Wikipedia talk:Verifiability

Citing a chapter

WP:BURDEN currently advises this:

Cite the source clearly and precisely (specifying page, section, or such divisions as may be appropriate).

I suggest that it say this:

Cite the source clearly and precisely (specifying page, section, chapter, or such divisions as may be appropriate).

Chapters can be particularly convenient in unpaginated ebooks, and not all chapters are further divided into sections. Citing chapters as an alternative to individual pages has long been accepted at Wikipedia:Citing sources#Books.

Naturally, it isn't always the best alternative. Just like it's sometimes appropriate to cite entire books ("Alice Expert says the Sun is really big" with a citation to her book, The Sun is Really Big), it is sometimes appropriate to cite a whole chapter, and it is sometimes appropriate to cite individual pages. The choice should depend upon whether you're summarizing a larger point, or just pulling an individual fact out of one sentence in the book. But I think that it is generally better to encourage people to consider citing a specific chapter than otherwise. Also, science-related books and reference works often have individually authored chapters, so editors should very frequently be naming the chapter and its authors regardless of whether a specific page number is also cited.

What do you think? Would this be an improvement, or not? WhatamIdoing (talk) 15:57, 15 May 2020 (UTC)

WP:BEBOLD. EEng 18:18, 15 May 2020 (UTC)

That would be a huge step backwards. Ebooks have alternative ways of identifying text, and it's often possible to find a page number for the text via Google Books or Amazon; if not WP:RX can usually help. And it's never appropriate to cite a whole book. Where does the idea come from that that would be enough? SarahSV ^(talk) 18:27, 15 May 2020 (UTC)

He's only suggesting adding the word chapter. What does your comment have to do with that? And while it's a narrow use case, citing a whole book isn't inconceivable (The first Random House edition ran to 1203 pages might be cited to the whole book. Like I said, it's a narrow use case.) EEng 18:39, 15 May 2020 (UTC)

It doesn't make any sense to write "Cite the source clearly and precisely", then suggest that a book chapter might be sufficient. Also, WAID, this is just odd. You wrote: "science-related books and reference works often have individually authored chapters, so editors should very frequently be naming the chapter and its authors regardless of whether a specific page number is also cited." Yes, of course, and we do and not only in science. Edited volumes are common. We cite author, chapter title, editors, book title, page range of the chapter, and the specific page reference. SarahSV ^(talk) 19:37, 15 May 2020 (UTC); 19:40, 15 May 2020 (UTC)

Chapters are already OK if appropriate: they're included in the phrase "or such divisions as may be appropriate". Sometimes they are appropriate. For example, a Wikipedia article cites an equation, and when you go to the source, you find that the equation itself, the definitions of the variables, and caveats about when it is valid are spread throughout a chapter; if you cited each page that was relevant, you'd cite the majority of pages in the chapter.

I prefer not to specifically mention chapters, because cases where just a chapter is appropriate are unusual, but we certainly shouldn't rule them out. Jc3s5h (talk) 20:00, 15 May 2020 (UTC)

Absolutely. Another example: article says X has been the subject of many popular songs, literary works, films, and dramas, you might naturally cite a book on X, and specifically the chapter in that book on X in popcult, but no particular page. I see what you mean about the possibility that mentioning chapters in the guideline might tempt people to cite them where a pg# would be more appropriate, but by not mentioning it you risk that, where a pg# isn't available/appropriate, an editor may fail to realize that a chapter# is at least helpful. EEng 21:40, 15 May 2020 (UTC)

You can cite a page range for the examples you give. What book would have no page numbers? It won't only tempt people to cite chapters; they'll do it. The requirement for page numbers for books has been in the policy for over 12 years (added here), but there was a requirement for page numbers on request before that. It would just be odd to remove it. SarahSV ^(talk) 03:43, 16 May 2020 (UTC)

Why would I ponderously cite pp. 434–513 when I can just cite ch. 7? EEng 14:59, 16 May 2020 (UTC)

There is no requirement to cite pages; other ways of specifying the portion of a work that supports a claim are clearly permissible. Since the requirement does not exist, adding "chapter" will not remove it. Jc3s5h (talk) 04:05, 16 May 2020 (UTC)

It says "specifying page, section, or such divisions as may be appropriate". It goes without saying that pages refer to books and other media that have them. If you're looking at a page, and you extract information from it to add to WP, why would you only cite the chapter? Sometimes Google hides the numbers but there are usually ways round that; Amazon usually offers a limited number; or someone at RX will help. "Clearly and precisely" does not mean chapters only. SarahSV ^(talk) 04:14, 16 May 2020 (UTC)

What if I'm not looking at "a page", but rather, the material to support the passage in the Wikipedia article is spread throughout a chapter? Suppose I want to summarize in a table various ways of writing angles in the field of land surveying, and I cite chapter 8, "Angles, bearings, and azimuths", of Wolf & Brinker's Elementary Surveying (n.p.: Harper Collins, 1989). The information about how to represent angles is spread all through the chapter. Jc3s5h (talk) 13:12, 16 May 2020 (UTC)

• Support this, it's a good addition and an increase in specificity when page numbers don't cut it, or in addition to those. Headbomb {t · c · p · b} 18:53, 15 May 2020 (UTC)
• Sort of support, but what we really want is for citations to be as narrow as practical. If it appears on one page or a few pages, give the page number(s). If it is spread through a chapter, give the chapter, and so on. The qualifier passim is a standard way to indicate that the information is distributed throughout the division indicated. Zero^talk 13:26, 16 May 2020 (UTC)
• Support I think this suggestion is useful. I appreciate that a chapter is already a "section" or "division" but it is a common one and the chapter title can often be extremely useful. I'm thinking of those huge professional medical books with a couple of editors and chapters written by selected experts: each chapter is similar to a comprehensive review paper with their own topic and authors. So the chapter is an important feature and worth explicitly encouraging. Although this doesn't discourage editors from being even more specific, sometimes the sentence or clause really is sourced to information dispersed throughout a chapter, and one can be too specific. The chapter can also be useful for when page numbers are unhelpful, such as when you have access to a different edition of the book. -- Colin°^Talk 10:15, 17 May 2020 (UTC)

How about if we said

Cite the source clearly and precisely, ideally giving page number(s) – though sometimes specifying a section, chapter, or other division may be appropriate instead.

EEng 15:05, 16 May 2020 (UTC)

It doesn't make sense to say "clearly and precisely" then "but sometimes not clearly and precisely; sometimes (we won't say when) citing a chapter would be okay". Even when an entire chapter covers something, you can still give a page range, or p. 60ff, or chapter 3, p. 60ff. But if the sourcing policy says editors may cite books but not give page numbers, that's what they will do, and when someone requests a page number, they will point to this policy and say page numbers aren't required. SarahSV ^(talk) 22:07, 16 May 2020 (UTC)

Gosh, you're right! I've struck the precisely, leaving the ideally giving page number(s) to make it clear to the reader what the, well, ideal is. As to sometimes (we won't say when) citing a chapter would be okay you're being silly. We leave a lot of things to editor judgment. Under my proposal we make it clear that we really want a page number if possible, but where that isn't possible or doesn't make sense a chapter or section would be good too. EEng 04:03, 18 May 2020 (UTC)

Just in case the discussion comes back to this side proposal, I've added the work instead for clarity. EEng 05:13, 31 May 2020 (UTC)

I begin to wonder how long the chapters are in your typical book.

I think there is no practical difference between saying "Chapter 7" and "pages 234–241", when chapter 7 happens to begin on page 234 and end on page 241.

I do think that specifying an ending point is often preferable to saying page 234ff (which means "and the following", which could include any page(s) after that 234 in the entire book). WhatamIdoing (talk) 19:59, 17 May 2020 (UTC)

A practical difference between citing a chapter and a page range that matches the chapter is that reference books often have many editions, and editions that are near to each other in order of publication often have the same, or nearly the same, chapters but different pagination. Especially if the title of the chapter is given, a reader who can't find the cited edition may be able to verify the claim in a similar edition. Jc3s5h (talk) 20:29, 17 May 2020 (UTC)

It looks like there is general support for including the word "chapter" in this sentence. Would someone like to add that to the policy? WhatamIdoing (talk) 18:43, 30 May 2020 (UTC)

I agree, and I'll be happy to do it, but since feeling has run high on this let's wait to see if everyone's in agreement, if not with the proposal itself at least with the fact that consensus has been reached. EEng 03:20, 31 May 2020 (UTC)

Oh wait. Are we talking about the original proposal, or mine a bit above here? But now that I look at mine I don't like it so much, so I guess it's the original we want? EEng 03:34, 31 May 2020 (UTC)

If you're going to remove that page numbers are expected, that will need an RfC. This is a core content policy. That has been in the policy for over 12 years, and in other forms before that. SarahSV ^(talk) 05:05, 31 May 2020 (UTC)

The current guideline doesn't say page numbers are expected but rather specifying page, section, or such divisions as may be appropriate so I don't know what you're talking about. EEng 05:13, 31 May 2020 (UTC)

To summarize... The old text was

(old) Cite the source clearly and precisely (specifying page, section, or such divisions as may be appropriate).

