Wikipedia talk:Verifiability

Reliability of sources

I've noticed that many right-wing sources on this site are considered unreliable; my question is: why? I'm politically neutral and, considering that the politics of Italy and the United States are different, I have asked myself this question. Thanks in advance. JacktheBrown (talk) 22:48, 28 September 2024 (UTC)

That's not a question with a simple answer, but whether a source is perceived as right wing or left wing shouldn't be a consideration in questions of reliability. If you look into discussions about sources the issue are rarely ideological. Instead the cause is that media organisation have diverged since the days of print media, and that divergence has impacted media on different parts of the political spectrum differently. That change in the real world has then had an impact on Wikipedia. -- LCU ActivelyDisinterested «@» °∆t° 10:47, 29 September 2024 (UTC)

@ActivelyDisinterested: the thing that seems strange to me is that extreme left-wing sources aren't banned (maybe I'm wrong), while extreme right-wing sources are.
I repeat: I don't know the political situation in the United States in detail. In Italy we have a right-wing government; it's a government that many Italians support, because Italy is quite conservative, which for the Italian context is a very good thing; for the US context, however, it seems not, since the American right has, for example, denied climate change, which unfortunately exists.

Returning to the main topic: the thing that seems strange to me is that extreme left-wing sources aren't banned (maybe I'm wrong), while extreme right-wing sources are. JacktheBrown (talk) 14:11, 29 September 2024 (UTC)

extreme left-wing sources aren't banned (maybe I'm wrong) yes you age +are+ wrong. Take for example in British news media The Canary and Skwarkbox are both considered unreliable, and are on the far left of politics. Also to be clear sources aren't banned so much as actively discouraged if the consensus is that they are unreliable.

Whether governments or voters are of a particular part of the political spectrum is also not a consideration in judging the reliability of sources. -- LCU ActivelyDisinterested «@» °∆t° 14:21, 29 September 2024 (UTC)

@ActivelyDisinterested: thank you very much for this detailed and very useful explanation. JacktheBrown (talk) 14:23, 29 September 2024 (UTC)

The primary reason media sources get Deprecated (which is not quite a “ban”), is that they have repeatedly been shown to not fact check (or, in some cases, shown to completely invent “facts” which they report).

As to why right leaning sources seem to be more likely to be deprecated than left leaning ones… this is due to the fact that right wing sources tend to get more scrutiny. It’s no secret that Wikipedia attracts academics, who tend to be a somewhat left leaning group. They notice (and complain) when right-wing sources don’t fact check… and they tend to be a bit more forgiving when left-wing sources do the same.

That being said, we have deprecated a few left-wing sources when enough evidence has been presented to show they are not properly fact checking. It’s difficult, but possible. Blueboar (talk) 15:04, 29 September 2024 (UTC)

@Blueboar: Recognizing and admitting this problem is certainly a big step forward, but actions are more important than words; I would like neutrality not to be compromised in the encyclopedia. JacktheBrown (talk) 20:12, 12 October 2024 (UTC)

With respect, Blueboar’s comment is not an admission of any problem, it’s an example of the long debunked myth of the liberal media, which Blueboar evidently believes is true. Virtually every aspect of their comment is untrue, IMO. For example, the vast majority of US sources are center to center right, not "left", and the leading opposition party in the US (D) takes policy positions that are considered center to center right in the rest of the world. What I’ve found most interesting about this is to look closely at the history of CNN and MSNBC, two of the so-called "leftist" bugbears attacked by the right as communists. As it turns out, both networks take center to center right positions on most issues and are run by pro-corporate, pro-big business leaders. The idea that reality has a liberal bias began in the early 1970s as a right-wing libertarian call to arms, which has created an alternate reality where right is center and center is left. This was intentional. It began as a way to limit government regulation and undermine democracy. This is a good example of what happens when people lose touch with things like facts and become a post-truth society. And this is the true reason right wing sources tend to be highly deprecated. They don’t believe in things like facts. And when facts no longer matter, democracy ceases to function. Apologies if this offends anyone. Viriditas (talk) 01:44, 24 December 2024 (UTC)

User:JacktheBrown - Outside of politics the banning also happens. I think it is just the mechanics or effect of WP:RSP. For the UK it seems more an elitism thing than a political leaning -- for example it seems most British press by volume is excluded as low class. RSP seems to interpret 'reliable' to mean 'respected' or 'truth' to that WP editor, excluding consideration of 'available' or 'accurate' from 'verifiability'. (You can elsewhere see discussions on a paywalled source or remote paper being preferred despite readers generally not having access to such or WP:VNT) The terms RSP uses are "Blacklisted" meaning mechanically edits including that site are blocked, or "Deprecated" meaning the guidance includes "generally prohibited" by automatic warning of such an editor and removal of such edits by third parties, or "Generally unreliable" meaning outside "exceptional circumstances" not to be used with removal by third parties and pings in TALK. If you want to see what discussions on what to ban are like, they are mixed in at WP:RSN. Cheers Markbassett (talk) 04:22, 13 October 2024 (UTC)

@Markbassett: I'm not convinced that this isn't political. What's the percentage between left-wing and right-wing sources that can be used? Perhaps an 80/20? JacktheBrown (talk) 18:43, 12 November 2024 (UTC)

I'm saying that banning sources via blacklisting and deprecated designations seems basically an elitism act, and that it exists in all venues or topics - so it's in effect at articles for sports or music or basically anything. I'm not talking politics, or whether some parties might exploit it for financial or personal reasons -- I'm saying it simply is always a direct issue to the policies for V and WEIGHT. You cannot have WEIGHT properly measured if you exclude considering any significant circulation because editors think those are labelled as lesser venues, and if you are selecting paywalled or tiny elite items as the one to cite, then it's preferring the not accessible ones which is contrary to it being V verifiable. Cheers Markbassett (talk) 22:25, 12 November 2024 (UTC)

@Markbassett: very interesting discussion. So is Wikipedia making a mistake in rejecting certain sources or is it doing so in good faith? In my opinion the second option is correct, but I would like to read yours. JacktheBrown (talk) 22:40, 12 November 2024 (UTC)

Well I tend to view RSP usage here as problematic, but my wider view is Wikipedia content is from editors that are just people - good people and bad people, informed one way among many, obsessive or ditzy, with good days and bad days, days they have time to be careful and days that are rushed. You can look over RS debates and RSP debates or just filter recent changes to the ones possibly vandalism and judge for yourself.

For RSP in particular, I was not a fan of the idea back when it started and events have not improved my views of it and the way mechanics of it work. It supposedly was meant to capture the RS conclusions from earlier "Perennial" RS questions, so it could help the RS consideration. But folks just propose and argue from whatever stance for banning - there is not necessarily prior RS considerations or policy criteria in play. And instead of informing a RS consideration it seems mainly a blanket forever judgement - or at least I have not seen any reversals up or down. My revision of your 80/20 question back to you would be more along the lines of if WP bans 80% of the UK press, then is it a 20% valid portrayal of views ? Cheers Markbassett (talk) 23:25, 12 November 2024 (UTC)

@Markbassett: is it possible to have an accurate statistic of how many left-wing and right-wing sources are deprecated? JacktheBrown (talk) 00:17, 13 November 2024 (UTC)

I think one can only get a statistic of the simple sort and count by their status like 20 banned, 44 deprecated, 133 generally unreliable, 100 no consensus, 119 generally reliable. Anything else would need a chosen filtering and categorizing. But I think RSP counts is less a concern than percentage of external, or individual articles being totally dependent on limited source.

The first concern is more how large a percentage of the total WP:WEIGHT of coverage for a POV is being excluded by RSP, because it seems the largest or most known sources are the ones to get excluded, so a simple count such as '5' banned needs the context of is that 5 of only 6 or is it 5 of 100, and does the 5 constitute 95% of the WEIGHT or what ? For example again, almost all of major circulation UK newspapers are banned, and the banning for reason for banning for being 'sensationalist' or 'just sports' then means not having the best available data or best known portrayals on topics that were UK scandals or UK sports events.

The second concern is that at some articles the cites are too limited by selections which perhaps excludes what the most common view is outside of WP, or at least will not show all the common views so be a failure of NPOV. Again, this is all categories -- for example one cannot have a music article that only shows the academic views and think that reflects the world opinions and complete story. Cheers Markbassett (talk) 21:53, 13 November 2024 (UTC)

@Markbassett: Wikipedia:Reliable sources/Perennial sources shouldn't have been created, perhaps Wikipedia is unnecessarily self-complicating. JacktheBrown (talk) 22:17, 13 November 2024 (UTC)

I don't think it's possible to get such a count, because you'd have to first agree on where to draw the line between left and right. For example, CNN is most commonly described as centrist, but some editors think it's leftist. The New York Times seems to be considered center-left, except that there are complaints that it's right-wing on trans issues. WhatamIdoing (talk) 00:33, 14 November 2024 (UTC)

I haven’t heard the NYT described as center left in polite company. I think that’s a talking point on the right. If you followed the last presidential election, it appeared that the NYT was rooting for Trump on the regular and frequently pushes center to center right policy proposals. I think there’s this misunderstanding about the NYT that has persisted for decades, that because they cover a wide variety of topics that somehow makes them leftist. It doesn’t. If you recall, the NYT has been the subject of critiques from the left in the US since at least WWII, when they pretended Hitler wasn’t a problem and the Holocaust wasn’t newsworthy. Many controversies surrounding their conservative to right wing coverage in the 1960s and 1970s. In the 1980s and 1990s, the NYT was frequently referred to by the left as Lies of our Times. In the 2000s, they were taken down a notch for supporting and promoting the Iraq War and for government stenography, failing to investigate the false claims of WMD. During the Trump, post-truth era, the NYT was described as giving Trump a free pass while also attacking Democrats who might have paid a parking ticket late, etc. One of the NYT’s lead reporters has been repeatedly accused of engaging in access journalism, replacing critical analysis with softball questions and empty analysis. I could go on. Viriditas (talk) 02:27, 24 December 2024 (UTC)

When I've seen people (usually inexperienced editors) complain about the New York Times (example), it's almost always because they think it's liberal/leftist. WhatamIdoing (talk) 07:54, 24 December 2024 (UTC)

Semi-protected edit request on 18 November 2024

The birth place of Sam Mills is not Neptune City New Jersey because there is not a hospital with maternity facilities in Neptune City and his parents never lived there. Therefore change Neptune City ,N.J. to Neptune Township ,N.J. He was born at Jersey Shore Medical Center in Neptune New Jersey. Parts of Neptune City is across the street from that hospital. There are too many people confusing Neptune City Borough with Neptune Township 2601:45:400:2CC0:8818:726B:923B:C7C9 (talk) 00:31, 18 November 2024 (UTC)

Is this about the Sam Mills article? You should ask at Talk:Sam Mills instead. (Also, not every baby gets born in a hospital, or in the place where their parents live.) WhatamIdoing (talk) 00:39, 18 November 2024 (UTC)

Are articles written by a publication owner/publisher reliable secondary sources, or are they self-published sources?

