When discussion has ended, remove this tag and it will be removed from the lists. If this page is on additional lists, they will be noted below.
This RfC is to determine the consensus about (1) whether the current explanation of "self-published" in WP:SPS generally serves us well, perhaps with small improvements, or if it should be revised in some significant way, and (2) how editors interpret "self-published," in order to help us revise the explanation if needed.
FactOrOpinion (talk) 21:02, 26 January 2025 (UTC)[reply]
RFCBEFORE discussions took place here (a disagreement about whether material published by GLAAD is self-published), here (a more general discussion of what "self-published" means), and here (an RfC: "Should grey literature from advocacy groups and other similar orgs always be considered WP:SPS and therefore subject to WP:BLPSPS?"). However, disagreements about the interpretation of "self-published" go back much further than the RFCBEFORE discussions; these examples from 2020 (here, here and here) and 2021 circle around many of the same issues. Notes from previous discussions is an attempt to summarize key issues raised in one or more of these discussions.
Notes from previous discussions
Notes from previous discussions
Sorry if this feels too long to read (though it's a lot shorter than reading the preceding discussions!). People raised lots of issues, and this is my imperfect attempt to capture the most salient. I've tried to remain neutral in the sense of including people's varied perspectives; however, specific views below may not be neutral, as people sometimes had strong views.
Categories of publishers
Some editors distinguished among different categories of publishers:
Organizations such as newspaper and magazine publishers, television broadcasters, non-vanity book publishers, publishers of peer-reviewed journals, record labels representing lots of artists. Some people call these "traditional" publishers, characterizing them as being in the business of publishing.
Organizations such as advocacy groups, universities, learned societies, think tanks, corporations, international non-governmental organizations, intergovernmental bodies, museums, foundations, charities, labor unions, and political campaigns. Some people call these "non-traditional" publishers. While they may publish quite a bit in the context of their main mission, they wouldn't describe themselves as primarily being in the business of publishing.
Governments might be sui generis. They have huge variations in size, differ both across and within countries, and they publish quite diverse types of materials.
Depending on how you interpret "self-published," a single publisher might publish a mix of self-published and non-self-published material. You might also conclude that some publishers have an arm that functions like a "traditional" publisher and another arm that doesn't (e.g., a government's publishing office versus its defense department, a professional society's peer-reviewed journal versus its advocacy arm).
General areas of consensus
There seems to be consensus about the self-publishing status of some kinds of publications:
Materials like the following are self-published: personal websites, personal or group blogs (as distinguished from newsblogs), social media posts, wikis, preprints, reader comments on websites, music/games released under the creator's own label, internet forum posts, vanity press books, patents, unscripted podcasts published by the podcaster, individual Substacks, Forbes.com "contributors" material, Kindle Direct Publishing books, user reviews, a paid promo in a newspaper, and personal YouTube videos. There is no barrier to the creator(s) publishing — or paying someone else to publish, print or host — what they want, even if it's sometimes removed after the fact via post-publication moderation (e.g., a tweet that's removed for violating X's terms of service). Sometimes the author is one person, and other times, two or more people are authors (e.g., co-authored research), but corporate authors aren't included. In most cases, there is no editor, but if there is an editor, the editor cannot prevent publication. (In the RfC, I called these "no barrier" materials for ease of reference, though that term wasn't used in previous discussions.)
Most material from "traditional" publishers is not self-published, though there might be a few exceptions (see below). These publishers sometimes host no barrier materials (e.g., a newspaper article is not self-published, but reader comments on the article are, even if they are moderated).
Areas where consensus is unclear
Previous discussions have not resolved whether the following kinds of material are always/sometimes/never self-published, and if it's "sometimes," what features distinguish the self-published materials from the non-self-published ones:
Material from non-traditional publishers and governments.
Material where the content is about the organization itself, even if it's edited by an employee who can block publication, and even if the organization mostly publishes non-self-published material. Examples include marketing material for the organization's products (including advertisements, where a newspaper or TV station effectively serves as a vanity press), press releases, political campaign material, annual investor reports, "About us" text, advertising rate info, information about employees, information about how to exchange a product, a government's explanation of how to use its services, and information about employment with the organization. (In the RfC, I've called these "organization itself" materials for ease of reference, though that term wasn't used in previous discussions.)
Material published by a "traditional" publishing company, but written by someone with control over publication (e.g., the company's owner, the journal's editor).
Words with multiple interpretations, and dictionary definitions of "self-published"
WP:V states that "Source material must be published, on Wikipedia meaning made available to the public in some form," with a footnote adding "This includes material such as documents in publicly accessible archives as well as inscriptions in plain sight, e.g. tombstones." In contrast, The Chicago Manual of Style considers some documents in public archives to be unpublished. "Publisher" can mean "any entity that publishes," or instead be limited to "an organization in the business of publishing." Some editors use "publisher" when referring to a printer (e.g., of a dissertation) or a host/platform (e.g., a social media site, Kindle Direct Publishing); other editors say that "publisher" is distinct from "printer" and "host/platform." The word "author" can also be used in different ways. "Author" might mean "the human being(s) who created the work," or instead be used in a way that includes corporate authors. For material published by an organization, someone's interpretation of "author" may depend on whether the person who wrote it is named. Thus, in a discussion, the intended meaning of a word may be ambiguous, and participants' interpretations may differ.
Dictionary definitions of "self-publish(ed)" include:
"issued directly to the public by the author rather than through a publishing company" (Collins)
"to publish (a book) using the author's own resources" (Merriam-Webster)
"to arrange and pay for your own book to be published, rather than having it done by a publisher" (Cambridge)
"to publish or issue (one's own book or other material) independent of an established publishing house" (Dictionary.com)
"publish (a piece of one's work) independently and at one's own expense" (Oxford American)
"publish by oneself or with one's own money" (American Heritage)
"That is or has been published by oneself; chiefly spec. (of a book or other work) prepared and issued for distribution or sale by the author" (Oxford English)
In the definitions that use "author," it's ambiguous whether it's meant to include corporate authors or only natural persons. Some definitions highlight (1) whether the author pays for the work's publication, some highlight (2) whether the author uses a "publishing company" or "established publishing house," and some highlight both. Although (1) and (2) intersect, they're not the same; for example, if material is written by an employee and published by the employer, the material is not self-published according to the first (unless you treat the employer as a corporate author), but may be self-published according to the second. Self-published material need not involve a cost, as with social media or wikis.
Other considerations
In reasoning about what is or should be considered self-published, people drew on diverse considerations, and a single person's reasoning often involved several considerations. Below are additional facts/opinions/questions that various people introduced. A single paragraph may include contradictory claims from different people:
a)Overview/ use: The meaning of "self-published" has significant implications for which sources can be used for WP content, especially for content about living persons. BLP content is currently sourced to materials that some people consider self-published (e.g., material published by advocacy groups, universities, and governments); maybe some editors are misinterpreting WP:SPS, or maybe there's a gap in the policy. At times, editors' assessments of whether a specific source is or isn't self-published seem to be based on whether they do or don't want the source/POV to appear in an article, an appeal to consequences. This seems especially likely to occur with contentious topics (e.g., gender, politics). The many debates about what is/isn't self-published show that the current explanation doesn't work well enough. A clearer explanation would help reduce the time and energy spent in these disputes, and would help us determine whether some article content needs to be removed as a BLPSPS violation, or whether a source that some people thought was excluded under BLPSPS can actually be used. Some worry that narrowing the interpretation of SPS would create a BLP "minefield." Although WP:NOTBURO, many editors quote WP:SPS and WP:USINGSPS when debating whether a given source is/isn't SPS, and new(ish) editors also turn to the policy and the essay to figure out what they're supposed to do, so we want these texts to be clear and to represent consensus.
b)The explanation as a whole: The characterizations of "self-published" in the WP:SPS footnote and WP:USINGSPS differ. Some would like the essay's characterization to replace the one in WP:SPS, and others disagree. The WP:SPS characterization is overly broad, and some (or many) things are characterized as self-published when they should instead be characterized as non-self-published. Alternatively, the WP:SPS characterization is overly narrow, and some (or many) things are characterized as non-self-published when they should instead be characterized as self-published. We should use a dictionary definition, not "wikijargon." Dictionary definitions are easy to apply to some kinds of sources (e.g., books), but WP editors use many kinds of sources, and it may be unclear how a dictionary definition would categorize these other kinds. Outside of WP, people often use "self-published" only for no barrier materials.
c)Reviewer: The footnote characterization refers to a reviewer. Depending on the source, it may be hard to know whether material is reviewed by someone, and if so, whether that reviewer is in a position to block publication. Some think that an organization can be assumed to have a sufficient review process based on features such as size and positive reputation. Others think that whether an organization has a sufficient review process cannot be assumed, and it has to be demonstrated with an explicit editorial structure.
d)Conflict of interest: The footnote characterization refers to conflict of interest. COI is distinct from bias. How do we assess whether a conflict of interest exists? Is one of the interests always "reliability" (which WP never actually defines, though it is linked several times to a "reputation for fact-checking and accuracy"), and if so, what is the other interest that might or might not be in conflict? For example, is it the interests of the reviewer's employer, and if so, how do we determine what those are? Does a reviewer always have a COI when checking content about the reviewer's employer, but seldom otherwise? Is there always a COI if the author and reviewer both get paid by the same entity? If "conflict of interest" remains in the characterization, should it be linked to the mainspace COI article?
e)Reliability: The footnote characterization refers to validating the reliability of content. It may be hard to know whether a reviewer is assessing the reliability of the material; a reviewer might instead only be checking things like grammar and organization. The reliability of a source depends on what WP statement you want to source to it. Whether self-published material is likely to be more reliable than non-self-published material depends in part on one's interpretation of "self-published." Even if self-published sources are less reliable on average, the characterization conflates "self-published" and "reliable," when a source might be one, the other, both, or neither. Policies highlight the presumed overlap of self-published status and non-reliability in several ways. For example, this is why most SPS cannot be used as sources, and why the EXPERTSPS and ABOUTSELF exceptions exist. WP:SPS appears in a section titled "Sources that are usuallynot reliable," the current characterization refers to the lack of an independent editor "validating the reliability of the content," and an early ArbCom conclusion said "A self-published source is a published source that has not been subject to any form of independent fact-checking ..." (A bit of history: the text from that ArbCom quote was introduced into WP:RS in 2006, and although there was text in WP:V about self-published sources at that point, the explanation of "self-published" was limited to examples, where the examples were all no barrier materials. The first text about SPSs was introduced into WP:V earlier in 2006. There was no equivalent to BLPSPS. The first text about SPSs was introduced into WP:BLP in late 2005, and there too, the examples were limited to no barrier materials. The current WP:SPS footnote (characterization + examples + quotes) was introduced in 2011. It was initially a footnote and only became a reference note in 2023.)
f)More on reliability: Sources might be creative work (e.g., music, games, fictional books/TV shows/movies, poetry). Most of the time, they're probably used as sources for statements about their own content or structure, so it's often not critical to assess whether they are or are not self-published; even if they're considered self-published, the ways they're used on WP would often fall under ABOUTSELF. Still, the current characterization doesn't work for them, as they're generally not sources in which reliability would/could be assessed; thinking about what would lead you to say that a creative work is/isn't self-published may be helpful in thinking about how you interpret "self-published" more generally. It's also unclear what it means for a reviewer to validate the reliability of things like opinion pieces and interviews.
g)Other features: In assessing whether something is self-published, some people consider who is responsible for distribution and marketing, and who is responsible for legal matters such as copyright, liability, licensing, and contracts, though the legal responsibilities might vary by country.
h)Examples: The examples in WP:SPS would provide better guidance if some of them were removed (i.e., they're not examples of self-published material) and/or if some other examples were added (e.g., examples of material that isn't self-published, examples that better illustrate where the border is for self-published / not-self-published). Maybe we shouldn't give a characterization of the sort that appears in the footnote, and we should focus on giving lots of examples: adding some examples of non-self-published sources (identifying them as such), and adding some self-published examples that are less obvious.
