I understand that WP:V is a content policy, but I and a growing number of other editors see major problems with "truth" in the Wikipedia namespace. Specifically, there are multiple instances of editors making claims that they believe to be true, but are in fact not true. These claims are made without presenting any evidence, and can (and do) mislead other editors. The most egregious example has been WP:OUTCOMES, which until recently made scores of claims (some untrue) about "common outcomes of AfDs." Many of these have been correct thanks to a few editors who strove to clean that area up. However, it continues to happen in other areas.

Some claims about Wikipedia are immediately verifiable and don't require any evidence (e.g. anything that states or quotes a policy/guideline, and anything that states the obvious). However some claims are not (e.g. "In the last 2 years, all articles about elementary schools have been deleted in AfD for being non-notable").

My question is: shouldn't claims about Wikipedia which aren't immediately verifiable include some sort of evidence to back them up? The essay WP:Inaccuracies in Wikipedia Namespace is about this issue, and covers it in more depth. I think a well-written guideline could solve this problem once and for all, but for some reason editors seem loathe to regulate WP namespace.

I welcome all thoughts and suggestions. ɳorɑfʈ ^Talk! 07:35, 8 April 2010 (UTC)

Good grief. I've yet to see that your "growing number of other editors" is anything more than new editors you've recruited or established editors whose support is at best nominal. Jclemens (talk) 07:44, 8 April 2010 (UTC)

Now that Mr. Negativity has had his say, I still welcome comments and suggestions. ɳorɑfʈ ^Talk! 04:33, 15 April 2010 (UTC)

For a long time we have had language to the effect that:

The threshold for inclusion in Wikipedia is verifiability, not truth,

with various explanations of what that means. When I joined Wikipedia the meaning was very clear: We don't say anything that cannot be verified. I have traced this back as far as February 2006, a time when two parallel versions of the policy were merged, and I guess it's not worth tracing it back even further:

The criterion for inclusion in Wikipedia is verifiability, not truth. This means that we only publish material that is verifiable with reference to reliable, published sources. [1]

This always meant that verifiability is necessary, not that it is sufficient. The word "criterion" was a bit misleading, so it was changed to "threshold". [2] A change in June 2006 preserved this:

The threshold for inclusion in Wikipedia is verifiability, not truth. "Verifiable" in this context means that any reader must be able to check that material added to Wikipedia has already been published by a reliable source, because Wikipedia does not publish original thought or original research. [3]

After some more editing, the paragraph looked like this at the end of the year 2006:

The threshold for inclusion in Wikipedia is verifiability, not truth. "Verifiable" in this context means that any reader should be able to check that material added to Wikipedia has already been published by a reliable source. Editors should provide a reliable source for material that is challenged or likely to be challenged, or it may be removed. [4]

Apparently this wording remained unchanged throughout 2007, but it was modified in April 2008:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether material is attributable to a reliable published source, not whether we think it is true. "Verifiability" in this context means that readers should be able to check that material added to Wikipedia has already been published by a reliable source. Editors should provide a reliable source for quotations and for any material that is challenged or is likely to be challenged, or it may be removed. [5]

After a period of some instability the text looked like this in May 2008:

The threshold for inclusion in Wikipedia is verifiability, not truth—meaning, in this context, whether readers are able via attribution to check that material added to Wikipedia has already been published by a reliable source, not whether we think it is true. Editors should provide a reliable source for quotations and for any material that is challenged or is likely to be challenged, or the material may be removed.

This was tweaked in June 2008:

The threshold for inclusion in Wikipedia is verifiability, not truth—that is, whether readers are able to check that material added to Wikipedia already has been published by a reliable source, not whether we think it is true. Editors should provide a reliable source for quotations and for any material that is challenged or is likely to be challenged, or the material may be removed. [6]

In January 2009 the paragraph was split:

The threshold for inclusion in Wikipedia is verifiability, not truth—what counts is whether readers can verify that material added to Wikipedia has already been published by a reliable source (see below), not whether we think it is true.

This policy requires that a reliable source in the form of an inline citation be supplied for any material that is challenged or likely to be challenged, and for all quotations, or the material may be removed. This is strictly applied to all material in the mainspace—articles, lists, and sections of articles—without exception, and in particular to information about living persons: unsourced material about living persons must be removed immediately. [7]

As a result of this split, the first sentence became vulnerable to a fundamentalist interpretation: That we have an obligation to publish even clear untruths just because there is a "reliable" source stating them. Once something passes the verifiability threshold, it can be forced into the encyclopedia.

This is not a theoretical problem, as admins can verify by reading Talk:Sam Blacketer controversy. Non-admins may get an idea from reading the AfD. There are situations where our internal processes are perfectly capable of proving a "reliable" source wrong, and it is extremely hard to deal with wikilawyers who insist that truth is irrelevant and that we are under an obligation to parrot what has been erroneously reported. In that particular case libellous information about an arbitrator was edit warred into the encyclopedia, and information that showed the libellous information was incorrect was edit warred out based on the technicality that it was easily verifiable by our own internal processes (edit logs etc.) but not through a published source.

As someone mentioned above, and as I can confirm, "verifiability, not truth" is typically used as an argument for putting eccentric claims into our articles, i.e. it is typically used the wrong way round, as if it described a sufficient condition. That was the case even when the wording made it clear that it is only a necessary condition. Now it has become worse. What can we do to fix this? Hans Adler 09:28, 10 April 2010 (UTC)

I agree that there are fundamentalists (or vandals) in Wikipedia who believe that all they need is a source for anything and it is fine to put it in here even if it is clear the "fact" is untrue. The unspoken "rule" is (or should be) that a piece of information must be "verifiable that it is true", if we could change the sentence to that it would solve many of our problems. Just as it is not acceptable to put something in an article that is true but unverifiable, it should be unacceptable to put something in that is verifiable but untrue. They go hand in hand. A reliable source for material can not be considered a reliable source in that instance if the piece of information is incorrect.Camelbinky (talk) 14:46, 10 April 2010 (UTC)

I think "not truth" should be deleted. Maurreen (talk) 15:17, 10 April 2010 (UTC)

I disagree... there is a valid reason for the statement. The fact that the occasional editor will try to twist it to their own ends is an annoyance, but worth it compared to what we would have without this statement (editors adding material because they believe "it's TRUE" or removing material because they believe "that's not TRUE").

we have to be careful in situations like this... We do not want editors to remove material simply because the there is a difference of opinion... we have to clearly demonstrate that the source is inaccurate. There is a difference between believing that X is wrong... and demonstrating that X is wrong.

As for situations where it can be clearly demonstrated that a source is inaccurate on some particular point... I would say that we can deal with it through consensus. If a consensus of editors agree that a particular source is wrong on a particular point, they can declare the source to be unreliable for that point. Blueboar (talk) 15:50, 10 April 2010 (UTC)

Can you clarify how you think "not truth" is helpful? Maurreen (talk) 15:54, 10 April 2010 (UTC)

Yes... it helps in two ways... the most important is that it makes the policy crystal clear to those who try to add material that they believe to be true (whether that belief is based on "I remember being told this by my professor when I was in college" or "my Pastor told me it's true"). It says... We don't care whether you think this is true... what we care about is whether you can cite a source for it. It can be absolutely 100% True... but if you can not cite a source, don't add it.

On the flip side, it clearly tells editors who might wish to remove information for POV reasons the same thing. We don't care whether you think the information if False... a reliable source says it, and that is good enough for us. If you want to remove this, you have to clearly demonstrate that the source is wrong... Or challenge it on some other grounds. There are lots of ways to legitimately challenge inaccurate cited material... but saying "it isn't true" is not one of them. Blueboar (talk) 16:20, 10 April 2010 (UTC)

I agree that we need to get the balance right, and removing "not true" is probably a bit too extreme. Do you have any ideas what we could do instead? I don't, unfortunately. Hans Adler 20:23, 10 April 2010 (UTC)

I agree with Blueboar that simply saying "this isnt true" or the reverse to keep/delete material is unacceptable; however there is still a problem of material that is cited but false. Most often someone with a little digging can uncover that the cited material is false. An example is in the Capital District, when I rewrote the article the cited material that said the Albany Times Union newspaper was originator of the term Capital District for the Albany-Schenectady-Troy area didnt seem right to me though it was a RS I did some digging in Google Books and found multiple references to the term Capital District decades prior to the RS's claim, I was able to prove that while we still dont know who or exactly when the term originated the sentence in the article had to go because it wasnt True. On the flip side I once saw an editor put a sentence into an article on a particular hamlet in New York that said it was a town, with a reference that called that hamlet a town. Towns in New York are political entities whereas the reference was referring to the hamlet in the generic "town" as in a dense built-up population. Ridiculous as it seems there was much arguing and the editor demanded a reference that flatly stated "X is NOT a town" be he/she had a verifiable reference for the place being a town and we didnt for the place NOT being a town. It is such arguments that make editors like me HATE the statement "verifiable, not the truth" (see the first section on my talk page for famous quotes that show why the continuation of "facts" as the truth leads to a decline in civilized intelligence).Camelbinky (talk) 21:37, 10 April 2010 (UTC)

Hans, just noting here that "the threshold for inclusion" language dates back to August 2005 after it was suggested on another page set up to discuss how to present the concepts of verifiability and NOR. Describing verifiability as the "threshold" seems clearly to describe it as a necessary condition. So long as we emphasize the attributable/attributed distinction nearby, which I did yesterday but I see someone removed it, it should be clear enough that (a) not every single thing that's sourced can be added and (b) not every single thing that's added needs a source. SlimVirgin ^{talk contribs} 22:22, 10 April 2010 (UTC)

Blueboar, thanks for your clear comments above. Very beautifully put!

Further up on this page I have just referred to Hans Adler's misunderstanding of this part of the policy. I have replied to his discussion against me here. That thread has just been closed by someone, but here is my reply to his misunderstanding and misrepresentation of this policy. (I also suggest that you all read my comment above.):

"Verifiability, not truth" is a very fundamental policy. If you (Hans Adler) want to change that wording, thus allowing editors like yourself to incessantly argue with editors who hold opposing opinions about the truthiness of a statement, and letting the majority determine what's allowed here, then change the policy. That would create a nightmarish situation of POV articles where the opposing POV, even if published in V & RS, would not be allowed any mention because a cabal of editors on one side determine a source should be eliminated because it doesn't jibe with their version of "truth".

Your version of Wikipedia would be "an Orwellian ministry of truth....That is a pass to which we ought not to come again." (From the British Chiropractic Association vs. Simon Singh appeal case.) BTW, there is nothing "absurd", "outdated or false" about the NSF/NSB statement. It's just as relevant today as it was in 2006. That pseudoscientific nonsense hasn't become scientific since then.

To really understand this, you all might want to take a quick look at the now closed thread. Note that all of Hans Adler's comments on this page need to be seen in light of his desire to change this policy to enable him in his edit warring against the use of the National Science Board's exact quote, because he thinks they are wrong. It seems to me that this wording is meant to deal with editors such as him who wish to force their version of truth in attempts to exclude sources with which they disagree. -- Brangifer (talk) 02:45, 11 April 2010 (UTC)

“

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.

”

“

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether some editors think it is true while other editors may think it is false.

”

The stark contrast offered by the unbalanced bolded phrase is necessary to indelibly implant in the minds of editors that Wikipedia is an encyclopedia, and not the many things that Wikipedia is not. If we are to shoot for clarifying balance in the lede sentence, the addition of true's opposite, "false", is all that is necessary. It might show that the "black and white" of "true and false" is made pale in comparison to the more desirable "verifiability". Just a suggestion.
 —  Paine (Ellsworth's Climax)  03:49, 11 April 2010 (UTC)

When we say "truth" we think about something that is trivial. If the truth is "the sky is blue", then who would dare suggest we shouldn't follow the truth when writing here? The problem is that the "truth" is actually a controversial thing at many topics. Even if we cut down the topics subject to opinion or analysis, there are many others where the lack, ambiguity or unreliability of available information does not allow to come up with a clear statement of which is "the truth", but many alternatives with varying levels of plausibility. Of course, some questions have a factual answer, "Are there aliens at Area 51?" can only be answered with either "yes, there are" or "no, there are not", but we don't have enough information to have a clear answer. "The truth" does not always equal "an universally accepted statement". And even more, there are also topics whose answer is factual, but is intrinsically tied to a bigger topic that belongs in the realm of ideas and opinions. Finally, we also have authors that follow a certain ideology or agenda and give more light to the information that benefits or harms something they agree with or oppose, or cherry-picking facts in order to do so ("POV warriors" are hardly an invention of Wikipedia, Wikipedia only designed this modern name).

In short: the potencial problem for wikipedia is that a "POV warrior" may be tempted to reject other viewpoints or underestimate the existence of real-world disputes about something, by using his sources to state that the other ones are "wrong" and are unreliable because they do not tell "the truth". To my understand, this is what this statement in the policy aims to prevent. MBelgrano (talk) 03:52, 11 April 2010 (UTC)

Paine, I like your second proposal. It gets to the heart of the matter for why we even mention "not truth" at all. The ending phrase should be included: ".... not whether some editors think it is true while other editors may think it is false." What is "truth" is often an unsettled matter, so both POV, if published in RS, need to be mentioned. They may often need to be mentioned as opinions with attribution, but one should not be excluded because a cabal of editors don't like it or think it's untrue. That would be wrong. While I'm a skeptic and generally oppose the ideas proposed by fringers and alties, I support the inclusion of their ideas if published in V & RS, not when stated as fact, but when stated to document that "this is what they believe". NPOV requires that we do this, while UNDUE and FRINGE require that we do this in a manner that makes it clear which is the mainstream POV.

MBelgrano, you seem to understand this manner in the same way I do. I'm wondering what you all think of what I wrote above to Hans Adler? -- Brangifer (talk) 04:43, 11 April 2010 (UTC)

(edit conflict). Thank you, BR, and just like many English words, truth can have several meanings. Some editors who read verifiability, not truth may at first be taken back by such a seemingly appalling statement. They think of "truth" in more of a philosophical sense rather than in the trivial sense in which it is really meant here. However, if they get through their "surprise" and read further, the policy does furnish them with what is really meant. Perhaps, though, by including the "not whether some editors think it is true while other editors may think it is false" in the lede sentence, this might help them get over their surprise more quickly, and keep them from "taking up arms for the cause of Truth".

As for your previous discussions with Hans Adler, just know that we all come here with certain agendas. There is no way around that. If we all keep in mind that we are all here on our free time with pretty much the common goal of improving this encyclopedia, then we can turn our disagreements into a learning process. Maybe that sounds too hopeful, but that, to me, is the bottom line.

 —  Paine (Ellsworth's Climax)  05:36, 11 April 2010 (UTC)

Thanks for replying. I have started a section below for comments. -- Brangifer (talk) 06:10, 11 April 2010 (UTC)

We should add that verifiable information from reliable sources (or a single reliable source) that can be demonstrated to be factually wrong on the basis of reliable sources should either be removed from an article, or (if it is WP:DUE) be properly contextualized and presented with inline attribution.  Cs32en Talk to me  14:00, 11 April 2010 (UTC)

That's a logical paradox. If both are reliable, how can be so sure that it isn't the one making the correction the one who is not holding the truth?

Of course, obvious mistakes, such as mispellings or talking about the Beatles as a 5-members band, must not be cited as facts even if a reliable source publish them, or used to present a non-existent dispute. But it's unneeded to write down that, that falls in the realm of the "occasional exception" where policies shouldn't be followed as if we were robots MBelgrano (talk) 14:16, 11 April 2010 (UTC)

Your Beatles example is a good one to show how truth is problematic... Yes, The Beatles had only 4 members in the band at any one time, but they did have more than 4 members over the band's entire history... thus saying they had 5 members can be considered 'true', depending on what you mean by "# member band". Truth is subjective. Blueboar (talk) 17:20, 11 April 2010 (UTC)

Unfortunately, no reliable source is reliable all the time (at least none that I would know of). And there are editors who insist on including something that has been reported, say, on CNN, even if other reliable sources have published reports that contradict the information and CNN has not included the information in relevant articles for several years. The occasional exception, unfortunately, is exactly the point where endless disputes may start if there is no guidance on how to approach the issue. (Common sense also is something that does not mean the same thing to everybody.)  Cs32en Talk to me  17:07, 11 April 2010 (UTC)

Proposed improvement

I would like to see the ending phrase of Paine's second proposal included: ".... not whether some editors think it is true while other editors may think it is false." -- Brangifer (talk) 05:27, 11 April 2010 (UTC)

It's awkward; it uses the "some/others" construction that we're always advising against; and it misses the point. The point is that V, not T, is the necessary condition for entry. No point in adding "which some editors might think is false, while others may be unsure, while yet another group may swing back and forth, or may even insist that reality is more fluid than that." :) SlimVirgin ^{talk contribs} 06:54, 11 April 2010 (UTC)

Your version is the awkward one as it's unnecessarily long and convoluted, but I suspect you wrote that with a humoristic/sarcastic twinkle in your eye. No, the simple formulation proposed by Paine does the job quite nicely. It makes it clear that we're talking about ideas that are disputed, not about absolute and unquestioned "truths" such as scientific facts (the Earth has one Moon, and the Earth isn't flat) and medical facts (in a normal spine there are five lumbar vertebrae), where MEDRS dictates what sources to use.

There are many issues where debate exists and where editors may disagree. This "truth" wording is designed to prevent debaters on one side from excluding sources because they express what is considered "truth" by other debaters. In fact it is designed to exclude such arguments from occurring at all, if they occur in a manner that diverts attention from actually building the encyclopedia. Many talk pages have a template which includes these words: "This is not a forum for general discussion of the article's subject." Discussions about the truthiness of a statement when used in an effort to completely exclude a source might well fall under this type of forbidden discussion. That type of discussion violates "verifiability, not truth". When editors understand that phrase properly, they should know better than to engage in that type of discussion for the wrong reasons.

There are obviously times to discuss the truthfulness of statements, but then it's for the purpose of determining whether a statement should be presented as unquestioned fact or an attributed and controversial opinion. The source is still used, even if it tells an outright lie. We have articles about very dubious persons where their deceptive statements are quoted, but they are only quoted to show their beliefs, not to make it look like their lies are unquestioned facts. -- Brangifer (talk) 14:52, 11 April 2010 (UTC)

I propose a second modification. After the statement, we can add "...with "truth" meaning "an obvious, indisputable or universally accepted fact", as not all topics can be completely described using only such statements. When there is no universal agreement between reliable sources as to which is the truth about something, Wikipedia will not attempt to settle who is right and who isn't, as it doesn't make original research. See the policy on neutral point of view to see how to manage disputes and polemic facts properly". MBelgrano (talk) 13:48, 11 April 2010 (UTC)

That seems pretty long, but there is a part that might be usable, and that's the last half starting with "When there is no....". See my proposition below. -- Brangifer (talk) 17:14, 11 April 2010 (UTC)

This doesn't solve one of the problems we had in the past: The truth was absolutely clear because it was in our editing histories and most of it could be verified by any editor; some details could only be verified by any admin, because some pages/files had been deleted. The problem was that what was uncontroversially (within Wikipedia) the truth had not been published by any reliable source. I can understand why we might not want to claim things that we know to be true but that only someone with knowledge of Wikipedia can verify; but it doesn't make sense for us to claim things that we know to be false, just because "reliable sources" got some Wikipedia internals wrong. In fact, the general public has a reasonable expectation that in matters concerning Wikipedia internals Wikipedia itself is the most reliable source and that we don't repeat untruths that others have written about us. If we refuse to do this very basic vetting, then we aer basically lying about matters concerning ourselves. And the Sam Blacketer controversy article was an example where for much of the time we did precisely that, libelling an ex-arbitrator under his real name. Hans Adler 14:55, 11 April 2010 (UTC)

Your suggestion seems to be dealing with other matters. Not that they're irrelevant, but they are different issues than the ones we're talking about right here. It just dawned on me that your comment belongs in the section above this one. -- Brangifer (talk) 15:10, 11 April 2010 (UTC)

No, I was responding specifically to MBelgrano's proposed formulation: "When there is no universal agreement between reliable sources as to which is the truth about something, Wikipedia will not attempt to settle who is right and who isn't". In the situation I described it would have supported the wikilawyers. In the end the situation was handled by means of WP:IAR, if I remember correctly (I can't read Talk:Sam Blacketer controversy, as it was deleted), but that wasn't entirely satisfactory. Hans Adler 15:37, 11 April 2010 (UTC)

Okay. Thanks for the explanation. -- Brangifer (talk) 16:52, 11 April 2010 (UTC)

The general principle – and I expect you disagree with it, but that's how it is – is that while we do not write anything that cannot be backed up by a reliable source, we have a lot more freedom when deciding (under this fundamental constraint) what to write and what not to write. Wikipedia is not trying to collect every last bit of information (such as the weekday on which a certain Jane Austen character proposed to another), so besides the verifiability threshold there is also a noteworthiness threshold. We mention a fact if it is independently noteworthy, or if it makes sense to mention it to flesh out an article. And there are additional constraints due to WP:NPOV: Otherwise many of our articles would be dominated by fringe positions, e.g. Earth by discussions of Hollow Earth and other similar nonsense about which there is a lot of information (but little debunking). What we put into our articles or not is not automatically determined by the sources; it's the result of consensus building which involves all available information sources, including editors' experience, Google searches, back-of-an-envelope calculations to check prima facies plausibility of claims, reputation-based arguments (if Erich von Däniken claims something it's probably false, even in the unlikely event that it sounds plausible), reading forum posts or Slashdot discussions etc. All of these methods are standard and perfectly proper. WP:NOR only talks about turning our original research into explicit claims in article space. (Actually, it also covers claims we make implicitly by juxtaposing statements from different sources in a suggestive way, but that had to be made explicit in WP:SYNTH.) Hans Adler 17:20, 11 April 2010 (UTC)

No, I don't disagree. There is a lot of truth to your statement. There are also many exceptions, but these are always considerations we keep in mind when determining content. This section happens to be dealing with a slightly different matter, not with UNDUE, SYNTH, etc.. -- Brangifer (talk) 17:35, 11 April 2010 (UTC)

Random break

Here is the current version:

• The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.

Combining Paine's and MBelgrano's suggestions we'd get this version:

• The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether some editors think it is true while other editors may think it is false.

  When there is no universal agreement between reliable sources as towhich is the truth about something's a statement's truthfulness, Wikipedia will not attempt to settle who is right and who isn't is wrong, as it doesn't make indulge in original research. (See the policy on neutral point of view to see how to manage disputes and polemic facts properly.) Wikipedia documents the real world using verifiable and reliable sources, it doesn't judge it. It is not an Orwellian "Ministry of Truth" which determines what is true. It only documents what reliable sources say about the matter.

I made some formatting changes and added a bit that may or may not be usable, ending up with this reading:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether some editors think it is true while other editors think it is false. When there is no universal agreement between reliable sources as to a statement's truthfulness, Wikipedia will not attempt to settle who is right and who is wrong as it doesn't indulge in original research. (See the policy on neutral point of view to see how to manage disputes and polemic facts properly.) Wikipedia documents the real world using verifiable and reliable sources, it doesn't judge it. It is not an Orwellian "Ministry of Truth" which determines what is true. It only documents what reliable sources say about the matter.