I think I asked this question over a decade ago somewhere and it's been lost in the recesses of Wikipedia talk archives. If the owner/publisher of a reliable source writes pieces for their publication, are they a self-published source? Like, for example, if Ian Danzig writes an article for Exclaim! (which he owns and publishes), or HM's founder and publisher Doug Van Pelt or Jesus Freak Hideout's owner and publisher John DiBiase write articles for their respective websites, or A. G. Sulzberger writes a story for The New York Times, are those articles self-published sources only or are they considered reliable, independent published sources?--3family6 (Talk to me | See what I have done) 22:00, 19 November 2024 (UTC)

… depends on the publication, publisher’s relationships with paper, and more. The discussions above as well as ongoing Wikipedia:Requests for comment/Grey Literature all suggest divergent opinions in community. Bluethricecreamman (talk) 22:27, 19 November 2024 (UTC)

Your use of "secondary" in the heading is out of place. An article can be published by a reliable publisher but be primary. A paper in a journal published by a learned society reporting new experimental results is an example of that. And there is nothing stopping a self-published article or book from being based mostly, or entirely, on other sources, which makes it secondary. Jc3s5h (talk) 22:44, 19 November 2024 (UTC)

Wikipedia:Reliable sources/Noticeboard/Archive 282#RfC: Quackwatch concluded that articles written by the owner/publisher of Quackwatch were self-published. WhatamIdoing (talk) 05:37, 26 November 2024 (UTC)

@WhatamIdoing I'm now also trying to figure out if the commentary and factual statements accompanying the premiere of an artist's work is essentially a primary source. For instance, on my sandbox I'm working on an article, and I cite some BLP claims (a musician converted to Christianity and his band released to Christian metal albums) to a premiere of new music by that band. The author of the magazine post premiering the material is the owner of the magazine (thus it's self-published), but in this case I'm assuming that the commentary of that author could essentially be considered closely affiliated with the band, that is, essentially a press release. I'll note that the faith of the musician is not in question, that's been stated by the artist himself, but the timing of his conversion (mid-1990s) and that two albums were released afterwards I haven't seen expressly stated in interviews or other press releases.--3family6 (Talk to me | See what I have done) 16:07, 26 November 2024 (UTC)

Of course it is a primary source. It is also non-independent ("Alice Artist tells you what Alice Artist thinks").

But depending on the structure, it might not be self-published; the most I could tell you is that the post could be argued to be self-published, and that it could be argued to not be self-published. The latter is because anyone who puts out a magazine is an established, traditional publisher, and some definitions exempt all traditional publishers. I would not worry overmuch about this one.

Remember that Wikipedia:Primary does not mean bad. WhatamIdoing (talk) 17:20, 26 November 2024 (UTC)

@WhatamIdoing oh, primary is fine, my concern is that if it isn't primary. Because then it's a secondary source (the publisher) making a statement that's essentially self-published but used to support BLP claims.--3family6 (Talk to me | See what I have done) 17:43, 26 November 2024 (UTC)

An Artist's statement (which is a case of "commentary and factual statements accompanying the premiere of an artist's work") is primary, non-independent, and often self-published (but not always, e.g., self-published on the author's website or a commercial gallery wall, but non-self-published if it is quoted in a scholarly article). It is also an excellent type of source to use (usually with WP:INTEXT attribution) for information about the art and the artist.

A magazine article that contains similar information is often primary (but not always, e.g., if it takes the form of a compare-and-contrast analysis). It is assumed independent unless there is some COI-type relationship between the artist/band/album and the author/magazine/magazine owner. It is usually not considered self-published. WhatamIdoing (talk) 22:59, 26 November 2024 (UTC)

@WhatamIdoing Right. My concern is that as the magazine is independent of the artist, and the content is a summary put together by the magazine's publisher, it's technically an independent source but a self-published one given that the publisher is the author and thus not under normal editorial review. In terms of the spirit of what not using SPS for BLPs is about (defamation), it's fine, as the subject states his faith elsewhere, but technically if the source is an independent SPS it's a violation of the stated consensus to never use such sources for BLP statements.--3family6 (Talk to me | See what I have done) 12:27, 27 November 2024 (UTC)

I think that "thus not under normal editorial review" is an assumption, and possibly a false one.

The usual practice seems to be to start off with "Well, is it obviously an ordinary newspaper/magazine/academic journal?" If so, then we're done: It's not self-published. WhatamIdoing (talk) 20:13, 27 November 2024 (UTC)

@WhatamIdoing well, that's why I brought up this issue. And as you point out, Quackwatch, for example, is considered to be self-published if the content is from the website publisher. With the example I'm working with, we're talking about something that was a print magazine but now is fully online only.--3family6 (Talk to me | See what I have done) 20:26, 27 November 2024 (UTC)

There are two ways to address this concern. I'll give you both:

1. If this magazine were still in print, would you still be worrying about whether this article is self-published? If not, then you shouldn't now. Being self-published has nothing to do with whether the publication is on paper or on a website.
2. What do you think is best for the article? If, using your own judgement as an editor, you think that citing this source makes that article better, then do it. If you think it makes the article worse – well, we wouldn't be having this conversation now, right? Because if you actually thought that, you'd have already rejected the source.

WhatamIdoing (talk) 21:33, 27 November 2024 (UTC)

@WhatamIdoing on point 1, yes, I've treated them the same in the past, and if I'm going to presume one form is reliable or unreliable, I'm going to presume the other.--3family6 (Talk to me | See what I have done) 04:16, 28 November 2024 (UTC)

Thank you for these replies, this was helpful and I'm already implementing some changes on pages I've created/worked on.--3family6 (Talk to me | See what I have done) 12:34, 26 November 2024 (UTC)

@3family6, you have made 26,000+ edits over the last 16 years, including about 150 articles. At this point, the only rule I really want to see you closely following is this one: Wikipedia:If a rule prevents you from improving or maintaining Wikipedia, ignore it.

As you have noticed, I'm happy to blather on about the technical distinctions between these things and the awkward gray areas, but when you're working on articles, you need to do what's right by the article, even if "the rules" suggest hurting the article. Total compliance with the rules is not the goal. Let me trust you to do the right thing. If you give me decent articles, I can figure out ways to make the rules align with community practices. Articles first. WhatamIdoing (talk) 23:05, 26 November 2024 (UTC)

@WhatamIdoing Wow, I really appreciate that comment, that is very kind. I guess my concern is your last part about community practices. I want to make sure that I'm actually going by community practice and not just unilaterally making content decisions. I was cleaning up one of my articles yesterday where I had include a source once way back that was not reliable. In general I think it's good for me to reflect on whether some of my editing practices are actually the best practice and actually improving the encyclopedia, or if they're bad habits that I fell into. I also just wanted to get some clarification on these gray areas not just for myself but for others. So there's a discussion I can point to.--3family6 (Talk to me | See what I have done) 12:31, 27 November 2024 (UTC)

Years ago, an editor used to say that if you haven't made 50 mistakes yet, you aren't a real Wikipedian. Mistakes are okay. Fixing them is great.

In general, I think it's important and valuable to conform to community practices, but sometimes the best practice isn't the popular one. The community needs some editors to be able to see and speak to best practices, instead of simply following rules, as if our policies and guidelines were holy writ. So I need – we need – you and editors like you to be thinking about what's best, rather than what's officially endorsed by The Rules™. WhatamIdoing (talk) 21:30, 27 November 2024 (UTC)

My take on the question is that most self-published sources are not reliable sources because we have no way of judging their reliability. We do make some exception; we accept blogs written by acknowledged experts when they are writing about the area of their expertise. In general, we accept sources as reliable when they (or their publishers) have a history of publication that has shown them to be generally reliable. It is not true that a self-published source can never be used. The point is that we should not use sources for which we cannot determine reliabilty, and we cannot make that determination for most self-published sources. If a self-published periodical (whatever that means) has enough history, however, we can make judgements about the reliability of that publication for topics it has published on. Donald Albury 22:20, 27 November 2024 (UTC)

@Donald Albury Right. I would presume that music reviews, if nothing else, could be considered editorials, if not more. Where it matters would be BLP statements.--3family6 (Talk to me | See what I have done) 04:19, 28 November 2024 (UTC)

Should there be a statement that accessible sources are nevertheless preferred?