The original proposal was to change this to

P1. Cite the source clearly and precisely (specifying page, section, chapter, or such divisions as may be appropriate).

However, I humbly submit that my alternative ...

P2. Cite the source clearly, ideally giving page number(s) – though sometimes a section, chapter, or other division may be appropriate instead.

... might be preferable because it emphasizes that page numbers are "ideal". Pinging Headbomb, Jc3s5h, WhatamIdoing, SlimVirgin, Zero0000, Colin. EEng 15:11, 1 June 2020 (UTC)

• I'm happy with either of the proposals. I do not see either of them as removing an expectation for page numbers (when page numbers exist and are a suitable way of identifying the content, of course). WhatamIdoing (talk) 15:46, 1 June 2020 (UTC)

  Yeah, we're past the red-herring idea that a anything we're discussing devalues page numbers. EEng 16:12, 1 June 2020 (UTC)

  @EEng, maybe you should just make a change (anything that seems reasonable to you) as a starting point. The first edit doesn't have to be the ideal one. WhatamIdoing (talk) 16:30, 24 June 2020 (UTC)

  [1] EEng 17:09, 24 June 2020 (UTC)

Not easily accessible sources - published sources that are not easily accessible as they once were

I propose this addendum: "This also means references to sources that had once been accessible, yet are not anymore, cannot be removed solely based on that reason. You have to prove the reference have been unpublished by its original author due to falsity or privacy issues, or is unreliable; in order to remove that reference."
The reason for this proposal is that websites and book copies go exhausted, websites go stale; there have been examples of Wikipedians trying to delete information because of removal, rewriting the history in effect. If we allow published sources that are not easily accessible, we also have to accept the situation of references going dead during the lifetime of this encyclopedia. This never happens with paper encyclopedias because they are fixed on paper and circulated; Wikipedia, however, constantly changes and its main form of consumption is electronic, via interwebs. Erkin Alp Güney 20:29, 3 June 2020 (UTC)

• I think "prove" is too strong a word. The level of standard for accepting an offline source is WP:AGF: we generally accept these sources without proof unless 1) they are used to support an extraordinary claim; or 2) the person who added them is a habitual liar. Reasonable doubt (i.e. challenging the source not solely because it is inaccessible, but for more substantial reasons) is sufficient to get an offline source removed. -- King of ♥ ♦ ♣ ♠ 21:36, 3 June 2020 (UTC)

This is fundamentally structurally flawed. Wp:ver does not make any requirement that restricts the presence of a reference. It makes reference requirement for the presence of text. There is nothing in wp:ver mandating removal of references. Sincerely, North8000 (talk) 01:06, 4 June 2020 (UTC)

This is wrong. Sources must be both Published and Accessible. It is no longer possible to verify that the cited source said ____ when every single copy of that source has been lost. The purpose of our verifiability rules is to make sure that someone can check the source; if no copy of source exists, then the purpose is not met.

It sounds like you may find the information at WP:DEADREF to be relevant. WhatamIdoing (talk) 02:43, 4 June 2020 (UTC)

Current rules allow sources that are legally publicly available but hard to access for some people (e.g. physically kept in a limited area, but accessible on demand). My proposal furthers this, and enables referencing "a last remaining copy" (after all, an encyclopedia is a historical artifact) or "a dynamic website gone down with no archives". Erkin Alp Güney 06:01, 4 June 2020 (UTC)

Erkin Alp Güney, I thought that I understood your proposal, and now I think I don't. Imagine that I have cited a paragraph to "a dynamic website gone down with no archives". I'm not quite sure exactly what the "dynamic" part means, but I assume that nobody has a copy of the webpage that I was citing. I went to that website, copied some numbers out of it ("As of April 2020, 4 people in this city had tested positive for COVID-19"). Now the website is gone. Nobody has a copy. Another editor thinks that number is suspiciously low, and thinks I meant to type "40" instead of "4". How will you verify whether I typed the correct information in the Wikipedia article? WhatamIdoing (talk) 15:33, 15 June 2020 (UTC)

Dynamic website means server-side content generation, depending on user profile and navigation state. How deep you verify accessibility depends on what grounds you get into doubt. If it is inconsistency of added content with what cited source actually claims, for a reference that was accessible at the time when added but not anymore, you have to do that in a reasonable time after it has been first added; after that, it is basically history. If you think that source has never existed, or a falsification for aforementioned sources, you also have to verify the fact for removing it, just like you did for insertion. Erkin Alp Güney 18:34, 15 June 2020 (UTC)

@Erkinalp9035, this may be a timely question. I'm looking at some websites that report the prices of some drugs. These websites do a quick search of several independent databases (e.g., looking up the price on various sellers' websites). This means that when I add the source on Monday, it will say that the usual price is (e.g.,) $23. When I go back on Friday, to make sure that I got it right, it may say that the price is $22.

Should people be using that kind of website at all? There is, effectively, no way for anyone to check that the price was actually what I said that it was, at the time that I added it. The most you could do is check what today's price is (which is probably going to be similar, but will probably not be the same). WhatamIdoing (talk) 16:37, 24 June 2020 (UTC)

Again, the reemedy is not removal of the reference. It is determination that the reference does not fulfill the verifiability requirement for the text, and eventually removal of the text. North8000 (talk) 12:54, 4 June 2020 (UTC)

Research Showcase on verifiability next week

Hello, all,

Next week's mw:Research Showcase is on the theme of "Credibility and Verifiability". It will be available at https://www.youtube.com/watch?v=GS9Jc3IFhVQ It will start on Wednesday, 17 June 2020 at 16:30 UTC (12:30 p.m. EDT), but it's recorded, so you can watch it later if you want to.

The first presentation is on a project called NewsQ, and I think will mostly about news and something like the circular sourcing problem. We are also promised some comments on what the Showcase's notes called the "US Perennial Sources list". It will be related to this paper.

The second presentation is called "Quantifying Engagement with Citations on Wikipedia", aka "Does anyone actually look at the inline citations?" This paper suggests that the answer is "mostly no", especially when the article is already well-developed.

Please share this with other editors who might be interested. Whatamidoing (WMF) (talk) 23:13, 12 June 2020 (UTC)

We should ban citing some social media: specifically TikTok but probably other garbage as well

I have been very concerned about citing social media like Facebook for 1.) the self-published nature of it, 2.) the frequently wildly inaccurate mis- and disinformation that it spreads, and 3.) tacitly encouraging our users to go to walled garden surveillance networks. I have recently seen citations for TikTok and I think this is too far: in addition to the problems that I have already outlined, this is malware from a totalitarian government. Under no circumstances should we encourage or even allow outgoing links to this. I think we should explicitly state that in the verifiability guidelines and have entries on individual apps and sites that are particularly egregious about this and should never be linked (I would definitely argue that Facebook belongs as well: it is totally inappropriate to encourage any outward links to their tracking ad network). ―Justin (koavf)❤T☮C☺M☯ 01:48, 30 June 2020 (UTC)

---

To be clear: This is specifically about TikTok and its unique problems but I situated it in terms of other social media as well. Please don't allow me to distract from TikTok in particular. ―Justin (koavf)❤T☮C☺M☯ 04:51, 1 July 2020 (UTC)

---

Would agree but it won't happen. Long ago recommendation to me was.. avoid pop culture junk and focus on academic topics that educate readers. Don't see Facebook used in ancient history or scientific articles.--Moxy 🍁 02:04, 30 June 2020 (UTC)

As to TikTok I would never ever see a case we'd ever need to link to it - can a TikTok post contain *anything* encyclopedicly useful? The new privacy concerns raised would make it worthwhile to eliminate it via blacklisting. Facebook's a different beast as there are informational posts made through there though they will nearly always be primary sourcing. Excluding Facebook would be too wide a net as an immediate issue, and if we start issuing concerns about linking to tracking ad network, that would make almost any offsite linking including to most reliable sources a problem. So this is really not an option. --Masem (t) 02:06, 30 June 2020 (UTC)

Masem, I have seen it cited on Nikki Blonsky. Very inappropriate and is basically posting a link to malware. ―Justin (koavf)❤T☮C☺M☯ 01:36, 1 July 2020 (UTC)

I believe this diff is what you're talking about which is extremely iffy. We do want self-sourced statements about "coming out" for sexual identity, but I would absolutely say its OR to presume that someoen using the song "Coming Out" via TikTok to do that isn't appropriate. And the weight of the problems of TikTok's security issues would not make it worthwhile to let that link through a blacklist through for this purpose. --Masem (t) 02:25, 1 July 2020 (UTC)

I'd support banning all social media links and making TikTok a BADSITE on account that they're Chicoms. The problem is that Wikipedia only exists for fans to read about their fandom, and for most of them that means using primary sources, especially social media. If this were a concerted effort for knowledge we'd ban all primary sources and force the use of secondary sources but we're not about knowledge, we're about high-minded fandom so SanFran gets donations. Chris Troutman (talk) 02:38, 30 June 2020 (UTC)

I think the subject's social media is acceptable for a limited number of cases, mainly for personal information:

1. Gender identity, sexual orientation, and pronouns, especially for marginally notable people where secondary sources may not regurgitate their self-declarations; and
2. Birth dates, ditto, unless there is evidence they may be lying.

