clearer is the strong and perhaps decisive influence upon the boy of his mother, Margaret Medley Fenwick.

Lord, Robert H. (July 1936). "The Organizer of the Church in New England: Bishop Benedict Joseph Fenwick (1782–1846)". The Catholic Historical Review. 22 (2). The Catholic University of America Press: 173. JSTOR 25013478.

Although this is enough to establish that the claim "Born in Maryland" is verifiable by Wikipedia's standards, we have yet to "check if any source has actually verified or supported any claim outside an appeal to authority". With this information, how would you verify the source text even further? — Newslinger talk 05:13, 8 September 2022 (UTC)[reply]

Do you honestly not know how to check to what extent a source is primary or secondary? 2600:4040:A23F:B200:BD2B:13A:31F2:5529 (talk) 13:29, 8 September 2022 (UTC)[reply]
I've already established verifiability per Wikipedia's standards, so my job is done. That question is for you to answer. — Newslinger talk 13:32, 8 September 2022 (UTC)[reply]
Okay, look in the book. Does the book say where it got its information from? Cite that. Does it not say where it got its information from? Do not cite it. 2600:4040:A23F:B200:BD2B:13A:31F2:5529 (talk) 13:43, 8 September 2022 (UTC)[reply]
The journal article does not, so according to your suggestion, the Benedict Joseph Fenwick article would not include the location Fenwick was born. If we applied your suggestion to the entire Benedict Joseph Fenwick article, it would be reduced to a fraction of its length, and may not even exist at all. If we applied your suggestion to every Wikipedia article, most of the content on Wikipedia would be removed, and for what benefit? When a corresponding primary source is named, how would you know that the citation is genuine, and how would you verify the primary source's authenticity? Your suggestion doesn't improve epistemic certainty much if all you are demanding is for a primary source to be named.
Your suggestion is not appropriate for an encyclopedia that strives to cover enough topics in sufficient depth to be suitable for general use. Wikipedia aims to maximize benefit for our readers, and drastically reducing Wikipedia's topic coverage in exchange for a marginal increase in epistemic certainty is not a trade-off that would benefit many readers. Wikipedia's purpose is "to benefit readers by acting as a widely accessible and free encyclopedia; a comprehensive written compendium that contains information on all branches of knowledge" (emphasis added). Our readers are our first priority, and our policies and guidelines – including the verifiability policy and the reliable sources guideline – are written to make our content most useful to them. — Newslinger talk 14:09, 8 September 2022 (UTC)[reply]
Or, at the very least, one could presumptively identify the source as "appeal to authority" until its status as a secondary source is confirmed. In this manner no articles need necessarily be shortened, but every reader will have available to them the justification that the citation gives for the claim.
It would also be helpful if you didn't take my argument in the worst light it could be taken and argue only against that. I outlined how this would be much easier with articles regarding logic, but for a reason I cannot understand you have conflated those two situations and are declaring based on that conflation that no justification could ever be useful anywhere in wikipedia.
Otherwise we should rename the "verifiability policy" the "appeal to authority policy," as no actual appeal to verifiability is taking place. 2600:4040:A23F:B200:499B:E1B5:77CA:A4D6 (talk) 14:48, 8 September 2022 (UTC)[reply]
Demonstrating what would happen to an article if we applied your suggestion to a claim that you selected is hardly putting your argument "in the worst light". We are on the talk page of the reliable sources guideline, a guideline that applies to all articles on Wikipedia, not just articles that exclusively contain non-empirical claims. Additionally, your very first comment clearly stated that your suggestion was also to be considered for empirical claims: "Let’s say the claim is that person X was at street Y at Z time. In order for wikipedia to reliably publish this claim, the editor should be forced to also source the justification for it, i.e. 'there is video evidence' or 'multiple first hand accounts place them there.'" Strong proposals for policy and guideline changes consider the impact that the changes would have on the entire encyclopedia, and also consider whether our readers would benefit from those changes. — Newslinger talk 02:47, 9 September 2022 (UTC)[reply]

Editors following this thread may find this article interesting: Is the Sky Blue? How Wikipedia Is Fighting for Facts by Redefining the Truth. End of message, carry on. Gråbergs Gråa Sång (talk) 06:41, 8 September 2022 (UTC)[reply]