Are we getting closer to something that makes it clearer? It must be clear that "verifiability" refers to one thing, and "not truth" to another, and both elements must be explained. -- Brangifer (talk) 15:10, 11 April 2010 (UTC)

As I explained above, I disagree with this version as it is fundamentalist. If The Register publishes an article that claims that Wikipedia user BlaBlaBla has uploaded pedophilia to Wikipedia, and also mentions their real name, and if our checkusers say that there is no doubt at all that the user B1aB1aB1a (note the digits replacing letters) who did do such uploads was from a different continent, then I don't want to have to invoke WP:IAR again (as in the Sam Blacketer case) to keep the libel out of the encyclopedia. Hans Adler 17:26, 11 April 2010 (UTC)

Just because a good formulation can be misused doesn't mean we shouldn't use it. We should deal with the misuse using arguments based on other policies. I have also refactored your comment. I'm not interested in my username being found by search engines in connection with something so distasteful and perverted. -- Brangifer (talk) 17:41, 11 April 2010 (UTC)

My example shows that it isn't a good formulation because it's too extreme. Any formulation that allows people to wikilawyer along the lines of "yes, I am not actually denying that we all know that the reliable sources are wrong, but that's what they are saying and therefore we must say it too" is too extreme. The purpose of the "not truth" part is to help in our everyday dealing with POV warriors who are convinced they are fighting for the truth. It is decidedly not to support POV warriors who are knowingly trying to make Wikipedia say what is to them as to the rest of us an obvious untruth, be it to get even with an arbitrator or just for the LULZ. If something is not a general principle ("truth is irrelevant even when it is obvious") then we shouldn't pretend it is one. We need to get the balance right. Hans Adler 17:55, 11 April 2010 (UTC)

I'm sure there must be other policies that can be brought to bear in such situations. No single policy rules in sovereign isolation to the exclusion of all others. It is true that one of the many "purpose[s] of the 'not truth' part is to help in our everyday dealing with POV warriors who are convinced they are fighting for the truth," but it is also there to prevent POV warriors from excluding sources because they believe they are "untrue", even when they are debatable matters without any universal condemnation from all V & RS, IOW not unquestioned falsities. Editors' personal beliefs about what is "true" about a matter aren't allowed to determine such matters. Wikipedia must not become an Orwellian Ministry of Truth. -- Brangifer (talk) 22:53, 11 April 2010 (UTC)

It just occurred to me we would never have any of this discussion or the problem at all if only we had a fully functional, stand-alone, and respected wp:commonsense POLICY along with IAR. We could keep the policy saying "verifiability, not truth" and understand that IAR and commonsense would be used appropriately in understanding that phrase. Unfortunately wiki-lawyers are not as endangered as wikidragons (if only wikidragons ate wikilawyers... we could feed them and save them from extinction!). Could we change the wording to be "verifiability, not truth; but please use common sense and understand that it is not acceptable to put false information in an article just because you have a reference" or something to that effect? Not that it would stop POV pushers anyways...Camelbinky (talk) 19:23, 11 April 2010 (UTC)

"Putting [unquestioned in all RS] false information in articles" is never allowable without proper attribution and framing. We are forced by NPOV to do this all the time because it's the only way to document those false POV. Even strongly debated opinions that aren't universally considered false should be attributed. We have plenty of other policies that deal with this problem. It doesn't require rewording this phrase. This phrasing applies to debated topics where opinions differ, or at least it should be made plain that this is the case. That's what the propositions above do, and I've combined them into a single proposed improvement. -- Brangifer (talk) 22:53, 11 April 2010 (UTC)

I agree with Hans that the proposed wording is unnecessarily strong and could easily lead to nonsense being forced into articles either to further a POV promotion, or simply for the wonderful LULZ that would ensue. The situation is tricky because I fully support the "verifiability not truth" dictum, yet a requirement to publish known-incorrect information is absurd. William Connolley gave another controversial example (discussion here): a prominent blogger wrote a piece that suggested WMC (when an admin) deleted 500 articles because he disagreed with the subject, and several argued that we (Wikipedians) absolutely know that the suggestion is not even wrong so it should be omitted, while others pointed out that "one of the key rules of Wikipedia was that the truth was not determined by us, the editors. Ergo, if a source meets WP:RS, then it's admissible for use...". I appealed to WP:REDFLAG to say that "exceptional claims require exceptional sources".

There will never be wording that covers all scenarios, and I do not want any weakening of "verifiability not truth", but there is no reason to strengthen it to a mechanical process. Also, we do "judge" sourced text in some senses (potential libel in a BLP requires strong and multiple RS; potential copyvios are deleted; WP:UNDUE text is deleted; WP:FRINGE blocks sources considered reliable by some). Johnuniq (talk) 03:45, 12 April 2010 (UTC)

The rather odd and exceptional cases cited don't undermine the need for the current wording, and I'm glad you don't feel the wording should be weakened. You have also pointed to a number of good policies and guidelines we do use to judge how we can deal appropriately with problems. I'm just interested in removing the excuse for POV warriors who wish to block the use of RS they feel are untrue, when there is external debate about the truthfulness of the statement, and especially when the only one protesting is the editor himself, rather than objections in RS. That's the background for Hans Adler's start of this entire thread. He's interested in excluding the use of an exact quote of National Science Board because he believes they made a mistake. The problem is that no RS make this objection, and plenty of other editors disagree with him, as well as the overwhelming results of two RfCs which also disagree with him and affirm that the statement and source are perfectly good to use here. He's trying to get this wording changed to further his edit war against the NSB, thus turning Wikipedia into an Orwellian "Ministry of Truth". That's why it needs to be strengthened a bit. Maybe not as much as I have, but just the addition of a few words would be enough: ".... not whether some editors think it is true while other editors think it is false." (The last seven words would be added to what we have.) This makes it clear we're discussing disputed statements, not obvious untruths. That misunderstanding needs to be prevented. -- Brangifer (talk) 05:55, 12 April 2010 (UTC)

The suggestions are too wordy and would introduce complications. In my view, the current formulation has strong consensus and is very clear. Verifiability, not truth, is the threshold for inclusion i.e. V is a necessary but not a sufficient condition. SlimVirgin ^{talk contribs} 03:59, 12 April 2010 (UTC)

See my comment above yours. I think the addition of about seven words would be enough to strengthen this phrase against misuse. It wouldn't be absolutely necessary to add all that I've suggested, although it's very useful content that could be tucked away in the footnotes as explanatory text, a practice we often follow on policy pages. -- Brangifer (talk) 05:55, 12 April 2010 (UTC)

The main problem is that people remember "verifiability, not truth", i.e. an incomplete version without the word "threshold", and when it is used it is normally used only to support inclusion of something. (To reject inclusion it's enough to talk about verifiability or missing reliable sources.) That's what is floating around everywhere, and what most editors encounter long before they first look at WP:V. Coming here with that background it's not unreasonable at all to read "The threshold for inclusion in Wikipedia is verifiability, not truth" as follows: "So long as it doesn't contradict other policies, we include everything that is verifiable, even if everybody agrees it isn't true. Our opinion simply doesn't matter because this is a project to summarise written knowledge, and verifiability is more than our approximation to truth."

This has always been a problem, and there may be a consensus that it's worth having this problem because pointing to "verifiability, not truth" is an effective way of making some POV warriors shut up. But this edit of yours has upset the previous balance by making the incorrect (meeting the threshold = sufficient as well as necessary) reading even more plausible. "Editors should provide a reliable source for quotations, and for any material that is challenged or likely to be challenged, or the material may be removed." This sentence stressd the necessity aspect of "verifiability, not truth" right after the principle was proclaimed. Without it a symmetric reading is much more plausible, and even the opposite reading (we must write everything that is verifiable, regardless of truth, although we may also write unverifiable things as well) would be possible if it wasn't so absurd and didn't contradict the rest of WP:V.

Also, the connotations that the word "threshold" carries according to you – necessity, not sufficiency – are part of our exegetical tradition, not inherent in the text. Hans Adler 05:45, 12 April 2010 (UTC)

I've noticed you often drive this to the extreme by stating "even if everybody agrees it isn't true." What about when it's only you who holds that opinion against the views of numerous others who disagree and think the statement and source are appropriate? Is edit warring in the absence of any external RS who disagree with the source proper? Is it then proper to use "threshold" to keep it out? I think not. It's that type of situation where quoting "verifiability, not truth", with emphasis on the "truth" part, should be enough to make contrary editors (who attempt to force their unsourced opinions regarding the "truth" of a statement on the encyclopedia) to pull back, stop objecting, and allow what many other editors consider to be a proper use of the statement and source. You can take the problem of unquestioned false statements and deal with them elsewhere. Here we're dealing with statements that might be debated, but they aren't unquestioned false statements. There's a world of difference how we deal with these very different situations. -- Brangifer (talk) 06:10, 12 April 2010 (UTC)

BullRangifer, what you apparently don't understand – and don't come with "AGF", because you said that my "comments on this page should be seen in the light of [me] wishing to change this policy to further the aims of pushers of fringe POV and to help [me] in [my] edit warring against the use of the NSB as a source", which is of course totally wrong – is that this is not at all about our current conflict. It is primarily about Talk:Sam Blacketer controversy. It's deleted, so I can't quote from it, but I am starting a new subsection below with some equivalent gems from WP:Articles for deletion/Sam Blacketer controversy. Hans Adler 09:45, 12 April 2010 (UTC)

It sounds like the Sam Blacketer case must have been pretty bad, even though I don't know the history. My suggestions are designed more to deal with fundamental problems that occur all the time, not just for some exceptional and rare cases like Sam's, so it's good you've started a separate section for dealing with those types of exceptional cases. We need to be careful that changes made to deal with rare cases don't create more problems for the common ones. That would mean even more trouble all the time. BTW, thanks for putting in the random breaks. This was getting pretty long and difficult to edit. -- Brangifer (talk) 13:52, 12 April 2010 (UTC)

Random break 2

Here is a version of the first paragraph that I could support:

The main threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.

This alerts the reader to the fact that there are other criteria, such as NPOV, that also may prevent addition. Hans Adler 06:00, 12 April 2010 (UTC)

That's the same as our current statement. It adds only one word without solving the part about truth. I'm not saying the addition isn't bad, but more is needed on the other end. -- Brangifer (talk) 06:10, 12 April 2010 (UTC)

Here's my suggestion, which combines yours and mine (actually Paine's):

The main threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether some editors think it is true while other editors think it is false.

Can we accept this? Brangifer (talk) 06:18, 12 April 2010 (UTC)

That's terrible. It posits "some editors think it is true while other editors think it is false" as a coherent inclusion criterion. —chaos5023 (talk) 06:24, 12 April 2010 (UTC)

The problem is that it's not the "main" threshold. What would the other thresholds be? And the true/false thing is just odd. That people think something is false is not the issue; we might want not to add something that's reliably sourced because we know it's an error, for example, or we know that another source has expressed it better or is more qualified. It's people thinking something is true and that it therefore belongs in WP that's the issue. SlimVirgin ^{talk contribs} 07:22, 12 April 2010 (UTC)

Some other thresholds: WP:REDFLAG. WP:NPOV, and specifically WP:UNDUE.WP:ONEWAY. Hans Adler 10:06, 12 April 2010 (UTC)

Not every edit can be neutral, and red flags can be okay if appropriately sourced. That sentence simply means that V is a necessary condition. Is there something about the wording that suggests to you that we're saying it's a sufficient condition? SlimVirgin ^{talk contribs} 10:23, 12 April 2010 (UTC)

Yes: The definite article ("the threshold") and the complete absence of any hint that it's only a necessary condition. The current wording is equivalent to the following:

What we require for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.

There is no indication that we also require something else. If a reliable source mentions Obama's shoe size we only need to find out where to mention it, because the information has passed the only threshold that could prevent its inclusion in Wikipedia. Hans Adler 12:03, 12 April 2010 (UTC)

Demonstration that editors misunderstand "verifiability, not truth"

The following examples are from WP:Articles for deletion/Sam Blacketer controversy. The colours highlighting some problematic parts are mine.

• Comment: "Reliable" sources whose accounts are wrong on almost every point. I know ... verifiability, not truth ... That is all very well as long as the article is about someone else ;) Is it really compatible with WP:BLP to have an article full of stuff which we know and can prove to be wrong, just by referring to our own archives? I have checked every edit Sam made to David Cameron going back to December 2007, when he became an arbitrator. Here is the edit apparently mentioned in the Daily Mail, where the Mail says he "tried to remove a reference to the Tories having a 'consistent' lead in the polls.". This is the only time I found Sam actually added content, rather than reverting vandals, since December 2007. Now, if you look at this edit, you will find that what he took out was running commentary on 2008 opinion poll results, cited to a 2007 (!) Reuters article http://uk.reuters.com/article/domesticNews/idUKL1310900320071013?sp=true – material which he replaced with cited material which noted that Cameron had appeared on the cover of Time Magazine, and was said by the Daily Mail to have been presented to the world as Britain's "Prime Minister in waiting". And Sam inserted the information that the Tories were "consistently" ahead. Some Labour activist! Sam actually put in the information these "reliable sources" accuse him of having taken out, just like he took out the unflattering attack picture these sources accuse him of having put in! Perhaps they don't know that if you look at a diff, it's the right side that has the new text, or that red text is text added, rather than deleted. What do I know. Our article here, citing the Daily Mail, says that Sam was "trying to adjust the description of the Conservative Party's lead in opinion polls over the Labour Party." This stupid innuendo and twisting of facts is unworthy of an encyclopedia, and it is unworthy of our project. JN466 22:08, 9 June 2009 (UTC)
  • Either WP:V matters, or it doesn't. Either it applies, or it doesn't. As you yourself pointed out, it's what's verifiable that matters. Much of the last 2/3 of your comment is simply original research and is prohibited, as I read policy. Unitanode 23:48, 9 June 2009 (UTC)

---

• Keep or failing that, merge the salient points to an appropriate article. Given the extended international coverage, even by such publications as the Corriere della Sera ([8]), our notability standards are certainly met. Yes, the coverage may be wrong, but WP:V's instruction to aim for "verifiability, not truth" does not contain an exception for issues about which we assume to know the (sadly unverifiable) truth, such as Wikipedia-related issues.  Sandstein  21:45, 10 June 2009 (UTC)
  • What is astonishing to me is how many people are willing to simply ignore both WP:V and WP:OR, simply because the original research comes from Wikipedia itself. Unitanode 22:36, 10 June 2009 (UTC)
    • It is not original research. It is primary sourcing. There is a difference. And Wikipedia -is- a reliable source on actions at Wikipedia, hence ArbCom can use diffs and the rest to determine appropriateness of rulings. So, information found on Wikipedia about actions on Wikipedia are enough to determine that the sources, if they contradict it, are unreliable. The same is a source saying that a bluebird is naturally red when all pictures of the bluebird shows that it is, indeed, blue. The Reliable Sources noticeboard look at credibility of reporting, especially when there is direct evidence that there is a mistake. Plus, newspapers can take up to a month to make corrections, if they even bother. Ottava Rima (talk) 23:05, 10 June 2009 (UTC)
      • • Can anyone else make any sense out of what Ottava Rima just typed? "It is not original research. It is primary sourcing." According to what I understand, the definition of original research is using primary sources instead of secondary ones. What you wrote makes no sense at all, and is not a justification for deleting this article in any way. Unitanode 00:02, 11 June 2009 (UTC)
          • I detect shortcutitis (citing shortcuts without actually reading the relevant policy). Try the relevant subsection of WP:OR, which is WP:PSTS. Primary sources are sometimes permissible. Disembrangler (talk) 00:14, 11 June 2009 (UTC)
            • I hardly know as many "shortcuts" as you do. I stumbled into this imbroglio, and am regretting every participating. Even still, a quote from your linked shortcut (does WP:IRONY, exist): "All interpretive claims, analyses, or synthetic claims about primary sources must be referenced to a secondary source, rather than original analysis of the primary-source material by Wikipedia editors." The relevance of this quote should be self-evident. Unitanode 00:19, 11 June 2009 (UTC)
              • The bit I was hoping you'd take away from PSTS was "Without a secondary source, a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by a reasonable, educated person without specialist knowledge." WP logs can therefore certainly be used in this way to back up the simple factual claims people have made about what actually happened vs what newspapers reported. Disembrangler (talk) 09:28, 11 June 2009 (UTC)
        • Original Research is not primary sourcing, and primary sourcing is not original research. Original research is to determine what is not readily available from a source of information. If an author writes a book, then you can discuss what the book says without saying what someone else claims the book says. Please look up the definition of "original". A "primary" source would not be original. This is readily apparent from actually reading WP:OR. Ottava Rima (talk) 00:31, 11 June 2009 (UTC)
          • To quote: "If no reliable third-party sources can be found on an article topic, Wikipedia should not have an article about it." That sounds like "don't use primary sources" to me. Reliable, third-party sources have been found here. We don't like their interpretation of the facts, so we want to delete the article? That makes no sense. Unitanode 00:36, 11 June 2009 (UTC)
            • Obviously, people applied the quote to why this should be deleted. The primary sources contradict the third party sources, thus making them unreliable. It has nothing to do with "not using primary sources". Primary sources are a source, but not a justification for notability. Don't dare confuse notability with verification. Ottava Rima (talk) 03:10, 11 June 2009 (UTC)

• • • • WP:IAR has been applied and this seems extremely sensible in this case. Smartse (talk) 23:54, 10 June 2009 (UTC)

---

[...]

But we aren't about truth, we are about verification. Those are the standards we apply to other articles, those must be the standards we apply here. --Cameron Scott (talk) 15:14, 11 June 2009 (UTC)

As per WP:BLP, it has to be both - "We must get the article right." That is one of our most important policies. A source on a BLP is not "reliable" unless it is extremely credible and not proven wrong. None of these sources meet the BLP requirements, as they hold factual inaccuracies that are blatant. Ottava Rima (talk) 15:36, 11 June 2009 (UTC)

Okay, I haven't been here that long, but even I know the "verifiability not truth" language. Do you really not know this? Unitanode 21:13, 11 June 2009 (UTC)

---

• Keep The deletion reasons are unconvincing. The deletion of the parent article does not affect this article the slightest.Many articles are unintelligible for people outside specific areas of interest but that in itself does not make it a good reason for deletion. Essjay controversy proves that it is in fact possible to write an article about a Wikipedia-related event based on external sources that is intelligible for people outside this community. Problems that the article might have in that regard can be solved via simple editing and while I understand the buzz the article creates here, we should apply our policies to all articles no matter the content. The article is sourced to multiple reliable sources (whether they got their facts right is not our concern, remember WP:V: Verifiability, not truth) establishing notability. Per WP:BLP1E this is not an article about the person but about the event. No other policy-backed reasons have been mentioned (NOTNEWS gets thrown around but this has continued for multiple weeks now). Regards SoWhy 10:24, 10 June 2009 (UTC)

Note that even an experienced editor who was arguing for deletion because of the inaccuracies (Jayen466) and a high-profile admin (Sandstein) understood the principle "Verfiability, not truth" as demanding that we publish what we know to be false, merely because it's in the "reliable" sources.

The situation was bad enough. I want to prevent that such a fundamentalist reading of WP:V will be even more defensible the next time something like this happens. Hans Adler 10:01, 12 April 2010 (UTC)

It's understood with all the policies that we have to use common sense in applying them. We could emphasize that, if you like. That if something is clearly an error, we shouldn't repeat it. But we have to be terribly careful, because just as people misunderstand the current policy, they will misunderstand anything you add, no matter how clearly you write it. They will rigidly apply a very literal understanding of it. So you could end up inadvertently making things worse.

This is why we emphasize that NPOV, V, and NOR (and BLP) must be read together, because they tend to offset one another. With the Blacketer issue, BLP clearly wouldn't allow that edit, because we knew it to be false. SlimVirgin ^{talk contribs} 10:29, 12 April 2010 (UTC)

I agree that we have to be terribly careful. The slogan "Verifiability, not truth", in spite of its occasional abuse, is a vital tool for stifling a certain type of advocacy.

Some problems with common sense:

• Some editors don't have it.
• Some editors think it isn't relevant (because principles are more important). (I think we create such editors when we fight against fringe, spam, POV pushing etc. using purely technical arguments and the editor at the receiving end thinks a common sense exception obviously should apply.)
• Some editors pretend to have no common sense, or pretend to consider principles more important, when it suits their purpose.
• You can't block someone for persistently violating common sense.
• You can't use any admin tools to enforce common sense.

My experience with discussions where it becomes necessary to explicitly invoke common sense is that ultimately it wins, but that can take many weeks of hot fighting.Hans Adler 11:55, 12 April 2010 (UTC)

• I agree about common sense. Problem is that, without it (deliberately or otherwise), the policies turned into sledgehammers, no matter how carefully written; rewriting to avoid problem A often introduces problem B. I think the community does usually do the right thing, but it can take weeks (and months and years) of RfCs and what-have-you. Sometimes the best thing is simply to wait. SlimVirgin ^{talk contribs} 15:39, 12 April 2010 (UTC)

Anyway...

Firstly, User:Hans Adler & User:BullRangifer should seek dispute resolution in another venue. A policy talk page and WP:RSN are not for content disputes. Secondly, I propose changing the first paragraph to read something like:

The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source. This does not mean false information should be added to an article; when reliable sources disagree, give due weight to the individual points of view.

I believe this would eliminate the misuse of the policy. I have not read the above discussion in full, as it is hard to cut through the content dispute to find what we are actually trying to fix here. Apologies if this has already been suggested. —Joshua Scott (LiberalFascist) 12:49, 12 April 2010 (UTC)

Firstly, this is not at all about my dispute with BullRangifer, although he has followed me here and is claiming that it is. Secondly, your version solves my problem, but the part after the semicolon doesn't really have anything to do with the rest. And I am not sure that the balance would still be right if we left it out. Hans Adler 12:59, 12 April 2010 (UTC)

Hans, my apologies, I was not meaning an attack on either of you, but I wanted to get back to the general issue at hand, rather than focusing on a specific instance. It wasn't my intent to slam anyone. —Joshua Scott (LiberalFascist) 19:38, 12 April 2010 (UTC)

Sorry for the delay (I had been writing a big article yesterday), but I want to notice my support for Brangifer's amendments over my text and the combination of it with the other proposals, at the green box up there. As for the futher development of the thread, I have another proposal: we may add a new notice at WP:NOR, stating that the fixing of obvious mistakes at reliable sources does not constitute original research. Even as reliable as a source may be, there are mispellings, typos or gross mistakes that slip by from time to time, and shouldn't be used to represent a dispute where there is none. However, it would be needed to discuss how to set apart an obvious case from a non-obvious one.

The problem with wikilawyering is when too much process is involved, such as creating policy for things that do not need any, either because the procedure is clear from already existing policies or because user's autonomy at such topic isn't harmful. When there are different perspectives about how to deal with a wikipedian topic, as in this case, anything that the policy might say has the potencial of being used for wikilawyering... and what is left unsaid may be used for wikilawyering as well.