Because I think there ought to be one. The present policy defends less accessible sources entirely but there ought to be a balance. If there is a more accessible source and a less accessible source for a given bit of information, and it's not desirable to cite both, the more accessible source should be preferred. If a piece of information with a citation from an inaccessible source is contradicted by a more accessible, equally reliable source, there should be a preference towards the verified piece of information unless the first source can be located. If I own the latest edition of a book on a topic and the text in question is the same as an older edition that is on archive.org, I should cite the older edition because editors can go look at the book themselves more easily. Doing otherwise might not be a problem in itself but might lead to issues down the line. Fangz (talk) 08:35, 4 December 2024 (UTC)

When Wikipedia was young the better sources were almost exclusively off-line. While it is true that more reliable sources are now available on-line (almost all scholarly journals and many books), I would say that more than 95% percent of what is available on-line is still not usable as sources in Wikipedia. Most of the best sources on-line are still those that have been published on paper before or simultaneously with the on-line version. Unfortunately, many of the on-line sources are behind paywalls, but that does not mean that we should accept poor sources simply because they are free. The WikipediaLibrary has helped with that, providing access to paywall protected sources to editors who meet the requirements. Indeed, it has allowed me to drop my private subscription to JSTOR. We should always strive to use the best reliable sources to support content in articles. I would say that more of the highest quality sourcing is available on-line now than was the case 15 to 20 years ago, and so there is even less reason today to ease our sourcing standards than there was then. Donald Albury 16:15, 4 December 2024 (UTC)

The nightmare case is editors misrepresenting sources, and doing it with obscure sources that are hard to check. I expect you can think of the particular editor I have in mind (I'll name them if requested but I don't know that it adds much). I don't have a real solution to this; such an editor would not really be stopped by a statement that high-availability sources are preferred. But it's something to keep in mind. --Trovatore (talk) 00:07, 5 December 2024 (UTC)

In many cases it is a good idea to cite an off-line or subscription source with lots of detail, or especially authoritative, also citing one accessible online, perhaps with less detail or authority. Remembering also that google books previews are only available in different countries or times entirely as the publisher chooses. Johnbod (talk) 16:24, 4 December 2024 (UTC)

I'm also seeing publishers or authors putting fairly recent but out-of-print books on-line for free, or as a free e-book. Not very common, yet, but I hope it is a growing trend. Donald Albury 17:55, 4 December 2024 (UTC)

All else being equal, where there are multiple high-quality reliable sources available for a claim, we can lean towards more accessible sources. But often all else is not equal. For example, citing an older source over a newer one can give the impression that the claim may be dated and no longer reflective of the literature. And where equally reliable sources disagree, we definitely shouldn't disregard one perspective based solely on how accessible it is. Nikkimaria (talk) 00:21, 5 December 2024 (UTC)

editing the text of WP:SPS

The text of WP:SPS currently says (1) Exercise caution when using such sources: if the information in question is suitable for inclusion, someone else will probably have published it in independent, reliable sources, and (2) Never use self-published sources as third-party sources about living people, even if the author is an expert, well-known professional researcher, or writer. Sentence (1) does not exist in the corresponding WP:RS/SPS text. Moreover, it implies that expert SPS cannot themselves be independent reliable sources, and it's not clear to me that the claim about probability is true. I propose deleting this sentence, or at least the portion of it that comes after the colon. Sentence (2) is a bit inconsistent with the corresponding text of WP:BLPSPS, which now includes the following exception: It does not refer to a reputable organisation publishing material about who it employs or to whom and why it grants awards, for example. WP:NPROF has similar exceptions in various places on that page; I'm not sure about other subject-specific notability guidelines for people. Should a corresponding sentence be added to WP:SPS? (FWIW, since editors disagree about what SPS does/doesn't encompass, I will likely start an RfC about that after the closure of the RfC on grey literature mentioned above. We could wait til that's over, as it may have broader implications for the wording of WP:SPS.) FactOrOpinion (talk) 22:19, 7 December 2024 (UTC)

In regard to BLP, it seems like "It does not refer to a reputable organisation publishing material about who it employs or to whom and why it grants awards, for example" is not saying that it is ok to use an SPS on a BLP, just that this does not count as an SPS. If that is the case does it provide a counter example to "Never use self-published sources as third-party sources about living people, even if the author is an expert, well-known professional researcher, or writer"? - Bilby (talk) 23:22, 7 December 2024 (UTC)

No, the recent clarification at WP:BLPSPS is saying that your employer is not a "third-party" to you. A press release is always self-published. A press release or a social media post saying "Bob's Big Business, Inc. is delighted to announce that they have hired Sam Sales as the new Vice President of Global Sales" is:

• non-independent (of the business; of the new VP of sales),
• self-published (written by the business, published by the business),
• primary (very close to the event, based on no prior publications),
• reliable (for sentences such as "Sam Sales was hired as Vice President of Global Sales in 2024"), and
• acceptable under BLPSPS.

WhatamIdoing (talk) 23:40, 7 December 2024 (UTC)

Bilby, people have varied views on what is/isn't self-published (see the SPS definition discussion above). AFAICT, everyone agrees that (a) things like personal blogs, wikis, and social media are SPS and (b) things like traditional newspapers and book publishers are not self-published, but people disagree about (c) whether [Edited to add: and under what conditions] publications from advocacy organizations, universities, companies, think tanks, museums, learned societies, governments, etc. are SPS. The community needs to come to some consensus about (c), and that's why I'm thinking of creating an RfC about it. Depending on one's view about what is/isn't self-published, the BLPSPS sentence might mean that those publications are not self-published (and so don't fall under BLPSPS) or that they're self-published but are exempt from BLPSPS. FactOrOpinion (talk) 00:43, 8 December 2024 (UTC)

No, "a press release is always self-published" is only true if the press release is published by a natural person, not an organization. It is, however, a primary source. Jc3s5h (talk) 03:03, 8 December 2024 (UTC)

You made the argument above that a natural person (e.g., me) can self-publish something, but that when an organization (e.g., "WhatamIdoing, Inc.") does exactly the same thing, it's magically not self-published. I'm still not buying it. WhatamIdoing (talk) 07:17, 8 December 2024 (UTC)

FactOrOpinion, rather than trying to keep everything perfectly in sync across multiple pages, maybe we should point to BLPSPS. That might involve adding works like "Per WP:BLPSPS," but would not necessarily involve removing any existing text.

About your (1): "Exercise caution when using such sources: if the information in question is suitable for inclusion, someone else will probably have published it in independent, reliable sources".

The change I'd suggest here is to change "published it in independent, reliable sources" to "published it in a non-self-published reliable source". The non-self-published part is the part that matters. That source will be independent 99% of the time, but it's technically not the problem we're talking about here.

About your (2): "Never use self-published sources as third-party sources about living people, even if the author is an expert, well-known professional researcher, or writer."

What we're trying to say is that even if the author is amazing, and even if the author has written a dozen non-self-published books on this exact subject, it's still 100% not okay to use their self-published works for BLPSPS purposes.

Imagine, e.g., that Gene Genealogist has written hundreds of articles and several books on genealogy. Gene has even written a whole book (HarperCollins, 2021, good reviews, decent sales) specifically on the subject of birthdates and is considered an expert in the field. Gene self-publishes this on social media: "Here's a fun fact: Paul Politician, Joe Film, and I all share exactly the same birthday. We were all born on the 32nd of Octember in 1960, making us all Baby Boomers. We are all 64 years old right now."

Using the list above, this post is:

• non-independent of self, but independent of the others,
• self-published,
• primary,
• reliable, and
• acceptable only for statements about Gene's own birthday/age/generation. It is unacceptable for statements about Paul or Joe.

The problem isn't that Gene isn't independent or reliable. The problem is that there was nobody to stop Gene if publishing this were a bad idea for some reason. And rather than say "Oh, it's never a bad idea" or "Do this only if it's a good idea" – and then have fights over whether this is a good or bad idea, then editors voluntarily placed a blanket restriction on ourselves: Don't use even independent expert SPS sources about other people. WhatamIdoing (talk) 00:01, 8 December 2024 (UTC)

Re: (1), why do you believe that "if the information in question is suitable for inclusion, someone else will probably have published it in a non-self-published reliable source" is true? How would one even go about testing that?

Re: (2), you quoted a different sentence, one that I'm not questioning. I understand the BLPSPS condition. I also understand that there's significant disagreement about what is/isn't considered self-published. But my question was whether something like It does not refer to a reputable organisation publishing material about who it employs or to whom and why it grants awards, for example, which is currently in the text of WP:BLPSPS, should be added to the text of WP:SPS. Instead of your Gene Genealogist example, it's instead something like Notable Academic got a Notable Award from Learned Society, as published in Learned Society's newsletter. If one considers a newsletter to be self-published (and you do, though some others might not), then that sentence is saying that it's still OK to cite the newsletter for text on the academic's BLP re: the academic having gotten the award. And the BLPSPS text makes that clear, but the SPS text doesn't. FactOrOpinion (talk) 01:18, 8 December 2024 (UTC)

(1) I don't think it's universally true, because "the information in question" could be basic information about the subject (e.g., all biographies should identify when and where the person lived; all articles about books should say what the book was about; all articles about species should say what kind of an organism it is), and sometimes part of that information might not be available anywhere else (e.g., a birthdate or birthplace; the author's explanation of what the book is "really" about; a scientific monograph from previous centuries).

However, outside of such basic information, if it isn't available in a non-self-published reliable source, it's unlikely to be WP:DUE for inclusion. Remember that this is an information statement rather than a rule (i.e., it does not tell you what to do), and it is not absolute, since it says this is only "probably" the case.