King of ♥ ♦ ♣ ♠ 02:54, 1 July 2020 (UTC)

King of Hearts, Including social media profiles that are spyware/malware? ―Justin (koavf)❤T☮C☺M☯ 04:52, 1 July 2020 (UTC)

If you wanted to make this about TikTok, then you should have led with that and not conflated it with general comments about the unreliability of social media. -- King of ♥ ♦ ♣ ♠ 13:48, 1 July 2020 (UTC)

If this is just about TikTok, WP:RSN is thataway. --Izno (talk) 13:55, 1 July 2020 (UTC)

"Wikipedia:UNSOURCED" listed at Redirects for discussion

A discussion is taking place to address the redirect Wikipedia:UNSOURCED. The discussion will occur at Wikipedia:Redirects for discussion/Log/2020 July 1#Wikipedia:UNSOURCED until a consensus is reached, and readers of this page are welcome to contribute to the discussion. Adam9007 (talk) 17:10, 1 July 2020 (UTC)

AfterEllen as a reliable source

Opinions are needed on the following matter: Wikipedia:Reliable sources/Noticeboard#AfterEllen. A permalink for it is here. Flyer22 Frozen (talk) 00:22, 3 July 2020 (UTC)

Bringing this article in line with Wikipedia:Identifying and using self-published works

This article is at odds with Wikipedia:Identifying and using self-published works. I edited this article to bring them in line. MichaelBluejay (talk) 15:23, 6 July 2020 (UTC)

• Reverted. The policy already tells people to read WP:RS, so repeating what RS says is duplicative bloat.

That said... if there is some sort of conflict between this policy and a guideline, then the better approach would be to amend the guideline to match this policy. WP:V is a core policy after all. It should take precedence. Blueboar (talk) 15:32, 6 July 2020 (UTC)

... Or to incorporate content from that guideline if we believe it appropriate for the policy, after reaching consensus. --Izno (talk) 16:14, 6 July 2020 (UTC)

MichaelBluejay made a number of undiscussed changes to WP:USESPS before coming here to "bring them in line". Schazjmd (talk) 18:25, 6 July 2020 (UTC)

First, I did try to discuss here, but I apparently didn't hit Submit after my last Preview, sorry. Yes, I edited Wikipedia:Identifying and using self-published works also, because it was internally inconsistent with itself and disorganized. If anyone doesn't like those edits, let's discuss them there. As for this article, editors need a concise definition of SPS and concise guidelines for when they're acceptable and when they're not, and that's what's missing here. Also, it's more accurate and appropriate to class SPS as "May or may not be reliable" rather than "Usually not reliable." Indeed, Wikipedia:Identifying and using self-published works characterized SPS as such in several places (before my edits). MichaelBluejay (talk) 18:51, 6 July 2020 (UTC)

Blueboar is correct: there is a conflict between this article and WP:USINGSPS, including the WP:LEADSENTENCE. Afaic, the definition there is clearly mistaken, and I've opened the discussion WT:USINGSPS#Definition is misleading on its talk page. In my opinion, the discussion here is dependent upon prior resolution there and cannot profitably go forward without it, or the discussions should be merged in an appropriate venue. Michaelbluejay cannot be faulted for attempting to foster consistency among policy pages, so discussion following this recent series of changes will hopefully lead to improvements. Mathglot (talk) 19:37, 7 July 2020 (UTC)

I agree, WP:USINGSPS should be cleaned up before bringing this article in line with it. I'll check the discussion there and see if there's anything I can contribute. -MichaelBluejay (talk) 10:29, 8 July 2020 (UTC)

@Michaelbluejay, at least for right now, I've reverted your re-ordering, because it made the changes seem more significant than what they really were. Here's a decent diff for anyone who is interested in seeing how few material changes he made to the wording. WhatamIdoing (talk) 22:10, 8 July 2020 (UTC)

The "bringing in line with" choice of words sort of confused matters. It sort of implies that making a core policy become consistent with a supplementary page is itself a reason to change the core policy.North8000 (talk) 12:49, 8 July 2020 (UTC)

• The point of “supplementary pages” is to EXPAND on concepts that are presented in the policy or guideline. That means a) the supplement should conform to the policy (not the other way around), and b) the supplement will contain DETAILS that are NOT necessarily included in the policy. That said... I welcome the opportunity to bring the supplement back into line with this policy. Blueboar (talk) 14:47, 8 July 2020 (UTC)
• The actual policy is that you make all the conflicting pages reflect the current community consensus, regardless of which pages say "policy" or "guideline" or "supplement" or anything else at the top. Also, relevantly, that page explains a concept that is found in multiple policies and guidelines, not just this one. We can't bring a page "in line" with just one of them. The definition of a self-published source needs to be the same in WP:V, WP:NPOV, WP:BLP, WP:N, WP:NOT, and more. The community will never stand for us claiming that a bio about an employee is self-published if it's at a government agency's website but non-self-published if it's at a university website (even though WP:NPROF would appreciate it), or self-published on a small business's website but non-self-published on a large business's website (even though that was tried on this talk page years ago). WhatamIdoing (talk) 22:15, 8 July 2020 (UTC)

  A discussion about changing the definition of self-published in any policy page should take place at WP:VPP, not at Wikipedia Talk:USINGSPS. Schazjmd (talk) 22:27, 8 July 2020 (UTC)

  Schazjmd, if there were actually a definition of that term in any policy page, I would probably agree with you (WT:V would be another reasonable location). But there isn't, so it logically follows that there can't be any discussion about changing it. There's no definition in any policy page to talk about changing. WhatamIdoing (talk) 18:26, 9 July 2020 (UTC)

  WhatamIdoing, my point is that a change to a policy (including adding a definition that isn't there) needs to be a community-wide discussion at WP:VPP. Schazjmd (talk) 18:31, 9 July 2020 (UTC)

  It is normal for discussions about this policy to happen on this page. (I haven't seen Michael proposing any changes to the definition, though.) WhatamIdoing (talk) 18:42, 9 July 2020 (UTC)

I don’t think it matters where the discussion is held, but we do want the wider community to be involved... especially if we are contemplating more than minor tweaks )and changing the definition is definitely more than a minor tweak). So... the discussion should probably be at least ADVERTISED at the pump. Blueboar (talk) 19:00, 9 July 2020 (UTC)

• To basically summarize, "self-published" should not only includes the situation where the person that wrote the material also controls the "publication medium" which is what WP:V already gives, but as we're finding out to be clearer, should also cover the situation when the the author controls the publication process. That is, this would be an approach that would cover the Forbes Contributor model that we know there's no editorial involvement between the author clicking "submit" and the appearance on Forbes.com - we currently don't call that self-published because, well, Forbes.com owns the site, but if we extend the idea to cover the process and not just the medium, then it covers that and things like open journals w/ no peer-review process, Medium.com and a bunch more questionalbe sites cleanly. --Masem (t) 00:16, 9 July 2020 (UTC)
  • I call those Forbes Contributor blog posts self-published, and I know that I'm not the only editor with that view. Whenever the same person/organization/entity controls what's written and when/whether/how it sees the light of day, then it's self-published.
  • https://www.lexico.com/definition/self-publish specifies "independently and at one's own expense", which is a formulation that may appeal to some editors, because it includes the Forbes Contributors/Medium/HuffPo bloggers, the ebook and invented-my-own-publishing-house publishers, and the business/org websites, but could be construed to exclude government websites and publications (because although the government agencies publish their materials "independently" in some sense, the taxpayers pay the expenses). WhatamIdoing (talk) 18:41, 9 July 2020 (UTC)

Just a quick note about the Forbes Contributor model... While it may be defined as a form of "self-publishing", it's important to keep in mind that editors do review each post after publishing, make minor tweaks, and occasionally remove them. There is some editor involvement within a certain period of time. Also, there is a threshold associated with a writer becoming a contributor in the first place. They are vetted based on their expertise and reputation. I get that the process of self-publishing is being called out here, but it should be clear that they are not in the same category as some enthusiast who creates their own site or publishes through a social media platform. --GoneIn60 (talk) 19:40, 13 July 2020 (UTC)