If there is a problem with people remembering only a page name or quote, and not the policy actual content, simply point so. The spirit of a rule overcomes the wording of it; but the complete wording of it overcomes the wording of titles or internal slogans MBelgrano (talk) 13:09, 12 April 2010 (UTC)

The proposal may solve one problem ... but does not address the problem that caused us to include the "V not T" statement in the first place... people adding information that they think it is true without a source. We do need something that tells editors: "we don't care if you think something is true... what we care about is whether a reliable source thinks it is true." That is the core to the entire concept of verifiability. Blueboar (talk) 13:14, 12 April 2010 (UTC)

I agree, but since we are not mind readers, we don't know what the source thinks. We do know what the source writes, however, so the modified statement would be "we don't care if you think something is true... what we care about is what a reliable source wrote about it." And that of course gets us back to the original key statement of principle: "the threshold for inclusion is verifiability, not truth." Except I would change it to "attributability, not truth", but that's another story... Crum375 (talk) 13:22, 12 April 2010 (UTC)

One thing I know is that attempts in the past to change V not T (or water it down in any way) have met with strong opposition from otherwise uninvolved editors. It's something that sounds counter-intuitive when you first read it, but it sums up what we do better than any other way of putting it.

We can't reduce all our concepts to the level where every single thing is spelled out—not least because the more things seem to be spelled out, the more confusion we actually introduce. SlimVirgin ^{talk contribs} 15:46, 12 April 2010 (UTC)

Absolutely. The best thing about V not T is that it is so surprising. The reader hits that and says... "Huh?"... which induces them to read the policy in more detail, or ask about it on this talk page (and we have several regulars who are very good at answering such questions). Yes, there are a few who misinterpret it (either intentionally or unintentionally)... that is true with all our policies and guidelines... it is fairly easy to correct them. The more I think about this, the more I think we should just leave the policy as is. Blueboar (talk) 16:13, 12 April 2010 (UTC)

Leaving it as it has always been is fine for me. My concern is that when the explanation of what it means was moved away from the first paragraph the delicate balance was upset. I didn't notice this at the time, or I would have mentioned the problem earlier. Hans Adler 16:52, 12 April 2010 (UTC)

Which sentence (or omission) is it that you saw as problematic when it was changed? I saw you gave a link earlier to a January edit (which I now can't find), but I didn't see how it had changed things. SlimVirgin ^{talk contribs} 17:07, 12 April 2010 (UTC)

If there is confusion among some that this is the only threshold for inclusion, maybe change the wording to "A threshold..." rather than "The threshold." Otherwise leave it as it is - it's fine. --hippo43 (talk) 17:51, 12 April 2010 (UTC)

Actually, I think a lot of this has more to do with trying to define the threshold for exclusion rather than with defining the threshold for inclusion. The question being asked isn't really about whether Truth should be seen as a valid reason to include material (I think we agree that it isn't) ... the question being asked is whether non-Truth should be considered a valid reason to exclude.

What makes the question so problematic is that the V not T passage is focused on inclusion, not exclusion. We have only one threshold for inclusion (verifiability)... what isn't said is that we have myriad thresholds for exclusion (OR, undue weight, relevance, triviality, etc.). Blueboar (talk) 18:23, 12 April 2010 (UTC)

I'm not sure that's quite right. These concepts are all thresholds for both inclusion and exclusion - it's just that material has to pass all these tests to be included, and needs to fail just one to be excluded. Being 'not verifiable' is grounds for exclusion in the same way as being 'trivial' (i.e. 'not notable') is, for example. I guess it's like a Venn diagram - only material that is verifiable and notable and published research and appropriately weighted and etc can be included.

I guess the problem is that this sentence tries to make two statements at the same time - it's really two separate, but related, points. The opening line of the policy should be a clear, simple statement about what the policy is. The current opening sentence should come later, as it gives a more nuanced explanation of what the policy is and is not.

I suggest something like

All material in Wikipedia articles must be verifiable. This criterion for inclusion is verifiability, not simply truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true.

--hippo43 (talk) 19:21, 12 April 2010 (UTC)

I like this version wordwise with the following small exceptions:

“

All material in Wikipedia articles must be verifiable. A main threshold for inclusion is verifiability, not simply truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true or not true.

”

 —  Paine (Ellsworth's Climax)  19:43, 12 April 2010 (UTC)

I don't think that's an improvement at all. Following the first sentence with "A main threshold" implies that the second sentence is not conected with the first. "This main threshold" would be clearer. However, I don't think we should be adding the word "main" here - maybe "this important criterion/threshold/whatever", if we need to add anything at all.

Second, the addition of the words "or not true" at the end doesn't make sense, adds nothing and makes the sentence less clear. If you want to add "or not true", you would need to do the same for the first half of the sentence - therefore "whether readers can or cannot check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true or not true." IMO this is pointless and ugly. The word "whether" takes care of "or not true", as it implies "whether or not". --hippo43 (talk) 20:07, 12 April 2010 (UTC)

There is a problem that can occur with the wording using "readers can verify" being taken too literally as "ALL readers CAN verify" which as Blueboar has said many times at RS/N here and other places that is not true, the idea of verifiABLE is that A reader MAY verify not that ALL readers CAN; we place no priority on a reference being on the internet, free, or easily accessible. Whether it is a document that is only on display at a naval museum in Texas (a real dispute occurred at the RS/N about that) or pay-site on the internet like JSTOR (which any college student can access for free at their college anyways), or a newspaper clipping only accessible on microfichem, or YES even a foreign language source, they are all verifiable because they do exist and SOMEONE can verify the claim; so sayeth the RS/N.Camelbinky (talk) 21:52, 12 April 2010 (UTC)

'So sayeth the RS/N'? I guess you mean 'so sayeth some contributors to the RS/N'; the reliable sources noticeboard does not decide this policy - the discussion here does. In any case, I'm not sure what you're arguing for - I think this is a tangent. If someone misreads the policy, more fool them - the wording "whether readers can check that material..." is fine. However, of course this means "all readers can check" - if they go find the source. Nowhere does it say the source has to be free of charge, online or whatever. --hippo43 (talk) 22:08, 12 April 2010 (UTC)

I have a new proposal, independent of the discussion going on. I suspect the root of the problem lies in the effords to explain this concept in a compact size, of just one or two sentences long. Why not link the essay Wikipedia:Verifiability, not truth in the statement, make all the explanations needed, and then turn it into a supplemental essay? There, we can provide all the explanations to all the concerns manifested here.

I should mention that the current wording of the essay is mostly my work, but feel free to rewrite it as needed. My proposal is the existence of a suplemental essay on this topic, it does not need to be this specific one. I provided the link just because it's the most natural name for such an essay to have (we may also move the current one to a new name and start a new one from scratch) MBelgrano (talk) 00:52, 13 April 2010 (UTC)

I'm against an essay. The last thing we need here is an essay which attempts to explain policy at even greater length, but without the weight of policy, or the scrutiny this talk page affords. If there is anything crucial that simply must be changed with this policy, let's change it. If it isn't crucial, let's not waste an essay on it. It might not be possible to say everything important in 2 or 3 sentences, but it should be possible in, say, 10 sentences max - we don't need an essay to solve this. --hippo43 (talk) 01:04, 13 April 2010 (UTC)

Essays exist for just the reason MBelgrano has stated, to be supplemental and expand on policy wording what those who work on the policy may mean but dont feel they need to spell out. The policy (and all policies) are meant to be broad in what they are dealing with and not take into consideration every possible problem or issue that may arise and solve it here in policy; thats why we have guidelines and essays and noticeboards, even this talk page is not a forum to deal with specific issues that arise from various interpretations of the policy, only a forum on how this policy should be worded. This proposed essay may in fact end up as a guideline when we are done with it. We shouldnt write into this policy "this how you should interpret verifiable, not truth; this is what we mean by it" etc; that's the role of an essay because there is no one Truth on interpreting policies nor on what a policy means, as evidenced by the conflicts that arise here and elsewhere; an essay to guide and show what a good number of "experts" understand it is a much better use of time. Of course the best use of time is to actually do some editing so off I go.Camelbinky (talk) 01:42, 13 April 2010 (UTC)

Essays on Wikipedia exist for whatever reasons the people who write them want them to exist. Quite frequently they exist to put forward a view that is somewhat different from the consensus established in a discussion page like this, or for people who like the sound of their own writing to let off steam. In most cases they will not receive anything like the scrutiny that a project discussion page gives, simply because fewer people can be bothered to carefully pay attention to what essays actually say.

This is a core content policy - it should be clear and concise enough that it doesn't need an essay to explain (some editors' views) of what it means. If it's not clear what "V not T" means in the policy, then we haven't done a good enough job of writing the policy. It's impossible to separate the wording of the policy and the meaning of the policy, therefore this discussion is absolutely about what the policy means and how it should be interpreted. I'm all in favour of free speech - anyone can write an essay if they want, but (please, God) don't let's try to use an essay as a solution to this problem. --hippo43 (talk) 02:05, 13 April 2010 (UTC)

I don't think the V not T statement is unclear. The vast majority of our editors get it. So far, all the suggestions put forward make it worse, not better. Blueboar (talk) 14:23, 13 April 2010 (UTC)

I agree. I was just trying to cover those few editors who don't get it right away and who take up arms in the great cause of TRUTH. The V not T statement has been argued about for a long time, as seen in the archives. The real bottom line seems to be to let the statement stand, while dealing with the few exceptions when they arise. Changing it would cause far more problems than leaving it as is. Also, I see nothing wrong with a brief essay to clarify the concept for those few editors who don't get it. But it should be very brief and well-focused. It can be linked to in the See also section, not in the lede.

 —  Paine (Ellsworth's Climax)  16:59, 13 April 2010 (UTC)

Completely agree with Blueboar and Paine Ellsworth. We dont have to make WP:V or any policy dummy-proof. I personally though dont see a problem with the many discussions by those that "take up arms" in favor of "TRUTH"; it leads to healthy discussions and allows editors to know they have a voice and its being heard, and sometimes new concepts, beliefs, and interpretations of policies can arise out of it or from a tangent which allows our policies to grow, evolve, and meet new challenges as they occur.Camelbinky (talk) 23:16, 13 April 2010 (UTC)

I must say, Camelbinky, that your words give me a new perspective on an argument that I previously lost on this Talk page (see here & here in Archive 37). There are many aspects of Wikipedia for which your words apply. The trouble with "taking up arms" in any "good" cause is that negativity easily sets in and can lead to bad feelings (or worse) all around. Of course, this is never the intended response or outcome, but things can get out of hand sometimes when one is blinded by one's predetermined ideas of Truth and of what is Right and Good.

 —  Paine (Ellsworth's Climax)  16:24, 17 April 2010 (UTC)

Currently, a source is defined as follows:

The word "source", as used in Wikipedia, has three meanings: the piece of work itself (a document, article, paper, or book), the creator of the work (for example, the writer), and the publisher of the work (for example, The New York Times).

To my mind, however, only the first (a document, article, paper, book, or other material available to the public) can properly be considered a source in the wikipedia sense. The creator of the work may be the originator, but we do not have access to his thoughts except as they are placed in documents, articles, or etc., so the creator is not a direct source. The publisher may have a reputation or a practice such as peer review which helps to affirm that the sources it publishes (documents, articles, or etc.) are reliable, but the publisher itself is not a source since it only presents the material, but doesn't produce it. This distinction is important: allowing creators as sources leads towards wp:OR, since editors will start trying to intuit what the creator meant rather than focusing on the available documents; allowing publishers as sources can sidestep wp:V by giving minor documents more credibility than they might otherwise have simply by virtue of having made it into a publication that 'normally' publishes reliable material. I'm suggesting a rewrite of the 'Reliable Sources' section along these lines:

The word "source", as used in Wikipedia, refers to any piece of published material - an article, paper, book, or any other document available to the general public. A source is reliable when there is good reason to assume that it accurately reflects a notable position or point of view on the topic in question.

There are several factors that can be used in assessing the reliability of published material. Material that is published by independent (third-party) publishers with a reputation for fact-checking and accuracy is preferred; such material is generally considered to be reliable, and avoids plagiarism, copyright violations, and unverifiable claims in articles. Academic and peer-reviewed publications are usually the most reliable sources where available - such as in topics related to history, medicine, and science - but material from non-academic sources may also be considered reliable in these areas, particularly if it appears in respected mainstream publications. University-level textbooks, books published by respected publishing houses, magazines, journals, and mainstream newspapers may also be considered reliable, with the consideration that these materials are not subject to the same level of review as academic works. Electronic media may also be used, subject to the same criteria. Other material may be considered reliable in specific circumstances, where encyclopedic demands require.

The appropriateness of any source depends on the context. Sources are reliable to the extent that they directly address and are appropriate to the material or claims presented in an article. Sources which make tangential claims about a topic, which mention a topic in passing, or which mention a topic as an example while discussing something else should not generally be considered reliable for that topic. Sources which themselves argue for a particular point of view - in particular, self-published material (see below) and material that comes from fringe sources - may be considered reliable sources for the purposes of outlining or describing a particular viewpoint, but should not be considered reliable for the sake of presenting mainstream or established ideas.

I wrote that quickly, but you get the basic thrust of it. comments? --Ludwigs2 00:01, 15 April 2010 (UTC)

1. We lack the power to make such a definition stick.
2. Anytime we adopt a non-standard definition of a word, not only must we consider whether our definition will be accepted by readers, we must also clearly specify the scope of the definition. "As used in Wikipedia" can never be an acceptable scope for a non-standard definition. Jc3s5h (talk) 00:11, 15 April 2010 (UTC)

(ec) Ludwig, you're right that only published or broadcast work can be a source for us. But it's still true that editors use the word source to refer to the writer, the work itself, or the publisher, and the reliability of each of these things is assessed.

We can ask, for example: (1) Should we use this well-known expert even if he's writing on his blog and what he wrote seems odd? (2) Is the article just what we were looking for, even if we've never heard of the writer and the publication is a very minor one? And (3) Is this New York Times article something we feel obliged to use, because the newspaper is so reputable, even though the writer is unknown and none of us agree with his conclusions? Because the weight of each issue (writer, article, publisher) is constantly weighed up to determine reliability and appropriateness, I'd like the policy to retain those three senses of the word "source". SlimVirgin ^{talk contribs} 00:13, 15 April 2010 (UTC)

@ Jc3s5h: well, with sufficient consensus, a new definition would certainly stick. whether that kind of consensus is possible is a different question, of course. no way to know except to ask.

@ SlimVirgin: I understand the weighting problem here (I think), and it's really why I opened this thread. I mean, just looking at your first example (a well-known expert writing something odd in his blog): if we restrain reliability so that it only refers to the document, then the question becomes an easy one - blogs are not peer reviewed, and are not part of the normal scholarly process of presenting or analyzing ideas, so we'd have to say that a comment like that wasn't particularly reliable for any scientific position. it might be reliable for an article about that expert (so-and-so believes X as written in his blog), but that's as far as it could go. On the other hand, if we use the current three-pronged approach to reliability, we suddenly have a problem - the blog is still not reliable for scientific concerns, but the author is reliable for scientific concerns, so how do we decide whether or not his comment in the blog is a reliable source for a scientific position? and what happens later if he changes his mind and writes something contradictory - technically we'd have to say that both statements are reliable for scientific positions (because we've assigned reliability to the author), despite the fact that the two statements are mutually exclusive.

I know that plenty of editors use 'source' to mean the document, the author, and the publisher in a kind of vague interchangeable way, and usually it doesn't make a lot of difference - the line between the statements "this document shows X said Y" and the simpler "X said Y" is thin - but there comes a point where you have to preclude that kind of sloppy equivocation and make it plain that it's the reliability of the document (with respect to its subject) that's in question. Otherwise we end up (as we've all seen) with a peck of wikilawyering that pits one criteria for reliability against another in a million fruitless ways. --Ludwigs2 05:09, 15 April 2010 (UTC)

The policy is written to make sure that editors can't exclude an expert voice just because they don't like where it's published e.g. on his blog. It would have to be a genuine expert, mind you, and we'd have to use common sense: I can think of one case (who shall remain nameless) of an expert who seems to have gone mad, and there's an unwritten consensus that his material will not be used, almost no matter where it's published. My own feeling is that sloppy equivocation is important to retain, because it allows editors a lot of leeway to use their judgment depending on context. SlimVirgin ^{talk contribs} 05:19, 15 April 2010 (UTC)

I agree with the general idea of confining "source" to the first of these: "The word 'source', as used in Wikipedia, has three meanings: the piece of work itself (a document, article, paper, or book), the creator of the work (for example, the writer), and the publisher of the work (for example, The New York Times)."

The creator and publisher would still be relevant. But for practical purposes, they can't be considered alone. The work itself is, and should be, the focus.

In a relatively recent AFD, an editor tried to use an unpublished lecture as a source. Regardless of who gave the lecture, the fact that it was unpublished made it moot for our purposes. Maurreen (talk) 05:32, 15 April 2010 (UTC)

Agree with Jc3s5h - as far as possible, we should avoid re-defining words and creating wikipedia jargon. People use the word "source" with all three meanings, and with good reason.

Also agree with SlimVirgin - it's not just the document we have to consider. Depending on the context, it can be all three. --hippo43 (talk) 09:22, 15 April 2010 (UTC)

Yes, I agree with SlimVirgin. Weight, though, should probably go with the old Hayakawa assessment. He wrote that people generally give weight in fairly equal portion to two things...

1. What is said (or written), and
2. Who said it.

The publisher should be given some weight, also, however not too much weight by comparison to the other two.

 —  Paine (Ellsworth's Climax)  16:03, 17 April 2010 (UTC)

The policy contains the sentence which begins "Articles should be based on reliable, third-party (independent), published sources with a reputation for fact-checking and accuracy; this avoids plagiarism, copyright violations..." This is only partly true. If a source contains plagiarism, and we include a fair-use quote or paraphrase in a Wikipedia article, we are spreading the plagiarism, so the sentence is right in that using high quality sources will minimize instances of spreading plagiarism this way, because high-quality sources will seldom contain plagiarism. However, using good sources, by itself, won't prevent editors from creating new plagiarism. Also, copyright violations are avoided not by using high quality sources, but rather by limiting the amount of material copied to what fair use allows, and by paraphrasing. Jc3s5h (talk) 00:20, 15 April 2010 (UTC)

I agree with those points. I think perhaps something got lost in the editing. Citing reliable sources could help to reduce plagiarism and copyright violations since the source can be more easily checked. But I don't see how just using those sources without citing them can help with those problems.   Will Beback  talk  01:13, 15 April 2010 (UTC)

If a source includes plagiarism then its not a RS from what I understand of the policy and appropriate guideline and therefore shouldnt have been used anyways right?Camelbinky (talk) 03:26, 15 April 2010 (UTC)

If we know about the plagiarism, most likely we shouldn't use the source. But what appears to be a reliable source might have undetected plagiarism. Jc3s5h (talk) 03:32, 15 April 2010 (UTC)

Agree with Will Beback - I'd change "this avoids plagiarism..." to "citing these sources avoids plagiarism...". Also agree with Camelbinky (!) and Jc3s5h - if editors know (and agree) that a source contains plagiarism, then it isn't good practice to use it. --hippo43 (talk) 09:09, 15 April 2010 (UTC)

I've been working on and off on a whole series of articles about film editing techniques, adding numerous citations.

The nature of the film industry tends to be one of word-of-mouth know how and picking up ideas by working with people. Many people with expert knowledge in the industry are also quite open, with information available on the Web which gives colour to the topic. While I currently avoid sources such as blogs (except as an occasional placeholder), what is the policy on providing (possibly multiple) blog entries to provide colour to an article - perhaps in addition to a traditional source?

For example, I am so far finding it hard to find referenced examples of establishing shots, as most reviews don't discuss the technical editing details of films, but there are multiple non-controversial examples, with thousands of establishing shots being broadcast every day. Recently, I have come across tvtropes.org which, as far as I can tell, seems to be as accurate as any newspaper article, and seems to be written by people in the industry. There are many sources of comparable nature, such as Masters Theses in film, which are secondary self-published sources when talking about films, and generally mutually self-consistent. The best blogs are written by professors in film studies and also seem to be reliable.

As film editing in not a black and white process (!), more opinions about how it should be done would seem to give the reader more context to assess the article contents.

An alternative would be to buy some books about films, and give only non-web accessible page numbers as references.

Most of the Wikipedia articles in this area seem to have been uncontroversial and written by experts or at least people who "knew what they were talking about" in the days when citations were a nice-to-have, so there is quite a lot of work to do tidying up the articles - not necessarily the actual content, but on referencing them in a way which allows the references to be checked against the text after the text has been re-edited over time.

Is it better to have an examples section with independent multiple self-published sources, or no section? Stephen B Streater (talk) 22:23, 18 April 2010 (UTC)

You have to use books. Since you have access to Masters' theses, these may be helpful in finding good sources, and doctoral theses are considered to be reliable. If you can establish the notability of blogsites then you can provide links to them. They may also be helpful in finding published articles. TFD 22:58, 18 April 2010 (UTC)

Blog entries are not reliable sources unless it is verifiable that they actually ARE from an expert, and even then they are only usable for voicing their opinions and in their own articles. They can not be used just to add "colour". Further, TV Tropes is not a reliable source at all (and it is NOT written by anyone in the industry, its user edited). There are more than enough reliable sources, especially print sources, without degrading any film industry article with low end non-reliable sources. There are also libraries if you don't want to buy. References do NOT have to be web accessible, FYI. And if no reliable sources can support the section, then it is better to have no section. -- AnmaFinotera (talk · contribs) 00:01, 19 April 2010 (UTC)

"TV Tropes... is NOT written by anyone in the industry, its user edited"

You seem to imply that those are mutually exclusive! We have some professional experts in various fields writing for Wikipedia (though obviously many of our editors are not), so there's no reason to assume that other user-edited sites are without expert contributors - or that they're any more reliable than Wikipedia!

How about something MediaCollege.com for basic references and possibly other references for "colour" so long as it's clear they back up the "reliably-sourced" information? Contains Mild Peril (talk) 00:38, 19 April 2010 (UTC)And yes, in the case of TV Tropes, it pretty much is mutually exclusive, since I doubt many industry folks would edit there. -- AnmaFinotera (talk · contribs) 01:15, 19 April 2010 (UTC)

Whether an editor is an "expert" is irrelevant. Wikipedia is not for the publication of personal research/opinion. Reliable sources are required.