Other things people mentioned, not about the characterization or examples of "self-published" per se:
The footnote includes a few quotes from sources, and depending on the RfC results, it may be time to update those.
About 10–15% of the sources listed in WP:RSP are currently identified as "self-published." Depending on the RfC's results, we may want to discuss the description of some individual RSP entries later. Depending on the outcome, either more or fewer RSP entries might be correctly described as self-published.
It's OK to leave the source quotes in a reference note, but the characterization should be moved into the body of the WP:SPS section.
We might think about changes to the WP:SPS text that aren't about the explanation of "self-published" per se. For example, should the EXPERTSPS text be modified to allow a group to qualify as "expert" in its field if academic and/or mainstream sources regularly treat the group as having expertise? (That may already be consensus practice, or perhaps people don't consider these to be self-published.) Should the text say that WP content sourced to EXPERTSPSs should always be attributed?
WP:ABOUTSELF and WP:BLPSELFPUB allow the use of self-published material in some cases, but it's unclear what is meant by "third party" in point 2 (e.g., if one considers a university's website to be SPS, can it be used for information about a professor, or does BLPSELFPUB preclude that?).
RfC questions
This RfC is solely about WP's interpretation of "self-published." It is not trying to assess whether a source is reliable, independent, primary, biased, etc., or whether its use is due or needs to be attributed or can help establish notability, as these aspects are distinct from whether the source is self-published.
Before you respond to the RfC questions, I strongly encourage you to open the Table below and think briefly about how you'd classify each of the examples. Options 2a-c will be explained below in Question 2, but even without knowing the specifics, you'll get a sense of how different ways of explaining "self-published" can lead to different results for how diverse examples might be classified. You can return to the Table later if it's helpful in understanding Options 2a-c.
Table illustrating differences in how the current explanation and options 2a-c categorize example sources as SPS or not
Table
Type of source
Current explanation
Option 2a
Option 2b
Option 2c
Personal social media post
SPS
SPS
SPS
SPS
Co-authored vanity press book
SPS
SPS
SPS
SPS
Government hearing transcript
SPS
maybe SPS speakers=authors, government=publisher or only printer?
maybe SPS might also depend on the topic
SPS
Foundation's webpage about its grant to an artist, plus artist's biographical info
would depend on the genre (fiction, memoir, poetry, etc.)
not SPS
not SPS
not SPS
The table shows that for some examples, all of the options classify the example as SPS or all of the options classify the example as non-SPS. These are examples of classes where editors' consensus is clear. The table also shows that for some examples, options vary in whether they classify the example as SPS or not. These are examples from the swath of publications where consensus is unclear.
Several of the cells for the current explanation are blank, as I've heard different editors express different opinions (for example, here are a number of quotes that WhatamIdoing collected from the WT:V archives where experienced editors stated contradictory views about whether Coca-Cola.com is self-published). Come to your own conclusion about each those cells: SPS, not SPS, or perhaps "the example's description doesn't provide enough information to know." You may want to say something about your conclusions in your !vote, especially if you chose answer 1a.
Question 1
WP:SPS explains the meaning of "self-published" with text in the body, supplemented by text in a longer footnote. The explanation as a whole is comprised of a link to the mainspace article on self-publishing, multiple examples, the statement "Self-published material is characterized by the lack of independent reviewers (those without a conflict of interest) validating the reliability of the content," and three quotes mentioning self-published material.
Full text of WP:SPS's explanation of "self-published"
Self-published material is characterized by the lack of independent reviewers (those without a conflict of interest) validating the reliability of the content. Further examples of self-published sources include press releases, the material contained within company websites, advertising campaigns, material published in media by the owner(s)/publisher(s) of the media group, self-released music albums, and electoral manifestos:
The University of California, Berkeley, library states: "Most pages found in general search engines for the web are self-published or published by businesses small and large with motives to get you to buy something or believe a point of view. Even within university and library web sites, there can be many pages that the institution does not try to oversee."
Princeton University offers this understanding in its publication, Academic Integrity at Princeton (2011): "Unlike most books and journal articles, which undergo strict editorial review before publication, much of the information on the Web is self-published. To be sure, there are many websites in which you can have confidence: mainstream newspapers, refereed electronic journals, and university, library, and government collections of data. But for vast amounts of Web-based information, no impartial reviewers have evaluated the accuracy or fairness of such material before it's made instantly available across the globe."
Consider issues such as whether the characterization of "self-published material" is a good way to characterize it, the explanation of "self-published" (link + examples + characterization + quotes) reflects consensus practice, the explanation provides sufficient guidance, and editors agree on how to interpret it. Which option best represents your view?
a) The explanation might benefit from small improvements, but it serves us well and we should keep it.
b) The explanation is problematic in some significant way(s), and we should figure out how to revise it.
If your answer is (a), propose small improvements if you want. If your answer is (b), please identify the main problem(s).
Question 2
The previous discussions show consensus that some classes of publications are self-published and other classes of publications are not self-published. But for a sizeable swath of publications, consensus is unclear. Options a-c describe three views from the previous discussions. Which view best captures the kinds of sources that you'd say are/aren't self-published? If an option represents your view pretty well but not exactly, just say how you'd modify it:
a) Self-published sources are those where there is no barrier to one or a few people (not organizations) publishing what they want, perhaps by paying some entity to publish, print, or host it. Examples include open wikis, internet forum posts, personal websites, music released by its creator(s), and preprints. Someone other than the writer/creator(s) may provide feedback or editing (e.g., an author hires an editor), but this other person cannot block publication. Everything else — including material published by diverse organizations — is not self-published.
b)No barrier materials are self-published. Sources are also self-published if they're published by an organization and the content is about the organization itself (e.g., "About us" text, an annual investors report, marketing material), even if these have been reviewed by someone who could have blocked publication. Everything else is not self-published. (Note: the fraction of an organization's publications that are about the organization itself can vary a lot from one organization to another.)
c) Material from "traditional" publishers (e.g., newspapers, books from a standard publishing company, peer-reviewed journals) is not self-published unless it's about the organization itself. Everything else is self-published, including material published by other kinds of organizations and any no barrier materials hosted by the traditional publisher (e.g., reader comments on a news article).
d) None of the above. Please describe your view, aiming for a description such that most of the time, other editors would say that it provides effective guidance for determining whether a given source is or isn't self-published.
Note: If the meanings of "no barrier" materials, "organization itself" materials, and "traditional" publishers aren't clear enough, there is more info in the Notes from previous discussions above (in the sections titled Categories of publishers, General areas of consensus, and Areas where consensus is unclear). The Table above also provides a number of illustrative examples.
Responses
A space for questions, in case people have any
For clarification, 2c would mean that GLAAD (what kicked off this whole argument), the Anti-Defamation League, and the SPLC (which we have debated the reliability of for years but are both marked green on RSP), would all be SPS and therefore it would be a BLP violation to use them on material relating to BLPs (the majority of their current use), correct? PARAKANYAA (talk) 08:02, 27 January 2025 (UTC)[reply]
Yes. (And based on my reading of the most recent discussions, the best editors to ask about what counts as a "traditional" publisher: WhatamIdoing and Void if removed.) FactOrOpinion (talk) 12:28, 27 January 2025 (UTC)[reply]
In a cosmic way, it will be very funny if we argued about the ADL in that months long hell RFC only to backdoor declare them GUNREL them this way (and yes, for the topics that they cover, that is basically what this is doing, we would have to remove them for probably 90% of all statements they're cited for onwiki). PARAKANYAA (talk) 12:40, 27 January 2025 (UTC)[reply]
I think the point PARAKANYAA is making isn't about reliability, but that such a result would have the same effect as declaring them unreliable for most of the ways they are currently used. -- LCU ActivelyDisinterested«@» °∆t°13:26, 27 January 2025 (UTC)[reply]
I'm saying it at the prospect, not that I think it will go that way. And no, it is about reliability, in a roundabout way - it is about whether we can use sources to source certain statements. A source that we cannot use for its scope even if correct is equivalent to unreliable, in practice. PARAKANYAA (talk) 13:27, 27 January 2025 (UTC)[reply]
For further context, my investment in this RFC is largely because as someone who has an interest onwiki in terrorism and hate groups, a *lot* of articles modern American hate group or terrorist organizations cite the ADL and SPLC. Probably all. The vast majority of these contexts at least partially involve BLPs. And if they are unreliable well someone (probably me) is going to have to remove hundreds of these citations and large swathes of article content, over article content that is perfectly fine, over what I view as a stupid technicality, over sources that have been repeatedly declared GREL at RfCs. This is stupid. PARAKANYAA (talk) 13:33, 27 January 2025 (UTC)[reply]
I haven't yet read people's arguments below, but as I understand your view here: the real issue should be whether an organization's publication are reliable, not whether they're self-published, and we shouldn't be excluding reliable publications on the basis that they're self-published. FactOrOpinion (talk) 13:43, 27 January 2025 (UTC)[reply]
Not really. I think editorial oversight is an important indicator of reliability. However I don't think "traditional publishers" have a monopoly on editorial oversight, and I think defining a lot of the things 2c would indicate as "self published" as self published is not something anyone outside of Wikipedia would do. PARAKANYAA (talk) 13:47, 27 January 2025 (UTC)[reply]
This isn't true. Saying, eg. A GREL source doesn't meet MEDRS doesn't mean the source is unreliable. It's saying certain topics apply a higher bar. BLP is no different. Void if removed (talk) 13:37, 27 January 2025 (UTC)[reply]
If it's for entire publications and it the publication is focused on the specific topic it cannot be used for then I would say it basically means that yes. If a medicine source does not meet MEDRS it is basically GUNREL because you can't use it for anything. If you can't use a source on hate groups for hate groups it is basically GUNREL because you can't use it for anything. PARAKANYAA (talk) 13:41, 27 January 2025 (UTC)[reply]
PARAKANYAA, just a quick follow-up: I don't know if you looked at the table, but when a few people looked at my draft, they said it was helpful. The SPLC publications would correspond to the two examples Advocacy non-profit's "About us" info and Advocacy non-profit's report on an anti-LGBTQ+ activist (but in the SPLC's case, the latter might be something like "Advocacy non-profit's report on a hate group" and/or "Advocacy non-profit's report on a neo-Nazi"). FactOrOpinion (talk) 15:40, 27 January 2025 (UTC)[reply]
@PARAKANYAA, if 2c is chosen (and I recommend it), then there are two paths forward:
We declare that specific individual advocacy groups are "traditional publishers" similar to newspapers, since their main business is to publish information.