While the Forbes.com contributor model uses a nominal amount of editor involvement, the lack of editorial oversight associated with Forbes.com contributors is well-documented by the Columbia Journalism Review, the Poynter Institute, BuzzFeed News, and The Outline. Since Forbes.com contributors rarely have their articles reviewed, these articles receive a similar level of editorial oversight as self-published sources; as a result, Forbes.com contributors are considered equivalent to self-published sources. — Newslinger talk 21:43, 22 July 2020 (UTC)

Let's make some progress on this. Can someone suggest a good workable definition of SPS? Currently, this policy doesn't offer any definition, which is not helpful to editors seeking guidance. -MichaelBluejay (talk) 17:07, 10 July 2020 (UTC)

Well, going off the way I'm thinking, possible language would be Self-published works are those where the author of the content has direct control of publication of that content. This may be when they directly through their own web site or social media account they control, or when they publish through a third-party publisher (such as Kindle Direct Publishing) or website (such as contributors on Forbes.com) without any editorial checks prior to publication. (This is just a staring point). I feel this doesn't disrupt how SPS are currently used nor affect how past RS decision on SPS would affected (eg cases like SPLC , Quackwatch, etc.) --Masem (t) 17:15, 10 July 2020 (UTC)

I'm not sure this is easier to understand/apply than the current definition, though I know the concern over the current definition isn't that it's unclear, it's that it's not accurate. What if we add a qualifier ("SPS are *generally* sources in which the author and publisher are the same"), keep the list of examples of SPS and non-SPS (which is the most helpful bit), and expand that list to deal with concerns about sources not covered by the stricter definition?

Also, I'd like to hear examples of sources which fail the current, strict definition, and why. -MichaelBluejay (talk) 13:42, 11 July 2020 (UTC)

I think that the problem isn't "failing" the definition, but people not understanding what it means for the author to be the publisher. WhatamIdoing (talk) 17:24, 11 July 2020 (UTC)

The question is: Is there a way to SUCCINCTLY explain what it means for the author to be the publisher? Blueboar (talk) 19:09, 11 July 2020 (UTC)

I don't know, but whether there is or not, a few acceptable/unacceptable examples will go a long way towards clarifying. Since we haven't made much progress, I will be bold and edit as follows: (1) Recast SPS as "may or may not be acceptable" (more in line with WP:USINGSPS), and (2) provide a starting definition (not really acceptable to not define SPS at all!), and (3) provide some acceptable/unacceptable examples. -17:50, 12 July 2020 (UTC)

The main issue is that the average editor probably presumes that when we say "author is same as publisher" that publisher refers to ownership of the place where it is published, and not "the person that does the process of publishing". Perhaps the way to say it "a self-published source is where the author of the material is generally the same person that directly processes the publication of that material." --Masem (t) 05:50, 13 July 2020 (UTC)

Speaking as a "learner"-stage editor who has been trying to make sense of this, I agree with your comment re: how typical editors likely interpret "publisher," but I might say something like "the person(s) who determine whether the work gets published," as there are situations where the publishing decision involves more than one person. I also think that an analogous situation applies to "author": a typical editor probably presumes that when we say "author is same as publisher" that "author" refers to the person(s) who wrote the text, not the person(s) or organization that controls the creation of the text; the writer may not be the author, as when someone is employed to write sales material for an organization, is given an assignment by a boss to create certain material, and the material is published with the organization's name identified and the writer's name not identified. That is, in that case author=publisher not because the writer determines whether the work is published, but because the organization is both author and publisher. But it's important to be clear that just because someone writes something as an employee, that doesn't always make the employer the author; for example, faculty at research universities are employed by the university and publishing research is part of their job, but the university isn't the author of a professor's research publications.

In trying to sort out when author=publisher, I'd also run into problems when more than one person has significant input into the writing and/or more than one person has control over whether a work gets published, which is why I raised a question at the RS/N about whether online forums with editorial boards are SPS (and now that I'm getting clearer on all of this, my take on my own question is that a given online forum might have a mix of SPS and non-SPS works, depending on the specifics of a given work that's published there). It may be that author=publisher invites misinterpretation, and it should be shifted to something like "person(s) controlling creation"="person(s) controlling publication," recognizing that because there may be more than 1 person on one or both sides, it's possible for there to be inequality even if person A is involved on both sides (e.g., if an editor at an online forum with an editorial board creates an article but may jointly make a decision about whether it can be published there along with 1 or more other editors there, then the article isn't SPS even though the editor who wrote it is also involved in the decision to publish it). And as Mathglot noted here [2], it's important to help WP editors not think of SPS in terms of where a work appears, with the "where" being uniformly SPS or not-SPS. I think that echoes this idea of shifting from "publisher" to "the person(s) controlling the process of publishing" or "the person(s) who determine whether the work gets published." -- FactOrOpinion (talk) 13:46, 13 July 2020 (UTC)

Something along the lines of Masem's definition would be most universal. If itis necessarily a bit abstract, the more common / prescriptive words can be given as an example. North8000 (talk) 19:48, 12 July 2020 (UTC)

My attempted improvements were reverted again. That's fine, as long as we're moving towards coming up with a working definition, and giving examples of acceptable/unacceptable use, but we don't seem to be doing that. Either my edits should stand, or others should step up to try fixing the problem. Right now, the guidance to editors on the topic of SPS seems pretty lacking. -MichaelBluejay (talk) 22:35, 12 July 2020 (UTC)

That's not how it works. Don't make non-trivial edits to this key policy page without consensus. Zero^talk 05:03, 13 July 2020 (UTC)

Yes, MichaelBluejay, that's not how it works. North8000 (talk) 12:26, 13 July 2020 (UTC)

Ultimately, we're going to have to come to an agreement on whether the definition of SPS needs to be revised. Because of the "Never use self-published sources as third-party sources about living people" rule, the definition of SPS has significant implications for which sources can be used as RS for statements that fall under BLP. As a simple example, if government publications are generally SPS, and if the Mueller Report is among them (in that the government is both the author and the publisher), then we cannot use the Mueller Report as a source for a statement about any of the people who were charged in the Special Counsel investigation. That strikes me as a well-intentioned policy (BLPSPS) having unintended and inappropriate side-effects because of the SPS definition. -- FactOrOpinion (talk) 15:38, 16 July 2020 (UTC)

I have sympathy for this, but there is also something to be said for the view that "Never use self-published sources as third-party sources about living people", while well-intentioned, always cast too wide of a net. The idea that the self-published views of experts should not be allowed in commentary within the BLPs of others within their fields of expertise, for example, has always struck me as nonsensical. Obviously comment based on professional rivalry or currying favor should be excluded (not to mention otherwise undocumented biographical material), but I am not convinced that these biases are less present in independently published material, anyway. So yes, too wide of a net - a general caution coupled with more precise guidance about situations never to use SPS would strike me as a more fruitful approach. Sigh. Newimpartial (talk) 15:47, 16 July 2020 (UTC)

I agree that this is more likely a case of "Never use self-published sources as third-party sources about living people" casting too wide a net. There isn't much discussion on the WP:BLP talk page. I did link from there to the discussion here and a couple of other relevant discussions (e.g., [3]). Perhaps the best solution is something like "SPS can be used for third-party BLP statements on rare occasions, as long as the author(s) meet WP's definition of an expert source for the statement in question and there's good reason to believe that it has been through some significant internal fact-checking" and maybe adding "and the claim should made in the source's voice." (I don't know what the "right" solution is, just trying to give some examples of how to cast a slightly wider net, as may be appropriate, while not making it so wide to create a lot of new problems). As an aside, as I continue to think about the Mueller Report (because I'm working on the United States v. Flynn page), it may be that I also misjudged that in some ways, as the Special Counsel's Office/DOJ carried out some of the interviews and is a party in the legal cases, so that document may be first-party rather than third-party for some statements. I have to say that as a newish editor, it's not easy to know whether I'm abiding by all of the rules properly; there's a lot to learn / think through. -- FactOrOpinion (talk) 16:29, 16 July 2020 (UTC)

At least to me, BLPSPS, when it comes to SPS by recognized experts in the field that BLP is involved with, should be limited to when it is discussing the person directly. Or to flip it, a recognized expert in the field talking about a facet that is related to the BLP but wholly distanced from being about the BLP should be okay. A hypothetical case would be the discussion of a fad diet populated by a BLP (but not notable for its own page) with a recognized nutritionist itemizing the faults of that diet but not touching any commentary about the person that created it, though ideally we'd want MEDRS sources to even being doing that (and to that end, we'd only be discussing the fad diet is so far as to explain broad claims and not any "scientifically" sound facts it may present. --Masem (t) 17:26, 16 July 2020 (UTC)