Thank you for the link to that article. What I find intersting is the way people use 'appeal to authority' as a way to reject the work of scientists and follow instead what some politician or journalist has written a book about or some presenter on television trying to raise their ratings. The scientists have peer review and the TV presenters are the 'authorities'. When Trump says to inject yourself with bleach to cure Covid and people believe him that is appeal to authority. If Wikipedia tried to verify claims in reliable sources we'd be a worse filter than the sources as we are simply not experts. NadVolum (talk) 10:45, 8 September 2022 (UTC)[reply]
An "appeal to authority" is when someone says "X is true because some authority says it is true." Scientists never use appeals to authority in their work because they are logically fallacious. They can be used in other situations to varying degrees of success, but "authority says X" never guarantees the truth of "X."
Otherwise if every wikipedia editor subjectively found right wing misinformation trustworthy, wikipedia has no actual safeguards protecting against this case. What I'm suggesting is that if one wants to claim "Injecting bleach cures covid," not only would they have to find a reliable source merely claiming that, but that source would themselves have to have presented some form of justification for that claim, probably in the form of evidence.
Look I get that this idea is upsetting to a lot of you so I'll stop responding soon. I was hoping this wouldn't be something fundamental to wikipedia, but my god. These roots go deep. 2600:4040:A23F:B200:BD2B:13A:31F2:5529 (talk) 13:41, 8 September 2022 (UTC)[reply]
I am sorry that I sent you here from the article Talk page, since this turned out to be a sort of universal time-waster.
It's pretty simple. Taken to the extreme, your demands lead to a sort of infinite regress. For every source, you can ask "how does the source know that", then ask the same for the source of the source and so on. You will end up asking if you are a brain in a vat. You have to stop somewhere, and the place where the Wikipedia rules tell you to stop is a reasonable place. If we demanded that everybody who adds a piece of information had to check the source of the source of the source, we would not get anything done, and Wikipedia would consist of a tiny amount of the articles it has now.
You say Scientists never use appeals to authority in their work, but actually, they do trust each other to a certain degree. When a scientist reads a paper by another scientist and finds reasonable content there, they just accept it as true. But they will not trust, say, a creationist, to get it right, because those are known to tell almost nothing but false rumours. Also, if the content is fishy and does not fit other data, people will check it.
Your idealism probably feels good to someone sitting in an ivory tower, but in practice, it cannot be done as extensively as one would do it in an ideal world. You have to balance the accuracy with the effort required to achieve it.
An expert is a person who has found out by his own painful experience all the mistakes that one can make in a very narrow field. [2] Experts know, roughly, what has to be checked and what can be trusted, who knows about what and who does not. They know where the tricky parts are. The "argumentum ad verecundiam" is about unsophisticated, blind acceptance, not about reasonable differentiation between knowledgeable and ignorant/mendacious sources. And a simple rule "do this, and you will end up with the truth" is impossible anyway. You'll always have to think. --Hob Gadling (talk) 14:30, 8 September 2022 (UTC)[reply]
That can be true for some forms of empirical information, but it is not the case for all forms of knowledge. It might be helpful not to argue against the straw man case and instead argue against the situation I presented. Either way, in that straw man case, it can still be good practice to leave at one source if it is an appeal to authority, a secondary, or a primary source. At least the reader will know which information on wikipedia comes from one of those "infinite regress" situations you describe.
And on the page for "Appeal to Authority" it notes how the appeal is never used in science. It's the difference between "E=mc^2 because Einstein said so" and "E=mc^2 because general relativity has been proven through empirical observation." Scientists will never use the first form, as that form is logically fallacious. 2600:4040:A23F:B200:499B:E1B5:77CA:A4D6 (talk) 15:00, 8 September 2022 (UTC)[reply]
the appeal is never used in science Wikipedia never uses it either. I challenge you find an article which says "X is true because authority Y said so." If you find one, that sentence will be removed because that is not how Wikipedia works.
You are the one erecting strawmen here. Wikipedia is never about what is true. People have told you that, again and again, but you do not seem to believe it and instead regard it as a sort of cop-out. "Verifiability" only means that when a Wikipedia article says "Source Y says X", you can verify it by checking the source Y and finding out if it really says X. If the article said "X is true" without any source, and X is not WP:SKYBLUE, then it should be deleted.
This is about the syllogism
  1. "Y says that X is true."
  2. "Y is an expert."
  3. "Therefore X is true."
Wikipedia says 1, and only 1. If 2 is true, Wikipedia is more likely to say 1, so it is also part of the system. But 3 is not. 3 is your invention, your strawman. If there is an argument from authority, it is there because you made it and put it in Wikipedia's mouth.
If there is a consensus among experts that X is true, Wikipedia should not say "Y says that X is true", because that would be misleading. It simply says X is true, but not because of an argument from authority. There is a solid reason for such a consensus, and we have competent users who know that reason. WP:CIR is an important page. I may be wrong, but my impression is that you start from a blank slate and basic axioms and to demand that anybody can derive the desired article content from that and from the world out there, just using logic and a recipe. That is not how it works.