We change BLPSPS to explicitly say that editors are allowed to accept a source for BLP purposes even if it is 'technically' self-published by demonstrating a consensus, e.g., via an RFC. (This is actually true now, but editors don't seem to believe it's true.)
I mean, sort of, but the way that is phrased is vague enough that there's still room to argue, which put us right back where we started. PARAKANYAA (talk) 10:01, 27 January 2025 (UTC)[reply]
@ActivelyDisinterested, I agree about the conclusion of the RfC on advocacy organization grey literature: people mostly thought that such lit. is not always SPS. But this RfC isn't limited to advocacy orgs (there are many other kinds of "non-traditional" publishers) or to grey lit. (these orgs may publish a mix of grey and non-grey lit.). In the previous RfC, people also weren't given the option of saying "never," and a few people who participated said "none of the above" or "bad RfC," as they didn't like any of the options. For the people who answered the advocacy org. grey lit. question with some version of "not always SPS," there are two main possibilities in my question 2: if they think that it depends on whether the organization is writing about itself vs. about something else, they'll choose 2b here; if they think that it depends on something else, they'll choose 2d and spell out what that "something else" is. Does that answer your question well enough? FactOrOpinion (talk) 13:16, 27 January 2025 (UTC)[reply]
I haven't read all of the arguments yet, but just skimming people's choices, many are 2b. That's a very different answer than the previous RfC, where people seemed to be saying that it varies from organization to organization based on the level of editorial review. FactOrOpinion (talk) 13:35, 27 January 2025 (UTC)[reply]
One thing to keep in mind… even if we deem advocacy group publications to be SPS, other sources (such as news media) reporting on what they say are not. To some extent, this is important in determining how much WEIGHT to give the advocacy group in BLPs… if the SPLC labels someone a racist, it is probably a good idea to check that someone else has taken note of their opinion before we mention it in a BLP. Blueboar (talk) 02:34, 28 January 2025 (UTC)[reply]
Or we could expand the note in BLPSPS about employers to say that editors may, from time to time, designate specific organizations or sources that are exempt from the "never use" rule. We made the rule in the first place; we can change it. WhatamIdoing (talk) 02:38, 28 January 2025 (UTC)[reply]
If the SPLC is unsuitable for BLPs on its own it being laundered without comment through the news media does not make it any more suitable. If it is an SPS, we should not be using it to call people such contentious things at all. PARAKANYAA (talk) 04:41, 28 January 2025 (UTC)[reply]
That is not true.
Imagine how silly it would sound if you said the same thing about an individual person: "If Dr Expert is unsuitable for BLPs on its own, then her quotation being laundered through the news media does not make it any more suitable" or "If Paul Politician's campaign website is unsuitable for BLPs on its own, then his quotation being laundered through the news media does not make it any more suitable".
I don't see why we would categorize individual statements sought out by the media the same, and it's a bit different when they are using the quotes to accuse them of what the SPLC accuses people of. If it is an SPS IMO we should not be using it for what we use it for at all. PARAKANYAA (talk) 05:31, 28 January 2025 (UTC)[reply]
Are you saying that if the media seeks out Dr Alice Expert, an expert in hate groups, who gets quoted as saying "_____ is the leader of a hate group", then that's supposed to magically be different from when the media seeks out an organization that has recognized expertise in hate groups and says the same thing about the same person?
No, I'm saying that most of the sources referencing the SPLC's opinion aren't in the form of seeking out quotes, they're in the form of referencing their designations of hate group. I do think that is different. And no, I think saying the SPLC is an SPS is dumb and it is probably GREL, but if it is an SPS we should not be using their designations at all. PARAKANYAA (talk) 06:04, 28 January 2025 (UTC)[reply]
So if the reporter e-mails the public relations office at the SPLC, or DMs them on Twitter, or runs into them in a party, and gets some information, which they write this way:
"Cal Communications, director of public information at the SPLC, said that ____ is a hate group."
then you think that's better than the reporter checking the SPLC's published information, and writing this:
"The SPLC has designated _____ as a hate group."
Do I have that right? It sounds like you believe that privately seeking out quotes is better than seeking out information whose accuracy can be checked against the original publication. WhatamIdoing (talk) 07:04, 28 January 2025 (UTC)[reply]
Discussion
@FactOrOpinion: RfCs are usually short and have discrete questions that are susceptible of simple answers. Asking people whether they think a policy as-written needs to be amended and how they would suggest amending it (question 1), or whether they ascribe to a particular philosophy that would guide how they would amend the policy (question 2), is something you should do before you start an RfC, so that you can propose those amendments for the community to !vote on. As of now, this is a massive wall of text that has no actionable items. I am boldly removing the RfC tag so that this can be workshopped further. voorts (talk/contributions) 22:22, 26 January 2025 (UTC)[reply]
@Voorts, I do recognize that RfCs are usually short. But not always; for example, this current one has 22 questions. I included 7 different BEFORE discussions that make clear that many editors interpret the meaning of "self-published" in different ways. Those discussions have not resolved what the consensus meaning is. I know that RfCs often have actionable items, but not always. I see this one as a mix: there is an actionable item (do we keep the current explanation, or do we significantly revise it?) and an information-gathering item (is there consensus about how to interpret "self-published"?). If the answer to the actionable question is "revise," then the info-gathering item puts us in a position to create an actionable revision. I ask that you replace the RfC tag. FactOrOpinion (talk) 22:49, 26 January 2025 (UTC)[reply]
@FactOrOpinion: Multi-question RfCs aren't an issue. I closed most of the RfC that created admin recall. The issue is that this RfC is far too open-ended (see WP:GOODRFC and WP:RFCQ). Many editors have already expressed differing views in the prior discussions, and consensus hasn't been reached. Asking people "does this need to change" without concrete proposals will just lead to the same debates being rehashed with no movement forward. All of that said, if you want to replace the RfC tag, go ahead. voorts (talk/contributions) 23:12, 26 January 2025 (UTC)[reply]
@Voorts, I agree that it's common to write an RFC question so it's susceptible to simplistic votes, but I'm not sure that's a good thing. But even if that were a requirement, this RFC meets it. My vote is 1b, 2c (as should surprise nobody who has read the many linked discussions). What's yours? WhatamIdoing (talk) 23:58, 26 January 2025 (UTC)[reply]
I don't think RFCs need to be simplistic. We already know that some editors take issue with the way SPS is written. A good RFC would propose a solution to that problem. For example, instead of framing question 2 as a set of opinions, it would include actual proposed changes to SPS. That could be accomplished by workshopping proposals and then putting them to the community. voorts (talk/contributions) 00:30, 27 January 2025 (UTC)[reply]
I can think of a half dozen ways to word each of those, but the problem has been getting people to agree on which general category is the real definition. They're not only disagreeing with the way it's written (though, if you think that, you should put yourself down as 1b for the first question); they're also disagreeing about what it is. It's "Oh, big organizations are totally self-published, because I don't want him to cite their website about this person" followed by "What do you mean, big organizations are self-published? This is a 100% reliable source about how that two-faced politician has ruined the lives of millions of people!"
They're not only disagreeing with the way it's written [...] they're also disagreeing about what it is
I second this, this is not a wording issue, this is a conceptual one. We cannot agree wording without overcoming the conceptual gulf, and that seems intractable.
The problem in practice is when editors work backwards from "I want to include this information" to "therefore the source isn't self-published". This happens because BLPSPS is clear policy, but the policy definition of SPS is woolier than USINGSPS, and USINGSPS is "only an essay", so in practice you can get around BLPSPS by arguing a very limited definition of SPS like 2a/2b. Void if removed (talk) 10:51, 27 January 2025 (UTC)[reply]
I don't think it is appropriate to categorize it as "getting around" the policy as when BLPSPS was written the idea was some guy's random website and Lulu.com. If you had asked an editor at the time whether the SPLC is a self-published source I do not think that would have been considered. PARAKANYAA (talk) 11:58, 27 January 2025 (UTC)[reply]
And this is exactly why the arguments derail. You want to use SPLC on 3rd-party BLPs, and in order to do so you have to argue it isn't self-published. Any criteria by which SPLC is not self-published also encompasses Lulu.com, but because nobody is strongly arguing we use Lulu.com for material on 3rd party BLPs, the fact that there is an enormous disconnect in how policy is being interpreted goes under the radar, until someone haplessly takes policy and WP:USINGSPS at face value, and the argument blows up on a CTOP talk page, again, or ends up in RFCs attempting "grey literature" carveouts, and so on.