@Masem: I've been working on the page for the court case US v. Michael Flynn (which arose out of Mueller's investigation), and in the background section, I'd like to be able to mention that in addition to the 1/24/17 interview that led to Flynn's false statements charge, he was interviewed 4 times in Nov. of 2017 before making his plea deal. He was also interviewed over a dozen more times during his cooperation period. There's oodles about the 1/24/17 interview in the media, but not so much about the later ones. I have substantiation for those post-January-2017 interviews from government sources (the Mueller Report for the first 4, and court documents for those and the later ones). These are claims about Flynn and the SCO/DOJ, not distant. I also know of MSM sources reporting things like "Mueller says in a sentencing memo that Flynn ... has provided 'substantial assistance' to the investigation, including meeting 19 times with Mueller’s team and Justice Department lawyers," but I'd like to be able to split out the pre-plea-deal from post-plea-deal interviews because of the way the article is structured, and I find it ridiculous that the SCO itself identifying the interviews in a court doc. may not be OK because it's SPS (unless it's judged to be first-party rather than third-party, since the SCO/DOJ is itself a party to the case), but a secondary MSM source simply repeating what the court docs say is OK. Or to take another example, someone introduced some quotes into the article from the court transcripts, and the BLPSPS rule says "Do not use trial transcripts and other court records, or other public documents, to support assertions about a living person," so I either need to go hunting for an article that quotes them or remove them. And if MSM are simply quoting from transcripts, why is the primary document unacceptable but the secondary MSM report OK? These rules simply don't make sense to me sometimes. -- FactOrOpinion (talk) 18:25, 16 July 2020 (UTC)

Yeah, I think there's an aspect of BLPSPS that has to do with "factual" coverage like a court case like that situation that BLPSPS can get in the way of a neutral - or in the case of Flynn, a more comprehensive/explanatory - article that may need to rely on SPS ( though I know you're looking at things like SCOTUSblog here which again, shouldn't be SPS). But that does rest on what answer we get at here. --Masem (t) 23:03, 16 July 2020 (UTC)

A related discussion that has come up that I think tells us we are in the right direction is over at WP:RS/N about YouTube (Wikipedia:Reliable_sources/Noticeboard#RfC:_YouTube) where some are arguing for a edit filter on it because the videos there are posted as an SPS, but as several oppose comments point out, YouTube is not really the "publisher" (again, back to the confusion). In other words, there's alignment on the thought processes here. --Masem (t) 23:03, 16 July 2020 (UTC)

Should this article have a definition of "Self-published" or not?

I'm creating a Talk section for each problem with the SPS treatment as I see it, to see if there's consensus that each issue is indeed a problem. If so, then after that let's move on to discussing fixes. If not, then there's no point in discussing fixes to something that editors think isn't broken (though I think an objective review will conclude that WP:SPS is currently pretty deficient in providing guidance to editors).

First issue: Should this article have a definition of 'self-published' or not? Again, here I'm not asking for proposed definitions, just an opinion about whether or not the article should include a definition. If your answer is no, then why do you think the article shouldn't have a definition? -MichaelBluejay (talk) 15:03, 15 July 2020 (UTC)

Yes. Now that I'm finally clear (I think) on what the definition means (see [4]), I find it very helpful as a learner-stage editor to have a definition. I think we should focus on making the definition clear, not delete it. -- FactOrOpinion (talk) 15:34, 15 July 2020 (UTC)

Yes. Adding my vote so we can move toward crafting a definition. -MichaelBluejay (talk) 21:12, 22 July 2020 (UTC)

Yes. It would be helpful to explain our jargon, but we would need another discussion to nail down the definition. — Newslinger talk 21:45, 22 July 2020 (UTC)

Should this article mention that SPS "includes almost all websites"?

One helpful bit on WP:USINGSPS is that it says something like, "SPS includes almost all websites, except those published by traditional publishers." That seems very helpful to me. While it doesn't cover every case, it covers *lots* of them, very efficiently/succinctly.

So, Should this article say something like, "SPS includes almost all websites, except those published by traditional publishers."? Again, here I'm not asking for exact proposed wording, just an opinion about whether the article should include wording similar to this. If your answer is no, then why do you think the wording is unhelpful? -MichaelBluejay (talk) 15:03, 15 July 2020 (UTC)

No. Is Google Books a traditional publisher? Many people would deny it is a publisher at all, rather just a platform that hosts works published by others. What about archive.org? Both these are used extensively and correctly. "Almost all personal websites" would be closer. Zero^talk 15:10, 15 July 2020 (UTC)

Only if it's true. Until we have a clear, agreed-on definition of "SPS" and "traditional publisher," how do we know whether a claim like "SPS includes almost all websites, except those published by traditional publishers" is true? I suggest that we postpone consensus on this question until we all agree on the relevant definitions and determine whether we have RS evidence whether "almost all websites, except those published by traditional publishers, [are SPS]" is true, or can come up with a related claim that is true. Personally, I doubt that there's good evidence re: the percentage of websites that are SPS vs. not, but I could be wrong, and of course if a source is a RS, we can cite its claim, even if the source itself doesn't provide good evidence for the claim (one of the problems with WP's definition for RS). -- FactOrOpinion (talk) 15:47, 15 July 2020 (UTC)

Even if the truth of the statement could be established, how would it help editors? It wouldn't provide guidance of the form "start by assuming it's unreliable", since that would require a completely different statement: "almost all websites you will encounter when you are searching for information on a topic are unreliable". Search engines do not provide you with a random website when you search for something. Zero^talk 16:02, 15 July 2020 (UTC)

Whether a source is SPS is a different question than whether it's RS. The intersection of the two categories is non-empty and sizeable, but so are the complements (the part of each set that lies outside the intersection). Recognizing that a source is likely to be SPS is useful in two ways: to heighten scrutiny about (1) whether it's a RS for the claim in question (though that assessment should also occur with non-SPS) and (2) whether it's a BLP claim for which thirty-party self-published sources can't be used, even when written by someone who meets WP's "expert" definition for that claim. I think it's helpful to point to the part of the RS discussion that notes that the acceptability of a source is context-specific, and if it's true that most websites and many paper publications are SPS, add the need for heightened scrutiny. But, it's certainly possible to address the latter issues without including a sentence like "SPS includes almost all websites, except those published by traditional publishers," so I'm open to changing my mind, and you've shifted me in that direction. -- FactOrOpinion (talk) 16:27, 15 July 2020 (UTC)

Yes. It is true, and (more pointfully), telling editors this helps them choose suitable sources. Look at List of most popular websites or any similar thing, and look at the types of sites.

• Blogs? Self-published.
• Corporate websites? Self-published (including all similar websites, such as non-profit organizations, fraternal clubs, etc.)
• E-commerce? Self-published.
• E-mail and mailing lists? Self-published.
• Internet forum? Self-published.
• Search engine? Irrelevant (because you shouldn't be citing the search results page anyway), but the non-results parts are self-published.
• Social media? Self-published.
• Software? Self-published (like any other corporate site).
• Wikis? Self-published.

The two types of websites that we agree aren't (usually) self-published are:

• Media organizations (including everything from the online copy of a daily newspaper to online-only magazines to television shows to film studios; the "corporate" part of the website [such as an "About us" page or an advertising rate sheet] is self-published, but the main contents are not)
• Service sites that provide online copies of publications that you should normally cite directly (e.g., Google Books, JSTOR)
  • Some of these "service" websites have mixed status (e.g., Amazon, Instagram or YouTube, where you can find both self-published videos of skateboarders recording their friends' tricks side-by-side with non-self-published official publications from media organizations such as Skateboarder (magazine)).