(2) My suggestion is that we solve the mismatch by having a single copy of the full BLP rules at BLPSPS (=not here) and that we indicate that WP:V does not have a complete copy of the full BLP rules. The alternative is that we have duplicated text, which will inevitably diverge over time. The available choices are:

• a single "official" text, and everything else points to it, or
• the maintenance hassle of resolving contradictions that arise between multiple copies.

The first approach is recommended in Wikipedia:Policies and guidelines#Content ("minimize redundancy"), but if you prefer ongoing maintenance hassles and the periodic need to de-conflict policies, then we could do that. WhatamIdoing (talk) 07:15, 8 December 2024 (UTC)

I forgot to address your Learned Society example before I posted this. It's the same as the employer press release:

The Learned Society publishes a newsletter. The things we call newsletters are usually self-published, because they are usually being written and published by the organization; however, in some cases, the organization only sponsors the publication, and the publication's staff has editorial independence similar to a magazine (e.g., Newsletters on Stratigraphy). For simplicity, I'm going to stipulate that this is a self-published newsletter.

One of their self-published newsletters says that Prof. Notable Academic got a Notable Award from the Learned Society. This newsletter item is:

• non-independent (of Learned Society; of Notable Award; of Prof. Notable Academic [at least wrt this award]),
• self-published (written by the org, published by the org),
• primary (very close to the event, based on no prior publications),
• reliable (for sentences such as "Learned Society awarded the Notable Award to Notable Academic in 2024"), and
• acceptable under BLPSPS.

WhatamIdoing (talk) 07:29, 8 December 2024 (UTC)

Re: (1), I still don't understand where you're getting the data from that allow you to conclude that "if the information in question is suitable for inclusion, someone else will probably have published it in a non-self-published reliable source." To me, it seems like a case where some editors believe this but don't have data to back it up and where it's unclear how one could gather such data. (Arguably, the vast majority of RS information isn't DUE, but that's a different issue, and is the case regardless of whether the source is SPS or non-SPS.)

Re: (2), your comment reminded me that there has been a similar conversation about the parallel texts in WP:SELFSOURCE, WP:ABOUTSELF, and WP:BLPSELFBUB. The people involved in that conversation decided that the best solution was "Remove the SELFSOURCE material in RS [and remove the BLPSELFPUB material in BLP]; retarget that shortcut to ABOUTSELF's location in V, and "advertise" the shortcut at that place instead; in RS [and BLP], summarize ABOUTSELF in a sentence and cross-reference it, without an unnecessary and potentially confusing devoted section."

Re: the Learned Society example, the only reason that it's acceptable under BLPSPS is because the exception is specified. It would not be acceptable under the current text of SPS. FactOrOpinion (talk) 17:09, 8 December 2024 (UTC)

For (1), this is not a statement about collected data. This is a statement about editors' collective experience.

For (1), there are two classes of information that are suitable for inclusion. Those two kinds are:

1. Basic information about a subject, which belongs in an encyclopedia article because of the requirements of the genre. This includes, for example, writing boring information like "George IV (George Augustus Frederick; 12 August 1762 – 26 June 1830) was King of the United Kingdom of Great Britain and Ireland and King of Hanover from 29 January 1820 until his death in 1830" instead of jumping straight into the subjectively attention-catching content, e.g., "King George IV, nicknamed Prinny, was known for being obese, profligate, prodigal, and promiscuous."
2. Non-basic information about a subject, which belongs in an encyclopedia article because it is in desirable sources.

For the second category, the existence of suitable sources is definitional.

For the first category, the "probable" existence of suitable sources is the collective experience of editors. You might not be able to find every common detail (e.g., a complete birthdate may be unknown, or it may only be available in a self-published source) but if the subject actually qualifies for a Wikipedia:Separate, stand-alone article, you will be able to find enough non-self-published sources to meet the requirements of the encyclopedic genre, e.g., to place the subject in the encyclopedic context of time and place.

In re the Learned Society example, it was accepted in many thousands of articles before the exception was specified. The exception was written down to make the written rules more accurately reflect the community's actual practices. Per WP:NOTLAW, the written rules themselves do not set accepted practice. Rather, they document already-existing community consensus regarding what should be accepted and what should be rejected. The accepted practice was set by the community for more than a decade. Recently, we updated the written rules to document the already-existing community consensus that an employer is not a True™ third-party from its employees (for the purposes of this policy) and that an award giver is not a True™ third-party from its awardees (for the purposes of this policy) and that therefore self-published statements from these sources are not excluded by BLPSPS.

Perhaps this is the fundamental misconception. The written rules follow the community practice. The written rules document the community practice. The written rules are not supreme. The community is supreme. WhatamIdoing (talk) 18:20, 8 December 2024 (UTC)

Thank you for having pointed out WP:NOTLAW. As much as I try to be familiar with (and abide by) the PAGs, I don't know that I've read all of them in their entirety, and I don't remember all that I've read.

Re: (1), I'd say that editors' experiences are a form of data, but it's unclear to me how the community determines what the collective experience is. There are ways to determine the consensus of a small number of editors (e.g., in an RfC), where those editors may make arguments based on what they believe about the behavior of editors not involved in the RfC, but that's still a tiny fraction of editors. I don't know how anyone could confirm in any reasonable timeframe that "In re the Learned Society example, it was accepted in many thousands of articles before the exception was specified." (You can't confirm it with any kind of straightforward search; instead, you'd have to individually look at the sources for that kind of info on many thousands of articles, and you also have no way of knowing how many editors left that kind of info out of yet other articles because they thought that the employer, awarder, ..., wasn't an acceptable source.)

It seems to me that in discussions, there's sometimes a tension between what PAGs say and what collective experience might be ("might" because I'm not sure how to determine "is"); that arose in the discussion about whether learned societies that are highly regarded within an academic field are wiki-notable, and is in play in the dispute about what is/isn't an SPS.

Re: your 1. and 2., I'd interpreted "if the information in question is suitable for inclusion, someone else will probably have published it in a non-self-published reliable source" as addressing both. And I've probably worked more on academic BLPs than other BLPs, which likely affects how I view all of this. FactOrOpinion (talk) 20:21, 8 December 2024 (UTC)

Widely advertised discussions, including but not limited to RFCs, are assumed to represent the view of the whole community unless and until disproven, or at least seriously challenged.

Every decision is made by a tiny percentage of editors. Even if we had a thousand editors respond to a question, which almost never happens, that's still only 0.12% of last year's registered editors. But it's enough; we don't need large numbers of editors to make the right decision. If the first decision is wrong, we'll discover that over time and adjust.

In many cases, editors are learning by watching. They see that he cited the Oscars website, and didn't get reverted; she cited the Emmys website and didn't get reverted; they cited the Nobel website and didn't get reverted; and then rationally conclude that award websites must be okay. They will also see the Oscars website sometimes get replaced by a book, the Emmys website sometimes get replaced by a music magazine, and the Nobel website sometimes get replaced by a newspaper article. This means that when a single experienced editor shows up for a discussion, they're describing what they have seen get accepted (or rejected) by dozens or hundreds of editors across many articles.

Academic BLPs are challenging, because the accepted criteria give little consideration to why we require significant coverage in independent sources in the first place. I would expect it to be difficult to much past the stub stage for many academic BLPs without veering into OR territory. WhatamIdoing (talk) 21:07, 8 December 2024 (UTC)

WhatamIdoing, thank you for your responses. You are helpful and extremely patient with all of my questions. FactOrOpinion (talk) 19:18, 9 December 2024 (UTC)

Proposal to adjust policy on self-published sources

Hi folks, I just wanted to discuss WP:SELFPUBLISH as I think it needs to be tweaked.

The proposed change

I propose that the current wording: Self-published expert sources may be considered reliable when produced by an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications. be changed to:

Self-published sources may be considered reliable when they are either:

• Produced by an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications.
• When an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications, uses the self-published source as a reliable source in a publication.
• When the source's claims can be clearly and non-controversially verified and deemed as reliable.

Why make this change? Currently we allow a wide range of sources without decent review standards. I'd be surprised if we've not all come across books containing claims that make us wonder if an editor ever held the book, seen podcasts given as sources, etc. This isn't to say that we shouldn't be highly suspicious of self-published sources, that I do not question, but there are other ways of determining a book's reliability than just checking if it was published by a third party.

As the policy currently stands, it does not matter how reliable a self-published book is. This is an issue as not only is this standard far above what we expect from other forms of media, it ignores the reality that book-editors often don't fact-check, and (most importantly) it limits us from using potential sources that can be demonstrated to be otherwise very reliable (even more reliable than other books published by a third-party, as I hope to demonstrate).