I don't know anything about film editing but, trying to be helpful, suggest that Google books is a good tool for locating citeable sources. See, e.g., this for finding sources such as Jeremy G. Butler. Television: critical methods and applications. Routledge; 2007. ISBN 9780805854152. p. 205. Wtmitchell (talk) (earlier Boracay Bill) 01:09, 19 April 2010 (UTC)

There are many books on filmmaking. The establishing shot goes back to the earliest days of movies, so it shouldn't be hard to find references for it. However it may require going to the library.   Will Beback  talk  19:54, 19 April 2010 (UTC)

Correct me if I'm wrong on this folks, but aren't blogs permissible in certain circumstances -- such as when the writer is notable, or when the blog really refers to the format, but it has behind it an RS? Also, aren't opinions (of notable people or RS's) things we are allowed to reflect (e.g., a NY Times editorial or op ed article) not as evidence of fact, but as evidence of a notable opinion? Best.--Epeefleche (talk) 17:19, 20 April 2010 (UTC)

Yes, some blogs are allowed. The question here is whether three is a better, more reliable source for the same information. Blueboar (talk) 22:15, 20 April 2010 (UTC)

Here is a link to the guideline on blogs. TFD (talk) 22:43, 20 April 2010 (UTC)

• Wikipedia talk:Words to watch#RFC. There's a proposal to merge several pages as part of a project to streamline the MoS. One part of the proposal is to merge Words to avoid, Avoid peacock terms, Avoid weasel words, and Avoid neologisms into a new page, Words to watch (W2W). Fresh input would be appreciated at the RfC. SlimVirgin ^{talk contribs} 00:41, 26 April 2010 (UTC)

I am disturbed by an attitude by one contributor at Wikipedia:Bots/Requests for approval/ValhallaBot‎ who seems to have low the standards for bots with the potential to make false changes. Looking through policies, I don't see a statement that lies are forbidden, nor do I see a statement that implementing automated processes with reckless disregard for verifiability is forbidden. I believe a new section should be added on this topic. Jc3s5h (talk) 17:14, 28 April 2010 (UTC)

Thought this quote from Jimbo Wales was interesting. [9]

I think that "If no reliable third-party sources can be found on a topic, Wikipedia should not have an article on it." is a good policy. To take the example of a popular book that receives no reviews, what kind of encyclopedia article could you write about it? You could write an original review, but that isn't an encyclopedia article. You could write a plot summary, but that isn't an encyclopedia article. You could do some kind of original research, but that wouldn't be an encyclopedia article.--Jimbo Wales (talk) 09:04, 22 April 2010 (UTC)

This reflects current policy. Is there anything else we can do to make this more clear? Arskwad (talk) 16:14, 22 April 2010 (UTC)

I don't think we can do more on this page (we make it very clear)... perhaps at WP:NOTE and WP:NOT?

WP:PROF and WP:ATH currently directly contradict "If no reliable third-party sources can be found on a topic, Wikipedia should not have an article on it." Gigs (talk) 19:10, 28 April 2010 (UTC)

A perfect example: Wikipedia:Articles_for_deletion/Colin_P_Flynn. Gigs (talk) 13:23, 29 April 2010 (UTC)

I don't think they contradict... these are sub-sections of WP:NOTE (people)... the need for reliable secondary sources that are independent of the subject (ie third party sources) is stated in the basic criteria section near the top of the page. Thus it applies even though it is not restated again in every sub-section. The question being wrestled with at the AfD is whether there are third party references or not. Blueboar (talk) 22:26, 29 April 2010 (UTC)

• an academic/professor meets any one of the following conditions
• 1. The person's research has made significant impact in their scholarly discipline, broadly construed, as demonstrated by independent reliable source
• The most typical way of satisfying Criterion 1 is to show that the academic has been an author of highly cited academic work

So a highly cited professor always gets kept at AfD, even if there's no biographical third party reliable source coverage. The effect of this is that we basically allow unverifiable autobiographies to be published in Wikipedia if the professor is highly cited, since these articles are invariably based on university bio pages written by the professor themselves or someone close to them. Gigs (talk) 18:52, 3 May 2010 (UTC)

The video is full of inaccuracies and has no business on a policy page. I could maybe see it as a "See Also" link. It amounts to a bad essay being transcluded into what is supposed to be a policy. Gigs (talk) 19:38, 28 April 2010 (UTC)

I agree with that 100%. I took if off NPOV a couple of days ago for the same reasons. It can't go in See also which is for wiki articles, but I wouldn't even include it as EL, because it's misleading and can cause confusion. Crum375 (talk) 20:04, 28 April 2010 (UTC)

I have transcribed it for easier review. Gigs (talk) 20:08, 28 April 2010 (UTC)

You've probably seen Wikipedia, it's a free online encyclopedia built by people like you and me around the world. You might wonder how thousands of people build an encyclopedia together. Well first Wikipedia is a wiki website, a website anyone can edit, so on Wikipedia, editing or creating a new article happens with the click of a button.

But for these contributions to remain in wikipedia, they have to follow two basic rules. The first is verifiability. With so many contributors, wikipedia articles must rely on information from published sources like books or newspapers, resources known for fact checking.

Requiring contributors to cite these resources in articles and quotations ensures that wikipedia articles are factual and high quality. If it's not verified, it can't be in wikipedia. For example, you can write that the US unemployment rate in 1935 was 20.1% but you must also cite its source for it to remain in Wikipedia. In this case numerous history books could be verifiable resources.

The second rule requires a neutral point of view. All wikipedia material must be presented fairly and without bias, just like any other encyclopedia. This means wikipedia is not a place for contributors to share their own opinions. Lets say you are an advocate for vaccinations, and you write, "Every parent should get their children vaccinated". Unfortunately, this is biased and certain to cause disagreement. It can't be in Wikipedia.

However, published opinions of experts can be included. And if these opinions differ, the article should present all the major opinions without endorsing one over the other. For example, writing that "Vaccinating all US children saves an estimated 33,000 lives" and citing a reputable source is a statement of fact that can be verified.

And if there is an opposing view, it should also be included. For example, a quote from a reputable source like "Critics claim that vaccinations have never benefited public health" helps to balance the article and keep it neutral.

By following these two rules, contributors can help respect one another and help create a free encyclopedia, the largest encyclopedia in human history. Learn more at wikipedia.org.

Okay, so the inaccuracy I see is that it says "verified" where it should say "verifiable". "If there is an opposing view, it should also be included" it maybe over-general, doesn't address "undue weight", but I don't know if that's more of an inaccuracy or a forgivable loss of fidelity in synopsis. What else are you identifying? —chaos5023 (talk) 20:14, 28 April 2010 (UTC)

(after edit collision) At first glance I don't have any problems with this video. Could you please point out the specific passages that you believe do not properly reflect policies and guidelines, so we have something to discuss. --SteveMcCluskey (talk) 20:17, 28 April 2010 (UTC)

The main NPOV issue I had with it is that it seems to equate minority and majority views, ignoring the fact that they need to be weighted by prevalence. Crum375 (talk) 20:20, 28 April 2010 (UTC)

The main V issue is that it ignores the important distinction between attributable and attributed. It seems to say that unless material is attributed, it must go, which is plain wrong. Crum375 (talk) 20:22, 28 April 2010 (UTC)

Someone is now reverting to keep it in NPOV. Crum375 (talk) 20:18, 28 April 2010 (UTC)

• "But for these contributions to remain in wikipedia, they have to follow two basic rules."

WP:V and WP:NPOV aren't supposed to be viewed in isolation of WP:OR.

• "If it's not verified, it can't be in wikipedia."

Actually, only information that is challenged or likely to be challenged needs citations. A Quest For Knowledge (talk) 20:23, 28 April 2010 (UTC)

• (e/c)We have three core policies, not two.
• Fact checking is only one factor in reliability, reliability is contextual to the claim its backing up and not automatically conferred on a book or newspaper
• "History books are verifiable resources" doesn't make any sense. Content is verifiable, not sources. Sources are reliable or unreliable.
• I'm not sure what a "published opinion of an expert on vaccine opposition" looks like, but I don't want to be the mediator in that case.
• "Critics claim" is a good example of using weasel words to appear superficially neutral
• But the fundamental problem is as you said, there's a "loss of fidelity in synopsis" that is entirely inappropriate for inclusion within the policy document itself. Writing nutshells is hard enough; they were very controversial when they were first put on, and are frequently subject to debate and refinement. This video produced by someone obviously not very familiar with Wikipedia policy is not an appropriate summary to include within a policy document. Gigs (talk) 20:25, 28 April 2010 (UTC)

I agree with the bottom line. The video shows a lack of understanding of the core content policies, but even if it were better, it would still be an essay embedded inside a core policy, which is a bad idea in general. Crum375 (talk) 20:28, 28 April 2010 (UTC)

Yeah, have to say that seems like a bad idea. It should be referred to in the way other essays are, as a "see also", and it should be clear that "essay" is its status. —chaos5023 (talk) 20:35, 28 April 2010 (UTC)

It is apparent to me that the video is intended as a short, broad summary of the policies on verifiability and NPOV. For a tutorial video, there is no great sin in committing "loss of fidelity in synopsis" (sometimes known as "oversimplifying" in English). It's fair to say that the core policy documents are not an appropriate place for an introductory video. But on the other hand, newbies get pointed at WP:V and WP:NPOV so often, these pages are arguably the best place to put them to guarantee that they will be seen by their target audience. I am happy to hear counterproposals. Tim Pierce (talk) 20:40, 28 April 2010 (UTC)

How about giving it its own essay page with a nice shortcut like WP:VVID and some text explaining that it's a simplified summary, so people can send newbies to it if they think it'd be helpful? —chaos5023 (talk) 20:45, 28 April 2010 (UTC)

How about starting with these three key points:

• "On WP everything must be attributable, though not necessarily attributed."
• "On WP all points of view must be presented neutrally, weighted by their prevalence among reliable sources, though tiny minorities should not be mentioned at all."
• "We may not include original research on WP. This means that anything we write must be supported by a published reliable source."

Starting with correct though minimal information, would be much better. And then, it could be included inside an essay. Crum375 (talk) 20:50, 28 April 2010 (UTC)

I don't see any consensus for this staying, so I'm pulling it. The discussion cycle is typically BRD, which would mean a bold insertion of the video, a revert, and then discussion here to see whether it should be re-inserted. I don't see a consensus for keeping it (or even any reasons to do so, until it's been fixed), so I'm removing it. -- Bfigura ^(talk) 20:59, 28 April 2010 (UTC)

I support removing the video, regardless of its accuracy. I appreciate the idea of providing a gentle overview, but having such media in a vital policy is a minefield: Do I have to watch the video to understand the policy? If I watch the video, can I ignore the text? What if the video appears to conflict with the text? What if wikilawyers argue using points made in the video? Further, it is too hard to maintain a video. Johnuniq (talk) 04:20, 29 April 2010 (UTC)

Please see this section of Wikipedia talk:NPOV. The NPOV policy currently contains two sections on specific topics: a 534-word section on pseudoscience and a 267-word section on religion. These sections were removed last month as being too specific after an RfC was posted on April 3. [10] The pseudoscience section was moved to WP:FRINGE, [11] and the religion section removed entirely. The sections have now been restored by others on the grounds that consensus was not established, or has changed. Fresh eyes would therefore be appreciated here on talk to decide whether to restore or remove the sections. SlimVirgin ^{talk contribs} 00:03, 2 May 2010 (UTC)

The place for discussing the merits of the RfC is the RfC itself, not the space behind a neutral pointer to the RfC. Hans Adler 08:32, 2 May 2010 (UTC)
The following discussion has been closed. Please do not modify it.
The purpose of this RfC needs to be clarified: it aims to demote WP:PSCI from policy to an ignorable guideline. Currently, WP:NPOV requires a neutral point of view, but makes an important exception allowing pseudoscience to be labeled as such. PSCI also ensures that articles can assert that science and pseudoscience are not simply two equal viewpoints. Johnuniq (talk) 04:34, 2 May 2010 (UTC)

From WP:V - "Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living persons or organizations, and do not move it to the talk page." There used to be a link to WP:BLP included in the statement, however WP:BLP#Legal persons and groups specifically excludes organizations from this protection. To bring the policies into agreement, I propose that the words or organizations be removed from this statement. There was a discussion at Wikipedia talk:Biographies of living persons#Groups about adding organizations to BLP protection, and it was rejected as WP:CREEP.  --Joshua Scott (LiberalFascist) 18:03, 8 May 2010 (UTC)

Think about the statement beyond BLP for a second... In terms of Verifiability, why should organizations be removed from the statement? I understand that they are not Living Persons... but I surely we do not intend editors to be free to add unsourced or poorly sourced material that might damage anyone... including organizations. The bar for what qualifies as a reliable source might not be set quite as high for organizations as it is for Living People, but I don't think we should remove them from the statement. Blueboar (talk) 21:47, 8 May 2010 (UTC)

BLP is clear that it is only speaking about living persons, and it is also clear that it can't be interpreted as making claims about other entities exempt from the normal policies. So it can't be used as an argument for crippling another policy. Hans Adler 22:55, 8 May 2010 (UTC)

Remember, "Any material lacking a reliable source may be removed". This sentence is in fact equating the protection of living persons and organizations. This policy mentions living persons 7 times, and only this once is "organizations" mentioned, so the policy is inconsistent with this wording included.  --Joshua Scott (LiberalFascist) 23:25, 8 May 2010 (UTC)

There is no contradiction at all. You are talking about WP:BURDEN. It just says that we should never leave potential libel (i.e. unsourced negative claims about a living person or an organisation) in an article. The fact that this is regulated further in a special case by WP:BLP, a policy whose sub-community is unwilling to make it cover organisations, doesn't imply that this obvious rule doesn't hold for the other case that is not regulated in such a way. With enough wikilawyering to the effect that libelling organisations is perfectly fine and should be allowed, we may eventually get a new "Claims about organizations policy". But the non-existence of such a redundant policy at this time doesn't make WP:BURDEN invalid and is not a reason to restrict its scope. Hans Adler 23:52, 8 May 2010 (UTC)

• 'Delete This is feature creep and a backdoor extension of WP:BLP. If we continue in this vein we'll soon have special rules to protect commercial products, political parties, and TV shows.   Will Beback  talk  04:42, 9 May 2010 (UTC)
• I see no reason at all to remove the statement, and every reason to keep it. It has nothing to do with BLP, and BLP does not control what WP:V says. If BLP doesn't want to cover organizations, that's fine, that doesn't mean V shouldn't. If BLP were delete tomorrow, that still would not in any way mean the statement should be changed here as well. -- AnmaFinotera (talk · contribs) 04:48, 9 May 2010 (UTC)

WP:V has quite a few happy consequences, as preserving high quality information is good for the world in many ways. One of them is avoiding harm to people, things, beliefs, religions, sensibilities, scientific theories, architectural wonders, animals plants and vegetables that might all suffer from ignorance and misunderstanding. However, the specific caution to avoid possible harm to living people is an exceptional matter relating to WP:BLP, and it tips the usual balance in favor of a higher standard of verifiability than for other things. This particular section is a reference to BLP, and as such we should accurately report what BLP says. At BLP, proposals have been advanced and rejected a number of times to apply the standard to groups of people, organizations, anonymous people, and so on. So adding organizations to the list goes beyond the purview of BLP. Beyond BLP, I think WP:V works just fine and we don't need to attach additional scrutiny to things that might hurt organizations, versus things that might hurt other things. After all that, in the spirit of "don't fix it if it isn't broke" I think WP:V is working just fine with "or organizations" in it, and it's been in there long enough that it's the status quo, so I wouldn't be in a hurry to remove it either. Just my $0.02. - Wikidemon (talk) 04:58, 9 May 2010 (UTC)

• Support leaving the wording in WP:V and WP:BLP unchanged. Of course BLP does not want to talk about organizations, and needs to make strong statements about BLPs which have the highest editorial requirements. However, WP:V is quite clear that all challenged material must have a good source, but it needs to qualify that and say to not delete "this 1000-year old vase was discovered at X" until a decent amount of time and discussion has occurred. But text like "SomeCorp cheat their customers[ref: dubious website]" needs to be removed immediately. It's just spelling out common sense. Johnuniq (talk) 05:09, 9 May 2010 (UTC)

Does anyone know if there was any discusion or consensus before this was added?  Will Beback  talk  05:27, 9 May 2010 (UTC)

Here's where it was added diff of add and state of talk page at that time. It is apparent to me that the original intent was to point to the BLP policy and nothing more. This was the BLP policy at the time. I find it very interesting that in footnote 8 on the BLP page, Jimbo is saying that BLP should apply to organizations. This is in contradiction to the current practice, which is what I want to resolve. Currently there is no agreement on whether or not BLP is applied to organizations, leading to edit wars removing sourced negative material about companies, and claiming BLP. If we say BLP applies to companies, I'm not going to be upset, I just want a consistent statement across the board from V and BLP.  --Joshua Scott (LiberalFascist) 06:27, 9 May 2010 (UTC)

Joshua may be correct in saying that the original intent might have been to tie the statement directly to BLP... but the question is, is this our intent now? The fact that the statement no longer includes a link to WP:BLP (even when talking about individuals) tells me that the current intent is to break this statement away from BLP. To no longer directly tie what is said in WP:BURDEN to WP:BLP.

I think a statement saying "do not add unsourced or poorly sourced statements" about anyone (individual or organization) stands on its own merits within WP:V, without needing to be a reference to WP:BLP. Blueboar (talk) 14:35, 9 May 2010 (UTC)

Exactly. When an editor who removes "Copsi Cola tastes approximately like horse piss with 10% added sugar {{citation needed}}" gets into an edit war, then that shouldn't be treated as a standard content dispute, perhaps even with protection on the wrong version. It should be an unambiguous case of applying a policy, so that the other party can be told they will be blocked if they persist. (This can be done even without such a policy, but then an admin must apply common sense, which is a rare commodity hereabouts.) Hans Adler 22:57, 9 May 2010 (UTC)

• I'm neutral on the issue at hand, but I should point out that this is exactly why policies should not redundantly summarize other policies like this. Gigs (talk) 18:57, 10 May 2010 (UTC)
  • Actually, I've changed my mind on this. We should not be giving special protection to corporations. The libel clause should only be about individual people. Gigs (talk) 21:23, 10 May 2010 (UTC)

This edit which gave emphasis to the idea of "attribution" seems to be a subtle attempt to give the failed policy proposal at WP:Attribution more traction, as is apparent from views expressed at the proposal to rename it so that we can reuse the title for something more useful than a failed proposal. Gigs (talk) 00:06, 29 April 2010 (UTC)

At WP:NOR an edit along the same lines was here. There was a brief discussion Changes have been made in the lead towards topic of WP:V. --Bob K31416 (talk) 23:02, 14 May 2010 (UTC)

There is discussion going on now at WP:NOR to revive WP:ATT... or to create something new but similar. Blueboar (talk) 00:34, 15 May 2010 (UTC)

Need a third party opinion about this. User:Alastairward has apparently begun to make statements that information from a military service record isn't a reliable citation for information in a Wikipedia article [12]. Forgive me, but a military service record is the absolute primary source for all military service information on Wikipedia military figure articles. I can't begin to think of how many articles have dates of rank, assignment histories, award citations, and countless other information obtained directly for official military files. As I pointed out to AW, we even have an entire article devoted to a service record: Service record of Reinhard Heydrich. So, need some third party advice in case AW begins removing cited material based on service record entires. -OberRanks (talk) 01:16, 11 May 2010 (UTC)

Looking at the conversation, it seems to me that he is saying that you can't just say "look at his service record, it exists", not that it isn't reliable. You must actually have seen it, read it, etc. and properly cite it. In the diff you posted, he seems to be explaining just that. The cite previously claimed just said "service record" with out being a valid cite. When you linked to the actual available copy, that was a proper source (as he notes). -- AnmaFinotera (talk · contribs) 01:23, 11 May 2010 (UTC)

I agree with AnmaFinotera but would add on the general topic of primary sources/service records- We PREFER third party published sources, BUT primary sources that can, by SOMEONE (not everyone) be verified from a legitimate source that is likely to have the correct information is acceptable. I think we can all agree that for information such as date of rank, assignment, awards, medals, commendations, courts-martial, etc the military is going to be reliable and correct. Primary source are invaluable to Wikipedia in making sure that secondary sources have gotten their facts correct, secondary sources are great on opinion and doing the synth that we cant, that's why we use them, but when it comes to FACTS like dates and hard numbers there is hardly any reason to put the secondary source above the primary or to throw out the primary all together especially if there is no easily accessible secondary source with the information (or only with false information).Camelbinky (talk) 01:25, 11 May 2010 (UTC)

IMHO military records are generally good primary sources, but that isn't universal. For instance, we wouldn't want to take Pat Tillman's Silver Star citation as a RS about the events leading to his death, and I don't know how far I'd trust a 1940s-era Soviet military record, given how much document-doctoring went on in that era. --GenericBob (talk) 22:59, 12 May 2010 (UTC)

I guess having a proper cite from a military service record would be a good idea. Is there a standard format? -OberRanks (talk) 01:26, 11 May 2010 (UTC)

{{cite web}} is fine.. Ideally include enough information that if the URL changes the source could still be easily located though. Gigs (talk) 01:36, 11 May 2010 (UTC)

The info was removed again from the Patton film article [13], under the same statement that the service record was not well cited. Interested users may wish to comment here. -OberRanks (talk) 22:21, 12 May 2010 (UTC)

OberRanks, you are misrepresenting me here. I never said that your information was removed solely because it wasn't cited. You provided a poor cite, one that didn't actually direct users to somewhere you can verify the information you added. This isn't an issue of reliability but actually letting users see where you got what you added to the article.

With regards what was removed, its a trivia issue and cannot be addressed here. Alastairward (talk) 22:57, 12 May 2010 (UTC)

Alastairward- OberRanks cite, while not the best, was valid in that it told you what he was citing and where to find it. I'm sorry if YOU can not go to the particular place and see the records for yourself. That does not make it a bad citation. A citation does not require that YOU be able to verify it, just that SOMEONE out there could if they were able to go to the particular place where you can see the records. As long as someone knows what center they need to go to and what records to ask for, then it is a valid citation. Can we close out this discussion now and stop the bickering?Camelbinky (talk) 05:34, 13 May 2010 (UTC)

The cite was pretty damned vague and besides which, the removal of the material it cited had nothing to do with how good a source was found, but whether or not it was trivia. That it was brought up here was quite the red herring with regards the discussion that was proceeding on the talk page of the article in question. Alastairward (talk) 18:48, 14 May 2010 (UTC)

Readers short of time are invited to read only the last few bold paragraphs at the end

Hi. If I'm raising this issue in the wrong place, or otherwise going about this in a manner that's sub-optimal, please enlighten and then forgive me. This will take a bit of typing to outline, so please be patient with me.

Verifiability policy indicates that "The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material added to Wikipedia has already been published by a reliable source, not whether editors think it is true."

However, we are also informed that "exceptional claims require exceptional sources" and so "claims that are contradicted by the prevailing view within the relevant community, or that would significantly alter mainstream assumptions [emphasis added], especially in science, medicine, history, politics, and biographies of living persons" require careful review of sources.

It seems obvious that an assumption is more or less something that readers (or editors) think is true. That's very different from something that is verifiable in the textual tradition of citation. Sometimes a lot of people share an assumption. On one hand the central guiding policy is verifiability, not truth. On the other, anything that challenges assumptions (as distinct from verifiable and reputable citations) is subject to special scrutiny. All this is more or less as it should be.

However there's an issue here in the gray area that any attempt to cover all circumstances invariably creates: cited and widely-held points of view that are not "mainstream assumptions" and that can be construed as facts (if anything can) are sometimes attacked as violating NPOV. Instead of citing sources to balance what is thought to be a non-neutral passage frustrated editors often cite policy (usually NPOV) to defend their assumptions.

I will provide an example here, and it will take more typing, so again, I beg the reader's patience. What follows might seem to be a bit of over-complicated analysis, but I see no other way to fully explicate the theme I'm trying to get my teeth into here. I should also explicitly say that in what follows I'm not arguing (here) for the issues in the example, rather that it is an example of a broader issue.