"Any criteria by which SPLC is not self-published also encompasses Lulu.com" - obviously not. An organization does not have to have commercial backing to have proper editorial review. PARAKANYAA (talk) 12:37, 27 January 2025 (UTC)[reply]
I agree with PARAKANYAA here, that "Any criteria by which SPLC is not self-published also encompasses Lulu.com" isn't accurate. I wasn't familiar with Lulu.com, but just took a very quick look and it looks like a self-publishing site. It's very easy to say that SPLC's materials aren't self-published (because it's an organization) but something published through Lulu.com is self-published (because that's an individual publishing their own book, and Lulu.com is operating as a printer rather than a publisher). Choice 2a takes that stance. FactOrOpinion (talk) 14:42, 27 January 2025 (UTC)[reply]
Lulu.com is a self-publishing site. So is any website anybody starts, including any group of people. There's nothing magical about some people saying "Let's be an organization" to make it not be written by those people and published by those people. WhatamIdoing (talk) 17:43, 27 January 2025 (UTC)[reply]
@WhatamIdoing, re: Lulu.com is a self-publishing site. So is any website anybody starts, including any group of people, that implies that the NYT website is also a self-publishing site. Also, isn’t Lulu itself mostly a host/printer rather than a publisher? If so, that’s distinct from websites with corporate authors. I know that you exempt the NYT because it’s a traditional publisher, and you turn to a couple of dictionary definitions for this exemption, but I’m curious what your response is to my comment in (c) here that I no longer believe dictionaries are attempting to address corporate authors at all. If you think dictionaries are addressing corporate authors, is there anything in particular that makes you think that? And how would you answer the question in my comment: What would it even mean for an organization to publish its website through an established publisher? That is, a book author can theoretically choose between self-publishing and approaching an established book publisher (recognizing that the manuscript may be rejected), but does an analogous choice exist for a corporate author that wishes to publish website content? FactOrOpinion (talk) 21:50, 27 January 2025 (UTC)[reply]
I don't think that dictionaries are primarily addressing corporate authors, but I've never seen one that excludes them.
Consider, e.g., the old-fashioned church cookbook, a perennial fundraiser of the pre-internet days. All the ladies submitted a recipe or two, and one of them typed them all up. The printing could be as informal as a couple of volunteers standing over the ditto machine or photocopier for a few hours to make as many copies as they thought they could sell. My mother had one when I was little. It was probably printed in the 1930s, and it had several different recipes for white cake. She once pointed out to me which recipe was the one from the lady who was most celebrated for her white cake. (I remember that she sifted the flour seven times before measuring it, but nothing else about it.)
Could anyone really describe that as anything other than self-published? There are no professionals involved, no editorial standards, not even anyone willing to say that we shouldn't run the recipe for Lime Jello Marshmallow Cottage Cheese Surprise for fear of hurting the author's feelings, and the main goal wasn't to be printing cookbooks or distributing any sort of information; it was just a means to an end. But there were lots of people involved, and usually even an official "organization" (probably named something like the Smallville Church Ladies' Auxilliary) sponsoring it.
IMO the fact that this is case of an organization writing and publishing it does not make it any less self-published than if a single woman had done everything by herself.
I agree with you that the church ladies (theoretically) had a choice between self-publishing and using a traditional publisher. Similarly, if Microsoft wants to write a manual about how to configure Active Directory, they have a choice between self-publishing it vs submitting it to a traditional publisher with an interest in technical content, such as O'Reilly Media.
However, that is not always the case. Some things realistically cannot be published using a traditional publishing model. That includes:
content you approve and post on your own website or on social media platforms
(But not content you hired a social media influencer to create and post without your involvement. If they get to decide what to write and post ["This is sponsored content, but it is entirely my own honest opinion"], then that's non-independent of you, and it might be self-published by the influencer, but it's not self-published by you.)
signs you post in the window of your business
pamphlets or advertisements you distribute to potential customers
e-mail messages you send to people inform people about a new product or an event
The fact that some models are always self-published does not mean that there's something wrong with the definition of self-published. It just means that some things aren't realistically non-self-publishable. WhatamIdoing (talk) 22:36, 27 January 2025 (UTC)[reply]
I think several dictionary definitions do exclude corporate authors, by using the phrases “your own” work, “one’s own” work/money, “one’s work,” “by oneself,” all of which indicate that the dictionary is only talking about natural persons. I can't think of a dictionary that refers to an organization as "you" or "one" rather than "it." Granted, it's possible to come up with examples of organizational website content that could theoretically be published by an "established publishing house" / "publishing company," as with the Microsoft example you gave. But I think there are many more examples of web content that cannot. For example, I doubt that most of the content on the Social Security or Department of Justice websites could be published using an "established publishing house" / "publishing company" (and I personally do not think of the government itself as a publishing house or company). Or consider searchable databases, like my county's library catalog or the Library of Congress's catalog. I do remember the days of card catalogs, but even then, I can't imagine any publishing company printing it as a book. FactOrOpinion (talk) 01:56, 28 January 2025 (UTC)[reply]
Few dictionaries use "you" at all, but "one" refers to any entity, not just humans, and not just if there is only one of them.
The problem with "individual humans" is that it's very easy to turn "a human" into "an organization". If I write a book, but before I print it, I create a small business, "Gas Station Press", and legally, it's my business that publishes the book, is that not still what you would call a self-published book? WhatamIdoing (talk) 02:37, 28 January 2025 (UTC)[reply]
Here’s a sample dictionary definition of "oneself" and sample dictionary uses of "one's." To me, those all look like references to people. Re: your question, that's not an example of a corporate author, which is what I was addressing, and it's a no barrier situation. But rather than focus on a specific example, my primary point is that I don't think dictionary definitions take a stance one way or another on the meaning of “self-published” when it comes to corporate authors, so it’s going to have to be wikijargon whether we want that or not. Editors can certainly make arguments for one or another choice re: what the wikijargon definition should be for corporate authors (2b-d). FactOrOpinion (talk) 12:16, 28 January 2025 (UTC)[reply]
I think you want mwod:one, from which the relevant definition is "an individual of a vaguely indicated group : anyone at all", with the example "one never knows".
A narrow reading of "an individual" would exclude any multi-authored publication, which would not be consistent with our inclusion of group blogs. WhatamIdoing (talk) 18:20, 28 January 2025 (UTC)[reply]
No, I don't want mwod:one, as none of the dictionary definitions used "one" (I wasn't careful enough when I wrote I can't think of a dictionary that refers to an organization as "you" or "one" rather than "it.") Three dictionaries used "author," but none of them includes corporate authors in their definitions of "author." One said "your own," one said "one's own," one said "one's work" and "one's own," one said "oneself" and "one's own," one said "oneself" and "the author." Agreed that none of them seem to acknowledge the potential for co-authored works. FWIW, I don't consider a group blog to be co-authored; instead, I consider the site to be shared but the blog entries to be authored singly, a bit analogous to comments here. FactOrOpinion (talk) 21:38, 28 January 2025 (UTC)[reply]
None of them exclude corporate authors (or co-authors), either.
The definitions don't explicitly reject corporate authors, they just all use words that generally refer to people rather than organizations. So I don't think dictionary definitions take a stance one way or another on the meaning of “self-published” when it comes to corporate authors. I didn't say that that link provided a definition of "one's." I said they were a sample of a dictionary's use of "one's." Each of those 777 phrases has a M-W dictionary entry, many of them idiomatic expressions, and the ones I skimmed referred to people (e.g., "flip one's lid," "get one's life back together"). FactOrOpinion (talk) 00:36, 29 January 2025 (UTC)[reply]
You have a banner somewhat hidden above stating This RfC is solely about WP's interpretation of "self-published." It is not trying to assess whether a source is reliable, independent, primary, biased, etc., or whether its use is due or needs to be attributed, as these aspects are distinct from whether the source is self-published. I suspect the real controversy is all about exactly those. How much of the debates is really about whether a source is "self-published" or not, and how much is about people trying to misuse a "self-published" label to win debates over those other things? Same as how many debates over whether a source is "primary" (or occasionally "tertiary") are really "if I label it 'primary' then I can skip having to give better reasons to exclude it". Anomie⚔12:40, 27 January 2025 (UTC)[reply]
I did a lot of thinking about the SPS explanation while reading the previous conversations to write up the Notes section. I think the current explanation needs revision. We need some kind of characterization to help editors understand what is/isn’t considered self-published, but that footnote characterization is really problematic. Some of my concerns:
a) It conflates whether a source is self-published with whether it’s been reviewed / is reliable. A given publication can be any one of the following: RS & non-SPS, RS & SPS, non-RS & non-SPS, non-RS & SPS. Sources may not be equally distributed among these four categories, but the distribution depends in part on how we interpret self-published and what WP content the source is used as a source for.
b) I understand the motivation for the characterization (we want RSs), but in practice, we can often only guess at whether there’s a reviewer, much less whether they have a COI and are assessing reliability. These reviewer/COI/reliability features are not part of any dictionary definition of self-published. The characterization doesn’t work for things like creative works (music, poetry, fiction), opinion pieces, and memoirs. Creative works are mostly used for material about themselves, but opinion pieces — though attributed — definitely aren’t. The characterization also leads to weird results for unreliable sources that normally wouldn’t be considered self-published (like Breitbart News).
c) I don’t think that the dictionary definitions address WP’s needs. I think they were only ever intended to apply to individual people (or small groups of people), and to situations where there is a choice is between publishing something yourself or submitting it to a traditional publisher (and the publisher then decides whether to publish it or reject it). Most of the definitions use words that indicate they’re talking about natural persons: “your own” work, “one’s own” work/money, “one’s work,” “by oneself” — none of these suggest “its own.” (Admittedly, “author” could be either a person or a corporate author.) What would it even mean for an organization to publish its website through an established publisher? (I don’t think there are such publishers for web content.)
d) When I look at the revision histories for WP:V and WP:BLP, the original examples for “self-published” were all no barrier materials.
e) I do understand the worries that if we limit the interpretation of self-published to no barrier materials, that creates the potential for an increase in problematic BLP content. But just because something isn’t self-published doesn’t make it usable. We can still rule things out if they’re not reliable, not due, primary, etc. FactOrOpinion (talk) 13:50, 27 January 2025 (UTC)[reply]
One comment: IMO, the conflation of reliable and non-SPS is unavoidable because from reading SPS, it's a clear implication that even if a source is an EXPERTSPS it is obviously intrinsically less reliable than a non-SPS source. The reason we restrict them from BLPs is because they are considered generally less reliable than sources with editorial oversight. It's a way a source can be less reliable but obviously any equivalent non-SPS would be prioritized over the SPS. PARAKANYAA (talk) 13:56, 27 January 2025 (UTC)[reply]
I don't think it's unavoidable. I think it's a product of how WP has chosen to characterize SPS. If you look at dictionary definitions, for example, there's nothing about reliability in the normal meaning of "self-published," outside of WP. It's easy to come up with unreliable sources that aren't self-published, like Breitbart News and tabloids. I think we'd be better off thinking about SPS and reliability as two distinct features than intertwining them, even if they often correlate. (For ex., if you look at hair color and eye color, you'll find that brown/brown and blonde/blue are more common than brown/blue or blonde/brown, but no one would say that hair color and eye color should be conflated.) Also, re: "an EXPERTSPS ... is obviously intrinsically less reliable than a non-SPS source," maybe, if you're comparing it to a non-SPS expert source writing about the same thing. But an EXPERTSPS certainly isn't intrinsically less reliable than any non-SPS. There are totally unreliable non-SPS, like Breitbart News. FactOrOpinion (talk) 15:01, 27 January 2025 (UTC)[reply]
Because editors tend to focus on single articles on websites as "the thing being published" the distinction between self-publishing and 3rd-party publishing isn't obvious, but as soon as you switch to other media, it becomes clear that this in no way refers to natural persons.