Why would this policy change lead to an improvement in Wikipedia? I raise this is as I've been working on Daisy Bates (author) and I've found that two of the recent biographies, often disagree with one another on very simple facts about Daisy Bates's life. My suspicion is that they both read autobiographical work written by Bates ("The passing of the Aborigines [sic]" (1936)), as well as a biography by Elizabeth Salter (titled "Daisy Bates: Queen of the Never Never" (1972)), and that they didn't do their due diligence in checking other primary sources. I'll give some examples in the collapsible table below:

Fact checking claims of de Vries and Reece

Some examples below: Bob Reece claims on page 37 that Bates camped at Ma'amba reserve in 1901 and was invited to meet a duke and duchess in that same year because she organised a corroboree to welcome them on their arrival in Perth. When I was trying to understand the situation and the reason for the conflicting account I found a journal article and a government report which contradicts Reece; they describe that what was organised in 1901 was not a corroborree and that it was not organised by Bates.[1][2] According Elizabeth Salter, Lomas and de Vries, Bates only went to Ma'amba after being hired by the government in 1904, this is backed up by correspondence we have between Bates and government officials (which is quoted by Lomas in his book). You only get the date of 1901 if you take Bates at her word in her autobiographical work. It appears that neither de Vries nor Reece actually read many of Bates's first articles. For example: They both disagree on what one of Daisy's first papers (published in Western Australia's journal of agriculture), "From Port Hedland to Carnarvon by Buggy" (1901), was about. de Vries thinks that it "covered her observations of the Indigenous people she had encountered" but this is plainly false if you read the journal article (as I have, at the J S Battye Library). Reece says that it was more a travel account. Both de Vries and Reece state that Bates started her enthographic work shortly after arriving in Perth, but there's no evidence of this. They both just trust Bates on this, but if you actually read her papers you see that they have nothing to do with anthropology. Again, Lomas was the only one (of the three) to actually read the original papers himself and point this out.

To summarise the reliability of de Vries and Reece, who are both published by a third party and have gone through some kind of review process:

• They both are often lacking in giving citations, this makes their reliability hard to judge.
• They both contain the errors that could have been avoided if they (or an editor) had read other primary sources.

This is not to say that they should not be used; I note all of this to contrast them with Lomas's book, which is self-published.

Coming to Brain D Lomas's "Queen of Deception" (2015):

Indicators that Lomas is reliable

Indicators of Lomas's reliability: Lomas's book is used as a source in Eleanor Hogan's "Into the Loneliness: The unholy alliance of Ernestine Hill and Daisy Bates". "Dr Eleanor Hogan is a 2023 National Library of Australia Fellow"[3] and her book was published by NewSouth Publishing at the University of New South Wales Press Ltd. This book won the "The 2019 Hazel Rowley Literary Fellowship".[4] Lomas has done a lot of original research on the topic: He has combed through old newspapers and government records to corroborate Bates's whereabouts; Old newspapers contained the names of folks who arrived at a port, Bates also developed a level of fame where people would report on her whereabouts in letters to papers, he also went through government records for her travel expense receipts she had sent to the government for reimbursement, etc. Reading through a ton of archived material spread between Perth and Canberra. This includes the just mentioned receipts for travel expenses, but also personal and government correspondence (who for a while was her employer). I haven't seen indication that the other biographers went through so much archived material. He regularly lets the source material speak for itself by giving quotations from primary sources. He is quite consistent in giving citations, making it very easy to verify his claims. As a result it has been far easier to fact-check Lomas than it has been to fact check de Vries and Reece. All of this indicates to me that Lomas is reliable.

Conclusion To summarise: Lomas is used as a source by an award winning non-fiction biographer, has done a ton of original research that involved reading through archived material, provides quotations from primary sources, and gives consistent and reliable citations. This results in Lomas being an arguably more reliable source on the topic than both de Vries and Reece. However, according to current policy, he cannot be used as a source alongside de Vries and Reece. I argue that current policy should be amended in the given or similar form. I am not arguing that self-published sources be allowed to be used without their reliability being demonstrated.

Happy to answer any questions, clarify any statements, provide quotations, etc. if people wish.

Criticism of Lomas and response to potential questions

A failing of Lomas fails is a lack of polish in terms of his grammar, typos and some of his citations' page numbers being off by a couple pages (which, while sloppy, is again better than the other biographers). This is where most an editor's attention would have gone to, and have been most useful. Why not use the other published secondary sources? They are being used, but in the interest of using as many reliable sources as possible, and giving a balanced article that avoids giving undue weight to misinformation, Lomas's book should not be barred from being used as a reliable source. Why is it self-published? If I had to guess I'd say that it is because he is retired and didn't want to go through the headache that is publishing. If you have experience with this you know that it can be difficult. This is only conjecture though.

FropFrop (talk) 07:32, 21 December 2024 (UTC)

@FropFrop to clarify, Lomas has never had any work published elsewhere?--3family6 (Talk to me | See what I have done) 12:42, 21 December 2024 (UTC)

As far as I'm aware, no. The author never states it explicitly but my judgment is that it was a retirement project. However, he does also speak French (and maybe dutch?) so he could have published somewhere that I haven't found.

FropFrop (talk) 01:05, 22 December 2024 (UTC)

• I haven't heard the other side of this argument (if there is one other than Wikipedia:Status quo stonewalling) but, so far, I find FropFrop's presentation well-written and convincing. - Butwhatdoiknow (talk) 16:26, 21 December 2024 (UTC)

• I'm a bit confused by "When the source's claims can be clearly and non-controversially verified and deemed as reliable." If the source's claims can be clearly and non-controversially verified and deemed as reliable, then why wouldn't you instead use the sources that reliably verify the information? FactOrOpinion (talk) 17:57, 21 December 2024 (UTC)

  In practice, that line amounts to "whenever we want to". WhatamIdoing (talk) 17:59, 21 December 2024 (UTC)

  I think that we're capable of taking due diligence when it comes to these things, but would a narrower wording help? Perhaps changing that line to "When the source's claims can be clearly and non-controversially verified and deemed as reliable. This is to be done by the editor/s providing quotations and the source's citations so that other editors can discuss the content and form consensus on its reliability." or something to that effect? FropFrop (talk) 01:14, 22 December 2024 (UTC)

  Sources aren't required to include citations, and most of them don't. If the source does have (good) citations, then why not find, read, and cite those sources instead? WhatamIdoing (talk) 03:20, 22 December 2024 (UTC)

  I could do that, but then I'd be relying on primary material (which I want to avoid) and it would require a lot more work on my part. A lot of this archived material only has physical copies available. I've verified some of it, particularly at the beginning when I was judging how reliable Lomas was but this was quite time consuming.

  FropFrop (talk) 01:10, 23 December 2024 (UTC)

  I could do that, but then I'd be relying on primary material (which I want to avoid) and it would require a lot more work on my part. A lot of this archived material only has physical copies available. I've verified some of it, particularly at the beginning when I was judging how reliable Lomas was. FropFrop (talk) 01:10, 22 December 2024 (UTC)

  I too think this would be a bad proposal since it would effectively open the door wide for OR. What FropFrop seems to argue for here (and as they are constantly arguing on Talk:Daisy Bates (author)) is essentially: "I did the OR, following up those sources and so I consider this credible." But for good reasons we don't allow OR here, because such judgements are better left to the subject-matter experts at hand. Without OR, "clearly and non-controversially verified and deemed as reliable" simply means: There's another reliable source that says so. But if that's the case, then we can just cite that source and no change to SELFPUB is needed. Gawaon (talk) 10:02, 23 December 2024 (UTC)

  That's not OR. If a source has WP:Published something, and an editor checks the source's footnotes just to be sure, then that's not an editor making up stuff that isn't in any published source. WhatamIdoing (talk) 01:17, 24 December 2024 (UTC)

  But if the editor decided based on this check that a self-published source is reliable and should be admitted, then surely it is OR – what else could it be? It's not something that could be decided on the base of the source alone. Gawaon (talk) 10:54, 24 December 2024 (UTC)

  @Gawaon, the definition of OR, from the first sentence of that policy, is:

  On Wikipedia, original research means material—such as facts, allegations, and ideas—for which no reliable, published source exists.

  So:

  • The word material means "stuff we put in articles". If it's not going in the article, then it's not OR.
  • If a published source says ____, and editors deem that published source reliable for saying ____, then saying ____ in the Wikipedia article is not OR.
  • Checking whether a source (self-published or otherwise) is reliable and should be admitted is required of all editors, every single time they cite a source. Sometimes this is quite easy (e.g., a newspaper you know is widely cited, a book you happen to know is reputable), and sometimes it requires effort or a trip to RSN, but determining for yourself whether the source is reliable is normal, expected, and desirable behavior.

  On a separate, related note, you might be interested in the old concept of source-based research, about which the NOR policy used to say Research that consists of collecting and organizing material from existing sources within the provisions of this and other content policies is encouraged: this is "source-based research", and it is fundamental to writing an encyclopedia.

  Source-based research is not about determining whether a source is reliable, but it does involve determining the strengths and weaknesses of sources, and sorting out contradictions, such as why some sources give −1.592 × 10⁻¹⁹ coulomb as the charge of an electron and others give −1.602 x 10⁻¹⁹ coulomb instead (e.g., different POVs? Different time periods? Different circumstances? One of them's just wrong?). If an editor decides that the best way to settle the question is to find and read the published+reliable+primary sources that the cited source names, then that's okay. We don't ban editors from reading the sources cited by our sources. Sometimes it even helps us get article content right (e.g., by figuring out whether our source's claim about "the average" should be linked to Mean or Median). WhatamIdoing (talk) 20:32, 24 December 2024 (UTC)

  Sure, but can an editor's own research promote a non-reliable (e.g. self-published by an unknown person) work into a reliable one? That's what we're discussing here, right? Gawaon (talk) 21:43, 24 December 2024 (UTC)

  A source is reliable if there's a consensus among editors that it's reliable. I've told this joke many times, but here it is again:

  • Three baseball umpires are talking about their profession and the difficulty of making accurate calls in borderline cases. One says: "Some are strikes, and some are balls, and I call them as they are." The next feels a little professional humility is in order and says: "Some are strikes, and some are balls, and I call them as I see them." The third thinks for a moment and says: "Some are strikes, and some are balls, but they ain't nothing until I call them."