In the caduceus article there was a passage that stated that the caduceus symbol is often used erroneously in the US to symbolize medicine or medical practice. This has been repeatedly challenged as being in violation of NPOV. However, those that challenge the statement do not provide any citations showing that its use is not erroneous.

The stance of those who wish to see the affirmative statement stay in the article (and I'm one of them) is that all of the current specialized studies of the caduceus cited thus far indicate explicitly that it was adopted in the US by "mistake", in "error", or as a result of "confusion". While perhaps three studies from prior to 1930 defended the use of the caduceus as a medical symbol, claiming it was not selected in error or by mistake, those studies today have been characterized (by an academic and professional medical source) as based on "flimsy and pseudo-historical research", and refuted or ignored by all modern researchers. Masochists are invited to review the discussion if they want to see a list of sources. They are all reputable (JAMA, Royal Society of Medicine Press, Oxford University Press, The Scientific Monthly, The Classical Journal, etc).

The stance of those who wish to remove the statement (or neuter it, rather than make it neutral, by attributing it as a point of view of "some" people) is that the statement violates NPOV. None of them produce citations showing that it's use is not an error, or indeed that whether or not something is an error is a POV issue in the first place. Basically the gist of their case is that it is a known fact that the symbol is used to represent medicine, therefore it cannot be an error. Though they could easily cite less-specialized sources which simply say it is a symbol of medicine (and little more) they don't even do that. They just point to the fact of its use as an indication that it could not be an error.

There are a number of points that make all this tricky, of which I'll note a few:

1. The sources that do affirmatively state it is a medical symbol are general reference works that don't go into detail, they don't discuss the issue of the mistaken adoption or the unknowing emulation of the initial error. Many indicate that it is today used as a medical symbol. Nothing more. Arguments based on the silence of sources are obviously less strong than one's made based on explicit, affirmative statements in sources.

2. The sources indicating that its symbolic valuation is based on a initial mistake and the perpetuation of the error are also the specialized studies dealing with symbolism at length. It would seem that specialized studies dealing with a subject at length carry more weight than short reference entries that mention a subject in passing, provided all such sources are verifiable, reputable, etc.

3. There is a low level of semiological sophistication in most of the sources, which does not address the thrust of the arguments made by those who want to take out the content they don't like, which amounts to the firm conviction that a semiological mistake is no longer a mistake if enough people make it. That's all fine and good except that none of the sources dealing with the caduceus show that degree of semiological sophistication (which I here note is entirely distinct from knowledge of their given subject, in this case the symbolism of the caduceus itself), and none of the editors involved have adduced citations from disciplines armed with the tools or theory needed to address such an issue.

4. Regardless of its use today, there is no doubt that as regards its initial adoption the view that it was not a result of mistakes, confusion or error was not only a minority view, but is now a dated one that does not receive serious attention in contemporary sources. Hence there is academic consensus that the initial adoption was an error. No source thus far presented gives any indication that an error can one day simply expire. What do we do when the clear consensus is that it was adopted in error, and many specialized studies by academic or medical professionals (and sometimes both) indicate that its continued use is an error, while general reference works make little or no mention of any of this, simply stating today that it is a symbol of medicine?

All that by way of showing that there are many gray areas in this example (as is often the case), gray areas which should be clarified by recourse to citation of sources (not policy), rather than obliterated by uninformed regurgitation of assumptions that "everybody knows to be true". To my mind, the most important issue here is that people are citing policy about NPOV and trying to bring the content they don't like into line with their view based on assumptions that they do not feel require citation.

This brings me to the general issue, which I hope justifies the time spent by the reader here: it seems that very often, in content disputes (though I don't doubt there are many important exceptions) editors should start with citations of reputable sources before they turn to citation of policy. Daily, I see the issue of putative NPOV violations raised to justify the removal of something that somebody doesn't happen to like. All this is an intricate matter of how NPOV intertwines with Verifiability as two core principles, yet the separate treatment of NPOV and Verifiability entrenches, to some degree, their independent application in practice (no matter how many times we are reminded that in theory all of this is knitted together). Is there any policy or guideline that explicitly suggests starting with citation of reputable sources before thumping the policy bible to defend one's claims about NPOV? If not, were I to try to craft some such guideline, would I do it here (which is the first place that comes to mind) or associate it with the NPOV policy, or craft it for both locations?

To be more explicit: It seems to me that claiming an NPOV violation as justification for removal (or ham-handed butchery) of cited, verifiable content from reputable sources is most often not acceptable in and of itself. An alleged NPOV violation often requires citations to establish it as a fact. Hence, a presumed NPOV violation is often only a basis, or a context for the presentation of citations to:

1. establish that the content in question is actually in violation of NPOV, and;
2. provide a non-arbitrary verifiable basis for modification of content to bring it into line with NPOV standards.

Basically, it would be helpful to be able to direct someone to guidelines for when a corrective NPOV edit can be made without recourse to (non-policy) citation and when a corrective NPOV edit requires citation to establish that NPOV guidelines are even being violated, and to provide a context for modification based on Verifiability.

Help?--Picatrix (talk) 19:50, 12 May 2010 (UTC)

My view is that "would significantly alter mainstream assumptions" should be taken as an indication that some statement may be "contradicted by the prevailing view within the relevant community", but that it is the latter which would make a claim extraordinary. Maybe a better wording would be "statements that contradict prevailing views within the relevant community", so that no explicit contradiction (i.e. a source explicitly stating that the view would be false) is necessary to make a claim extraordinary. I'd drop the former characterization ("mainstream assumptions"), but add a sentence that the fact that something would alter commonly held assumptions should be seen as an indication that a claim may be extraordinary. Editors who challenge the inclusion should, however, point out how the statement contradicts prevailing views within the relevant community; pointing out that a statement would alter a (perceived or real) mainstream assumption would in itself not make a claim extraordinary. All claims of policy violations should be backed up by sources, as the personal opinions of us, as Wikipedia editors, are not relevant in this context. Often, those sources are already included in the article. If they are not, this should be taken as an indication that the article urgently needs better sources, that the mainstream view is not as clear as it may have been assumed to be, or that it may be plainly wrong. (While it is a commonly held view that gravity is a force, the verifiable prevailing views within the relevant community, i.e. physicists, actually differ from that assumption.)  Cs32en Talk to me  21:39, 12 May 2010 (UTC)

Thanks very much for your reply. I wonder if you have any thoughts on the specific question about qualifying when and how NPOV has to work together with Verifiability through more explicit policy or guideline statements as proposed in the (now) bold paragraphs at the end of my previous post. If this is not warranted (e.g. if the preference is to avoid outlining specific courses of action in favor of more general statements of first principles) can you direct me to policy that would support the assertion that "All claims of policy violations must be backed up by sources" beyond more general assertions such as "Verifiability is one of Wikipedia's core content policies, along with No original research and Neutral point of view. Jointly, these determine the type and quality of material that is acceptable in articles. They should not be interpreted in isolation from one another, and editors should therefore familiarize themselves with all three." Thanks. --Picatrix (talk) 13:21, 14 May 2010 (UTC)

Picatrix, would you mind very briefly summarizing your original question? It's a lot to read, but I'd like to address it if I can. SlimVirgin ^{talk contribs} 01:11, 15 May 2010 (UTC)

Thanks. Certainly:

It seems to me that claiming an NPOV violation as justification for removal of cited, verifiable content from reputable sources is most often not acceptable in and of itself. An alleged NPOV violation often requires citations to establish it (even arguably) as a fact. Hence, a presumed NPOV violation is often only a basis, or a context for the presentation of citations to:

1. establish that the content in question is actually in violation of NPOV, and;
2. provide a non-arbitrary verifiable basis for modification of content to bring it into line with NPOV standards.

It would be helpful to be able to refer to guidelines for when a corrective NPOV edit can be made without recourse to (non-policy) citation vs. when a corrective NPOV edit requires citation to establish that NPOV guidelines are even being violated, and to provide a context for modification based on Verifiability. Is there any policy or guideline that explicitly suggests starting with citation of reputable sources before thumping the policy bible to defend one's claims about NPOV? If not, were I to try to write some such guideline, would I do it here (which is the first place that comes to mind) or associate it with the NPOV policy, or mention it for both locations? All the other material above establishes an example context for this question. --Picatrix (talk) 08:53, 15 May 2010 (UTC)

Wikipedia's policy reliance on published works discriminates against politically incorrect persons and philosophies, such as Men's Rights -- as in the case of the recently deleted entry on Rich Zubaty. This is because Feminists are powerful enough to, and actually do, hinder the publication of anti-Feminist and Men's Rights works, with the result that there is little of this available to support a claim to Wikipedia article status. The media are similarly strongly influenced by Feminism and censor anti-Feminist views, so that media articles are also relatively rarely available to support a claim to Wikipedia article status. And even self-published sources are not given credibility by Wikipedia policy unless the self-publisher has previously been published by third parties. This results in the under-representation of politically incorrect views, issues and persons amongst Wikipedia articles.PeterZohrab (talk) 08:22, 15 May 2010 (UTC)

See : WP:Righting Great Wrongs. TFD (talk) 09:50, 15 May 2010 (UTC)

Says Zubaty: The publishing industry is terrified of women. When my first book came out the Fems went to the distributor and said: if you don't drop his book we will boycott your other 600 books.

The distributor dropped me like a stone. That violated our contract and I offered to sue the distributor. They put me back on the sales list, but years later I found out that anyone who called to order books was told they didn't stock them any more -- when they had 700 sitting in their warehouse!

Sabotage then, sabotage today. Deleting the Rich Zubaty wiki page is sabotage. —Preceding unsigned comment added by Lew Loot (talk • contribs) 21:00, 15 May 2010 (UTC)

Can anyone tell me exactly how to insert a non-English reference? Where do I put the original link, the transalted text? Or just direct me to a similar case and I'll see from there. Thanks Căluşaru' (talk) 19:06, 17 May 2010 (UTC)

It's up to you, but if you use the citation templates, e.g. {{Cite news}}, you can use the trans_title parameter to include a translated title. Here is a sample article which makes extensive use of trans_title. For additional translations, you can use the footnotes if they are too long to fit in the main text. Here is a sample article which has a translated quote in the main text, with the original in the footnotes (#22). Crum375 (talk) 19:21, 17 May 2010 (UTC)

You should follow the citation style already in use in the article (if there is any consistent style). Style guides explain how to deal with translated titles, but they won't all do it the same way. Jc3s5h (talk) 19:27, 17 May 2010 (UTC)

Okay, I was told that this policy should not address topicality of something in an article. Can anyone tell me which policy this would fall under? The de facto rule is that any references cited have to at least make mention of the article topic in some way as a ground rule, and that any references being cited for some other purpose, that don't actually refer to the article topic at all, probably are in the wrong article. But the closest I've found where this is spelled out is WP:SYNTH; maybe there is another section to look in? Til Eulenspiegel (talk) 15:22, 22 May 2010 (UTC)

It's incorrect to focus on the article's topic. A source should directly support the material in question, but that material need not be the article topic. For example, the article may be on a bird species X, which was discovered by Professor Y. There is no prohibition to provide material and sources about Professor Y, for example relating to his other discoveries, even though the article's topic is bird X. And so on. Crum375 (talk) 15:38, 22 May 2010 (UTC)

Really? Are you sure? That is news to me. On most articles I have seen, I think the editors would say all material certainly does have to be about the article topic, or else it is "off topic". They would ask that the references about Professor Y (not mentioning species X) be moved to his article, and just link Professor Y. Til Eulenspiegel (talk) 15:59, 22 May 2010 (UTC)

It may be news to you, but I suspect that if you take any WP:FA, or any other fully developed article, you'll find lots of examples. Even if Professor Y has his own article, we could still include a section about him summary style, and in that section include some (sourced) highlights of his career, or other interesting discoveries. And of course Prof. Y may not have a separate article. The point is that you can't muzzle an editor who wants to expand on the core topic of the article and develop the various subjects touched on. And even if you could, it would be a style or neutrality issue, not "verifiability" or "attributability". Crum375 (talk) 16:06, 22 May 2010 (UTC)

Let me emphasize that the above notwithstanding, per WP:SYN, one may not combine sources to advance a position, unless that position is directly supported by a reliable source. So when discussing Mr X, I may not bring in a source about Y (which doesn't mention X), and then use that latter source to disparage X in some way. Crum375 (talk) 16:12, 22 May 2010 (UTC)

This is more within the scope of WP:NOR... However, it is true that NOR is purposely a bit vague about what it means by "directly related to the article topic". In many ways topicality is a judgment call... something that has to be assessed on an article by article basis. Certainly an article on topic Y would need to establish why it talks about person X (ie it would have to make the connection between topic Y and person X, and support that connection through reliable sources)... but once that has been done, once we have established that there is a connection... it might be appropriate to discuss person x in more detail, and support that detail with sources that are more focused on person X (and may not even mention topic Y).

In any case, I would agree that this policy should not get into the topicality issue. Blueboar (talk) 16:28, 22 May 2010 (UTC)

Yes, anything discussed in an article has to be connected to the article topic via some logical chain, but that's not a verifiability/attributability issue. The attributability focus is on making sure that any material we add to the article must be directly supportable by a reliable source. Crum375 (talk) 16:35, 22 May 2010 (UTC)

(e/c) Please see WP:TOPIC, the undue weight section of the neutral point of view policy, Wikipedia is not an indiscriminate collection of information, (the essay) Wikipedia:Handling trivia#Connective trivia and Wikipedia:Summary style. These are variously or in some mix depending on the specifics I was addressing, what I would cite to argue against off-topic material, which can take the form of truly unrelated to the topic, a side note to the topic, or something that should be mentioned in the article on the topic, but not with the level of detail it is. Whether the source mentions the topic is irrelevant. In an article on sunspots, where one has cause to mention why the sky is blue, the citation for that information need not have any mention of sunspots (but it must verify what it is cited for, i.e., why the sky is blue). You should be focusing on is whether the information belongs in the article, not whether the citation for the off-topic information mentions the topic of the article, which is besides the point.--Fuhghettaboutit (talk) 16:45, 22 May 2010 (UTC)

This policy is about attribution, which requires all material in an article to be attributable to a reliable source to be eligible for inclusion. There are other requirements, such as neutrality, notability and style, but those issues don't belong here. Crum375 (talk) 16:51, 22 May 2010 (UTC)

Since your post indents one level past mine I suppose you are addressing my post. If so I'm a bit baffled. I was responding to the OP's question about what relevant policies apply other than this one, and confirming that verifiability does not apply directly to the matter with which he is confronted. Explaining to me how verifiability works in response... You must have completely misread my post and its context. Please see teaching grandmother to suck eggs.--Fuhghettaboutit (talk) 17:07, 22 May 2010 (UTC)

Not sure about grandma, but my indent was not an indication that I disagree with you. I think we agree; I just used different words to summarize my view. Crum375 (talk) 17:11, 22 May 2010 (UTC)

Then the person misunderstanding was me. Sorry. The etiquette of indentation level may me read it as a response to me, rather than a follow-up on my post to the OP.--Fuhghettaboutit (talk) 17:16, 22 May 2010 (UTC)

Thanks Fuhghettaboutit, thst gave me the answer I was looking for, I think WP:SS will especially help me to have something to point to. Til Eulenspiegel (talk) 17:20, 22 May 2010 (UTC)

The policy currently uses some redirects to sections within policies, piped to appear as "below" or "above" -- when they actually redirect to whole other policies.

Wikipedia:SOURCES, for instance, is referenced in wikilinks that are piped to "above" and "below", but it really redirects to Wikipedia:Verifiability#Sources. Geo Swan (talk) 21:43, 22 May 2010 (UTC)

Seems to me the "above" and "below" links are all strictly internal piped links, starting with "#". So whatever shortcut you're using it ends up in the same place except labeled differently. For example, if you've used the shortcut WP:SOURCES and the target is Wikipedia:Verifiability#Self-published_sources_(online_and_paper), the link ends up in the same place at WP:SOURCES#Self-published_sources_(online_and_paper). Did I miss any examples of piped links that aren't going to the proper section? ... Kenosis (talk) 22:22, 22 May 2010 (UTC)

Exactly... WP:SOURCES is not a separate policy, but a section of WP:V (as WP:SPS is, in turn, a sub-section of WP:SOURCES). This means that when we refer to these section in other parts of the WP:V policy, it is appropriate to say "above" or "below". This would not work when linking to these sections in other policies or guidelines. Blueboar (talk) 23:24, 22 May 2010 (UTC)

Could someone clarify the status of such blogs as reliable sources? Footnote #4 in the policy says "Some newspapers host interactive columns that they call blogs, and these may be acceptable as sources so long as the writers are professionals and the blog is subject to the newspaper's full editorial control. In March 2010, the Press Complaints Commission in the UK ruled that journalists' blogs hosted only on the websites of news organizations are subject to the same standards expected of that organization's print editions (see Plunkett, John. "Rod Liddle censured by the PCC", The Guardian, March 30, 2010). " To me, that says that a professional journalist's blog, posted on a UK newspaper's web site is reliable. Is that incorrect? —Preceding unsigned comment added by Momma's Little Helper (talk • contribs) 13:42, 18 May 2010 (UTC)

This is a good question but will probably get more attention at WP:RSN (under a separate topic maybe). — e. ripley\^talk 17:16, 18 May 2010 (UTC)

This is not a question about a particular source, but a request for clarification regarding a general policy. I appreciate SlimVirgin's response, below, which is my understanding as well. Momma's Little Helper (talk) 23:59, 18 May 2010 (UTC)

Yes, it's a reliable source because a professional journalist's blog on a newspaper website is just another word for a column. It's not a personal weblog. SlimVirgin ^{talk contribs} 17:17, 18 May 2010 (UTC)

Just to be clear, at least some of the genesis for Momma's Little Helper's question stems from questions about whether an article published in Yediot Aharonot about a living person is reliable or not (specific discussion here). Supporters of including the information argue that it should be included because some newspaper columnists from accepted reliable sources have written blog entries that summarize/recount the original article, but which contain no new reporting themselves. I contend that in that instance, a blog entry that does nothing more than describe another publication's story can't be used as a WP:RS. — e. ripley\^talk 17:23, 18 May 2010 (UTC)

Of course some reliable sources are better than others. Have modified the Newspaper and magazine "blogs" section to reflect the source about the PCC ruling more accurately – note that The Spectator is a magazine, not a newspaper, and that the PCC expects the same standards in newspaper and magazine blogs that it would expect in comment pieces that appear in print editions. Not as useful as news pieces, but usable as comment pieces, presumably with in text attribution. . . dave souza, talk 19:04, 18 May 2010 (UTC)

Newspaper blogs aren’t directly comparable to columns in print. The difference for our purposes is in the lack of editing. As a rule of thumb: Anything in a print newspaper is reviewed by at least two people after the writer and before publication, but newspaper blogs are not edited.

See this article from the American Journalism Review (Dec 06 – Jan 07):

• "Most newspaper blogs are self-edited."
• "Nobody edits what I write before it goes online." – Daniel Rubin, full-time blogger at the Philadelphia Inquirer
• "Some newspapers try to edit at least some of their blogs." Maurreen (talk) 06:55, 19 May 2010 (UTC)

Your last point seems to be the case in the UK , per the footnote (now part of the article itself) I quoted: "journalists' blogs hosted only on the websites of news organizations are subject to the same standards expected of that organization's print editions". Momma's Little Helper (talk) 13:38, 19 May 2010 (UTC)

This is all important to know, but we're missing a larger point: Even if a newspaper blog doesn't go through a proofreader or copy editor (which blogs at my publication did) or a top editor, they still carry the imprimatur of the newspaper or magazine. Professional journalists of high-caliber, WP:RS periodicals write to a necessarily high, professional standard and know that whatever they write is going to represent the newspaper or magazine.

I agree with Maurreen that an edited print column is a higher-quality source. I'm not sure there's a practical difference between a newspaper/magazine print column or online column a.k.a. blog, though: When you're talking about, say, the real-estate blog of Jay Romano of The New York Times, whether there's a top editor on that or not, it's still going to be up to the professional level of The New York Times. --Tenebrae (talk) 00:02, 25 May 2010 (UTC)

For anyone interested, based on recent discussions here and on WT:NOR, I have started a draft version of a new proposed ATT policy under my user page, which includes the current versions of WP:V, WP:NOR and some material from WP:ATT. It is very much a works-in-progress, but comments would be appreciated on its talk page. Thanks, Crum375 (talk) 00:29, 24 May 2010 (UTC)

I'll take a look.

Some advice... if you really are serious about moving this forward... SHOUT it to the rooftops... post notice after notice about it in every venue you can... repeatedly. Keep people updated on the progress with yet more notices. One of the more common comments about the old ATT page was "I never knew this was in the works"... even though several hundred editors contributed to the drafting, a lot of people were still taken by surprise when it actually went live. You have a chance to avoid that error this time around.

Also, please remember that people have become attached to the existing policy pages... especially the core policies that they cite every day. You can expect knee-jerk opposition to the idea of a merger, even if the proposed policy says exactly the same thing using the same language. Take this opposition seriously.

I wish you luck, and offer my support. Blueboar (talk) 00:53, 24 May 2010 (UTC)

As far as SHOUTING, I fully agree, but it only makes sense to do so when there is a shoutable product. At the moment, it's just a first draft, and it can benefit greatly from comments and ideas. Thanks for the support. Crum375 (talk) 01:02, 24 May 2010 (UTC)

Offering my support as well. I'll look at the draft in detail later, but at first glance it seems to be going in very much the right direction.--Kotniski (talk) 06:37, 24 May 2010 (UTC)

Thank you. Any input would be much appreciated. Crum375 (talk) 14:51, 24 May 2010 (UTC)

In an earlier thread I proposed making the following change to the wording of WP:SOURCES:

At the time, most of the objections to this amendment were in relation to using autobiography as a source in articles. Having thought about this issue, I am not against using autobiographical sources, but I realise that they are a potential minefield, in the sense that, they are a form of self-published sources, and for that reason, are not strictly reliable in any case. For instance, if I were to quote an autobiograhical source in an article about a living person, I would do in a way that made it absolutely clear that I was doing so (e.g. "XYZ said in his book that...") to alert the reader to the fact that a person speaking about their own life may not be the unbiased source of information about events that affected them.