If we ignore real-world examples like this and stick to the idea that it is only single, natural persons who can self-publish, then absolutely any group blog is not self-published.
If we accept that it isn't just natural persons who can self-publish, then the "no barrier" aspect makes it clear that barriers internal to the publishing entity itself are "no barrier". If someone is commissioned by their employer to write a white paper, and a manager reads it and posts it on the employer's website with their employer's approval, it is self-published by the employer, who is the entity in control of the entire authoring and publication process. Void if removed (talk) 14:52, 27 January 2025 (UTC)[reply]
Re: "it is only single, natural persons who can self-publish," option in 2a in no way limits it to a single natural person. It also includes groups of natural persons that aren't an organization, such as when a bunch of people co-author an article and post a preprint, or a group blog where each person chooses for themself to hit "post," or WP, where millions of people each make an independent choice to hit "publish."
As for your last paragraph, consider "If someone is commissioned by their employer, [a newspaper], to write a [newspaper article], and [an editor] reads it and posts it on the [newspaper's] website with their employer's approval, it is self-published by the [newspaper], who is the entity in control of the entire authoring and publication process." The process is the same, yet you claim that the newspaper's website isn't self-published, but some other employer's website is self-published. You're not distinguishing between the two actual publication processes, you're distinguishing between business models (e.g., the newspaper business makes its money selling what they publish, but a non-profit isn't a business and may make its money though donations and grants, or a corporation may make its money by selling soft drinks, or a government gets money through taxes rather than in direct exchange for the services it provides and its publications are generally in the public domain). FactOrOpinion (talk) 15:19, 27 January 2025 (UTC)[reply]
A corporation – an actual, legal corporate entity – can, in practice, be one person. I know multiple people who are sole proprietors of multiple corporations. (I'm in the US. People from other countries are usually astounded by this).
It is not unusual for "an organization" (e.g., a retail store) to be an entity that is just a couple of people. They're called mom and pop businesses for a reason. The US has 30 million businesses with zero employees, which works out to about one of these organizations for every 10 adults. A formal business is "an organization", but I would still consider anything put out by "WhatamIdoing's Gas Station" or "Foo's Fine Fashion Shoppe" to be self-published. Wouldn't you? WhatamIdoing (talk) 20:35, 27 January 2025 (UTC)[reply]
@FactOrOpinion I go into this in my answer to question 1, but the source of your (very reasonable) confusion -- What would it even mean for an organization to publish its website through an established publisher? -- is that the quoted text from the footnote was written in 2000, about the Internet as it existed in 2000, and has not been altered since. The person who wrote it, Joe Barker, was apparently a "recognized expert on Web searching" -- at least according to Berkeley's obituary, which is ironically the exact kind of source that this whole WP:SPS discussion is about. But he died in 2009, so by definition he is not an expert on evaluating websites on the modern Internet.
@Gnomingstuff, Yeah, there's a short comment in the Notes from previous discussions section that it may be time to update the quotes (and I think there are other reasons to update them besides the fact that they're old; I don't think they really help to clarify what it means for something to be self-published, and if you look at the context for the first two, these universities were focused more on when websites can be used as sources of info in students' written work — and perhaps how to cite it — than whether material is/isn't self-published per se). The question of mine that you quoted wasn't actually based on the footnote quotations at all. It was a response to the dictionary definitions quoted in the Notes from previous discussions section. In previous discussions, some people argued that we should be using dictionary definitions, but people chose to emphasize different definitions (and I explained the main two interpretations under the definition quotes), leading them to very different views of whether corporate authors are/aren't primarily self-published. After thinking about it, I've concluded that the dictionary definitions just don't serve our needs, because they're mainly trying to address situations where entity X (perhaps a person, perhaps an organization) wants to sell item Y for profit, where Y is some kind of publication (e.g., Y is a book, a video game, music), and X has a choice between two business models: selling it themselves (doing the marketing, etc., and making their profit through direct sales) or selling it to some other entity (which then takes on the marketing, etc., and pays X a fixed amount and/or a percentage of sales). I don't think the dictionary definitions are in any way trying to address the wide diversity of publications used on WP, nor corporate authorship. So we need to come up with our own explanation of what kinds of sources are self-published. FactOrOpinion (talk) 23:46, 28 January 2025 (UTC)[reply]
Apologies, misinterpreted. (I think arguments from dictionary definitions are almost always pedantic and silly, this discussion being no exception.) Gnomingstuff (talk) 00:33, 29 January 2025 (UTC)[reply]
I just wanted to note that I've moved the Table above the Questions to encourage people to at least glance at the Table before answering the Questions. When I was working on my draft, the people who responded said that they found it helpful. I also added a few words, since it has columns for Options 2a-c, which aren't explained until Question 2. I have not changed the wording of any of the Questions. FactOrOpinion (talk) 16:17, 27 January 2025 (UTC)[reply]
@Voorts, quick questions (if you’re willing to answer even though you think it’s a Bad RFC): you said “per my other comments here and per AD's comment.” I understand the “per my other comments” part, but I don’t understand “and per AD's comment,” as AD answered both questions and didn’t say that it’s a bad RfC. Would you mind saying how your Bad RFC response is connected to AD's responses? And if you look at the Table, would you say that there's no option that corresponds to how you interpret the meaning of "self-published"?
@JoelleJay, do you have thoughts about Question 1? Why I’m asking: my impression from our previous exchange (about UC Berkeley’s website, in this RSN discussion) is that your answer would be 1a, though perhaps I’m wrong about that. But for Question 2, you chose 2c, while @Espresso Addict and @Aquillion (both of whom chose 1a) reject 2c, and I'm curious about different people interpreting the footnote characterization so differently.
Espresso Addict, I agree that “reliability is more important than whether it meets some definition of self-published,” but reliable SPS generally can't be used for BLP content. For example, we could not use Timothy Snyder’s Substack statements about Trump and fascism, even though Snyder’s a historian of fascism whose work is widely cited in his field; we had to wait for the New Yorker to publish his essay about this. FactOrOpinion (talk) 21:50, 27 January 2025 (UTC) Pinging @Voorts and @JoelleJay, since I forgot to sign my earlier comment and the pings didn't go through. FactOrOpinion (talk) 01:06, 28 January 2025 (UTC)[reply]
[You didn't sign so your ping did not go through.] As I wrote elsewhere, reliability always needs to be calibrated against how controversial the statement is. A highly controversial claim needs a hyper-reliable source (preferably multiple). Whether it is self-published seems to be a side issue. Espresso Addict (talk) 21:56, 27 January 2025 (UTC)[reply]
Thanks for the heads up re: the ping. I've added my signature. One of the things that puzzles me in this: if what we really care about it reliability, why are we using SPS as a proxy for unreliability rather than just assessing reliability directly? The current explanation would say that Breitbart should be an SPS because the editors there are unlikely to assess reliability, and I think that's a crazy result. It's not an SPS and is unreliable. FactOrOpinion (talk) 22:03, 27 January 2025 (UTC)[reply]
That's a very common problem. When editors mean "I think this is a bad source", they aren't always willing to just say that. Instead, they try to dress it up in some kind of official WP:UPPERCASE reason: "This is a {primary|self-published|non-independent} source". Common sense is not encouraged; WP:WEWP:MUSTWP:BEWP:CORRECTWP:ANDWP:OFFICIAL, WP:ORWP:ATWP:LEASTWP:SOUNDWP:LIKEWP:IT.
That said, reliability is generally judged according to a relatively agreed-upon set of standards, most of which are fact-based (e.g., "Is it self-published?", which is a question that will have a correct answer, even if we don't have the information needed to figure out what the correct answer is). See WP:NOTGOODSOURCE for an informal list of the standards. A source that has all the good characteristics and none of the bad characteristics is more likely to be judged reliable than one that's the other way around. Breitbart has some of the desirable characteristics, but fails utterly on the "fact checked and accuracy" part. A social media post by an accused person to deny all guilt misses most of them but is still deemed reliable. The list of desirable qualities is not an absolute requirement, but it is statistically likely to produce the desired results. WhatamIdoing (talk) 01:32, 28 January 2025 (UTC)[reply]
@PARAKANYAA, re: I think editorial oversight is an important indicator of reliability. However I don't think "traditional publishers" have a monopoly on editorial oversight, I’m inclined to agree, but I’m curious: unless an organization makes its own editorial process clear (e.g., via a masthead naming editors) and/or belongs to a category where we generally know something about the editorial process (e.g., it's a newspaper), how do you assess whether there’s editorial oversight in any specific organization? For example, do you assume this based on reputation?
@Anomie, re: people trying to misuse a "self-published" label to win debates over those other things, do you have any thoughts about whether there’s a way to clarify what it means for something to be self-published, so that people directly address those other things instead of misusing “self-published”?
@Void if removed, re: When talking about other media, this is common terminology that nobody is confused by, it seems that your examples are all situations where entity X (perhaps a person, perhaps an organization) wants to sell item Y for profit, where Y is some kind of publication (e.g., Y is a book, a video game, music), and X has a choice between two business models: selling it themselves (doing the marketing, etc., and making their profit through direct sales) or selling it to some other entity (which then takes on the marketing, etc., and pays X a fixed amount and/or a percentage of sales). But many of the organizations we’re talking about — such as governments, advocacy groups, universities, learned societies, think tanks, international NGOs, intergovernmental bodies, museums, foundations, charities, labor unions — are not for-profit entities; they derive much or all of their income in other ways, so they’re not making a choice between whether they do direct sales vs. sales through a publishing company. Also, even if we focus on for-profit entities X, the thing they’re selling might never be thought of as a “publication” (e.g., Z is a food item, a car, clothing), so the question of SPS vs. non-SPS just doesn’t apply to the thing they’re selling. For all of these reasons, your reference to other media just doesn’t make sense to me as an argument about how to treat publications by these other kinds of organizations. @WhatamIdoing, this also touches a bit on your point about mom and pop businesses: my guess is that they’re generally for-profit and often are not selling things that we’d consider publications. FactOrOpinion (talk) 17:38, 28 January 2025 (UTC)[reply]
Many of the weak but plausible sources we see (e.g., content farms on websites for small law firms) are for-profit entities. However, non-profits do the same: a theater non-profit will stage ("publish") a new play, or a museum will mount an exhibition of selected artwork, and they will also post information on their website about the context and meaning of the art they're bringing to the world.