  This is how Wikipedia works: A source isn't "non-reliable" until editors make the call. Editors might well decide that Lomas is reliable for a given bit of article content. They also might decide that it's not. But nobody's "promoting a non-reliable source into a reliable one", either through their own hard work to establish whether the source is one that we should rely upon or through any other means, because it's the community that ultimately makes the call, not just one editor. WhatamIdoing (talk) 02:27, 25 December 2024 (UTC)

  It's also essentially impossible for other editors to verify, without repeating all the research by themselves. Gawaon (talk) 10:54, 24 December 2024 (UTC)

  So? See Wikipedia:Reliable sources/Cost.

  But: It doesn't matter. If one editor decides that they want to make sure that a source is really, truly, absolutely backed up by the sources that it cites, then they're allowed to do that. You can decide whether you trust them; you can decide for yourself whether you think they're lying; you can decide whether you think their evaluation is incompetent. But you can't say "You did a lot of work, but I'm unwilling/unable to replicate it, so we have to ignore that". Some people spent a lot of time learning other languages or studying advanced math or obsessing about the reputation of academic journals. I can't verify their conclusions without repeating all that work myself, but that's okay. I don't have to. They are not limited or restricted to the level of work that I choose to do. WhatamIdoing (talk) 20:36, 24 December 2024 (UTC)

if there is one other than Wikipedia:Status quo stonewalling Thank you for that! Another editor has been quite frustrating and has been citing WP:SELFPUB and has not engaged with any of my requests or offers of broader discussion. Going so far as to remove all citations of Lomas (white not editing the content to remove information I gathered from him, saying that it'll be done later).

FropFrop (talk) 01:09, 22 December 2024 (UTC)

Same here. In particular, I find the “Indicators that Lomas is reliable” convincing. — Charles Stewart (talk) 05:36, 22 December 2024 (UTC)

@FropFrop, I don't think your second point ("When an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications, uses the self-published source as a reliable source in a publication") will work.

Imagine that someone self-publishes (e.g., posts on social media) a manifesto for some obviously wrong view (e.g., Flat Earth, tinfoil hats, coffee tastes good, whatever you want). Alice Expert cites this manifesto in a high-quality reliable source as evidence that this view exists (e.g., "These views appear to be genuinely held by some people. The 'None of This Nonsense' Manifesto, which went viral in 2023, lays out four primary reasons for believing in magical dragons, including...").

That would be an instance of an expert who "uses the self-published source as a reliable source in a publication", and we still don't want editors to use the self-published manifesto directly themselves. WhatamIdoing (talk) 17:57, 21 December 2024 (UTC)

Speaking of manifestos, imagine this being applied to murder/suicide notes. The UK news yesterday had a story about someone sending a scheduled social media post announcing his suicide. So: the self-published announcement got "used" as "a reliable source in a publication" (several, actually), and this would therefore make the self-published announcement 100% okay to use in articles (except that to the extent that it would violate MOS:SUICIDE). That's what we don't want. WhatamIdoing (talk) 18:01, 21 December 2024 (UTC)

This is a very fair point.

To avoid such potential scenarios, what do you think of the following:

• When an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications, uses the self-published secondary source as a reliable source in a publication. [edit put in italics]

Would the addition of "secondary" and the (already included) "as a reliable source" be sufficient? Or would the latter need greater specificity?

FropFrop (talk) 01:25, 22 December 2024 (UTC)

The problem I have with this proposal it that it would mean that a single mention in a reliable source would be enough to promote all of a self-published work as "reliable/useable" for us, including (and especially) the parts what weren't mentioned in the reliable source. Specifically, in this case: the only real argument (besides OR) I have read from you re Lomas's reliability is that he's cited in Eleanor Hogan, Into the Loneliness (NewSouth, 2021) which is indeed a reliable source. But in that whole book Lomas is cited just once, with two sentences sourced to him (for a full quote, see Talk:Daisy Bates (author), my comment from 09:35, 23 December 2024). Based on that his work is mentioned just this single time in her whole book on Bates I'd rather conclude that she doesn't consider him particularly credible, otherwise she would certainly have cited him more often. By contrast, de Vries and Reece are both cited more than a dozen times – clearly they are more reliable sources for her. In our article on Bates, largely rewritten by you, Lomas is cited about 35 times, and you want to keep all those references based on this single mention by Hogan. However, if we want to cite the single fact which Hogan attributes to Lomas, we can just cite her book without having to mention him at all. As for the other 35 facts you would like to cite, there's no evidence that Hogan considers any of them credible, and so I don't see how this single mention by her could be a basis for the extensive use of his book which our article currently makes (in clear violation of SELFPUB as it currently stands). Gawaon (talk) 10:23, 23 December 2024 (UTC)

Yes, it is only once. I made this point as part of a larger argument. If this were the only point that support's Lomas's reliability, then I wouldn't make the larger argument.

Perhaps then the amendment could be worded like so (taking some of the adjustments made by @FactOrOpinion):

When assessing reliability, be especially wary when the source is self-published, and doubly so if it's being used as a source about a living person. Self-published sources are more likely than non-self-published sources to be unreliable sources for WP content, though there may be mitigating considerations, for example:

• If the content comes from a reputable organization and involves information such as who works for them or who they gave an award to.
• The content falls under ABOUTSELF.
• The content is produced by a subject-matter expert whose work in the relevant field has previously been published by reliable, independent publications.
• When the source's claims can be clearly and non-controversially verified and deemed as reliable. Editors should provide sources' quotations and sources citations so that other editors can discuss the content and form consensus on its verifiability.
• When an established subject-matter expert, whose work in the relevant field has previously been published by reliable, independent publications, uses the self-published secondary source as a reliable source in a publication.

If a self-published source meets one or more of the above examples, that does not automatically warrant its use as a reliable source; careful consideration of its reliability should still be made. Keep in mind that if the information in question is suitable for inclusion, someone else will probably have published it in a non-self-published reliable source.

FropFrop (talk) 02:11, 24 December 2024 (UTC)

Your points 1+2 essentially seem to describe or refer to WP:ABOUTSELF, since that already exists as an independent section, there's no reason to repeat it here. About your points 4+5 I have already explained in other comments here why I think they would be unworkable and weaken the basis of reliability at which the aim. Nothing has changed in that regard. Gawaon (talk) 10:59, 24 December 2024 (UTC)

When the source's claims can be clearly and ... If a self-published claims can be verified using other reliable sources, use those sources instead.

...uses the self-published secondary source as a reliable source in a publication This is treading on the toes of WP:USEBYOTHERS. If a source is widely cited in other high quality sources it could be considered reliable. -- LCU ActivelyDisinterested «@» °∆t° 11:20, 24 December 2024 (UTC)

Overall, looking at the specific situation, I wonder whether WP:IAR might be a better solution. We generally try to avoid changing overall rules just to deal with a situation at a single article. Perhaps an RFC explaining the situation? Or just asking "Is Lomas' book reliable for <this exact sentence>?" WhatamIdoing (talk) 18:07, 21 December 2024 (UTC)

Most of what I posted originally I took from an ongoing discussion on the article's talk page. An editor has been very firm on this being a gross violation of WP:SELFPUB and does not take WP:IAR or WP:ADHERENCE seriously. Neither have they engaged with my arguments for why Lomas should be an exception to WP:SELFPUB. They've been editing the article to remove all citations of Lomas (white not editing the content to remove information I gathered from him, saying that it'll be done later). FropFrop (talk) 01:18, 22 December 2024 (UTC)

It sounds like it's time for Wikipedia:Dispute resolution procedures. Maybe an RFC? I think I'd pick something small, like "Can we cite this book for this one sentence?" If you get agreement for the simplest/strongest single instance, then it will be easier to expand from there.

There have been several questions about self-published sources recently, including the one from Void if removed at #SPS definition and from 3family6 at #Are articles written by a publication owner/publisher reliable secondary sources, or are they self-published sources? and from FactOrOpinion at #editing the text of WP:SPS. The volume of questions, plus the rise of more "respectable" self-published works (this one, but also things like established fiction authors self-publishing books they like but their publisher didn't want to bother with) makes me wonder whether we need to re-think this concept entirely. WhatamIdoing (talk) 03:29, 22 December 2024 (UTC)

WhatamIdoing, FWIW, I've been trying to come up with wording for an RfC on the definition of SPS (checking the consensus on what people consider self-published, and whether the current definition and examples communicate that well). I think I understand the main views (I'm in the midst of rereading the discussions to check), but so far, I haven't been able to come up with something short; maybe there's just no way to make a short RfC about what I think is important to get at, or maybe it's a problem with how I'm thinking about it. I've also been waiting to see what the closer of the grey literature RfC says, though I gather that the timeline for that is open.

A couple of questions that came up for me vis-a-vis your view (I feel like I'm a font of never-ending questions; feel free to ignore them):

• You've said that you sometimes put governments in the traditional publishers category. For example, you've called the Census Bureau a traditional publisher and wrote "little-g government(s) ... are, in some respect, traditional publishers (e.g., of laws and reports)." I'm curious what guides your view about when the government is a traditional publisher vs. when it isn't.
• Both you and Void if removed have said that you think of traditional publishers in terms of the business model. Void if removed elaborated one traditional model: "an arrangement with a separate publisher is a business setup where both parties bring something (marketable content vs marketing infrastructure, connections, brand recognition etc)," where the publisher pays the author (content creator) and "If the publisher rejects the [content], then the author is free to sell it to a different publisher." That more or less works for books, freelance journalism, peer-reviewed journals, and probably movie documentaries. But it doesn't capture the structure of most print, electronic, radio, and TV journalism publishers, and probably not TV documentaries (assuming that you consider radio and TV journalism to fall in the traditional publisher category). How would you describe their business model(s)? And would you say that a business model is simply irrelevant to governments as traditional publishers?