I feel that we should revisit this proposal, because independence is an important principle in the real world, that when compromised, can have catastropic effects (the Enron Scandal comes to mind in this regard) for those who regard reliable sources as an important form of external verification. It seems to me that independence and reliablity are two vital characteristics of high quality sources, and to ignore one or the other would fatally compromise this policy on verifiability. Would anyone care to support this proposed amendment? --Gavin Collins (talk|contribs) 08:54, 25 May 2010 (UTC)

This works fine for most articles... but not all. How would the proposal affect an article that is about the beliefs and practices of a specific group... say a religious group. I would think that the best (most reliable) sources would be those from within the religious group itself... sources that can speak with authority as to what the beliefs are (or are supposed to be). In other words, there are at least a few situations where articles should be based primarily on dependent sources. Blueboar (talk) 13:06, 25 May 2010 (UTC)

I agree. There are many situations where the best and most knowledgeable sources are the ones directly involved. We should of course strive to present, based on the reliable sources, all prevailing views about the topic, per WP:NPOV, but the majority of the sources and the highest quality ones may sometimes be closely related to the subject matter. Crum375 (talk) 13:52, 25 May 2010 (UTC)

I agree, but the problem is not solved by turning down this proposal, because "independence" and "third-party" mean almost the same thing. So by leaving "third-party" the problem remains. Jc3s5h (talk) 14:00, 25 May 2010 (UTC)

Also agree with Blueboar. Independence is important, but it's not a key criterion for inclusion of all information. We have guidelines indicating when it's appropriate to use self-published sources as sources on themselves; writing this into verifiability would call for eliminating that practice entirely, which would be highly problematic. —chaos5023 (talk) 14:01, 25 May 2010 (UTC)

I gather that this idea stems from recent discussions at WP:NOTE... I fully agree that notability needs to be established through independent sources (to show that someone other than those directly involved have taken note of the topic)... but that is a different issue from what we are talking about here. As a general rule, independence is good in a source... but not always. It depends on the topic. Blueboar (talk) 14:03, 25 May 2010 (UTC)

Yes. Articles should be "based" on third party sources (and also not on primary sources) to establish notability, but we don't require all sources, or even most sources, to be third party. To add the word "independent" and delve into its definitions would create a false impression. Crum375 (talk) 14:31, 25 May 2010 (UTC)

I have to disagree with this proposal. For some topics, particularly historical type stuff, it works great. I agree on Blueboar's concern, that while independence is important, it is not a requirement for all information that should be in an article, and would be concerned on how it could effect some topical areas. And how does one determine if something has a "reputation for independence"? I've never seen a source that does, really. Almost every media outlet is considered "biased" by some party or another, even academics. It wouldn't leave much sourcing available. -- AnmaFinotera (talk · contribs) 14:26, 25 May 2010 (UTC)

In answer to Blueboar, at present this policy says that "articles should be based upon reliable, third-party published sources with a reputation for fact-checking and accuracy". Although it is silent on the role of primary sources, essentially this policy says that articles should be based on third party, not primary sources alone. This point is the subject matter of Wikipedia:No original research#Primary, secondary and tertiary sources. That does not mean you can't use primary sources, or sources which are not striclty reliable such as self-published sources, but what WP:BURDEN makes clear is that without third party (secondary & tertiary) sources, a topic should not have its own article. I think Blueboar is wrong to say that there are a few situations where articles should be based primarily on dependent sources - my reading of policy says that can never be the case.

My proposal goes beyond the issue of weight that should be given to primary sources, to what makes a third party source reliable: if a source is not independent, then is not reliable either: they are two sides of the same coin. Simply put, a lack of independence may compromise, or may be reasonably be expected to compromise, a source's reputation for reliablity. --Gavin Collins (talk|contribs) 15:01, 25 May 2010 (UTC)

That's not what our policies say. Per WP:PSTS, an article may not be based purely on primary sources. And per WP:N, an article must be based on reliable third party sources to establish notability. But as I noted above, "based on third party sources" does not mean that all sources, or even most sources, should be third party. And in many situations, the best and highest quality sources are connected to the subject matter. So in summary, we use third party sources to establish notability for the article's subject, and may use any reliable sources to fill in the details, per WP:V, WP:NOR and WP:NPOV. Crum375 (talk) 15:04, 25 May 2010 (UTC)

Gavin.collins is creating a useless tautology. First he defines independence as, in part, "a source is free from pressures...that may compromise, or can reasonably be expected to compromise, the source's reputation for reliablity". Then he states "if a source is not independent, then is not reliable either". Only if you are using Gavin.collins definition of independence. Many sources may be closely connected to subject matter, but still have a strong reputation for accuracy and fact checking. An example would be the United States Supreme Court. These sources might be biased, or may be constrained to work within a biased framework, yet can still be reliable. Jc3s5h (talk) 15:25, 25 May 2010 (UTC)

I believe that editors should normally and automatically evaluate sources for independence—because someone with a significant conflict of interest really isn't a "third party"—but I think that explicitly spelling it out on this page might be just a bit WP:CREEPy. WhatamIdoing (talk) 18:01, 25 May 2010 (UTC)

If editors automatically evaluate sources for independence, then should we not say so in this policy? --Gavin Collins (talk|contribs) 21:34, 25 May 2010 (UTC)

There are at least three problems with this suggestion ... The 1st, and most important, is the term "compromise" which potentially could lead to libel lawsuits against WP if a talk page says a source's "reliability" (which could be considered synonymous with "personal integrity") has been compromised. The 2nd is the fact that the vast majority of subject matter experts are directly connected to their subjects of expertise by vocation and few outside of certain vocations even understand the subject matter. The 3rd is that nearly ALL modern published sources are constantly subject to economic pressures which determine what subjects and what perspectives of said subjects will be allowed (funded) to go to press. 66.102.198.179 (talk) 04:35, 1 June 2010 (UTC)

The proposal says that independence may compromised, or can reasonably be expected to compromised, where there is a strong connection, so that excludes most modern published sources. For example, it is one thing to write commentary, crticism or analysis about a topic, but this is totaly different from being paid to write promotional content. I think you can set aside these concerns. --Gavin Collins (talk|contribs) 08:27, 3 June 2010 (UTC)

WP:NEWSBLOG currently says "Posts left by readers may never be used as sources." But is this not like letters to the editor? Does it not depend on whether the person or organisation writing can be identified as a notable expert on the subject under discussion? For example there was a magazine published during the Victorian period Notes and Queries Online, which was a sort of editorial overseen paper blog. Information from N&Q was cited and included in academic publications. In most of the cases I have seen academic selection of information extracted from N&Q was from from known experts in their field. -- PBS (talk) 22:18, 26 May 2010 (UTC)

Letters to the editor are screened in some way. Posts on blogs aren't, so we have no idea who's writing them. SlimVirgin ^{talk contribs} 22:42, 26 May 2010 (UTC)

SV is working from incomplete information. Posts on some blogs are screened; posts on others are not. Carefully moderated blogs might have approximately the same level of editorial oversight as the letters to the editor page. (Which is to say: neither of these are particularly strong sources.) WhatamIdoing (talk) 23:12, 26 May 2010 (UTC)

Can you give an example of a newspaper blog that screens posts in that way? SlimVirgin ^{talk contribs} 23:50, 26 May 2010 (UTC)

The BBC isn't quite a newspaper, but Stephanie Flanders has a blog here that I would certainly accept as a reliable source.—S Marshall T/C 00:08, 27 May 2010 (UTC)

I think you may be confused. Blogs on respected media are actually columns, and are considered reliable per WP:NEWSBLOG. What is at question here are reader comments posted to blogs or online articles. Crum375 (talk) 00:19, 27 May 2010 (UTC)

Oh. Yes, I see that; I took SV's remark in isolation and replied to it. But now that I've read it, I'm not sure that PBS's question is about reader comments on blogs. I think it's more about these.—S Marshall T/C 00:28, 27 May 2010 (UTC)

Saying that a reader post to a blog "might have approximately the same level of editorial oversight as the letters to the editor page", even if true (and that is unlikely) is hardly a ringing endorsement - quite the opposite, in terms of being a reliable source. In any event, posts to newspaper blogs are only screened, at best, in the most rudimentary way; that is, if they contain obscenities, or are overtly racist, or are simple spam, or contravene some law, or the like. Letters to the Editor are carefully screened, if for no other reason than the fact that it is expensive to print newspapers, and the Letters to the Editor page have very limited space. On the other hand, space for user comments on blogs is essentially free, so just about anything is allowed. Jayjg ^(talk) 06:50, 27 May 2010 (UTC)

If I were to write a letter to editor of The Times and forge the signature of Karl Marx it would be incumbent on the editor to a) check the postmark to ensure it hadn't somehow been caught up in the postal system for decades, b) decide based on content whether it deserved publication anyhow, and if it did, c) add an editorial comment to alert any unsuspecting readers to the improbable attribution. Few open blogs make even basic attempts to verify the author's identity. For a newspaper-operated open blog, it would be straightforward to use customer account numbers or require an automatic confirmation email, but this is seldom done. For registered contributors there may be a karma system, but it only goes so far. LeadSongDog come howl! 13:58, 27 May 2010 (UTC)

Such trusting faith.  ;-)

In a large newspaper, the editor probably won't even see the envelope, and s/he'll assume that the letter is written by some reader whose name really is "Karl Marx" — a Karl Marx, e.g., [14], not the Karl Marx. They don't know, or really even care, if you're using your real/everyday name. You can — and people do — use middle names, maiden names, and made-up names. Typed, unsigned, plain-paper letters that are supposedly from local politicians usually get double-checked, and letters that are obviously from the same crank (every newspaper has at least one) get discarded, but that's really about it. WhatamIdoing (talk) 17:01, 27 May 2010 (UTC)

As an example, consider the New York Times blogs: They screen do comments (at the minimal level), but they also select comments for highlighting in their "Comments of the Moment" section. That section receives the same level of active editorial selection, and suffers from the same limitations on space, as the letters page of a typical newspaper. IMO the "Comments of the Moment" section is (1) still reader comments and (2) selected by an editor. WhatamIdoing (talk) 17:01, 27 May 2010 (UTC)

Let us suppose that a opinion is published in a newspaper and added to the newspapers bloc site, in which the newspaper columnist claims that something in a recently published book was wrong, and that the author of the book replies on the blog with a clarification. It would be incumbent on the newspaper to check that it was indeed the author of the book and the author of the acknowledgement was one and the same person (if not they would leave themselves open to legal action). I'm thinking along the lines of David Irving's letter to the Times in 1966, but published in this day and age as a reply in a blog on www.timesonline.co.uk/tol/comment/blogs/. -- PBS (talk) 23:22, 27 May 2010 (UTC)

The external form of a medium is not the definitve factor. Obviously most of what appears in most such blogs is worthless, but this is not universally true. Posts left by readers stand on the authority of their authors. In most cases this is not much, in some cases it is. Each case of this sort has to be justified, but it's become a regular ode of expression by the most highly reputed authors and publications. DGG ( talk ) 05:15, 28 May 2010 (UTC)

Quite! But WP:NEWSBLOG currently says "Posts left by readers may never be used as sources." (my emphasise). -- it comes down to the old distinction we make about reliable sources, "the piece of work itself (a document, article, paper, or book), the creator of the work (for example, the writer), and the publisher of the work". Usually a Newspaper readers comment to a blog fails on two of the criteria. The content and the creator. But occasionally, as with letters to the editor, a reader's posting will tick all three boxes. As such the blanket ban imposed by "may never" is not always going to benefit the project. Should the wording be modified, or should editors rely on WP:IAR? -- PBS (talk) 10:36, 31 May 2010 (UTC)

I need advice on the implementation of this rule when the challenged fact is common knowledge. The example above that "Paris is the capital of France" is a trivial example that you would hope would never be challenged. But for the sake of example, what if it were challenged in good faith? How do you select a source for information that widely known?

Or to take a slightly more complex example, how do you source Pulley when all the facts in the article have been common knowledge (at least among engineers and the equivalent) since the time of Aristotle and can be found with trivial ease in any high school textbook? Unlike more modern findings, there is no breaking source or authoritative dissertation proving the mechanical advantage equations. Pulling out my highschool textbook seems inappropriate. Why should it be given precedence or credibility? Websites covering that level of basic information are equally speculative, often appearing to have been written for elementary students and containing no source information themselves.

By the way, sourcing the opening definition of "pulley" to the Oxford English Dictionary seems to me to be a massive waste of readers' time and a distraction rather than an addition to the page. Thoughts on a general solution to this problem would be much appreciated. Rossami (talk) 22:10, 2 June 2010 (UTC)

Most obvious facts like that would have been sourced incidentally elsewhere in the article. One of the references in Paris will already confirm that fact. When it first became the capital is presumably cited in the body of the text, negating the need to cite it in the lead. Leads don't need to be cited if the information they summarises is cited elsewhere.

As for pulleys and other basis STEM topics, why not use a decent textbook at a source? Fences&Windows 22:38, 2 June 2010 (UTC)

Rossami, as in virtually all cases when something is challenged by someone, which seems "obvious" to someone else, it's far easier to provide a source than to argue about it. If something is so obvious, there must be tons of good sources, probably textbook level, and many online. So just find one, provide an inline reference to it, and move on. Crum375 (talk) 22:46, 2 June 2010 (UTC)

I think perhaps I did not communicate the point well. There are hundreds of textbooks covering mechanical advantage, none any better than any other and none pointing the reader to the actual source of the information. Arbitrarily choosing one creates a false sense of authority and distracts rather than helps the reader. But perhaps I'm just being obtuse. How would you source pulley? (Because I and dozens of other editors have looked and that article and been unable to solve it for several years now.) Rossami (talk) 00:05, 3 June 2010 (UTC)

With a few minutes of googling, I came up with this book which defines 'pulley'. I am sure many more can be found. Crum375 (talk) 00:55, 3 June 2010 (UTC)

The book referenced by Crum375 is a fine example of the PROBLEM user Rossami is talking about. That book is a revised version of work published in 1878 ("...series of elementary text-books in science, written by the late Dr. J. Dorman Steele...") which in turn references a large number of older texts:

The following works, to which the author acknowledges his obligation for valuable material, will be useful to teacher as well as pupil, in furnishing additional illustrations and in elucidating difficult subjects, viz.: Tait's " Recent Advances on Physical Science"; Arnott's "Elements of Physics" (7th ed.); Stewart's "Elementary Physics," also his "Conservation of Energy," and "Treatise on Heat"; Atkinson's " Deschanel's Natural Philosophy"; Lockyer's "Guillemin's Forces of Nature"; Herschel's "Introduction to the Study of Physical Science"; Tomlinson's "Introduction to the Study of Natural Philosophy"; Beale's "How to "Work with the Microscope"; Schellen's "Spectrum Analysis"; Roscoe on "Spectrum Analysis"; Lockyer's "The Spectroscope," and "Studies in Spectrum Analysis"; Airy's "Geometrical Optics"; Nugent's "Optics"; "Chevreul on Colors"; Thomson and Tait's "Natural Philosophy"; Maxwell's "Electricity and Magnetism"; Silvanus Thompson's "Lessons in Electricity and Magnetism"; Faraday's "Forces of Matter"; Youmans' "Correlation of Physical Forces"; Maury's "Physical Geography of the Sea"; Atkinson's " Ganot's Physics"; Silliman's "Physics"; Tyndall's Lectures on light, Heat, Sound, Electricity, also his "Forms of Water"; Snell's "Olmsted's Philosophy " (revised edition); Loomis' "Meteorology "; Miller's " Chemical Physics"; Urbanitzky's " Electricity in the Service of Man"; Cooke's "Religion and Chemistry"; Darnell's "Principles of Physics"; Anthony and Brackett's "Text-book of Physics," and also numerous works named in the "Reading References" at the close of each general division. They may be procured of the publishers of this book. The pupil should continually be impressed with the thought that the text-book only introduces him to a subject, which he should seek every opportunity to pursue in larger works and in treatises on special topics.

I think what Rossami wants is possibly the oldest published (therefor the most original) description of a pulley, or else a published work that is somehow become considered as the "bible" of basic mechanical physics. 66.102.198.60 (talk) 00:11, 4 June 2010 (UTC)

Well, this is a wiki. So start out with a couple of textbook sources, and over time add more, if better ones can be found. Remove old sources if they don't add any value to the list. The idea is simply to refer the reader to various high quality reputable sources, not to find one single ultimate "killer source". Crum375 (talk) 00:24, 4 June 2010 (UTC)

Also related to the same conversation at WP:N, we now have an assertion that www.coca-cola.com, published by Coca-Cola, Inc., is not self-published. I think the rationale is that multinational corporations are too big to be capable of self-publishing a website.

Would anyone object to adding "corporate websites" to the list at WP:SPS? Or perhaps it would be more appropriate to provide a definition of non-self-published (i.e., something with both editorial independence [from the business side] and editorial control [of the reporters]). WhatamIdoing (talk) 17:54, 25 May 2010 (UTC)

A sufficient majority of the editors who regularly edit source-related policies and guidelines insist on talk pages (but do not reveal in the guidelines and policies themselves) that "self-published" refers to the actions of a single individual, or small band of individuals. Self-published, in the minds of this cohort, cannot refer to large organizations. The guidelines and policies have been written with this covert definition in mind. Thus, a change in the definition to include large organizations requires a serious re-write of all the source-related policies. Jc3s5h (talk) 18:00, 25 May 2010 (UTC)

"Self-published," as the policy makes clear, refers to individuals or small unprofessional groups publishing things like personal blogs. It doesn't refer to The New York Times or Coca Cola publishing material about itself. SlimVirgin ^{talk contribs} 18:03, 25 May 2010 (UTC)

Then it's (past) time to do that serious re-write. Corporate websites are basically advertising published by and for the corporation, with zero oversight by any other party (much less an independent one), and it is stupid to pretend that advertisements aren't published at the direction, cost, and discretion of the advertiser. (And just who do we think publishes coca-cola.com, if not Coca-Cola, Inc.? Martians? The US government?)

(And if anyone claims that the newspaper ads BP is running are properly published in the same sense that the news stories next to them are, I might actually scream.) WhatamIdoing (talk) 18:13, 25 May 2010 (UTC)

"Self publishing" in the Wikipedia sense does not refer to an organization publishing its own material, because that definition would include virtually all sources, including the New York Times. By "self publishing" we mean individuals or small groups who have no or minimal oversight levels. Coca Cola and other large corporations are extremely careful in what they publish, and have many vetting layers, so they are clearly a reliable source, though normally a primary one. This is clearly covered in WP:SOURCES and WP:SPS. Crum375 (talk) 18:19, 25 May 2010 (UTC)

No, it wouldn't. Stuff NYT published about itself (e.g., its subscription rates) should be considered self-published. Stuff it publishes about others should not. There is no oversight on a company's own website: Only Coca-Cola itself decides what goes on coca-cola.com.

And again: If Coca-Cola, Inc. isn't the legal person who is publishing that website, then just who is? WhatamIdoing (talk) 18:27, 25 May 2010 (UTC)

I think you have your terminology confused. The NYT does not magically become "self published" when it describes itself, and not "self published" for other topics. Its vetting mechanisms, as well as the physical publishing and printing mechanisms, are substantially the same, regardless of what they publish. The difference is that if they describe themselves, they become a "primary source" for that information, since they are directly involved in the topic being covered. This is true for all sources: if they describe a topic they are involved in, they are a primary source for it, otherwise they are a secondary source. Therefore, Coca Cola or other corporations are a primary source when describing their products. They can be very reliable, but they are still primary. Please read WP:SOURCES and WP:PSTS for more info. Crum375 (talk) 18:33, 25 May 2010 (UTC)

I don't have my terms confused: I'm using the standard, plain-English definitions, and I'm arguing that Wikipedia should, too.

I do think that you've never worked for or had any experience with a newspaper or other traditional form of media. Here are some facts that you don't seem to possess: Even at The New York Times, a distinction is made between advertisements (including self-advertisements) and editorial content. Circulation departments do not employ journalists to draft their published subscription terms. Publications like advertising rate cards are not considered editorial content and are completely exempt from journalistic editorial control: No editor named on the masthead at NYT has any authority over what the ad department publishes about the ad department's policies or fees.

I fully agree that coca-cola.com is a primary source for statements about Coca-cola, Inc and its products. The website is also a self-published source, because it is published by Coca-Cola, Inc. about Coca-Cola itself. Nobody except Coca-Cola, Inc. controls the publication of their corporate website; it is therefore self-published. The classic classification system (primary, secondary, and tertiary sources) does not include a fourth point called "self-published". Whether something is self-published or not is an entirely separate consideration from whether it is a primary, secondary, or tertiary source. To give a simple example, ISBN 9781556435218 (a dictionary) is (1) a tertiary source and (2) self-published. (The same can be said, BTW, for Noah Webster's works.) WhatamIdoing (talk) 20:09, 25 May 2010 (UTC)

(edit conflict)

One might reasonably argue that self-published speaks to the lack of a discernable and functional distinction between the author and the publisher. If one is employed by or otherwise subordinate to the other then they effectively have a single identity and consequently reduced reliability. The NYT editorials should be considered self published. "Staff writer" items about the NYT should too. Byline attributed articles by individual journalists gain some degree of additional credence, and very reputable journalists may rise to be arguably independent. One simple test is whether they are syndicated to papers that are not under the same ownership. There is however no case in which one would expect Coca-cola's advertising agency to publish material that Coca-cola was opposed to. That rather is why it is called an agency.LeadSongDog come howl! 20:25, 25 May 2010 (UTC)

(ec) When we say on Wikipedia "self published", we mean individuals or groups with no (or minimal) vetting layers, publishing content. When it comes to larger organizations with established formal mechanisms for accuracy checking and multiple individuals involved in content vetting, they are no longer "self published" by our definition. In any case, there are two independent criteria for source evaluation: reliability, which depends on vetting layers and reputation for accuracy checking, and type, primary vs. secondary, which depends on whether the published material is related to the source. To determine if a particular source is appropriate in a given situation, we need to decide its reliability and whether it's primary or secondary. A large corporation is a good source for technical details about its products, for example, but not an acceptable source for their notability, nor how they compare to competing products. Crum375 (talk) 20:31, 25 May 2010 (UTC)

Just to keep the pot boiling, what about audited accounts? Johnbod (talk) 19:48, 25 May 2010 (UTC)

It's obviously a primary source for the kind of statements we might be making (e.g., about a company's revenue during the previous year). I believe it would be better described as self-published than as non-self-published: It's usually part of a larger report (that the auditors don't sign off on), and it is the company itself that makes the audited statements available to the public. WhatamIdoing (talk) 20:09, 25 May 2010 (UTC)

That seems a very specious argument to me; it is obviously verifiable that it is a secondary source (by the auditors), and the auditors do review the whole report, although I accept there is a difference. It is of course also normally the only source for any figures in it; when a report is on the web there is little benefit in instead referencing a newspaper report for a simple figure, which will be reported without the extra detail in the notes to the accounts & so forth. In the same way newspapers are apt to ignore the differences between the audited and unaudited parts of the report when reporting. Obviously no professional analyst etc would prefer to take their figures from a newspaper report than from the actual accounts, and there are reasons for this. Johnbod (talk) 10:14, 26 May 2010 (UTC)

Self-published is not a synonym for unreliable. I have not said that the audited statement (or the annual report that contains it) isn't reliable. I have said that it is a primary source and that it is published by the author (the company), not by someone else (an independent/third-party publisher, or the CIA, or Martians). WhatamIdoing (talk) 17:17, 26 May 2010 (UTC)

It is part-written (most of the text in fact), audited, and if necessary re-written by the auditors (the only ones to sign the audited part), who are required by law to be independent, and published as a legal requirement. The authors of few other sources are required under pain of heavy legal penalties to be independent. Johnbod (talk) 19:18, 26 May 2010 (UTC)

Let's try this in smaller steps, with an example:

1. Who wrote Enron's 1999 annual report?
2. Who published Enron's 1999 annual report?
3. Are the answers to Question 1 and Question 2 better described as being "the same" or "different"?
   • If "the same", then it is self-published.
   • If "different", then it is non-self-published.

This really isn't that hard: If the entity that writes (part or all) of something is the same entity that publishes it, then it is self-published. The fact that Enron hired (supposedly) independent auditors to help write, or at least sign off on, parts of Enron's annual report, which was clearly published directly by Enron, does not mean that Enron is not, in both legal and practical terms, both an author and the sole publisher of its 1999 annual report.