I'm not sure how we would classify such a thing. Theaters and art museums are (sort of?) in the publishing business. The play is likely to be somewhat fictionalized, so we could treat it the same way as we treat a novel, but the art exhibition could be closer to news (e.g., iconic photos from a recent war). WhatamIdoing (talk) 17:56, 28 January 2025 (UTC)[reply]
Publishing has, traditionally, been a business. A publisher takes on the cost of distribution, marketing etc in return for a share of the profits. Some publishers are run in a not-for-profit sense, but they still take on the role of the traditional publisher.
To be "self-published" is to skip that relationship and distribute yourself. The era of digital information means that the barrier to simply placing material online is nil - so the vast majority of materials online are self-published, because few people need a publisher in order to throw up a website.
It is hard to say what a publisher does, when all you care about is text, freely available on the internet.
The relationship is much clearer with commercial activities and other media because there are issues with rights and remuneration that are a hassle, and a publisher takes care of as part of the relationship. A self-published game developer can operate freely in the US, but not China, where a Chinese publisher is a requirement. A self-published band can easily sell material on bandcamp, but don't benefit from the relationship a publisher might have with other media channels in licensing their songs for other commercial use. A self-published visual artist might sell material themselves online, but struggle without the backing of a publisher to legally defend their artworks from theft.
When it comes to physical publications - books, newspapers, magazines, journals - again, the publisher/publication relationship is obvious, and again it makes sense to view the online equivalent as a digital copy. Even so, some material is online-only, and it still has a publisher because it describes a relationship, not a process.
Ideally, we'd be building an encyclopedia from the best sources - books, newspapers, magazines, journals. That just happens to overlap with traditional publishing. But it isn't a perfect overlap and I think we need to get away from SPS as a blunt proxy for reliability and think about different gradations. Void if removed (talk) 18:19, 28 January 2025 (UTC)[reply]
So another edge case - sometimes a blogpost or a substack is reposted in its entirety on another site. @FactOrOpinion do you think these still count as SPS?
For example, the Lemkin Institute for Genocide Prevention has a feed of news stories. Irrespective of their reliability generally, this feed is merely trivial reposts of stories published elsewhere. As it states at the top: We post articles from media outlets around the world - and they seem to expire after a certain period or number have been posted.
These posts are just the entire content of the source, copied and reposted, with copyright attribution. It appears to be nothing more than a very dumb news aggregator. Sometimes, these posts are reposts of blog posts or substack articles. Does the act of simply taking and reposting an obvious SPS by this trivial process somehow turn it from an SPS to a non-SPS, and thus acceptable for 3rd party BLP claims? I would say of course not, because that is no barrier at all. Void if removed (talk) 21:17, 28 January 2025 (UTC)[reply]
As long as "self-published" can be used as a heuristic for "avoid using, maybe not reliable", people will argue that something is self-published to avoid having to directly address reliability and the like. Anomie⚔23:31, 28 January 2025 (UTC)[reply]
I agree with @Void if removed that in an ideal world we move away from "SPS as a blunt proxy for reliability." Some SPS is completely reliable (an election website that shows how many votes a candidate receives), and some SPS is more unreliable or subject to bias. We should be mature enough to know the difference. --Enos733 (talk) 22:34, 28 January 2025 (UTC)[reply]
Question 1 survey
1b: What we have is not working. If it were working, we would not have an apparently endless string of discussions at WT:V, WT:BLP, WT:RS, and other pages in which editors wonder whether a source is "really" self-published. We have been talking about this for years. It is broken and needs to be fixed. WhatamIdoing (talk) 00:00, 27 January 2025 (UTC)[reply]
I have never had an issue with the definition of SPS, despite using it somewhat often at AfD – there, I know an SPS when I see one, and obviously an organization reporting on itself doesn't count towards notability. On the other hand, I do not edit BLPs very much, and I am inclined to believe WhatamIdoing here, since she tends to (despite the name) know what she is doing. Putting most of the definition in a footnote seems like a bad move, and I can see some need for clarity, so I'll land vaguely in the 1b camp. Toadspike[Talk]09:14, 27 January 2025 (UTC)[reply]
I've never had a problem with the definition in SPS, put it's clear that it needs sorting me work. I'd agree with Toadspike that having some much detail in a footnote isn't helpful either. 1b. -- LCU ActivelyDisinterested«@» °∆t°09:28, 27 January 2025 (UTC)[reply]
1b: I agree this is fundamentally broken, as shown by the endless circular discussions about what it actually means and why we even use this standard and how it impacts reliability generally or BLPs specifically. There is no clear consensus what it even means or why we use it, and there is a tension between how some editors apply it in practice (if a source is from an organisation that enough editors believe to be useful on some matter of BLP, it is invariably treated as not self-published) and how it seems to have come into being in the first place (to simultaneously prevent BLPs from filling up with unvetted defamatory material or self-serving material). There is wide difference of opinion of whether there should there be a blanket ban or blanket endorsement of using the websites of partisan advocacy orgs as single sources of information on BLPs, and this ends up centred on the meaning of SPS. It can't be fixed by tinkering, as there is no consensus on the direction of that tinkering. Ie. saying explicitly that websites of advocacy orgs either are or are not self-published would be a small change in terms of text, but an enormous change depending on interpretation of existing meaning. Void if removed (talk) 10:26, 27 January 2025 (UTC)[reply]
1a. SPS is fine and is totally clear; it describes a specific type of non-WP:RS and provides limited exceptions that allow them to nonetheless be used under certain circumstances. Some minor tweaks to make this more clear wouldn't be amiss, but the core problem is that people are trying to use it as a substitute for the broader WP:RS policy, which it is not. Questions about what structural aspects are required of sources, or what the threshold is for one to be cited, or what sort of editorial controls and fact-checking are needed - these are covered by the rest of RS policy, and come down to things like whether a source has editorial controls and fact-checking, which grant it a reputation for fact-checking and accuracy. If someone is going "alright yes it meets that threshold but it is a SPS and therefore shouldn't be used" then they're talking nonsense, and I'd support minor tweaks to make that more obvious if people are really facing confusion over that point; but I don't think any substantiate policy changes are required. The meat of WP:SPS is, essentially, "here are some types of sources that usually fail WP:RS and the limited ways in which you can use them anyway." --Aquillion (talk) 14:40, 27 January 2025 (UTC)[reply]
If it is "totally fine and clear", then why have editors spent tens of thousands of words disagreeing about what it means over the last couple of years? WhatamIdoing (talk) 17:53, 27 January 2025 (UTC)[reply]
Aquillion I agree with your point about people substituting SPS for broader RS policy, but I'd argue that's evidence that something is wrong with the SPS policy. Do you have a suggestion for how else to fix it? Do we remove it entirely? Is there even a way to fix this? Genuinely curious to hear your thoughts as an editor who's been around a while. CambrianCrab (talk) 01:06, 28 January 2025 (UTC)[reply]
1b. Not because the idea is problematic, but because the wording sucks. If we interpret the current wording literally -- and if there is wording, then some people will interpret it literally -- then we are saying that books are SPS because some books are self-published, which is both insane and not our actual practice (because it is insane). It's also behind the times. The line between "newsletters" and news blogs is almost completely blurred now; some news organizations have "newsletters" that are just the equivalent of special-edition inserts, and many Substack pages that call themselves "newsletters" are essentially journalistic operations -- which is becoming increasingly common as traditional media organizations shutter. Similarly, many news organizations have podcasts now, with the full gamut of research, editorial oversight, and fact-checking. For example, under the current wording the Sold a Story podcast series, which won the Alfred I. duPont–Columbia University Award (one tier below a Pulitzer) and has named editors, researchers, and fact-checkers, is SPS because it's a podcast. Which is, again, insanity. Gnomingstuff (talk) 19:19, 28 January 2025 (UTC)[reply]
Update: I just went through the footnotes more closely and GOOD GOD are they even more of a mess. The most recent guideline is from 2011 (14 years ago). Then there's one from 2003 (22 years ago). And that quote from the Berkeley guide? It is completely unchanged from the 2000 version of that website. In 2000, the big news in search was that Yahoo was partnering with hot new startup Google. Like, holy fuck. Why is this in policy? Why are we using guidance from the AltaVista era to determine whether a website is self-published? Gnomingstuff (talk) 20:09, 28 January 2025 (UTC)[reply]
It's really difficult to change some parts of our policies. Sometimes this happens because of https://xkcd.com/1172/ (Don't take that away; I've been wikilawyering with that line for years!). In this case, we sometimes have trouble convincing editors that the problem actually exists. I doubt that anyone here, if asked what a self-published source is, would come up with a line like "someone without a conflict of interest has validated the reliability". If you proposed it today, they'd all tell you that you were wrong, and that you have confused SPS, INDY, and fact-checking. But once it's on the page, it's sometimes difficult to get it removed. WhatamIdoing (talk) 23:44, 28 January 2025 (UTC)[reply]
Yeah it was a rhetorical question, I know how this stuff happens, this is just an especially stupid one. (There should be a Muphry's Law equivalent that any statement about the reliability of sources will use unreliable sources to back it up.) Gnomingstuff (talk) 00:48, 29 January 2025 (UTC)[reply]
1A although I'm getting the feeling that what I consider small improvements others might consider to be large ones... So maybe we're talking past each other? Not sure I'd say bad RfC, but its close. Horse Eye's Back (talk) 23:05, 28 January 2025 (UTC)[reply]
I think more or less the same minus "as books, patents, newsletters, personal websites, open wikis, personal or group blogs (as distinguished from newsblogs, above), content farms, podcasts, Internet forum postings, and social media postings" because it is both ambiguous and unwieldy (a position which seems to mirror Gnomingstuff's but they voted 1B). Horse Eye's Back (talk) 00:55, 30 January 2025 (UTC)[reply]
1b. I explained some of my reasoning in the Discussion section above (especially here). I agree with others who say that the existing explanation is broken. It doesn’t provide sufficient guidance / is to open to significantly different interpretations (e.g., how do we know if the material has been reviewed? do we just guess on the basis of the organizations reputation, or does it need to have a masthead, etc.?). I also don’t believe that the footnote’s characterization makes sense as a way to explain what it means for something to be self-published, as self-publishing is distinct from reliability, even if there is some negative correlation. Somehow, we need to come up with a better explanation. The main explanation should also be in the body of the policy, not in a footnote, and if we’re going to include quotes, they should be more recent and should truly help explain the meaning of self-published. FactOrOpinion (talk) 00:52, 30 January 2025 (UTC)[reply]
Question 2 survey
2c. An ordinary newspaper is not self-published. An ordinary corporate website is self-published. Corporate authorship and corporate publication are real. If Coca-Cola, Inc. hires employees to write and publish content at https://www.coca-cola.com/, then we should not pretend that them having 80,000 employees makes them work like an independent newspaper or book publisher. They're writing their content and publishing their content themselves. That makes them self-published. WhatamIdoing (talk) 00:05, 27 January 2025 (UTC)[reply]
To expand on my prior remarks, 2c is not a perfect fit, but I think it's the closest. I tend to assume that "formal" government publications are non-self-published. Reports such as the United States census and the Morbidity and Mortality Weekly Report are clearly non-self-published for our purposes. An "about this agency" or "about this politician" page (such as https://www.mcconnell.senate.gov/public/) might be self-published but its ordinary uses would fall under ABOUTSELF exemptions anyway. Someone answering questions at a press conference (a politician, a police chief, a disaster response official), on the other hand, might be self-published.