As for FropFrop's proposal, it strikes me as more about who counts as an expert than about what counts as self-published. FactOrOpinion (talk) 15:07, 22 December 2024 (UTC)

I don't think "free to sell it to a different publisher" is a given. 3rd party publishing can still prevent the author from self-publishing or selling it on to another publisher. This is pretty much standard practice in music publishing, with all sorts of high profile cases of artists contractually unable to release their own material, while the publisher refuses to publish finished work.

The thing about traditional publishing is that it encompasses more than just "making material available", but for sourcing purposes, we consider "making material available" to be all that really matters to be "published". So the distinction is sometimes unclear - its on a website either way, so what's the difference? But a publisher will invariably be responsible for matters like advertising, promotion, legal due diligence, complaints, distribution in physical media, and so on.

And really, the only reason it matters is because of WP:BLPSPS.

So after all this I think the real question is: what level of sourcing do we need for contentious claims about 3rd party BLPs?

Much of this debate around the edges of what is or is not an SPS has that goal in mind. Being traditionally published (ie a book, newspaper, magazine or journal) is to my mind a reasonable proxy for "you can use this to make a contentious claim about a 3rd party BLP, if its due", because a publisher is on the hook for defamation as much as the author. And that is probably the extent to which I care about whether a source is SPS or not. Void if removed (talk) 15:37, 22 December 2024 (UTC)

Thanks for your "free to sell it to a different publisher" point, agreed, though I don't think the SPS restriction matters much for something like music, since I don't see that being used for fact or opinion content, only for content that falls under ABOUTSELF. Ditto for things like TV dramas/comedies and non-documentary movies, though I'm still unclear about whether you'd say those are published by traditional publishers. Although you and What am I doing have highlighted business models, I don't recall others doing so, so that's why I'm trying to understand how you describe traditional publisher business models. Re: the other things that a publisher is responsible for, don't they also apply to publishers that you consider non-traditional (e.g., if GLAAD publishes something defamatory, isn't it on the hook too)?

I think your description of the "real question" is a significant part of the real question, but not all of it, as FropFrop's proposed change shows. (And I've encountered similar questions on RSN.) Also, right now, SPSs can't be used as BLP sources even if the claim is non-contentious.

As best I can tell, SPSs are singled out because editors think SPSs are much less likely to be RSs. For example, the WP:SPS section appears in a larger section titled "Sources that are usually not reliable," the current definition refers to the lack of an independent editor "validating the reliability of the content," and an early ArbCom conclusion (referred to by What am I doing above) said "A self-published source is a published source that has not been subject to any form of independent fact-checking ..." (That text was introduced into WP:RS in 2006, and although there was text in WP:V about self-published sources at that point, there was no attempt to define self-published, and the examples were totally limited to situations where one or a few individual persons had total control over whether their own work was published.) I assume that this is why the expert source and ABOUTSELF exceptions exist: self-published content written by experts in their area of expertise is much more likely to be reliable than other SPS content, and SPSs are often reliable sources about the author (but not about others or if the content is too self-serving; for that matter, we also have to beware of non-self-published sources producing self-serving content, as might occur in their marketing material, though as best I understand, you and What am I doing always consider marketing material to be self-published).

So I think the underlying issue is assessing whether a SPS is reliable for the content in question. When it comes to the SPS policy, it may be sufficient to say something like

When assessing reliability, be especially wary when the source is self-published, and doubly so if it's being used as a source about a living person. Self-published sources are more likely than non-self-published sources to be unreliable sources for WP content, though there may be mitigating considerations, for example, if the content comes from a reputable organization and involves information such as who works for them or who they gave an award to, or the content falls under ABOUTSELF, or the content is produced by a subject-matter expert whose "work in the relevant field has previously been published by reliable, independent publications" or falls into the situation that FropFrop introduced. Keep in mind that if the information in question is suitable for inclusion, someone else will probably have published it in a non-self-published reliable source.

However, that still means that we need to be clearer about what does/doesn't constitute self-published material. FactOrOpinion (talk) 19:43, 22 December 2024 (UTC)

This all sounds good to me (with the amendments added after WhatamIdoing pointed out some potential issues).

FropFrop (talk) 02:30, 23 December 2024 (UTC)

Void if removed, if you don't mind a couple more questions ...

In the RSN discussion of SBM, you said "We have multiple highly debatable and contested terms, at the heart of a core policy, and radically different interpretation of them."

• I'm guessing that in addition to "self-published," you're thinking of "author" and "publisher." Are there any other terms (besides "self-published," "author," and "publisher") you think are contested?
• For each of these terms, if you have a sense of the different meanings attributed to them, would you say what they are? (For ex., I know that you've distinguished between a publisher as "the person who publishes" vs. as "a business whose business is publishing.")

Thanks! FactOrOpinion (talk) 15:34, 23 December 2024 (UTC)

So, some examples gleaned from the various debates:

• Author could mean anything from a single individual who wrote a blogpost, to the company whose unnamed employees were instructed to write content.
• Publishing could mean the trivial act of placing some content online, or a business arrangement whereby a publishing company takes responsibility for the distribution of content, along with marketing and various other concerns.

And as I mentioned in the previous debates, in English "publisher" (a publishing company) and "publisher" (a person who did the act of publishing) are the same word, and this leads to confusion. Depending on how you interpret all the above, self-published can mean anything.

By the narrowest definition of self-published, only something like a single-author blog where one identifiable person both wholly wrote and trivially placed content online is "self-published".

By a broader definition, a corporate website where company employees write content at the behest of the company and it is placed online at the sole discretion of that company, is also self-published, because the author (the company) is the publisher.

By a maximal definition, anything where there is not a traditional, commercial publishing structure, ie a book publisher, a journal, or traditional news media, is self-published.

Many of these debates get derailed by discussions about why we should care, and because many of us are laser focused on "published" as meaning "I can read it online", the commercial aspects of traditional publishing arrangements seem like a total irrelevance.

So, with that aspect usually ignored, most seem to argue that we care something is self-published because we hope for some level of independent oversight, and so editors might point to "editorial oversight" as proof something is not self-published. This is the case for SBM, where a group blog set up and run by individuals who are also presently its sole editors somehow has been decided to be "not self published".

But that, to me, seems like a hack. Editorial approval is not what defines "self-published", its just one component of how we assess a source's reliability and accountability. I think this is part of the incorrect conflation of SPS with "unreliable", when SPS and RS are really parallel concerns, just as PRIMARY and SECONDARY are. Trivial approval steps between someone writing copy and another person ticking a box or pressing "approve" on a blog are not sufficient to make something not self-published. As I've given as an example previously, by this minimal definition, a celebrity social media account run by a PR agency is therefore not "self-published", which seems to be a nonsensical interpretation.

Another common issue is publishing uncontentious information (like scientist X won Y award), and so different interpretations of "self-published" emerging to get round the BLPSPS restriction.

By my reading of past discussions we care very specifically about third-party BLPs because of the interplay between a) the low threshold for anyone to put any nonsense they like online and b) the possibility of that being defamatory.

And my interpretation then is a traditional, commercial publishing arrangement is the thing that provides some level of confidence that Wikipedia is not going to be on the hook for defamation, where "a blog owner approved their mate's guest post before pressing submit" does not.

So if I were to propose something it would be something straightforwardly restrictive like:

• Any non-trivial, negative or otherwise opinionated claim about a 3rd party BLP must be sourced to something traditionally published (presumably with lawyers who checked this claim before going to press), ie a book, magazine, journal or newspaper.

When it comes down to it, I think this is what BLPSPS is all about, and I think if you settle it that way all the argument about what is or is not self-published ceases to be quite such a concern. Void if removed (talk) 16:27, 23 December 2024 (UTC)

Lawyers don't normally check any claims during the publication process. A large publisher will have some on staff, and a medium-sized one will have some on call, but (a) there is no routine legal review of content and (b) even when something gets flagged to their attention by other employees, they're not actually engaged in fact checking. They're only evaluating the information that's handed to them, to determine how much risk the company will be exposed to.

This should be obvious if you think about it. Imagine the ordinary working of an ordinary daily newspaper: The city had a meeting, and the newspaper sent a reporter to it. The reporter comes back with notes and writes an article. The editor looks it over. The editor might ask some questions or make suggestions. After any agreed-upon changes, the article appears in the next morning's paper. There is no "get approval from a lawyer" step in the editing process (unless the editor deems it necessary, and that's going to be an unusual circumstance). There's not even an independent, pre-publication fact checking step. WhatamIdoing (talk) 01:31, 24 December 2024 (UTC)

FropFrop, the RFC I have in mind would be posted at Talk:Daisy Bates (author) and would say something like this (pick any sentence you think is appropriate):

Can we cite this book:

Lomas, Brian (2015-10-29). Queen of Deception: The True Story of Daisy Bates. CreateSpace Independent Publishing Platform. p. 38. ISBN 978-1-5170-5385-7.

to support this (currently uncited) sentence in the article: "Eventually arriving in Broome, she boarded the Sultan with Father Martelli and arrived in Perth on 21 November 1902"? The relevant page says "<insert direct quotation here>" and cites a letter in the archives at Big University. This might seem like overkill, but while I was at the archives last month, I found the letter Lomas cites and verified that the letter supports Lomas' claim.

When you are in the middle of a content dispute, changing the policy so that you can win the dispute is not usually the best approach. Settle the local dispute first. WhatamIdoing (talk) 23:14, 22 December 2024 (UTC)

When you are in the middle of a content dispute, changing the policy so that you can win the dispute is not usually the best approach. Settle the local dispute first. That's what I tried to do prior to writing this up, as I imagined that my post here wouldn't be taken well due to it being an active dispute. However the other editor said that I should try and change policy before trying to make the argument for inclusion. Because the conversation seemed to be going nowhere, I followed through with that suggestion.