NB that "self-published" is not a fancy way of spelling "unreliable". It is a concise way of spelling "the author is the publisher". Self-published things are frequently quite reliable. WhatamIdoing (talk) 20:55, 26 May 2010 (UTC)

I would say that for the sake of WP, a company's own website is not self-published, as in practice we treat self-published sources and primary sources (such as companies' web sites) differently. Most self-published sources are a synonym for "You probably shouldn't use this", while primary sources are mostly a synonym for, "You can use this, but be careful with it." Angryapathy (talk) 20:45, 25 May 2010 (UTC)

I think the point is that when we say "self published", we mean that the individual human author publishes the material himself, with no formal vetting layers. Once multiple other independent people (even if employees of the same company) are involved in vetting the material, in some formal mechanism, it is no longer a "self" project. Crum375 (talk) 21:05, 25 May 2010 (UTC)

Crum, you have just defined The Mulberry Advance (a respectable, traditional, dead-tree, small-town newspaper with a single employee, and consequently zero vetting layers) as a self-published source, and press releases(!!) from multinational corporations and political campaigns, which tend to be signed off on by legal, marketing, and financial departments, as non-self-published sources.

This is not fundamentally a difficult concept: If "you" publish it "yourself", then it is "self-published". If "you" publish "their" work, then it is non-self-published. When "you" refers to a corporation or other non-human entity, then any employee or agent of the corporation is part of "you": The actions taken by Coca-Cola's own legal department really are considered the actions of the company, not the actions of individual, independent humans. WhatamIdoing (talk) 21:15, 25 May 2010 (UTC)

You're conflating "self-published" with "primary source", again.

They are different things. "self-published" refers to secondary sources, and refers to lack of editorial process and rigor. A newspaper which is accepted by a community and newspaper publishing peers, even if it has one employee, still meets a minimum definition of editorial process and rigor. A website, which nobody vets and can give feedback on, does not.

"Primary source" is where information comes from to start with. A company press release announcing a new product is a primary source, a company financial statement is a primary source, etc.

The distinctions between primary and secondary sources are a key factor in understanding information reliability. Georgewilliamherbert (talk) 21:25, 25 May 2010 (UTC)

WI, regarding your Mulberry Tree paper, if it's run by one man, yes it is technically "self-published". But, if it has a good reputation for accuracy and fact-checking, and reports on things other than the author/publisher himself, it could be used "carefully" to support some small-town events, for example. And as GWH noted above, once a paper like that is in long term circulation, it gets vetted by peers and readers. So in its favor are the facts that it has a good reputation for accuracy, and it's a secondary source, while in its disfavor is the fact that it's a one man show. So we probably won't use it to support contentious issues, but for births, weddings, and funerals, carefully, perhaps. Crum375 (talk) 21:45, 25 May 2010 (UTC)

The part of your definition that disturbs me most is the part in which you declare that any press release — say, the ones BP is putting out about its oil spill problems — are non-self-published simply because multiple employees of the company were involved in writing it.

(As a point of fact, The Mulberry Advance may not be self-published: the sole employee may not also be the publisher. Usually, with a small-town newspaper, the publisher is the business owner, rather than an employee. With a large outfit, there's more division; e.g., the publisher of the NYT is Arthur Ochs Sulzberger, Jr., and the owner is The New York Times Company.) WhatamIdoing (talk) 22:02, 25 May 2010 (UTC)

Observe that it is quite possible for a published tertiary source to be based on primary and secondary sources that never got published if those primary and secondary sources were reliably archived. In a forensic case for instance, a bullet or DNA sample is a primary source. Images and analyses of them may become the first published source on the topic when they get to court, but that should not be confused with it being a primary source of information. LeadSongDog come howl! 21:42, 25 May 2010 (UTC)

I disagree. A report which analyzes bullets and DNA samples, given lab data, is still primary, if it performs the original analysis of that data and shows how the lab results match (or don't match) the suspect. In other words, if it's a report produced by a professional investigator to make an original incriminating (or exculpating) case against a suspect, then it's primary, since that investigator is involved in analyzing the lab data, and making original conclusions about them. If someone else, like a news journalist, reads that report and describes it to the public, it would be secondary. Crum375 (talk) 21:55, 25 May 2010 (UTC)

George, I'm not conflating "self-published" with primary, etc.—but you are. Self-published does not "refer to secondary sources": it refers to publications that are published by the author (whether that author is a human or a corporation).

1. Who published something and
2. Whether the publication is primary/secondary/tertiary

are completely separate issues. All of the possible combinations exist. Consider these examples:

	Example of self-published source	Example of properly published source
Example of primary source	Grandma posting at Blogspot about her house burning down.	The first-hand, "eyewitness" report in the local newspaper written by the reporter who was dispatched to the scene of the fire.
Example of secondary source	A meta-analysis posted on a researcher's own website.	A meta-analysis printed in a scholarly journal.
Example of tertiary source	The dictionary mentioned above.	The current version of Merriam-Webster

Whether a source is self-published is completely separate from whether it is primary, secondary, or tertiary. Self-published sources can be any of these types. I think it would help a lot if people stopped pretending that Grandma's blog either doesn't exist, or isn't a primary source, or isn't self-published (and those are your three options, if you keep insisting that self-publication can only apply to secondary sources). WhatamIdoing (talk) 21:54, 25 May 2010 (UTC)

WhatamIdoing - I'm not confusing the two. I'm using "self-published source" in the same sense that it's used throughout Wikipedia WP:RS and WP:V discussions - as shorthand for "self-published secondary source". Your comments that you believe that www.coca-cola.com is a "self-published source" indicate that you don't understand the local terminology used by the policies. Primary sources are primary sources, no matter where they're found. Georgewilliamherbert (talk) 22:15, 25 May 2010 (UTC)

I see: you are talking about whether something is WikiJargonSelfPublished (that is, a special kind of self-publication, defined nowhere [e.g., notice that your claimed criteria of applying solely to secondary sources is not mentioned at SPS], only applicable to certain rarefied aspects of Wikipedia, and that excludes many self-published sources). I'm saying: Let's stop caring about whether something is WikiJargonSelfPublished. Let's write our policies so that recommend that editors notice whether a source is RealWorldSelfPublished. WhatamIdoing (talk) 22:25, 25 May 2010 (UTC)

Break 1

WI, there is no wikijargon here. When we say "self published", we mean it in the most common^{[according to whom?]} way, i.e. an individual publishing stuff on his own, without help, or perhaps with a couple of his buddies. To take this simple definition and stretch it to cover Coca Cola, so a giant corporation becomes a "self publisher", that's a huge stretch, and we don't do it. Crum375 (talk) 22:34, 25 May 2010 (UTC)

Crum, there's a whole lot of WikiJargon here, if George's claim that primary sources can't be considered self-published on Wikipedia is true. For your made-up definition based on size -- a definition that is also not mentioned in SPS, I see -- I'd like to say that nobody else in the world -- nobody except a few stalwart preservers of this page on Wikipedia -- thinks that major corporations are too large to publish their own materials.

Who do you think is writing BP's press releases? Who do you think publishes ("makes them available to the public") them?

I can tell you what the answer is, according to the long-established norms of the publishing industry and all copyright laws: They are written as a work for hire for a very large corporation (this actually does make the employer/corporation legally the author), and they are directly published by the same very large corporation. They are neither written by, nor published by individuals, or Martians, or anyone other than their corporate author. WhatamIdoing (talk) 22:59, 25 May 2010 (UTC)

Again, you are missing the point. When we say "self published", we don't mean it literally, because by the literal meaning, virtually all sources do their own publishing, including the New York Times, Nature, and Scientific American. By "self published" we mean simply that an individual human being published the material himself. We don't include corporations in "self", as that would include nearly everyone and everything, and make the term practically meaningless. Crum375 (talk) 23:05, 25 May 2010 (UTC)

Again, you keep saying that stories in The New York Times publish themselves, and I keep telling you that The New York Times is published by Arthur Ochs Sulzberger, Jr. (an individual human) for The New York Times Company (the publication's owner). Lining up Wikipedia's definition with the definition that has been used, with substantial success, in the real world for literally centuries does not result in the alleged slippery slope.

If you mean "stuff published by an individual or very small group", then you need to say "stuff published by an individual or very small group". You'll still be left with the difficulty of explaining why stories written by a couple of professional journalists should be outranked by the excretions of a large committee made up of corporate lawyers and publicists, but at least people reading the page will actually know that you don't actually care about the relationship of the author and the publisher. WhatamIdoing (talk) 23:29, 25 May 2010 (UTC)

Yes, you are correct that that's the general meaning. The current verbiage is: "self-published media—including but not limited to books, newsletters, personal websites, open wikis, personal or group blogs, Internet forum postings, and tweets—are largely not acceptable." Can you suggest an improvement? Crum375 (talk) 23:44, 25 May 2010 (UTC)

Sure, I can give you several suggestions for improvement. One option is to start the section with an actual definition, e.g., "Self-publishing is the publication of books and other media by the authors of those works, rather than by established, third-party publishers." This, of course, has the effect of defining advertisements, corporate press releases, and spam from political campaigns as self-published works, even if the organization is quite large, and defining most newspaper articles as non-self-published, even if the newspaper is rather small, which I understand that you rather oppose.

If you don't like that, another option involves removing the misleading term "self-published" entirely, and substituting what you think the intent of the section is, i.e., "Sources published by individuals or small groups—including but not limited to books...—are largely not acceptable. Sources published by large groups, including self-promotional materials from major corporations, are largely considered reliable (although they may be subject to restricted use as primary sources)." NB that I don't think this is a good standard, but if you're right that this is what Wikipedia intends to communicate, then Wikipedia should have courage of its convictions and say so plainly, rather than camouflaging it as an issue of who issues the publication rather than an issue of how many humans have been involved in it. WhatamIdoing (talk) 00:22, 26 May 2010 (UTC)

We can't change established core policies on a whim, and I see no need to, as the core concepts have worked fairly well up to now. And I am sure you will see that even with your own suggestion, it's tricky to make changes in something that works. You suggest: "Sources published by individuals or small groups—including but not limited to books...—are largely not acceptable." The problem with saying outright "small groups" is that there are some fairly small groups which may be, to use your own example, a small town paper, with 4 employees, which do a great job publishing their local highly reputable newspaper, while there are other larger groups which have a website full of junk. So if we tried to nail this down, we'd run into all sorts of problems. This is why we leave it more flexible, by trying to convey the meaning of "self published" through the examples. Not perfect, but nothing ever is, and this works. Perhaps an essay can help, or a wiki glossary. Crum375 (talk) 00:35, 26 May 2010 (UTC)

IMO, the core concept has worked very badly, if we mean to say "Small groups are less reliable than large groups", and instead we tell editors, "Things published by the author are less reliable than things published by an independent publisher." If we mean to say that David is a worse source than Goliath, then we should say it. If we don't -- and I believe that we don't -- then let's not say "Self-published sources are worse than properly published sources" on the actual policy page, but then surprise everyone by claiming that these words actually mean that "Small groups are worse sources than large groups" instead of what they actually say. WhatamIdoing (talk) 01:32, 26 May 2010 (UTC)

There is no surprise, and no discrepancy. When we say "self published", we mean published by the authors themselves, as opposed to going through a vetting process and having other independent eyes inspecting and verifying the material prior to publication. We clarify here and elsewhere that reliability of sources relates to the number of vetting layers included in the publishing process. So it all fits in, and there are no surprises. Crum375 (talk) 01:44, 26 May 2010 (UTC)

Really? Are you firmly committed to that position?

Because I've been saying that "self-published means published by the author" all day long, and George has told me that I'm wrong, because self-publication can't happen to primary sources, and you've told me that I'm wrong because self-published means that it's only things published by an individual author or small group of authors, i.e., that there is some magic number above which a group of authors becomes too big to publish what they write themselves.

Perhaps you would now answer one of the questions I asked above: For the purpose of SPS, who is the author of these publications? And who is the publisher? And if you conclude that they are the same, are we agreed that these press releases are actually self-published, regardless of the fact that lots of individual employees were (presumably) involved in the publication?

And if you actually, finally believe that "Self-publishing is the publication of books and other media by the authors of those works, rather than by established, third-party publishers" (i.e., the basic dictionary definition is correct), can we include that statement in WP:SPS, so that people will quit saying things like "Self-publication is the self-publication of secondary sources" or "Self-publication is any kind of publication by individuals or small groups"? WhatamIdoing (talk) 03:38, 26 May 2010 (UTC)

Yes, when WP says "self published", for example in WP:SPS, we mean published by the authors, with minimal if any vetting layers. This is independent of their primacy: they can be primary, secondary or tertiary, but generally they are not a reliable source, except for information about themselves, if it's not unduly self-serving. As far as your BP press release, a press release from a large corporation, like any other publication it produces, is not "self published" in the WP meaning, because the authors of the releases have to run them by other employees (e.g. lawyers and executives) who vet them before release. Typically, such releases are primary sources, since they discuss subjects with which the authors and their employers are involved. But again, the issue for reliability is the number of vetting layers, not whether it's a corporation selling widgets or newspapers. Crum375 (talk) 03:55, 26 May 2010 (UTC)

So in your world, corporate authorship doesn't exist, right? So ten guys who publish a book are self-published, but the same ten authors, if they have the foresight to incorporate their business, publishing the same book, are not self-published? WhatamIdoing (talk) 03:59, 26 May 2010 (UTC)

When we say "self publish", we focus on the individual human authors who actually compose the words, and the number and quality of vetting layers between those original words and their final publication. If it's a corporation, odds are they would be more concerned about liability, and would be more likely to have a formal policy in place for vetting their published documents. When we say "self published material is unreliable", we mean material written and published by the authors, with minimal if any vetting, is unreliable. If there are 10 guys who incorporate, they would likely also set up formal corporate policies, including one for vetting their published materials. But the bottom line again is the number of vetting layers. The more known or expected layers, the higher the expected reliability. And corporate liability influences the number and quality of the vetting layers. All of this is independent of source primacy, though corporations typically write about themselves, and are therefore likely to be primary sources for that material. Crum375 (talk) 04:10, 26 May 2010 (UTC)

I don't think that you know what you're talking about. Incorporation makes these ten authors less liable for what they write. Incorporation means that if they screw up, nobody can take their homes away from them. If they screw up as a partnership, their personal assets are jointly and severally on the line.

But I have specified that we are publishing "the same book". Are you certain that "the same book", written by "the same ten authors", is published by the authors if they are a partnership, and is not published by the same ten authors if they have incorporated? WhatamIdoing (talk) 04:35, 26 May 2010 (UTC)

Incorporation protects the individual owners' homes, but it leaves their combined work effort, where they focus and invest their time and energy, very much open to lawsuits, a libel one in this case. So when they set up their corporation, they would normally take steps to minimize its legal risks, and one of the first steps would be to set up formal policies to ensure the quality of their product(s). This is especially true if there are more than a couple of employees. So yes, if the same book is written in a corporate environment where a quality management system is likely to have been instituted, including formal quality assurance policies, that book, along with any other product, is likely to be of higher quality and therefore more reliable. So in general, individuals publishing their own material without formal vetting layers are less reliable than corporations doing so. But if we have specific knowledge that an unincorporated group or organization has a formal quality assurance mechanism in place, with multiple pairs of eyes regularly vetting published output, that would count towards reliability too. The bottom line again: the more vetting layers, the more eyes expected to routinely inspect the output before publication, the more reliable it is. Crum375 (talk) 12:44, 26 May 2010 (UTC)

Yeah, you're not getting it.

Incorporation means that only the work product is on the line. Non-incorporation means everything is on the line: work product, homes, cars, bank accounts -- everything. Why would any (rational) person take more care to protect a small fraction of their assets than all of their assets, including that small fraction?

Incorporation of a small business in most of the US costs about $200, requires you to fill out one or two simple forms, and takes about an hour. That's it: Once the state processes the paperwork, your business is incorporated. Sloppy, lazy, careless people are perfectly capable of doing this. People who fully expect to get sued do this as a matter of course. None of this paperwork involves creating a quality management system, formal policies, or anything else. There are millions of corporations in the US that have only a single person involved in them. There are thousands of individuals who have created and maintain multiple corporations all by themselves.

But to drag you back to the subject: The question is not, "Is a corporate press release a reliable source on Wikipedia?" The question is, "Did the business both write and publish it?" (Or, was it written by Martians and published by the CIA, or something like that?)

Again: The question is not reliability. The question is only self-publication (is the author the same as the publisher). WhatamIdoing (talk) 17:35, 26 May 2010 (UTC)

---

And let me reply to your point of journalists vs. corporate lawyers. That's apples and oranges, because normally journalists would write a story about a third party, and be a secondary source, while the legal eagles would write about their own corporation, and be primary sources. Independent of their reliability (though we assume both groups are reliable), the latter would generally be used only to help describe the corporation itself, while the former could be used as a general source. Crum375 (talk) 23:53, 25 May 2010 (UTC)

Eyewitness reports are always primary sources, even if they are written by professional journalists, and this section makes no distinction, or even any mention at all, of whether a self-published source is primary or secondary. WhatamIdoing (talk) 00:26, 26 May 2010 (UTC)

Yes, if a journalist gives an eyewitness report, or tells about his family, that would be primary. This is why I used the word "normally", in "normally journalists would write a story about a third party." And we don't tell you if a self-published source is primary or secondary because it can be either one, and in general it has no effect on its reliability, since they are normally not accepted except to describe themselves (where they would typically, but not always, be primary). There is more detail about primary vs. secondary in WP:PSTS. Crum375 (talk) 00:41, 26 May 2010 (UTC)

Break 2

WI, grandma's blog is self published, but may well be secondary, if she describes things independent of herself. Reliability is independent of source primacy: a primary source can be highly reliable, like a crime lab result, while a secondary source can be highly unreliable, like grandma's blog. Crum375 (talk) 22:01, 25 May 2010 (UTC)

It's true that a blog could be a secondary source, but I have specified Grandma's posting about her own house burning down, and posting about an experience that directly happened to you is always a primary source. WhatamIdoing (talk) 22:04, 25 May 2010 (UTC)

Perhaps this dispute has to do with the fact that we use the notion "self-published" for at least two purposes, and in the case of large organisations publishing about themselves it makes sense to consider them as self-published for one of them, but not for the other:

• The stricter rules that govern self-published material are a mechanism for discarding irrelevant stuff: Things that nobody wants to read but which the author absolutely wants published, for whatever reason. Such authors and their few loyal fans have a tendency to appear on Wikipedia to promote these sources. To keep Wikipedia clean, we try to minimise our use of such stuff.
• When something has gone through a third-party publisher there is a presumption that usually there aren't any really bad POV excesses. Reasons include an editor who takes care to keep the quality of the publisher's publications high; that authors who get formally published are generally of higher quality; a higher degree of self-control when writing for a large audience; and that pure advertising usually doesn't get published in this way.

When the New York Times writes about itself we can assume that the first point doesn't apply. The second point actually depends on the media they are using: An article about the NYT in the NYT will probably be written to the newspaper's normal standards and therefore should not be regarded as self-published in either way. (Or if it is, there should be a common sense exception for such sources.) When the NYT self-publishes a book about itself it's a different matter. In this case there is a good chance that it crosses the line to advertising. The first problem does not apply in this case, but part of the second does. The situation isn't very different from a self-published paper by a respected scientist: Many of the problems of self-publishing don't apply, but we still consider it self-published.

Like almost everything, whether something is self-published or not is not purely a matter of black or white, or even of shades of grey. The world is full of colours.

There can be communication problems if one editor argues from the plain meaning of the word "self-published", one argues from its wiki meaning as literally defined in policies/guidelines, and yet another argues from the spirit of the policies/guidelines. All these approaches are valid, and they may lead to opposite results. Hans Adler 22:43, 25 May 2010 (UTC)

I'd like to refine your example: When Arthur Ochs Sulzberger, Jr. (the publisher) runs a story in The New York Times (the publication) about The New York Times Company (the owner), the first point (probably) doesn't apply. When that corporation publishes its advertising rate card, or an advertisement that is attempting to attract subscribers, then the first point does apply. WhatamIdoing (talk) 23:05, 25 May 2010 (UTC)

Self-published acceptable for "existence" or attribution?

The SP policy says that the rationale is that anyone can easily self-publish and "claim to be an expert in a certain field". But what if the work is referenced not to substantiate a fact, but rather to demonstrate the existence of an opinion?

For example, I publish a website critical of a cult I used to belong to, which includes original source documents (MSM excerpts, advertisements the group published, etc.), as well as the stories of other former members. I think that the site would be a suitable reference for WP article statements such as, "Former group member Michael Bluejay now publishes a website critical of the group, asserting that it is actually a mind-control cult", or "Former members of the group now say that the group operates as a mind-control cult." The site isn't used to substantiate some special fact, only to show the existence of a claim being made. So I don't think this use goes against the intent of the policy, since the intent of the policy is to prevent using self-published sources to justify facts, not the existence of claims.

Taking this a step further, I'm hoping to quote another former member who described his cult experience as a factor in the failure of his marriage. Here again, the article wouldn't claim as a fact that ex-members' marriages failed as a result or their being involved in the group, only that they made that claim. Is the site acceptable to show that the claim has been made?

Incidentally, the overwhelming majority of sources we're using for the article aren't self-published, but in a couple of cases my site is the only source available for the bit in question. I know my site isn't not the best source, but I think it's better than nothing. The quote above by the former member about his marriage failure hasn't been published anywhere else. Hence my problem.