I think the key point for organizations is to think about whether the (or at least a) main purpose of the organization is to publish information, and whether you therefore deliberately do so with the kinds of responsible oversight that we expect from traditional publishing houses (e.g., ordinary newspapers, magazines, books, academic journals...).
A small handful of people who create a small press and set up ordinary industry controls (e.g., external authors) are non-self-published, but a small handful of people who create an organization to publish whatever they write themselves is self-published.
A small handful of people who create an organization to produce widgets and happen to also publish some information about widgets (e.g., in the hope that providing good information will lead to increased sales) is self-published. An equally small handful of people who create an organization to produce a trade magazine about widgets and the widget industry, operating recognizably like any other trade rag, is non-self-published.
2b. Whether or not something is self-published is about if it is reviewed by an editor before publication. No barrier materials are definitively self-published. Non-traditionally published materials can go either way. Corporate authors publishing about themselves are pretty much always going to be SPS. Advocacy groups that have an editorial structure to review, edit, and potentially reject material prior to publication may present other concerns (like bias, notability, etc.), but are not SPS. CambrianCrab (talk) 02:31, 27 January 2025 (UTC)[reply]
2c. Initially I was going to recommend some nuance for major organizations/governmental bodies (e.g. WHO) that might release influential position statements, but then I remembered that such sources are influential because they're loaded with experts who would naturally meet expert SPS. Grey literature from orgs like GLAAD may or may not be reliable depending on the author's expertise, but they are still SPS. JoelleJay (talk) 06:15, 27 January 2025 (UTC)[reply]
2b (or 2a, I think categorizing things "published by an organization and the content is about the organization itself" unnecessary because ABOUTSELF). Exempting "traditional publishers" is wholly arbitrary, and fails everything in the margins - also no one besides us would ever class as self-published. How long does an online publication have to exist before we declare it a "traditional" publisher? EXPERTSPS is not enough to overcome that because you can't use EXPERTSPS on many significant pages. Though, if we do go with 2c, it will be kind of funny to have to remove the ADL/SPLC/GLAAD from every single BLP claim it is used for (imagine if after years of arguing whether they are reliable, we backdoor declare them unreliable for 95% of its uses through a technicality. only on wikipedia) PARAKANYAA (talk) 07:53, 27 January 2025 (UTC)[reply]
I think there's a difference between "this is SPS but falls under ABOUTSELF" and "this isn't SPS." For example, consider political campaign materials. If you choose 2a, you're saying that they aren't self-published, in which case they might be used as a source for an attributed statement about the political opponent (though such content might be rejected as undue). If you choose 2b, we'd likely conclude that all of a campaign's materials are about the campaign/politician themself, even when the campaign is ostensibly commenting on a political opponent. In that case, the statement about the political opponent is excluded under ABOUTSELF condition #2 (the opponent is a third party) and probably also under condition #1 (unduly self-serving). FactOrOpinion (talk) 14:08, 27 January 2025 (UTC)[reply]
Hm. If you say it that way, I see the issue. I will think on that more and may comment again, though I think that is a less pressing distinction to make given what we actually tend to argue about than 2b/2c. PARAKANYAA (talk) 14:11, 27 January 2025 (UTC)[reply]
Political campaign materials would be GUNREL anyway, so they would still need to follow ABOUTSELF. If we actually adopt 2b as written I feel using "this is mostly about the organisation itself" (which... is the organisation the politician or the campaign?) to exclude a claim that is not about the organisation would be less useful than using an actually assessment of reliability to exclude. If that is something that is actually intended rather than something that people think might be a useful side effect, I would suggest that the exact wording could use some more work to pin things down and leave less room to argue one way or another. Alpha3031 (t • c) 15:56, 27 January 2025 (UTC)[reply]
2a/b per CambrianCrab and PARAKANYAA. The "traditional publishers" thing is ridiculous and absolutely not how Wikipedia should be distinguishing reliable and unreliable sources. I believe WhatamIdoing's !vote above does not justify why 2c is necessary, as it completely ignores all the edge cases that make 2c problematic. There are also cases like this, a BBC article reporting on a scandal at the BBC, where I am fully in the 2a camp – the BBC is a reliable source and the fact that they are reporting on themselves in a negative light only reinforces this. Such an article should be fair game for the article Richard Sharp (banker), which does in fact cite the BBC several times – citing a reliable source like this shouldn't be questioned for somehow violating BLPSPS when it clearly does not. Toadspike[Talk]09:27, 27 January 2025 (UTC)[reply]
Basically, my reaction to Sources are also self-published if they're published by an organization and the content is about the organization itself depends very heavily on which organization we're discussing – if it's a reliable news outlet publishing an article, it's not SPS. If it's a reliable news outlet's corporate press release (e.g. [1]), it's SPS. If it's some other organization's about page or press release, it's probably also SPS, but might still be reliable per the expert clause. Toadspike[Talk]09:33, 27 January 2025 (UTC)[reply]
I think that having an agreed-upon definition – whatever it is – is what's necessary. I think 2c is sensible. None of the options (2a, 2b, 2c, or even 2made-up-on-the-spot) is necessary. Once we settle on approximately what the definition is supposed to be, we can sort out the details and make all the policies work. WhatamIdoing (talk) 20:38, 27 January 2025 (UTC)[reply]
Yes, I agree that revising our definition of SPS is necessary, but none of these specific options is. Sorry if this came across as a jab at you specifically; I did not mean it that way. Toadspike[Talk]11:03, 28 January 2025 (UTC)[reply]
I think there is a distinction between news reports about an entity by that entity and announcements about an entity by that entity and on behalf of that entity. JoelleJay (talk) 00:12, 28 January 2025 (UTC)[reply]
2c Corporate, charity and advocacy group websites and reports are self-published. Simply putting an additional person within a company to act as "oversight" does not change a publishing arrangement, which is a structural one, with commercial separation between a publishing organisation and the titles it publishes. By 2a and 2b, any celebrity twitter account where an intern reports to a supervisor is somehow not "self-published", which is no standard at all. The sole impact of this is use on BLPs, and this comes up time and again IMO because of a desire to use partisan advocacy websites as sources of contentious political material about 3rd party living persons. SPS creates a blanket ban against such things unless a secondary, non-SPS picks it up. Void if removed (talk) 10:13, 27 January 2025 (UTC)[reply]
@Void if removed What do you think of the BBC example I gave above – do you think this article could be used on this BLP? I believe it would run afoul of the "about the organization itself" clause in option 2b (and thus also 2c). I don't mean to badger, I am interested in hearing the reasoning of 2c supporters. Toadspike[Talk]10:52, 27 January 2025 (UTC)[reply]
I don't think that should fall under this definition - this is a news report about events involving the BBC that goes via the normal publishing channels, whereas the "about the organisation itself" clause is intended to reference material like this. It could perhaps be worded better to make that distinction clear. Void if removed (talk) 11:07, 27 January 2025 (UTC)[reply]
@Void if removed "Simply putting an additional person within a company to act as "oversight" does not change a publishing arrangement, which is a structural one, with commercial separation between a publishing organisation and the titles it publishes" - by this definition, would newspapers not be self published? If not, why not? Other than an arbitrary carveout for them as traditional there is little practical distinction in the age of the internet. PARAKANYAA (talk) 11:04, 27 January 2025 (UTC)[reply]
I don't want to rehash the same arguments yet again but no, newspapers are not self-published, and yes there are practical distinctions. Again, the confusion arises because in English we use the same word "publishing" both for the business of publishing, and the act of making material available, but they are not the same thing. I can publish something on a website but that doesn't make me a "publisher" in a business sense, and that's the difference between whether something is or is not self-published. We tend to ignore the business meaning of "publisher" and everything an actual publishing company does in terms of promotion, advertising, distribution etc etc because all we care about is if the information is available digitally so we can use it on wikipedia. To us, a newspaper website and an advocacy website look the same, and the commercial distinctions are of no obvious importance. But we have this policy, SPS, that arguably makes that distinction and another policy, BLPSPS, that depends entirely on this distinction, and thus we have to care. Void if removed (talk) 11:24, 27 January 2025 (UTC)[reply]
About "the idea that a source being owned by a business enterprise": I don't think that being in the business should imply being a for-profit business. ProPublica is a non-self-published non-profit organization. The Guardian is a non-self-published non-profit organization. Many individual academic journals are sponsored by non-profit organizations. To avoid being self-published, we do need some sort of "organization", but we don't need a profit motive. WhatamIdoing (talk) 19:00, 27 January 2025 (UTC)[reply]
I mean if we're going to go down the route of "being owned by a business enterprise" then we are eliminating virtually every newspaper, magazine, or publishing house in the world that isn't a worker-owned co-op. (Or, ironically, government-owned media.) Gnomingstuff (talk) 19:07, 28 January 2025 (UTC)[reply]
2a, with 2b as a very very distant second if it's necessary to block the absurdity of 2c. Oppose 2c in strongest possible terms. That said, none of these are quite accurate, and I'm only supporting even 2a (let alone 2b, which has its own problems) in order to put a stake in the heart of 2c. A SPS is a type of non-WP:RS - more specifically, it's the policy that covers the limited an exceptions that allows us to use an otherwise unreliable sources in certain contexts. Reliability is contextual, but broadly speaking, to say that something is a SPS in a particular context is to say that it is a non-RS in that context. This means that even 2a is not quite right (or at least leaves vital points out), because what matters isn't really whether they have an editorial process or fact-checking (though those are of course required for all RSes) but whether those grant it a reputation for fact-checking and accuracy. 2b is fairly absurd because, again, SPS is mostly an exception allow to non-RS sources in certain limited contexts; and the biggest part of that exception is WP:ABOUTSELF (is the intent to do away with that? It seems unclear.) But 2b's serious flaws are still relatively minor compared to 2c, which is anathema - it would make our judgements about reliability of sources depend on entirely superficicial aspects of their stated structure and organization, in ways that have nothing whatsoever to do with actual reliability. It would simultaniously disfavor sources with strong reputations, editorial controls, and fact-checking processes that obviously pass WP:RS via non-traditional publishing methods, while favoring sources that ought to fail RS just because they put on a cheap suit and pretend to be a newspaper. Nothing about this is workable. "Newspapers, journals, and books only" has never been the purpose of RS or SPS - there are obviously a large number of websites of various non-traditional types that pass RS, and this has been universal practice throughout most of Wikipedia's history. --Aquillion (talk) 14:35, 27 January 2025 (UTC)[reply]