When people here say that yes, we do use appeals to authority, they do not mean 3 above. Using one single authority to determine whether we say X is true or not, should not happen, and those users will not allow that. They mean that we use the consensus. --Hob Gadling (talk) 16:32, 8 September 2022 (UTC)[reply]
Another thing. If the way articles are written did involve checking the logic of the sources, and whether what they say really follows from their sources, by Wikipedia editors, that would be asking too much of many of them. I have met legions of users, new ones and old ones, and there are quite a few who are absolutely hopeless when it comes to logic, especially among the new ones. All of them, including me of course, are hopeless when it comes to most subjects. I can check if a philosopher wrote what we ascribe to him, but I cannot determine whether what he writes follows from his own sources because I lack the expertise. That is the case with most other fields. And that is true for everybody. With your method, we would have to depend on those editors who are experts themselves. Leading to - appeal to authority! But how would we even know whether an editor is really an expert? With your method, the appeal to authority would have to get much worse than it is now. We do need knowledgeable users to check that we do not write complete bollocks, but if somebody had to check the internal logic of the sources, we would need loads of real experts investing loads of time. It's not only not feasible, it would also stop being "wiki", which means "quick". It's just a naive, unrealistic idea born from a huge distance between the thinking and the actual doing. --Hob Gadling (talk) 17:46, 8 September 2022 (UTC)[reply]
In the article on Arguments from Authority, the quote "Historically, opinion on the appeal to authority has been divided: it is listed as a non-fallacious argument as often as a fallacious argument in various sources," is cited by a source which itself is citing another source merely making that claim. There is currently no other justification for this claim appearing on wikipedia.
What is the justification for claiming that "Historically, opinion on the appeal to authority has been divided?" It is merely an author saying that an authority said that. Nowhere in the text does the author ever justify that claim outside that single appeal to authority.
Within that same source there is a quote indicating that appeals to authority are standardly treated as fallacies. Which position should be able to be quoted by wikipedia editors in what way given that the source outlines contrary positions? This seems like a case where editors being able to comprehend logical concepts such as arguments or claims being contradictory might be useful. If editors themselves are not capable of this action, then I still do not see why merely asking them to quote the justification would involve doing any sort of logic as you claim. They are doing the literal exact same thing as before, they are just citing one more point of reference.
Again, I am in no way claiming that editors need to be experts themselves or even need to evaluate the justification themselves. I am making the suggestion that editors cite the justification. In this way, verifiability would be placed at the forefront of wikipedia's values rather than appeal to authority/consensus. 2600:4040:A23F:B200:870:20BC:5611:CDE4 (talk) 19:56, 8 September 2022 (UTC)[reply]
What is the justification for claiming that "Historically, opinion on the appeal to authority has been divided? Now that is something that belongs on the Talk page of that article, not here.
I am making the suggestion that editors cite the justification. Still requires expertise. Justifications can take several pages or whole books. One would need to summarize it. Yes, it is nice when someone does that, but this is about the guideline. If the guideline said one always has to add the justification, everything would become less feasible. Lots of sources would not be useable anymore because the justification cites another, more difficult-to-obtain source. All articles would become longer and more boring. Wikipedia articles are for giving a short summary. Those who are interested can go the sources and look up the justification. --Hob Gadling (talk) 06:48, 9 September 2022 (UTC)[reply]
Well I'm not upset at all. And appeal to authority is in effect how most science manages to proceed. Scientists in general don't pore over what others do to check everything is okay - they depend on publications in reliable sources. The sources are counted reliable becuse they've gained that reputation ensuring work they publish is peer reviewed. That is what Wikipedia is doing but rather less rigorously - we count newspapers reporting on the sayings of politicians for instance without requiring a reputation for peer review! The standard depends on the topic. Do you really think you are qualified to do what you seem to expect others here to do? We are not expected to have any special qualification except a fairly basic competence. NadVolum (talk) 14:31, 8 September 2022 (UTC)[reply]
"And appeal to authority is in effect how most science manages to proceed."
I continue to be flabbergasted at the positions outlined by not only one, but seemingly most wikipedia editors. But it does explain why the issue is an issue in the first place. No, appeals to authority should never be used in hard science.
And surely editors have the capacity to read citations? Again, I am NOT suggesting that editors do original research. Merely that they quote the source's justification for the knowledge. 2600:4040:A23F:B200:499B:E1B5:77CA:A4D6 (talk) 15:07, 8 September 2022 (UTC)[reply]
(Edit conflict; I was writing almost the same thing about infinite regress.) None of this means that we don't apply a critical eye to sources. In my book, a good editor is one who is always looking for the best source for each substantial item, and this investigation can include checking how well the sources justify their claims and what the expertise of their authors is. A historian who wrote a whole article on a historical event is preferred over one who just mentions it in passing, and so on. But once the decision has been made, we give the information and cite the source. It isn't practical to ask for more. The article talk page is there if detailed discussion is needed. Zerotalk 14:49, 8 September 2022 (UTC)[reply]