Regardless, I'm happy that my post seems to be going well and has sparked some genuine conversation on the broader issue. I'll try and see the dispute through.

FropFrop (talk) 01:18, 23 December 2024 (UTC)

FactOrOpinion, I don't think that we should focus on wording for an RfC on the definition of SPS. Specifically, I think it's a bad idea for Wikipedia's definition to diverge from the ordinary definitions. I encourage you to think more about clarifying "Wikipedia's rules about when it's acceptable to cite self-published sources" instead of "creating a Wikipedia-specific definition of the word self-published".

For your specific questions:

• I think that it is not unreasonable to consider a government agency to be a traditional publisher if its main duties revolve around publishing information. This means, e.g., that the US Census would be considered traditionally published, but an announcement from that same government agency that they're having a public meeting, or that they have job openings, would not count as traditionally published. Compare:
  • The book publisher Bloomsbury traditionally published Harry Potter and self-publishes current job openings on their corporate website.
  • The government agency United States Census Bureau traditionally published 2020 United States census and self-publishes current job openings on their government website.
  • An elementary school does not traditionally publish anything, and self-publishes current job openings on its government website.
• It's true that the structure of most print, electronic, radio, and TV journalism publishers does not always allow for selling rejected works to others. Journalists (and musicians, as mentioned above) are allowed to negotiate contracts that allow a publication exclusive control over whether their work gets published. In the case of ordinary journalists, they get paid the same salary whether the article runs or not. I'm not sure why this question has arisen. The inability to take a document elsewhere for publication isn't necessarily proof of anything, but it's associated with traditional publishing.

WhatamIdoing (talk) 00:21, 23 December 2024 (UTC)

Thank you for your answers to my questions. Re: "I'm not sure why this question has arisen," it's because you've said things like "Being a traditional publisher is about the business model," and it seems to me that traditional publishers have more than one business model, and I'm trying to understand which business models count as traditional publishing. For ex., when it comes to the government as a traditional publisher, arguably the idea of a business model doesn't even apply.

I do understand that you think it's a bad idea for Wikipedia's definition to diverge from the ordinary definitions. You'd already said as much earlier (e.g., "Yes, it's true that our terms sometimes diverge from ordinary words, but self-published in this policy is supposed to be the ordinary dictionary definition. It is not supposed to be some kind of wikijargon").

Some problems with that approach:

• Dictionaries don't all define "self-published" in the same way. Looking across a number of different dictionary definitions, I've seen two main features highlighted: (1) whether the author pays for the work's publication, and (2) whether the author uses a publisher (with variations such as "publishing company" and "established publishing house"). Some definitions highlight (1), some highlight (2), and some highlight both. Although (1) and (2) intersect, they're definitely not the same (e.g., material written by an employee is not self-published according to the first, but may be self-published according to the second). Also, some definitions refer to an "author," which may be ambiguous (e.g., by author, do they only mean the specific person(s) who wrote/created something, or are they also including corporate authorship?).
• A number of editors clearly don't agree with your preferred definition / don't think it represents practice.
• You've said things like "most traditional publishers have ... [some] self-published content (e.g., marketing materials, investor relations reports, advertising rate sheets)," but I don't see that carve-out in any dictionary definitions. (Maybe I've missed it.)
• To the extent that WP:SPS defines self-published (in a Reference note at the bottom of the WP:V that says "Self-published material is characterized by the lack of independent reviewers (those without a conflict of interest) validating the reliability of the content"), that definition already diverges from dictionary definitions. So if your goal is to get people to use a dictionary definition, and especially to choose your preferred dictionary definitions over other dictionary definitions, you'd need an RfC to align policy with your preference.

FactOrOpinion (talk) 02:46, 23 December 2024 (UTC)

I believe that there are some government agencies that sell documents they publish. The US federal government has some limitations on exclusivity, so the costs are usually limited to reimbursing the cost of printing and distributing, but there's nothing inherent in a government publishing office that would prevent them from using the same business model as other traditional publishers. WhatamIdoing (talk) 06:05, 23 December 2024 (UTC)

Point taken. But it seems to me that traditional publishers do not all use the same business model. One model is the one I originally quoted from Void if removed, where there are a couple of variations (the person can take their work elsewhere if a given publisher rejects it, or they sign a longer term contract giving one publisher exclusive rights to future work even if the publisher decides not to publish it). Another business model is the one used with non-freelance journalists, and if a government sells documents created by employees, perhaps that falls under the same model. I'm not sure if you'd say that those are the only two business models that a traditional publisher might use. FactOrOpinion (talk) 14:42, 23 December 2024 (UTC)

The business model is mainly: I publish this (book, news article, whatever) because I think people will pay me for it. WhatamIdoing (talk) 01:33, 24 December 2024 (UTC)

That seems to apply to some self-published material too (e.g., some self-published books, an individual's Substack with subscribers), though it clearly doesn't apply to other SPS. FactOrOpinion (talk) 01:57, 24 December 2024 (UTC)

Yes, but the theory is that authors are less likely to judge the market correctly. Everyone believes their first novel to be a masterpiece, until they've written ten more. WhatamIdoing (talk) 07:57, 24 December 2024 (UTC)

"We generally try to avoid changing overall rules just to deal with a situation at a single article." This doesn't strike me as a single article issue. If it was, we'd just cite wp:KUDZU and be done with the discussion.

Instead, it appears to be an example of a fundamental change in the publishing industry issue.

Or maybe not a change. FropFrop's example shows that publishers are not the same as fact checkers, which may have been true all along. In that case Fact or Opinion's "more about who counts as an expert" is a better way to look at it. - Butwhatdoiknow (talk) 15:39, 22 December 2024 (UTC)

To my mind, the “fact vs opinion” question has always been the key here. I have long thought that SPS material should always be presented as the author’s opinion (ie with in-text attribution). This shifts the debate from WP:V to WP:DUE. The citation reliably verifies that the author said what we say he/she said… the more important question is whether it is appropriate for us to note this author’s opinion in X specific article. Blueboar (talk) 22:46, 22 December 2024 (UTC)

FropFrop, I must admit I'm somewhat annoyed you started this discussion behind my back. It's true I had suggested you might try getting SELFPUBLISH changed, but I wasn't monitoring this page and had no idea you had actually decided to do so. Discussions shouldn't silently be forked and continued elsewhere without the involved editors being informed, so I don't think it was fair what you did here. Gawaon (talk) 09:53, 23 December 2024 (UTC)

Now that you know about this discussion, what is your opinion regarding the substance of FropFrop's proposal? - Butwhatdoiknow (talk) 17:17, 23 December 2024 (UTC)

See the two comments I wrote above regarding specific details of the proposal. In a nutshell: I think the suggested changes are unworkable and would make SELFPUB considerably worse. And, thought that discussion rather belongs on Talk:Daisy Bates (author), Lomas seems to be a doubtful and considerably biased author, hence I'd say that his is just the kind of book SELFPUB is meant to protect us against – and that's how it should remain. Gawaon (talk) 19:13, 23 December 2024 (UTC)

SELFPUB isn't really meant to protect us from bias (and biased sources can be reliable anyway). It's meant to protect us from things like authors being reckless with the facts, and to focus due weight considerations on content that's newsworthy/non-self-published. WhatamIdoing (talk) 01:42, 24 December 2024 (UTC)

I agree to some degree, which is why I meant that discussion belongs on the article talk page rather than here. Still I think some overlap is likely – biased authors are more likely to distort the facts in order to get their point across, and I'm pretty sure that that sometimes happens in Lomas. Indeed FropFrop, in an earlier comment on Talk:Daisy Bates (author), already admitted that his bias sometimes leads him to problematic conclusions ("He does get overly critical at times (often assuming that Bates outright lied when it could have been an honest mistake)"). Academic peer review should normally help to reduce such issues, forcing authors to be more honest with themselves and their audiences; and the editorial process though which nonfiction books go with non-academic publishers should have the same effect, at least to some degree. I know very well that these processes will never work perfectly, but I see them as good "filters" that should make our work better, and hence I'm weary of self-published sources, which didn't make it through these filters. Gawaon (talk) 11:09, 24 December 2024 (UTC)

Non-academic book publishing often encourages overblown claims, because controversy results in free publicity, and publicity improves sales.

I agree that traditional publishing is a good filter for us. If we say that 50% of traditionally published books are bad for us, then I'd start with an assumption that 95% of self-published sources are bad for us. But it's also possible that sometimes, probably rarely, a self-published book will be an okay source for us. WhatamIdoing (talk) 20:47, 24 December 2024 (UTC)

I agree, and there's already an exception in SELFPUB for some such works. I doubt, however, that the additional exceptions discussed here would be great at finding additional okay sources, and rather fear that they would allow slipping a lot of not-okay sources. Gawaon (talk) 21:41, 24 December 2024 (UTC)

For the general case, I agree with you. For this specific case, however, it's possible that Lomas is one of the few "okay" sources instead of the many "not-okay" sources.

What I'd suggest to you is that you try to disentangle your worries about bias (How dare he call her "the queen of deception", just because she told so many lies about herself!) from your worries about reliability. What matters for that article is not whether this long-dead person told lies, or whether the recent sources called it what it was. What matters for that article is much more mundane: Is Lomas correct when saying that she arrived in Place A on Date B? WhatamIdoing (talk) 02:34, 25 December 2024 (UTC)

Ah, apologies. As you suggested that I try and amend the policy I didn't think this would be an issue.

FropFrop (talk) 02:14, 24 December 2024 (UTC)