Just going over WP:RSP, sources that currently have a clear consensus that they're WP:GREL and which people might reasonably argue should be categorized as self-published under 2c (and therefore disfavored) include Amnesty International, the ADL (outside of I/P), Aon, Atlas Obscura's articles, Behind the Voice Actors, Bellingcat, Climate Feedback, Common Sense Media, The Conversation, Metacritic, Rotten Tomatoes (the conclusion specifically separates out the user reviews as self-published, making it clear that there's a consensus that RT's own aggregations and articles are not), SCOTUSblog, Snopes, and the SPLC. I skipped over videogame ones because you already asserted that you believe them to pass 2c but for many of them I'm not seeing it - they're just described as websites or blogs. I suspect you'd also argue that some things on this list should be SPSes, or that others pass somehow, but honestly to me that underlines the problem - it's a vague vibes-based standard based on editors' personal feelings about superficial aspects of how a source is structured, so it'd inevitably lead to confusion. And for the ones that we should be exerting caution for in certain contexts, we already have much more clearly-worded policies (WP:INDEPENDENT and WP:RSOPINION.) SPS was never intended for the use you're trying to turn it to here and isn't a good fit for it. --Aquillion (talk) 18:48, 27 January 2025 (UTC)[reply]
@Aquillion, what makes you think that a publication like Atlas Obscura ("American magazine and media company", according to its WP:SHORTDESC) would be considered self-published under 2c? Magazines are traditional publications, and they are published by traditional publishers, right? WhatamIdoing (talk) 20:42, 27 January 2025 (UTC)[reply]
I don't interpret "traditional" as meaning "formally classified as a traditional newspaper" etc., but rather as having to the editorial controls that make something a traditional publisher. I also don't see how 2b would eliminate ABOUTSELF or any other SPS exemption? JoelleJay (talk) 00:33, 28 January 2025 (UTC)[reply]
@JoelleJay, your response makes me curious: do you agree with how I classified the 2c examples in the Table, or would you instead say that some of the things that I classified as SPS for 2c might instead be non-SPS because you consider them to have the same editorial controls as a traditional publisher (or, for that matter, that some things I classified as non-SPS are actually SPS because they lack such editorial control)? You're correct that publications that are considered self-published under 2b might still be acceptable under ABOUTSELF/BLPSELFPUB, but it depends on what one wants to source to them, since ABOUTSELF/BLPSELFPUB excludes content about third parties and unduly self-serving content. FactOrOpinion (talk) 22:37, 28 January 2025 (UTC)[reply]
I would be ok with considering releases by specialist governmental bodies on topics under their purview, such as consensus medical guidance like the essential medicines list by WHO, as non-SPS. I think "advocacy org websites" is far too broad a group to apply blanket categorization to directly; rather, if such an org does have traditional media editorial controls then it may be non-SPS and therefore qualify as such under 2c. JoelleJay (talk) 02:07, 29 January 2025 (UTC)[reply]
I'm fine with a/b-ish, though I do partially share the concerns raised by Aquillion. I don't think it would make too much of a difference in practice, but the distinction for self publication in current practice, at least in my opinion, seems based on the effectiveness of the "barrier" or publishing process. A GENREL source that derives its reliability from their editorial and publishing process is not self published. If it derives reliability solely from the EXPERTSPS exemption for the author, then we can consider it self-published. If it's not reliable either way, then I suppose a/b might classify some things that are self-published as non-self-published, though I don't really care too much since if it's not reliable we shouldn't be using it anyway. Alpha3031 (t • c) 16:28, 27 January 2025 (UTC)[reply]
For example, do you want a model that would let you say the publication-focused advocacy organization that was responsible for the original publication of Dressed to Kill (book) was self-publishing, but the far more reputable, equally publication-focused advocacy organization that's responsible for https://www.splcenter.org/resources/extremist-files/ is not?
Do you want no definition at all, so editors can apply the label whenever they want, even if the results are contradictory and unpredictable? I'd like to have a definition, but I understand that not everyone values consistency. WhatamIdoing (talk) 18:17, 27 January 2025 (UTC)[reply]
I think no attempt at a one-size-fits-all definition is ever going to work well. What always matters is whether the source is sufficiently reliable for the information it covers. Beyond the obvious cases the whole question of whether a document is self-published or not just appears to be the wrong question. It should be possible to reference WHO to say that they employ Tedros Adhanom Ghebreyesus as their Director General, or the British government to say that Keir Starmer is the UK's prime minister, or the University of Cambridge to say that Deborah Prentice is their VC, even though all are living. Further, it should be possible to reference WHO to say that WHO estimates the number of cases of monkeypox in the UK to be x, but the British government estimates them at y, without waiting for either number to come out in a peer-reviewed journal. Espresso Addict (talk) 20:15, 27 January 2025 (UTC)[reply]
Whether a source is self-published is one of the many factors we have traditionally used to determine whether a source is reliable. If we disagree on whether it's self-published, we are likely to disagree on whether it's reliable.
All of the examples you give (WHO/UK/UC) would be permitted under 2c. The first set would be explicitly permitted under WP:BLPSPS (which says it's fine to use self-published materials from a reputable organisation publishing material about who it employs), and the latter only needs to pass WP:MEDRS (which will generally accept it). WhatamIdoing (talk) 20:48, 27 January 2025 (UTC)[reply]
The moment we label, say, all WHO publications as "self-published", someone will start saying we can't use them to support some perfectly vanilla piece of information. Espresso Addict (talk) 20:58, 27 January 2025 (UTC)[reply]
Given that the WHO's raison d'être is collecting and sharing information, I think there's a good argument to be made for considering them to be a traditional publisher. Look through https://www.who.int/publications They publish books, they sponsor academic journals – they are effectively in the business of publication. WhatamIdoing (talk) 22:49, 27 January 2025 (UTC)[reply]
2d. I lean closest to 2b but this is just a mess of edge cases, and besides obvious cases like open wikis, how much of a barrier there is to publication is not information that is usually divulged to the public. This works in all directions -- organizations claiming there is more of an internal barrier than there actually is, organizations simply not mentioning the internal barrier that exists because they have better things to do, and internal barriers changing because of new hires/departing employees/shifts in internal org structure/etc. The stuff at the bottom of the table is also very strange. Why are small presses (as opposed to vanity presses) even in this conversation? What on earth would one be verifying with a music album (do you mean the liner notes)? And live TV broadcasts are actually a pretty good example of what I mean: the live real-time broadcast is "self-published" but the version that goes up on the Internet -- which is the one that we would probably be citing -- might have been edited in the interim. Gnomingstuff (talk) 19:05, 28 January 2025 (UTC)[reply]
@Gnomingstuff, re: your choice of 2d, would you be willing to say how you'd explain to a new editor what kinds of publications are self-published? Because ultimately, we need some way to explain this / capture consensus practice.
The small press example is there because WhatamIdoing asked me to add it. She's in a better position than I am to explain why, but my guess is that she wanted an example demonstrating that something can be a traditional publisher without being a large organization. The music album is there to demonstrate one way in which the existing characterization fails, as it cannot distinguish between SPS creative work and non-SPS creative work (though perhaps in practice that distinction doesn't matter, since it can be used under ABOUTSELF even if it's SPS). An opinion column — which we do use for attributed statements, and which would likewise fail the current characterization — might have been a better example. There are many examples that would be good to include, but I figured that I could only make it so long. Personally, I disagree that a live TV broadcast is self-published. I'd say that it's published by the TV station rather than the reporter, and that it's not an example of the station serving as a vanity press for the reporter. True, a reporter could say something that the station doesn't approve of, but I think that if a reporter started to do that, the live broadcast would quickly be cut off by the station. FactOrOpinion (talk) 23:05, 28 January 2025 (UTC)[reply]
Not sure it's even possible because, as you said, in practice, we can often only guess at whether there’s a reviewer, much less whether they have a COI and are assessing reliability. (To which I'd add: whether the reviewer is competent, whether they're usually competent but this one time they had a rough day, whether the reviewer is just rubber-stamping, whether someone went in after the reviewer and changed stuff, whether there are 10 reviewers and 9 of them are competent but 1 of them "fact-checks" with ChatGPT...)
I think the two basic questions here are "is there a barrier to publication?" and "is the barrier bullshit?". But I don't think it's possible to make a hard-and-fast rule that covers all or even most cases, except obvious stuff like open wikis and vanity presses being self-published.
At any rate I think question 1 is the more pressing one, because the current text of SPS does not align with 2a, 2b, or 2c -- or, for that matter, the "current explanation" column in the table. My preference is 2b > (very slightly) 2a > (by a large margin) 2c, but honestly, any of those would be better than what we have right now. Gnomingstuff (talk) 00:18, 29 January 2025 (UTC)[reply]
We could just write in to the definition "is the barrier bullshit" and leave the actual determining of whether it's bullshit to essentially the same process we'd use to determine whether a source is reliable or unreliable (talk page or RSN). Our policy and guidelines don't actually need to painstakingly call out every single possible edge or corner case, just establish the general principles we should work on, in the same way that the current definition of reliable goes through general concepts of reputation rather than having editors scrutinise whether a source uses a specific format of review process. Alpha3031 (t • c) 07:55, 29 January 2025 (UTC)[reply]
This one I think is actually a bad RfC, I don't think that it is possible to make such sweeping statements about whole classes of sources (at least not in a way which is genuinly helpful), that is exactly what editorial discretion is for... That people constantly argue over the nuances of a given source isn't a bug, its a feature. Horse Eye's Back (talk) 23:10, 28 January 2025 (UTC)[reply]
2b. A corporation's statements about itself are self published, whether it's a corporation of one person or one with a PR department of hundreds. Yet I don't want to muzzle the Southern Poverty Law Center because they're not a traditional newspaper. --GRuban (talk) 01:26, 29 January 2025 (UTC)[reply]