There is the one category that we haven't really settled on, which is whether a government website is "self-published" (because the same agency writes and makes their content available to the public) or not (because they're not publishing at their own expense). The Mueller Report could be classified either way, depending upon the definition you choose. Mostly, in practice, I think that we treat typical government websites (e.g., the sort of page that tells you whether you need a permit to cut down a tree) as if they're self-published, and more formal publications (e.g., annual reports of school statistics, or the decennial US census) as non-self-published. WhatamIdoing (talk) 18:08, 18 July 2020 (UTC)

I've reconsidered and now think it's important not to make claims about "websites," as a website may include a mix of SPS and non-SPS. For example, Yahoo.com is on the list of most popular websites, and a lot of Yahoo pages are self-published, but Yahoo News is a mixture of news aggregation (e.g., from the Associated Press) and original edited news articles that are similar to news published in more traditional sources. If "more formal publications" from the government are considered non-SPS, then government websites (e.g., justice.gov) also have a mix, and some will have a mix anyway (e.g., loc.gov has a mix of material written by Library of Congress staff and material originally published elsewhere, as does nih.gov). Ditto for university websites. Personal blogs are SPS, but some academic blogs have sizeable editorial boards and are probably a mix of SPS (for articles written by their editors) and non-SPS (for articles written by guest authors and submitted for publication, where the article may be rejected or go through significant editing). I think that a claim about "websites" encourages people to overgeneralize, and if we're going to include this kind of statement, it makes more sense to say something like "Almost all websites include some self-published material, and a vast number of websites are exclusively SPS. However, a given website may have a mixture of self-published and non-self-published material, and WP editors should focus on whether the specific material being used as a source for a claim is or isn't self-published." -- FactOrOpinion (talk) 02:11, 19 July 2020 (UTC)

Yes, over-generalization is common in source discussions. North8000 (talk) 11:26, 19 July 2020 (UTC)

I think the claims about "websites" is (a) helpful overall and (b) recognizes that, in practice, when editors are trying to make a case for citing a 'website' as a reliable source, that they're almost never trying to make a case for citing 'this book/newspaper/academic journal I happen to be able to read online'. WhatamIdoing (talk) 17:02, 19 July 2020 (UTC)

Yes. I agree with WhatamIdoing that the statement is true. Including the statement helps editors understand that the websites cited on Wikipedia are not a representative sample of websites on the Internet. Editors who are aware that the average website is self-published are likely to evaluate websites more carefully before citing them as sources. — Newslinger talk 06:14, 24 July 2020 (UTC)

Should the article include some examples of acceptable and non-acceptable use?

Examples are very helpful to editors. WP:USINGSPS provides several helpful examples, but that supplement is unofficial and so not authoritative, and I doubt there's any interest in elevating the supplement to be official policy. WP:USINGSSPS is too long to be included in this article, but simply including some good/bad examples of SPS wouldn't take much room and would go a long way towards clarifying how SPS should/should not be used.

So, Should this article include some examples of acceptable and non-acceptable use? Again, I'm not asking for which examples should be included, just whether it's a good idea to provide this kind of guidance to editors. If your answer is no, then why do you think that examples are not helpful? -MichaelBluejay (talk) 15:03, 15 July 2020 (UTC)

Yes, they're helpful. -- FactOrOpinion (talk) 15:50, 15 July 2020 (UTC) [Edit: I've reconsidered. I think it's important to have examples, but I don't think it's important to have them here rather than at USINGSPS. -- FactOrOpinion (talk) 22:59, 22 July 2020 (UTC)]

I don't think so, or, if we include them, the examples should be very brief. USINGSPS is a much better place to put examples and 'worked examples' of how we concluded that these sources were(n't) self-published. WhatamIdoing (talk) 18:12, 18 July 2020 (UTC)

No, I think policies are more suitable for explaining broader concepts rather than specific examples. Wikipedia:Reliable sources § Self-published sources (online and paper) (WP:RSSELF) is the guideline that elaborates on WP:SPS, and any examples would be more appropriate there. WP:USINGSPS would be even better. — Newslinger talk 21:31, 22 July 2020 (UTC)

Newslinger, WP:SPS links to this policy (Verifiability), not Wikipedia:Reliable Sources. So, any definition and examples (if there are to be any) should go here, not there. Also, WP:USINGSPS isn't a good place to have the only examples, because it's not policy. When I refer to WP:USINGSPS, the bad editors I battle will surely dismiss it as unofficial. Either the examples should go here, in Verifiability, or WP:USINGSPS should be elevated to be official policy. -MichaelBluejay (talk) 00:44, 25 July 2020 (UTC)

Should this article characterize SPS are "usually not reliable" and "largely not acceptable"?

The article currently characterizes SPS and "usually not reliable" and "largely not acceptable". In fact, it seems that's not true; there are a whole host of reasons in which SPS are both reliable and acceptable. WP:USINGSPS lists several.

So, Should this article characterize SPS as 'usually not reliable' and 'largely not acceptable', or should we instead direct editors to exercise caution with SPS, and then provide some guidance on acceptable/non-acceptable use? -MichaelBluejay (talk) 15:03, 15 July 2020 (UTC)

No, as it's uncertain whether the assertions of "usually" and "largely" are true, and phrasing it this way increases the likelihood of editors dismissing SPS as a category, even if a given SPS is reliable and acceptable for the claim for which it's cited. We should encourage editors to make case-by-case determinations for SPS. -- FactOrOpinion (talk) 15:56, 15 July 2020 (UTC)

I think it's premature to be asking this question. SPS should be dismissed as a category for some types of claims (e.g., anything medical), and embraced for other types of claims (e.g., Joe Film announced his engagement on Twitter). Twitter appears (very often as a ref) in more than 40K articles right now, and Facebook in 60K, so I'm unconvinced that people are rejecting them inappropriately. WhatamIdoing (talk) 18:16, 18 July 2020 (UTC)

WhatamIdoing, I think I understand your point that some SPS are good and some are not, but I don't see why you think it's premature to ask whether we should characterize SPS are "largely not acceptable" or not. From your discussion, it seems like you agree that "largely not acceptable/usually not reliable" is too broad, but I don't want to put words in your mouth.

As for whether valid SPS are rejected inappropriately, come on an editing walk with me and I can show you examples. That's what prompted my attempts to get SPS defined. I can't edit in my area of expertise in part because I'm battling bad editors who are quick to revert valid SPS. I'm on an editing holiday until that's resolved. -MichaelBluejay (talk) 00:55, 25 July 2020 (UTC)

No. I think it's far better to say that SPS should be scrutinized for acceptability, rather than broadly paint them with a "largely not acceptable" brush. So, with two No's and no Yes's, does this count as consensus to change the wording? -MichaelBluejay (talk) 00:55, 25 July 2020 (UTC)

Let's define self-published sources

Consensus above (no objections) was that this policy should define SPS (which makes sense, because it's impossible to apply the policy if we don't know what SPS is and what it's not). So let's hear some suggested definitions, please. (I'd offer mine, except that I don't have strong feelings about what the definition should be (just that we should have one), and when I copied the one from WP:USINGSPS (publisher=author), it was rejected.) Please suggest an actual definition, not just a discussion that doesn't include an actual definition. If no actual definitions are proposed, I intend to try again to insert the "author=publisher" definition from WP:USINGSPS. About that, Whatamidoing thinks WP:USINGSPS is a good place for acceptable/non-acceptable examples of SPS, so if WP:USINGSPS is trustworthy, then it ought to be trustworthy for the definition. If it's not trustworthy for the definition, then it's not trustworthy for acceptable/unacceptable examples either, and the acceptable/unacceptable examples should be included in this policy. -MichaelBluejay (talk) 21:23, 22 July 2020 (UTC)

My proposal: "Material is self-published if the person(s) who controlled the material’s creation = the person(s) who controlled whether the material was made available to the public."

And then in clarification, note things like the following (or, if it seems more appropriate, leave some or all of these things for the USINGSPS page):

1) “Control” is important, as a person writing promotional materials for a company ultimately isn’t the person(s) who controls the creation; rather the employer controls that, and the writer is employed to create work to specification. As part of assessing this, the existence (or not) of independent editorial staff plays a role, and we’ll need to clarify the meaning of “Independent” (e.g., without a conflict of interest for the material in question, able to reject that material for publication).

2) “Material” means the specific work that’s being used to substantiate the claim (a specific webpage, book, article, piece of music, etc.) rather than an entire website, journal, etc., as it’s possible that a given website, journal, etc., might include a mix of content that’s self-published and content that isn’t self-published. For example:

• an article in an edited newspaper isn’t self-published, but the comments posted below the article on the paper's website are self-published; a standard op-ed isn’t self-published, but an editorial by the editorial board is self-published, as are ads
• an article on a website with independent editorial decision-making isn’t self-published, but a live-blog on the same website is self-published
• an article on a site like Forbes.com might be self-published, while another article originated in Forbes magazine and is not self-published
• a government webpage at nih.gov may be self-published (e.g., the overview for grant applications) or not (e.g., the PubMed pages that provide information about journal articles)
• a professor’s own page on a university website is likely self-published, but an edited article on the university’s website isn’t

That said, a website, book, etc. may consist entirely of self-published material (e.g., Twitter, Facebook) or may have no self-published material (e.g., books by a traditional publisher). -- FactOrOpinion (talk) 22:42, 22 July 2020 (UTC)

Thanks, I like it! -MichaelBluejay (talk) 22:46, 22 July 2020 (UTC)

This definition covers many self-published sources, but I think it excludes some types of sources that are commonly considered self-published. Let's use Wikipedia itself as a counterexample. Wikipedia consists primarily of user-generated content, which is a type of self-published material. Wikipedia does have Articles for creation, a process in which content created by editors is reviewed by another editor who was not involved in its creation; this uninvolved editor determines whether the content gets published. However, articles that are published through AfC are still considered self-published and user-generated because the AfC review process is not rigorous enough to ensure that the published content is reliable. I think it would be more helpful for the definition of self-published to focus on the comprehensiveness of the source's editorial oversight, since this is the factor that directly influences the "reputation for fact-checking and accuracy" that is demanded of a reliable source. — Newslinger talk 09:07, 24 July 2020 (UTC)