On a side issue, "Appeal to authority" is a term which has different meanings in different contexts. It's main meaning is from the context of talking about arguments, and in that context it's considered a bad thing/ defective as an argument. Also, in the above threads, it's discussed in the context of being a potential substitute for the scientific method in areas where the scientific method is used, where it is of course also a bad thing. But if you misplace the term into other realms, it can applied to considering something coming from a trusted, proven, careful source that is expert in that area to be accepted as fact based on who it came from. And that is how much of the world successfully operates, and there's nothing wrong with it.

2600 you may be arguing against the other extreme which is blindly accepting something simply because it meets Wikipedia's definition of a "Reliable source" despite the fact that those rules often don't select for actual reliability. Wikipedia could use some evolution there. But you trying to help in that area by promoting requiring the procedure that you specify is not an effective way to start. And to try to do it in just a talk page conversation without learning even the basics of how the alternate universe of Wikipedia operates (which I don't think that you have) is only going to be frustrating for you and the others involved in the conversation. If my/that guess is right, may I suggest doing a thorough read of WP:V and also WP:NOR, in the context that they are on even the shortest list of Wikipedia's most influential policies. Also look at a sampling RFC's most of which including sourcing questions. Your proposed procedure is really only applicable to areas of objective fact and I think that would see that only a tiny fraction of sourcing questions involve that. Sincerely, North8000 (talk) 20:08, 8 September 2022 (UTC)[reply]

"because it meets Wikipedia's definition of a 'Reliable source' despite the fact that those rules often don't select for actual reliability. Wikipedia could use some evolution there."
Thank you for letting me have some input. All the best. 2600:4040:A23F:B200:F4A0:E9EB:101F:12F3 (talk) 21:10, 8 September 2022 (UTC)[reply]

Emails as references

Is it possible for an email by a reliable scholar to be used as a reference in an article; particularly for resolving ambiguity about an earlier, already published work of his/her? I am almost certain that i have come across a couple of such references in the past, but unfortunately i cannot recall exactly in which articles i saw them. I initially asked RoySmith via email about this, and he forwarded me here. I thought there was no guideline or policy that explicitly touches upon this subject; however, having been forwarded here, i noticed that there is some relevant information in the FAQ template, at the top of this page. Specifically we read:

  • Is personal communication from an expert a reliable source?
No. It is not good enough for you to talk to an expert in person or by telephone, or to have a written letter, e-mail message, or text message from a source. Reliable sources must be published.