@Newslinger: What's your proposed definition? (Michaelbluejay is asking us for "an actual definition, not just a discussion that doesn't include an actual definition.") -- FactOrOpinion (talk) 10:30, 24 July 2020 (UTC)

Moderated content doesn't change self-published, because its post-process. And we can distinguish this from , say, redactions and errata made by RSes with editorial control as there, we expect those sources to say what was changed and why, where with moderation, that's just done without comment. So this all still works in the framework of sites like Wikipedia, IMDB, Know Your Meme, etc all being self-published with moderators. The key with the self-published part is the lack of editorial review and intervention before the proverbial "PUBLISH" button is hit. --Masem (t) 14:40, 25 July 2020 (UTC)

Not quite (a key, maybe) because even if someone on Wikipedia, has an editorial review and intervention before the post-button is pushed (even if the button is pushed by the editor, not the author), it is still self-published/user generated. Alanscottwalker (talk) 14:49, 25 July 2020 (UTC)

I mean, whether an editorial review can be done on WP before it is posted, that's optional, sure, but there is no minimum requirement or limitation in the publishing process chain that brings in another other editorial review before the user that wrote the material hits the publish button. Whereas at a reliable newspaper, the author sends to editorial review, and after any back-and-forth, that's sent to a publishing department or to the website department that prep and then push to print/publish at minimum; those are required steps and thus eliminate the self-publishing factor, if you get what I mean. A self-publishing route that could employ editorial control if desired will still be called self-publishing for our purposes if that control is not a mandated feature. --Masem (t) 15:05, 25 July 2020 (UTC)

Again, not quite. There was actually a discussion recently raised about the Signpost, which is a "newspaper" (some say) and published on Wikipedia and it asserts some kind of mandated editorial review, and one or two people vaguely tried to argue such was not self-published, although the consensus was clearly that it is. It would not become, not-self-published merely by taking all review off Wikipedia before putting it in the edition. Alanscottwalker (talk) 15:15, 25 July 2020 (UTC)

We'd still be looking at the lowest-common denominator - that being the whole of WP, which is self-published. But what key is here is to arrive at a broad definition that works 99% of the time and allows for common sense discussion for outliers, and not to get a mechanical exact definition to account for all possibilities. The current SPS definition is far too narrow to reflect where it should be and makes discussing what is an SPS difficult to argue (eg with things like Forbes contributors and YouTube videos, broadly, because of the impression that "publisher" must be the website owner). Getting use close enough and letting contested cases be determined on talk pages as necessary is fine than the current status. --Masem (t) 15:37, 25 July 2020 (UTC)

Well, that probably arises from the tendency of some to take every section of a policy and read it out-of-context. If someone cannot or will not get that what we really want are professionally published books and journals from entities with a good reputation (as in WP:SOURCE), and the farther you get away from that the more problematic and grey it becomes, they just won't get the point, no matter what we write. Alanscottwalker (talk) 15:51, 25 July 2020 (UTC)

This will always be the case for any policy or guidelines. What's core is that too many editors, due to how the current statement on SPS here is written, walk away presuming that SPS extend only to case where the author controls the publishing medium, like their blog. That's a common case of an SPS but it omits the broader scope that should be included for purposes of using SPS to evaluate sources, particularly for BLPSPS evaluations. There will still be the BURO applications of that, but at least we are starting from a point that's more appropriate. --Masem (t) 16:50, 25 July 2020 (UTC)

No. First, you point to an essay, the reason it's an essay is it does not have general consensus. Whatever happens, this should not be done here, it's at best a guideline issue, if you think you can come up with a workable guideline that is always going to have play in the joints (sometimes on Wikipedia, as in life, you do have learn to live with uncertainty). The idea of framing it in concrete policy stone is quixotic at best, and unworkable in detail -- such detail issues among other things are why we have a whole noticeboard to discuss specific pieces of information. In short, try to actually buy into, we are not here to (re)publish or give undue prominence to anyone's or groups' original thoughts whether Wikipedian's or otherwise, nor string them together in articles. -- Alanscottwalker (talk) 14:05, 25 July 2020 (UTC)

Alanscottwalker, I'm confused about what you're saying "no" to (to defining SPS? to my proposed definition? to something else?). WP:V explicitly refers to SPS [5] without defining them. Most of us think that a definition is helpful, and I attempted to improve the WP:USINGSPS definition based on issues people raised here and on the talk page there. I'm not wedded to my proposal. Do you have a different definition to propose? FWIW, as a newish editor, it wasn't at all clear to me where to get help determining whether a given source is a SPS, as compared with whether it's a RS, and part of the problem that I've encountered is that people tend to overgeneralize and say that an entire publication (website, journal, ...) is SPS or not, rather than focusing on whether a specific article in the publication is SPS, where that specific article is the one that someone would like to use for a citation. (Granted, sometimes we can say that an entire publication is exclusively SPS or exclusively non-SPS, but other times publications have a mix of self-published and non-self-published material.) -- FactOrOpinion (talk) 16:07, 25 July 2020 (UTC)

Masem and Alanscottwalker, I'd be delighted if both of you would propose a working definition of SPS. -MichaelBluejay (talk) 18:31, 25 July 2020 (UTC)

PinkNews as a reliable source

Opinions are needed on the following matter: Wikipedia:Reliable sources/Noticeboard#PinkNews. It's a reassessment matter. Flyer22 Frozen (talk) 02:49, 10 July 2020 (UTC)

Does Footnote 9 still have consensus?

Footnote 9 currently reads, in part (emphasis added):

Sources that may have interests other than professional considerations in the matter being reported are considered to be conflicted sources. Further examples of sources with conflicts of interest include but are not limited to articles by any media group that promotes the holding company of the media group or discredits its competitors; news reports by journalists having financial interests in the companies being reported or in their competitors; material (including but not limited to news reports, books, articles and other publications) involved in or struck down by litigation in any country, or released by parties involved in litigation against other involved parties, during, before or after the litigation; and promotional material released through media in the form of paid news reports.

Recently, there was a discussion at RSN, "Is CNN usable as a source for unflattering information about Fox News?", which closed with "The consensus is that CNN is a usable source for unflattering information about Fox News. Editors largely do not accept the financial COI argument to disqualify CNN." (permalink)

Should Footnote 9 be modified, e.g. to remove the part about "discredits its competitors", or in some other way? Levivich ^{[dubious – discuss]} 22:57, 22 July 2020 (UTC)

Courtesy pings to opener and closer of that discussion, Aquillion and starship.paint. Levivich ^{[dubious – discuss]} 23:03, 22 July 2020 (UTC)

• @Levivich: - thanks for the heads up. I would suggest “discredits its competitors” be removed, purely based on the WP:RSN discussion. Anyone wishing to re-add the phrase should perhaps create an RfC at Wikipedia:Village pump (policy), with alerts to editors who participated in the WP:RSN policy and also any editors involved in the original discussion to add in “discredit its competitors”, if such a discussion existed. starship.paint (talk) 04:06, 23 July 2020 (UTC)
• I would remove the "discredits its competitors" language as well. I do not think it is consistent to allow a source when it covers its competitors positively, while disallowing that same source when it covers its competitors negatively. We currently have WP:UBO, which allows us to consider usage by other reliable sources as a positive indicator of a source's reputation. If we are allowed to consider a reliable source's endorsement of another source, we should also be able to consider the source's rebuke of another source. The "promotes the holding company of the media group" language is valid, in my opinion, and should stay. — Newslinger talk 04:17, 23 July 2020 (UTC)

  Since this question is likely to be of broad interest, I have upgraded this discussion to an RfC. — Newslinger talk 06:34, 24 July 2020 (UTC)

• I agree that "discrete its competitors" should be removed. -MichaelBluejay (talk) 05:16, 23 July 2020 (UTC)
• If a media source deserves to be discredited, where else are we going to read about it other than competitive media sources? Out of necessity, we have to accept such sources. But in this case, I would not rely on a single competitor; I would want to see multiple sources which are independent of the disreputable source and each other carry similar negative stories. Jc3s5h (talk) 11:07, 23 July 2020 (UTC)

  Alternatives to competitive media sources (e.g., CNN/Fox) include academic journals, books, and noncompetitive media sources (e.g., CNN/BBC). Levivich ^{[dubious – discuss]} 06:50, 24 July 2020 (UTC)

  @Levivich: How about New York Times versus Washington Post? Associated Press versus Reuters? Are they competitive? starship.paint (talk) 07:13, 24 July 2020 (UTC)

  @Starship.paint: AP/Reuters, yes. That means if there was, say, a scandal at AP, we shouldn't use a Reuters article as the only source for content about the scandal (and perhaps not use it at all, but at least not as the only source), because Reuters has a COI when it comes to its main competitor, AP. NYT/WaPo is a bit more grey-area for me. NYT/WSJ are in the same city and thus would be "conflicted" in my view. NYT and WaPo are not in the same city, so they don't compete for the same local market, but they do compete for the same national market, in the same way as AP and Reuters. So if there were a scandal at NYT, I would avoid using WSJ as a source at all, and I'd probably avoid using WaPo as the only source, although if it had a particularly in-depth article (as it probably would), I might use that along with other, less-conflicted, sources.

  I don't want to bludgeon this discussion so I'm going to make another related comment now and then shut up: one thing that the earlier RSN thread made me realize is that my own view of CNN and Fox News is likely outdated. I remember when Fox News launched as a direct competitor to CNN, and it became the main alternative for 24hr cable news. For example, as I remember it, people watched the Monica Lewinsky scandal unfold either on Fox or on CNN, depending on their political leaning. Some twenty-five years later, they're really not head-to-head competitors in that way anymore. I was persuaded by the argument that several RSN !voters made that the audience for Fox News, if Fox News disappeared, would not go to CNN, they'd go somewhere else. So perhaps when judging which media groups are competitors, and in which markets they compete, it should be more about more than geography. Levivich ^{[dubious – discuss]} 07:37, 24 July 2020 (UTC)

  Thanks for your detailed reply, Levivich! starship.paint (talk) 09:02, 24 July 2020 (UTC)

• I don't think it makes sense to generalize from a discussion about a specific CNN claim/article (in the RSN discussion) to a claim about media sources in general. Rather than simply removing “discredits its competitors” from footnote 9, I think it would make more sense to keep it but soften the statement a bit, saying something like "Further examples of sources with that may have conflicts of interest ..." and then add something to note that the details of a possible conflict of interest case should be assessed to determine whether it's an actual conflict of interest or is instead professional reporting on a negative fact about a competitor. Or if people think that the rest shouldn't be softened (e.g., that "news reports by journalists having financial interests in the companies being reported or in their competitors ..." should still be prefaced by "with" rather than "may have"), then reword it as two sentences, separating out and softening the claim about "articles by any media group that promotes the holding company of the media group or discredits its competitors," but not softening the rest. -- FactOrOpinion (talk) 13:56, 23 July 2020 (UTC)
• Using an examples CNN and Fox, I honestly wouldn't want a claim from CNN to be used to "discredit" Fox. A book or three on bias in the media from an expert on such things using Fox as a prime example probably already exist and should clearly be preferred by our policies. I think the full close from the other discussion goes this way as well. You can use CNN, but it's probably not a good idea, and if you do, attribute. I tend toward leaving the phrase in the note. --Izno (talk) 14:19, 23 July 2020 (UTC)

  I don't think Izno's post at 14:19, 23 July 2020 (UTC) allows for the speed at which corporate takeovers can change the nature of a source; waiting for a book may not be reasonable. Also, since such discussions are apt to be contentious, relying on books, which are less likely to be widely available for free, may present a problem. Jc3s5h (talk) 20:12, 23 July 2020 (UTC)

  We are not here to right great wrongs. --Izno (talk) 20:29, 23 July 2020 (UTC)

• A source writing about it's competitors is as egregious of a degredation-due-to-COI as any other type of COI referred to here. I see no reason to remove it based on some discussion about CNN. North8000 (talk) 21:46, 23 July 2020 (UTC)

Notified: Wikipedia:Village pump (policy), Wikipedia:Reliable sources/Noticeboard, Wikipedia talk:Reliable sources, Wikipedia talk:Independent sources, Wikipedia talk:WikiProject Reliability. — Newslinger talk 06:41, 24 July 2020 (UTC)

• The footnote doesn't make sense. What does "articles by any media group" mean? Press releases? Or does it mean articles by authors/journalists in publications (ultimately) part of a media group - which would be casting the net super wide and would seem to think that all media organizations exert total top-down editorial control, which is a crass over-simplification: the BBC (say) and RTV are different in this respect. I think the text in question is unhelpful; if a reliable source comments on an the unreliability of an unreliable source that is good for us; vice-versa not so. Trying to re-frame this as some kind of relativist play-off of COIs is unintelligent. Alexbrn (talk) 07:12, 24 July 2020 (UTC)
• Added by Wifione, who was indef banned for POV problems and sockpuppetry. If Wikipedia ever gets to the point where that level of detail on the quality of sources matters, we'll have come a long way indeed. But we're so far from being there that this footnote seems detrimental. --Hipal/Ronz (talk) 15:09, 24 July 2020 (UTC)

Good point. That's an argument for removing the entire #9. North8000 (talk) 15:19, 24 July 2020 (UTC)

• Remove the footnote. This application is clearly incorrect, as we can't even agree on which sources are competitors. Furthermore, it seems questionable to apply this only to media sorces. By the same standard, we should also discount authors who have written similar books and might be competing for sales, academics who may be competing for the same position, etc. I think that COI should be defined more narrowly to affiliation rather than worrying about competition. (t · c) buidhe 15:41, 24 July 2020 (UTC)
• Remove - I would not only remove the bolded part, but the entire footnote, for the reasons stated above. It is badly worded (it is unintelligible in part) and makes little logical sense. Neutrality^talk 22:54, 24 July 2020 (UTC)
• Remove (or substantially rewrite) I'm building on the excellent point made by @Buidhe:. Conflicts of interest and definitely far broader than media sources. it is always the case that one should give due consideration to the possibility that there is a conflict of interest. it is quite easy to imagine a conflict of interest if one media outlet is criticizing another media outlet, but importantly a is not necessarily a conflict of interest and b there are many many other examples. The first point means we should not be making the absolute statement that there is a conflict of interest

  are considered to be conflicted sources

  in the second point means that we should not privilege media sources as so much more likely to be conflicted then others that they deserve special mention and everything other than media are swept up in almost an aside

  include but are not limited to

  . The first point is more important. the footnote as written basically states that there is a conflict of interest simply by being a media source. This gives far too much power to editors who might be interested in excluding criticism of media sources. As written, any criticism of one media source by another media source can be simply removed on conflict of interest grounds without needing to assess whether it might be well grounded.--S Philbrick(Talk) 18:31, 25 July 2020 (UTC)

Listed at WP:CENT. Levivich ^{[dubious – discuss]} 19:56, 25 July 2020 (UTC)

• Remove the discredits its competitors clause. The New York Times discredits the Korean Central News Agency.[6] The Times is still a reliable source. — BillHPike (talk, contribs) 23:05, 25 July 2020 (UTC)
• Remove the discredits its competitors clause per Newslinger and Billhpike. One important thing to remember is that this is a list of examples — removing the phrase doesn't mean every possible instance of a media source covering another source negatively is acceptable, but rather just that the issue is sufficiently nuanced that it's not appropriate to use it as a clear-cut example of something not okay. I generally trust editors to be able to decide on a case-by-case basis whether a media outlet is operating with a sufficient editorial firewall when covering another source for it to be considered reliable. {{u|Sdkb}} ^talk 06:39, 26 July 2020 (UTC)

  Oops, seems like we don't have an article for editorial firewall. Someone should probably write that... {{u|Sdkb}} ^talk 06:40, 26 July 2020 (UTC)

  It's covered (briefly) at Chinese wall. Calidum 07:11, 26 July 2020 (UTC)

  Thanks, I've created a redirect from Editorial firewall to Chinese wall § Journalism. — Newslinger talk 07:51, 26 July 2020 (UTC)

• Remove discredits its competitors for the reasons others have highlighted as well as the related or in their competitors in the following sentence. I also believe the litigation part should be rewritten, because, based on my reading, if a person sues a newspaper for libel we could no longer use that newspaper as a source for that person, even if the case is thrown out. (My concerns about that section go beyond that, but that is just one example.) I'm ambivalent about the rest of the paragraph, except the paid news report part. Calidum 07:03, 26 July 2020 (UTC)
• Remove. This is premised on on two myths that are, unfortunately, broadly considered gospel. Namely that (a) human (and journalists', specifically) behaviour is largely driven by financial considerations, and (b) that any single journalist would compromise their work for the infinitesimal difference some story critical of a competitor may have on their future salaries. A CNN journalist will not criticise FOX News because management told them to, or because they own put options. They'll do it because they hate FOX and everything it stands for. 193.27.14.72 (talk) 20:20, 27 July 2020 (UTC) — 193.27.14.72 (talk) has made few or no other edits outside this topic.

"Sources that may have interests other than professional considerations in the matter being reported are considered to be conflicted sources." Wouldn't "I hate the other news agency" be as much a non-professional consideration as a financial interest is? Ken Arromdee (talk) 21:33, 27 July 2020 (UTC)