Furthermore, the information page WP:PUBLISH – that is wikilinked above – includes the following:

  • All reliable sources must be both published and accessible to at least some people[; ... however, the definitions of these terms] are separate from the idea of "reliable". [...] All reliable sources are published, but not all published sources are reliable for encyclopedic purposes. [...] It is necessary for the information to be made available to the public in general, not just to individuals or selected groups of people. [... An example would be a] broadcast email, including email-lists if they are archived and public—but not email messages or other forms of personal communication sent only to you or a small number of people[.]

The idea of reliability is not really an issue here, as self-published sources by an established expert on a subject matter, whose work in the relevant field has been previously published by reliable, independent publications can be considered reliable; per WP:RSSELF. The necessity of publication and accessibility could also be addressed if the email was to be published in the talk page of the corresponding article, that is accessible to everyone. The only real issue that i see, is how to provide confirmation for the authenticity of the email, in order to show that it indeed originates from the established expert in question? A possible solution to this could be a direct communication of the expert with the Wikipedia:Volunteer Response Team, explaining his/her view about a topic or reporting content that is erroneously attributed to him/her (already covered by the team's activities via reports of article errors); then the confirmation of his/her identity by an agent (assuming this is even possible), and the publication of the email and the respective VRT ticket number in the talk page of an article. As for the article itself, a properly written explanatory footnote meant to clarify an ambiguity of a published work that is referenced, would be sufficient in my opinion. Demetrios1993 (talk) 23:03, 9 September 2022 (UTC)[reply]

There's a related conversation about being published above - in my opinion, no, it would not be published if it was a private email. However, perhaps the policy should be clarified because, if published just means "made available to any subset of the public," I guess an email would qualify. In my opinion though, a test of being published is whether it was made available to the general public at large, either through a posted website, publication of a book or journal or article or periodical etc, or arguably, a public sign or public archive, but I do not think a private email should be included. Andre🚐 23:19, 9 September 2022 (UTC)[reply]

Daily Mail and sports

It seems absurd that I can't add a section to an article about a current event that dozens of association football sources are talking about [3][4][5][6][7][8][9] because they are all citing the Daily Mail, and Daily Mail is considered unreliable by WP. Clearly the rest of the published world does not agree with WP in that regard, at least not on the topic of sports. - Keith D. Tyler 05:03, 13 September 2022 (UTC)[reply]

This is the talkpage for discussing changes to a specific WP-guideline. At Wikipedia:Reliable sources/Noticeboard you'll find the page for "posting questions regarding whether particular sources are reliable in context." Gråbergs Gråa Sång (talk) 09:12, 13 September 2022 (UTC)[reply]
This process is extremely obfuscated, as if to dissuade such discussion. There should be a discussion page for every source, not just an endless unorganized list of archived talk pages for all sources. I can't find the link where it is considered "correct" to discuss a given source. Keith D. Tyler 17:52, 13 September 2022 (UTC)[reply]
WP:RSN is usually the place, but sometimes the talkpage of the article where you want to use it can be a good place. You can find links to old discussions about DM at WP:DAILYMAIL (More than 50, from 2007 and onwards). That link goes to a page which is a list of sources that has been discussed multiple times, usually at RSN. Gråbergs Gråa Sång (talk) 18:21, 13 September 2022 (UTC)[reply]
Its been brought up dozens of times at the RSN. Ultimately the Daily Mail is incredibly libellous and allowing it even for something seemingly trivial/non-controversial such as football would open the floodgates. DM is just not a good source period. It's not even worth the risk of some of the legal issues wiki could get embroiled with because of DM's dangerous reporting. >> Lil-unique1 (talk) — 19:20, 13 September 2022 (UTC)[reply]

WP:HEADLINES

While the premise of this guideline is clear, I think it should be expanded to make explicit that the rule of thump here applies to all kinds of titles, not just in the news. The way it is currently outlined, the emphasis is on news headlines, but books titles are just as subject to sensationalization by publishers as news headlines are by editors. Similarly with academic writing, titles often oversimplify and eschew the most precise terminology. I think book and academic titles should be explicitly incorporated into the text, if not already understood, to make it clear that all headlines and titles are fundamentally flawed as an informational resource. There is of course a due weight argument underpinning all of this in that a title appears just once, while a work's subject will be mentioned again and again within the body of the work, thereby making the mentions in the body automatically more relevant than any singular titular mention. However, I was recently engaged in a discussion where an editor did not appreciate this implicitly. Hence my sense that this to be made explicit. Iskandar323 (talk) 15:27, 19 September 2022 (UTC)[reply]

Correct spelling of name

How can I or somebody correct the spelling of a name on a Wikipedia page? Page: the lovin' spoonful Under past members: David Jayco should be David Jayko Thank you, Roxane Rclbuss (talk) 19:31, 19 September 2022 (UTC)[reply]

That would be every mention of him in the article, they're all Jayco. You can, see WP:TUTORIAL, but I'm not sure you should. Afaict, everyone Google has heard of spells it like WP does.[10]. Could be citogenesis, of course, but I'd like to see a WP:RS that supports the spelling. Gråbergs Gråa Sång (talk) 20:15, 19 September 2022 (UTC)[reply]

Semi-protected edit request on 26 September 2022

Dear Wikipedia, I am Fiery Cushman's mother and I am asking that the Entry for Fiery Cushman be changed from "his mother taught psychology at American University." to "his mother was a clinical child psychologist practicing in Washington, DC and Maryland." While I did teach child psychology for a semester at American University that was not my primary occupation, and I was not on the faculty there. Sincerely, Lynnwood Andrews 2601:18C:4280:7130:1C2E:6936:44B6:5B44 (talk) 23:00, 26 September 2022 (UTC)[reply]

 Not done: this is the talk page for discussing improvements to the page Wikipedia:Reliable sources. Please make your request at the talk page for the article concerned. MadGuy7023 (talk) 23:04, 26 September 2022 (UTC)[reply]

"Generally acceptable" USERG

A concern was brought up at WT:ALBUMS regarding a difference in language around WP:USERG between here and the project page. Over here it says USERG are "generally unacceptable", whereas over there we have it as "should never be used". A counterpoint was made in that discussion that "generally unacceptable" may be too soft and could lead to arguments about how any given source should be a specifically acceptable exception. I happen to agree with that point and think changing the phrasing here could be helpful in avoiding that. Thoughts? QuietHere (talk) 15:20, 30 September 2022 (UTC)[reply]

I was the other one thinking it sounded kind of soft. In what scenario do we actually cite USERG content? I know it's acceptable when a reliable source reports on USERG content (for example, the New York Times does a story on users "review bombing" music/film/games on Metacritic's user reviews. But in scenarios like that, you'd be citing NYT, not MC. Sergecross73 msg me 15:27, 30 September 2022 (UTC)[reply]
The wording in question at the discussion in WT:ALBUMS - user-generated content should never be used (emphasis added) - appears to be addressed already in the FAQ at the top of this talk page:
Are there sources that are "always reliable" or sources that are "always unreliable"?
No. The reliability of a source is entirely dependent on the context of the situation, and the statement it is being used to support. Some sources are generally better than others, but reliability is always contextual. (emphasis added)
Is there another proposed re-phrasing we should consider? --N8wilson 🔔 15:43, 30 September 2022 (UTC)[reply]
The problem I see with Wikipedia:WikiProject_Albums/Sources#Unreliable_sources is that the way it summarizes WP:USERG fails to make it clear that Wikipedia:Reliable sources, which contains WP:USERG, says that if a claim on a site with mostly user-generated content can be verified as belonging to a particular human author (not just a screen name) then it can be treated as a self-published source. Self-published sources can be used if the restrictions are satisfied. The Albums project subpage appears to say user generated sites can never be used, even if the author can be verified and the author satisfies the criteria for using a self-published source, which is not true. Jc3s5h (talk) 16:28, 30 September 2022 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Reliable_sources&oldid=1113264296"