I'm keen on hearing others' thoughts about all this. Thanks, MichaelBluejay (talk) 07:11, 26 May 2010 (UTC)

My own personal view is that curent policy on self-published sources is fundamentally flawed. The idea that "Self-published material may in some circumstances be acceptable" conflicts with the specific issue of independence, which I have set out above[15]. If the publisher and the author and one of the same person, then this source of information can never be classed as reliable, in the same way that I am writing here on this talk page cannot be classed as being reliable either. A source without some form of oversight just does not operate the weakest forms of checks or balances that provide a modicum of assurance that the source can be relied upon. I think that a source that is self-published is just one form of source which is not indpendent, because the author is the same person as the publisher. --Gavin Collins (talk|contribs) 08:20, 26 May 2010 (UTC)

Thanks, but I'm not sure you're listening to me. First, I gave an instance in which the publisher is *not* the author, but more importantly, you seemed to ignore my whole point that using a source to show the *existence* of thought (the source *itself* is such thought) seems to be quite a different matter from relying on a source to justify *facts*. Agreed that self-published sources are usually unreliable in the latter case, and the policy says as much, but doesn't really address the former. That's what I'm seeking some feedback on. MichaelBluejay (talk) 16:10, 26 May 2010 (UTC)

Really, Gavin? You're absolutely sure that Barack Obama's political website cannot possibly, under any circumstances, be a reliable source of the fact that Obama ran for President, or that his slogan was Yes we can? Is this on the theory that 100% of campaign staffers with access to the website are too stupid to know their own slogan? WhatamIdoing (talk) 17:22, 26 May 2010 (UTC)

Really, WhatamIdoing. Remember Watergate. It seems to me that Presidents can be good or bad sources, but when they publish their own stuff on their own site, I would not trust them with a barge poll. --Gavin Collins (talk|contribs) 21:59, 26 May 2010 (UTC)

Okay guys, you're really hijacking my section here. I had some very specific questions that I opened up to discussion. I'm hoping to get some comments on them. Thanks, MichaelBluejay (talk) 02:03, 27 May 2010 (UTC)

The article seems to have multiple reliable sources. I wouldn't think your self-published site should be used. Due to the sorts of claims made on your site, I don't believe Wikipedia can link to it for any reason. Nothing against you, or your site, but I think that sort of thing would be a bit of a problem for Wikipedia for legal reasons. Have you been recognized by a reliable third-party as being knowledgeable on this subject? If so, maybe your site could be used, but I still wouldn't think so: not with the subject matter of your site.  Chickenmonkey  02:35, 27 May 2010 (UTC)

Well, I've been quoted in newspaper articles as a former member of the group who is now critical of it, and I think at least one of them has mentioned that I run the website. I have a hard time believing that WP could ever be held accountable for content generated by its users. Has that ever happened? Finally, I'm really hoping to see the issue I raised discussed in a broad context, not just about my own website. That is, can a self-published site be a good source to show the *existence* of thought (i.e., the source *itself* is the evidence of that thought, inherently)? I know that a self-published source is usually insufficient to back statements of *fact*. What I'm asking about is entirely, completely different. MichaelBluejay (talk) 05:24, 27 May 2010 (UTC)

I think it can be acceptable. The problem is showing that this opinion is WP:DUE, not that the opinion exists. WhatamIdoing (talk) 05:29, 27 May 2010 (UTC)

I'm quite sure using a self-published website to show the existence of an opinion/thought is acceptable, but only if that site is reliable -- which would mean third-party sources have recognized the site as knowledgeable and reliable.

I don't know of any instances in particular where Wikipedia has been held accountable for content generated by its users, but potentially libelous material is advised against in WP:BLP -- which doesn't apply to groups, but I think specifically of Scientology in this particular situation. That doesn't mean we shouldn't address criticisms of groups. That just means we should take special care to ensure adherence to policies like WP: Undue weight, WP: Notability, WP: No original research, and this one WP: Verifiability. In situations like yours, where a group is being labeled a "cult", we must be especially careful to rely solely on very reliable sources.  Chickenmonkey  05:57, 27 May 2010 (UTC)

In any case you would have a WP:COI issue. I can't conceive of circumstances where a WP editor should be citing or discussing a website that he publishes, particularly one advocating a (self-professed) POV. Even making a talkpage argument for its citation seems pretty questionable. Leave it up to others to write that article. At the very most, I would ask other editors on the talkpage if they thought the source was useful, while being sure to disclose my conflict there.LeadSongDog come howl! 12:48, 27 May 2010 (UTC)

Thank you. We do have an admin leading the rewrite of the article and policing the sources. But anyway, again, I'm really hoping to see the issue I raised discussed in a broad context, not just about my own website. That is, can a self-published site be a good source to show the *existence* of thought (i.e., the source *itself* is the evidence of that thought, inherently)? I know that a self-published source is usually insufficient to back statements of *fact*. What I'm asking about is entirely, completely different. MichaelBluejay (talk) 05:04, 28 May 2010 (UTC)

To repeat what I said above, a self-published source can support the existence of an opinion, but it can't support adding that fact (=that the opinion exists) to an article.

Imagine that the world believes bread is a kind of food. Imagine that someone self-publishes a statement that he, personally, believes that bread is far better suited as paving material than as food. The mere existence of this self-published source does not give us a good reason to include this tiny-minority opinion in an article about bread. The statement is simultaneously WP:Verifiable and WP:Undue. WhatamIdoing (talk) 20:27, 4 June 2010 (UTC)

Correct... you know, this keeps coming up, both here and at WP:IRS. People keep asking if a self-published source can be reliable for a statement of opinion. The answer is yes (assuming the opinion holder is the author of the SPS)... But there is more to inclusion in Wikipedia than just verifiability and reliability. As WP:Undue makes clear... not every opinion is worth mentioning. I am thinking we may need to make this clearer in the policy. Blueboar (talk) 20:58, 4 June 2010 (UTC)

I disagree. Its not the source as a tiny-minority opinion that is the problem, it is the fact that the opinion has not been subject to fact checking or peer review that is the issue. Remember, the three issues regarding self-published sources are lack of editorial oversight, lack of independence, and lack of a "permanent" channel or medium of publication. Minority opinion/fringe theory is an entirely different issue. I think there is a lot of confusion surrounding the problems associated with self-published sources. --Gavin Collins (talk|contribs) 21:13, 4 June 2010 (UTC)

We would need fact checking and peer review if we presented the opinion as being accurate (or as being inaccurate)... but to simply state that the opinion exists (and that a given person has stated it) is not something that requires third party fact checking or peer review. A person's stated opinion could be full of factual errors, but it is still the opinion they hold. Your typical conspiracy theorist holds all sorts of nutty views... most of which will not hold up to fact checking and peer review. But a statement that simply says that he holds these views can be verified by looking at his conspiracy theory rant blog. As far as WP:V goes... the rant blog is a reliable, verifiable source for a statement as to what the views of the conspiracy theorist are, even if the views themselves are crap. This is one of those rare cases where a primary source is better than a secondary one.

To put this succinctly... a self-pubilshed source is fine for saying an opinion exists... it is not fine for saying anything else about the opinion. Whether to mention the opinion in a given article is not within the scope of this policy. Blueboar (talk) 21:34, 4 June 2010 (UTC)

Bottom line

The bottom line is that we can reasonably assume of our reliable sources that they have a professional structure in place for checking facts and legal issues before publication. They have professional people who are paid to say "no, don't publish this." That is absent with personal websites and blogs where an individual or a small unprofessional group is publishing straight to the website. That is what we mean by self-published: that no one stands between the writer and publication. There are no checks and balances. No one is paid to say no.

That is never the case with The New York Times or Coca Cola or the White House, whether they're writing about themselves or something else. When writing about themselves, those organizations are primary sources of information about themselves, but they are never self-published sources within the meaning of this policy. SlimVirgin ^{talk contribs} 17:45, 26 May 2010 (UTC)

Then we need to quit using the words "self-published" (meaning: the author is the publisher) and start saying what we mean, e.g., "Bureaucratic procedures improve reliability" or "If a corporate press release is vetted by the same number of humans as a story in The New York Times, then these sources should be considered equally reliable."

Right now, we're saying that we care whether the author is the same person as publisher on the page, but SV and Crum believe that we mean the number of humans involved in writing and publishing the source is the major issue. This is, at best, misleading. If you really believe that this is the point of this section, then I dare you to change it to say, in plain language, what it allegedly means. WhatamIdoing (talk) 18:38, 26 May 2010 (UTC)

The term "self-published" has been used in this policy to mean the same thing for years, and I haven't seen it cause confusion in articles, so I can't see any benefit in changing it. And the policy already says what you recommend it ought to say: "In general, the best sources have a professional structure in place for checking or analyzing facts, legal issues, evidence, and arguments; as a rule of thumb, the greater the degree of scrutiny given to these issues, the more reliable the source." SlimVirgin ^{talk contribs} 18:43, 26 May 2010 (UTC)

I've always intepreted the policy in the same way as WhatamIdoing. You can have 27 layers of vetting, but if the vetters belong to the same organization as the author, and they are vetting information about that organization, then I think that counts as self-published. Do we really trust a corporation to be objective or even 100% accurate in what information they post on their own website or in their own press releases? Heck no. A corporate website is a marketing tool, not a bastion of journalistic ethics. This doesn't exclude those websites from being used as sources in some cases; the corporation is clearly an expert on itself and thus some of the information could be appropriately included. Karanacs (talk) 19:11, 26 May 2010 (UTC)

I agree, Karanacs, but that just means they're primary sources. It doesn't make them self-published. There's no benefit in extending the definition of "self-published" to include anyone or any company writing about itself. And in any event when the White House, for example, writes about itself, it still has multiple layers of professional oversight to get through. Plus, we don't require of any of our sources that they be objective or 100 percent accurate. SlimVirgin ^{talk contribs} 22:02, 26 May 2010 (UTC)

If this is truly how Wikipedia defines "self-published":

By "self publishing" we mean individuals or small groups who have no or minimal oversight levels. Crum375 (talk) 18:19, 25 May 2010 (UTC)

The way Wikipedia defines "self-published" is absolutely correct. By "absolutely correct" I, of course, mean "completely wrong". See? I can misuse words, too (I apologize for being frank). If a corporation puts out a press release, that's self-published, because the entity itself wrote it. No matter how many vetting layers it has, all of those vetting layers exist within one entity. Whether something is "self-published" is completely separate from whether it can be used as a reliable source. If some actor publishes, on their website, the movies they've been in: that information is "self-published". That would also be a "primary source". Does that mean we inherently can't trust this information? No. If that same actor publishes, on their website, a blog that says "The director called me the best actor ever": that information is "self-published". That would also be a "secondary source". It's obviously not reliable. The same "self-published" source, both reliable at times and unreliable at others. With that said, if a corporation, on their website, publishes an indisputable fact (when they were founded, for instance), that's "self-published" and a "primary source". You get the point.

I do completely understand the intention of saying "we try not to use self-published material, such as blogs, etc, because they may not be reliable", but that has nothing to do with material being "self-published". At the very least, there needs to be a discussion on correcting how Wikipedia is telling its editors to interpret "self-published". Why would it be acceptable for Wikipedia to redefine a word?

As was said at WT:N:

There's no sense in a small number of editors making up its own definitions. These terms are used in a certain way in the academic world, so we should try to stick to that usage. It varies a little between subjects, but what's being suggested here is something I don't recognize at all. SlimVirgin ^{talk contribs} 17:32, 25 May 2010 (UTC)

By the way, "By 'self publishing' we mean individuals or small groups who have no or minimal oversight levels." So, how many oversight levels is enough? How many individuals is too many?  Chickenmonkey  20:36, 26 May 2010 (UTC)

Like everything in life, it comes down to editorial judgment and commons sense. The New York Times have many vetting layers, while Podunk Weekly may have just a couple, but the point is that when we say "self published" we mean the author can press a couple of buttons and have his output published. When it's not self published, it means the author needs to have several other people, ideally paid professionals, vet and/or approve his version before it's published. When legal vetting for potential liability is involved, it's even better. If it's a couple of guys and their dog publishing something, it's self published. If it's a big corporation, like the NYT or Coca Cola, it's not. In general, corporations are more reliable than a loose bunch of individuals, but we need to evaluate the number of vetting layers and their quality, and decide whether the author is able to just press the "Save" button and publish, or whether he's got a bunch of hoops to jump through. The former being "self published", the latter generally not. Crum375 (talk) 21:04, 26 May 2010 (UTC)

I should clarify, I'm not arguing that the policy should change. It reads plainly that "self-published media [...] are largely unacceptable." That's fine, because most "self-published" media shouldn't be accepted as reliable. What I am/was commenting on is the discussion that has taken place on this talk page where "self-published" has been misunderstood. That said, the phrase "what we mean is XXX" should not have to be used. If it does, the policy should probably be more clearly written: perhaps to include "corporate publications" as S Marshall has stated below. Currently the policy doesn't say not to use self-published sources, but that they are "largely unacceptable". It seems, and I may be wrong, that many editors believe this policy to be meaning that we should not use self-published sources.  Chickenmonkey  22:19, 26 May 2010 (UTC)

• "Published" is where someone independent of the author assumes the risk, responsibility and profit involved with releasing the material. "Self-published" is where the author takes the risk, responsibility and profit. "Corporate publication" is a separate category. It's like being "published" in that the author isn't the person who takes the risk—the author's boss does—but because the author and the person taking responsibility aren't independent of each other, it resembles self-publication as well.—S Marshall T/C 21:29, 26 May 2010 (UTC)

No, in the wiki meaning "independence" is not really an issue. You can have the chairman of the New York Time write an opinion column, and his words would be still vetted by other employees including corporate lawyers, before publication. They are clearly not "independent", but they are still different pairs of eyes, and they are paid to review and vet material, no matter whose, before it goes out. And that's the essence of "self publishing": if you can press a "Save" button and publish your material whenever you want to, it's "self published". If you need to run it by a bunch of people whose job it is to review publications before they go out, then it's not self published. Corporations are generally not self published, because they have professional vetting mechanisms in place. And it doesn't matter who is the "boss", only whether the vetting mechanisms are in place to correct mistakes and help reduce liability. Crum375 (talk) 22:12, 26 May 2010 (UTC)

S Marshall is right, and every single reliable source supports his definition. WhatamIdoing (talk) 23:14, 26 May 2010 (UTC)

(edit conflict)Well, Crum375's view has the benefit of being simple and easy to understand, but I think that when we're discussing verifiability in terms of the reliability of sources, "independence" is an issue. I think it's overly simplistic to disregard it. My position is that the trustworthiness of a source isn't a binary, yes-or-no thing. There's a continuum.

For example, let's imagine an article on a soft drink, where there's a dispute about the potential health effects of drinking it. The sources might be: (1) a full series of experiments by a professor at Oxford University, published by Oxford University Press; (2) a corporate press release by the manufacturer; and (3) a television interview with someone who got sick after drinking a can of Moxi-pop. Technically, all of these sources have been seen by someone who's paid to edit them, but you wouldn't give them equal weight, would you?

This is why I say it's about risk, responsibility and profit. The television interview gets least weight because the TV station is not to be held responsible for what the person being interviewed has said, even if there's been so much editing that the interview has been cut from half an hour to two minutes. The corporate press release gets the next most, but there's no independence and there's a profit motive. What you'd believe, and what you'd want the article to be based on, is the academic study.—S Marshall T/C 23:34, 26 May 2010 (UTC)

PS: I want to clarify that I'm not saying a corporate press-release is to be treated the same as a self-published book or a blog post. I don't think that's right. I think that for many matters, we can take what a large company says about itself at face value. If the company's website says it has 325 employees and a turnover of $40 million, then we can report that as fact.—S Marshall T/C 23:46, 26 May 2010 (UTC)

If you have a soft drink manufacturer writing about its soft drinks, it would be a primary source. Such sources, though often reliable, cannot be used to support contentious claims, and cannot be interpreted or analyzed, and otherwise must be used very cautiously, per WP:PSTS. We would normally require a secondary source, which by definition should not be involved with Coca Cola, to report on its health effects. Also, if it's a scientific issue, we'd prefer to rely on high quality peer-reviewed publications or mainstream media. Crum375 (talk) 23:52, 26 May 2010 (UTC)

Then I don't think we're disagreeing. :) Up until that post, you seemed to be saying that a press release was as reliable as an independently-published article, and I do understand why WhatamIdoing would want to challenge that.—S Marshall T/C 00:01, 27 May 2010 (UTC)

Reliability is not a blank check. A primary source can be very reliable to tell us the temperatures in Moscow over the last century, but not about their significance. A press release, which is also primary, can be very reliable to tell us that Corporation X has decided to add widget Y to its product line, but not to compare it to its competition, or evaluate its health effects. Crum375 (talk) 01:22, 27 May 2010 (UTC)

Bottom line, redux

Experienced Wikipedia editors and admins understand what we mean by the policy and believe it's sufficiently clear. Objections complaining about jargon or denotations of particular words are worthwhile to clear up individual confusion regarding points. However - now that we have cleared up those points, the policy stands as it has stood for some time. It has withstood the test of time and multiple challenges by those seeking to abuse the site in some manner or another.

All Wikipedia policy is subject to ongoing review and evolution, but the proper venue for this sort of review is the Village Pump. I predict that a discussion there will with near unanimity support the existing policy, but anyone who believes that it's fundamentally flawed should feel welcome to take it up there and attempt to change people's minds. Georgewilliamherbert (talk) 21:35, 26 May 2010 (UTC)

I think it is flawed. We refer to sources as being self-published, but what we really mean is that the source is not independent. I think this is the venue to discuss this issue, as I feel that if we have make it clear what independence is, then this policy will be clearer as to what is a self-published source, and why they are not considered to be reliable. --Gavin Collins (talk|contribs) 22:17, 26 May 2010 (UTC)

No, we don't mean by "self published" not independent. See also my reply above. What we mean by "self published" is that the author can just press a "Save" button and publish his stuff whenever he wants. Once you introduce a professional vetting layer, and legal liability scrutiny, by multiple people, and every author (even "the boss") needs to jump through these hoops before his stuff is published, then it's not self published. Crum375 (talk) 22:25, 26 May 2010 (UTC)

Which is to say that Wikipedia has invented its very own definition of "self-published", that has absolutely nothing to do with whether the author is the publisher, but instead is being used as a secret code word for "big organization" or "editorial control". This purported definition of "self-published" directly contradicts every reliable source and introduces concepts totally outside reliable sources.

I don't mind praising editorial control on this page. I just don't want editorial control to be spelled s-e-l-f-p-u-b-l-i-s-h-e-d, to the predictable confusion of every editor who isn't in on the joke. WhatamIdoing (talk) 22:37, 26 May 2010 (UTC)

I still can't see the point of wanting to change the way we use "self-published." Sources writing about themselves are primary sources, and that somewhat limits the ways in which they may be used. What is the benefit of attaching a new term to them? SlimVirgin ^{talk contribs} 22:44, 26 May 2010 (UTC)

The issue of independence is broader than whether a source is self-published or not. Sometimes groups of commentators can club together to give the impression that their views are independent, say by starting a magazine that endoreses their viewpoint. A good example of this is the article Socionomics; the sources for this turn of phrase are basically a group of stock market analysts trying to promote their proprietary brand of predicting new stock market trends, despite the fact that accepted economic theory (e.g. Random walk hypothesis) says you can't. Its an article topic that has been deleted several times, but basically because it fails the independence test as it is defined in WP:N. Now where does that test come from? All of the other principles contained in the notability guideline originate from one or more of Wikipedia's content polices. I think independence is a topic that has been omitted from WP:V, or is not set out explicitly, and has become subsumed by WP:SPS. Perhaps I am completely wrong on this issue, but this is what I suspect to be the case. --Gavin Collins (talk|contribs) 22:59, 26 May 2010 (UTC)

Gavin, my question is why the concept of a primary source isn't enough. A primary source is an involved source. I'm wondering why that isn't sufficient for the purpose of judging independence. SlimVirgin ^{talk contribs} 23:53, 26 May 2010 (UTC)

You can have a bunch of people get together and produce a wonderful, accurate, highly reliable and reputable product, or they can produce pure junk. This is why we specify that a reliable source must have a reputation for accuracy and fact checking. If the reputation is that a source produces junk, then it's not reliable, regardless of whether it's self published or not. Crum375 (talk) 23:10, 26 May 2010 (UTC)

The point? How about so that it's intelligible to people who don't already know the secret code words? How about so we live up to your (quoted above) claim that Wikipedia shouldn't make up its own definitions of basic terms? So that Wikipedia's policies don't look like they were written by (or interpreted by) people too stupid to figure out a dictionary definition? So that editors won't keep saying "it's self-published" when they actually mean "it's not independent" or "there's no editorial control"? So that Wikipedians won't waste time with conversations like "That's self-published" — "No, it's not, the author is X and the publisher is Y" — "No, I mean that it's 'WikiJargonSelfPublished', which the real world calls 'written by one person and published by another person, but there aren't very many humans altogether'".

(Gavin is right that self-publications can be independent, and non-self-publications may be non-independent.) WhatamIdoing (talk) 23:09, 26 May 2010 (UTC)

Could you stop being so rude, please? We have our own use of lots of terms on WP. Original research, for example. Verifiability is another example. When I cautioned against making up terms I was talking about primary/secondary sources. As for self-published, the White House would laugh to hear themselves called self-published. They wouldn't laugh to hear a blog called that. I think our use of it is quite normal. But I've completely lost track of what you're arguing or trying to achieve. SlimVirgin ^{talk contribs} 23:57, 26 May 2010 (UTC)

There are no "code words": self publishing means the author presses a "Save" button and publishes the material himself. If he has to run it by multiple vetting layers before it can be published, then it's no longer "self" published. No codes. Crum375 (talk) 23:13, 26 May 2010 (UTC)

Every single reliable source disagrees with your made-up definition. WhatamIdoing (talk) 23:15, 26 May 2010 (UTC)

Here is the very first definition Google came up with for "'self publishing' definition": "Definition # The simplest definition of self-publishing is when an author produces and publicizes his own book for public consumption." (on www.ehow.com/facts_5017513_definition-selfpublishing.html) Crum375 (talk) 23:27, 26 May 2010 (UTC)

And you'll note that there's not one word in that definition that says "unless he had a dozen lawyers sign off on it" or "except when a corporate author is involved" or "except when there are multiple vetting layers" or any of the other unsupportable claims you've been adding to it. An author who runs his publication through ten million vetting layers, and publishes it himself, is still self-published -- according to the very definition that you quote. WhatamIdoing (talk) 23:33, 26 May 2010 (UTC)

Not at all. If you work for the New York Times, write an article with some other reporters, and have your corporation publish it, it is not self published. That's because the article would have to go through vetting layers and probably legal scrutiny before it can go out. Crum375 (talk) 23:42, 26 May 2010 (UTC)

Okay, let's go back to a more basic concept. Do you know what "publication" means? It means "producing and distributing" materials (publicity, mentioned in your definition, is helpful to the distribution, but not actually essential to the act).

Do we agree that if the author is the publisher, it is self-published — full stop, no exceptions, no further conditions, that whenever the author=publisher, then the material is self-published? WhatamIdoing (talk) 23:47, 26 May 2010 (UTC)

The words "author", "publishing", "publication" and "publisher" mean different things in different contexts. This is why we should focus on the process. If the human individual writing a document can press a button at will, and distribute the document to the public, then it's "self published". If the document must be scrutinized by different professionals before it can go out, then it's not self published. Trying to make this more complicated than this will only add confusion and gain nothing. Crum375 (talk) 23:59, 26 May 2010 (UTC)

The process is spelled e-d-i-t-o-r-i-a-l o-v-e-r-s-i-g-h-t. If that's what you actually want to address in this particular section on this page, then you need to "correct your spelling", as it were.

NB that I (strongly) favor editorial oversight; I just don't agree that it is spelled anything like s-e-l-f-p-u-b-l-i-c-a-t-i-o-n. WhatamIdoing (talk) 00:10, 27 May 2010 (UTC)

I don't entirely agree, no. There are exceptions. For example, if The Times publishes an article that's about itself, then that's "published", not "self-published". And besides, it doesn't seem to address the basic division between you two. I think Crum375's view is about whether the author and the publisher are the same person, and I think your view is about whether the author and the publisher are the same corporation.—S Marshall T/C 23:55, 26 May 2010 (UTC)

The publication does not publish itself.

I agree that Crum is having difficulties with the concept of corporate authorship, but he is also apparently convinced that "did the same person/entity write it and publish it" is the same issue as "was there any editorial oversight"? WhatamIdoing (talk) 00:10, 27 May 2010 (UTC)

Oh, I'm sure Crum375 understands corporate authorship. I'm concerned that both of you have at times seemed to be taking binary, on-off positions, where one was saying "Corporate authorship isn't self-publication!" and the other was saying "Corporate authorship is self-publication!" Neither of those are true. Corporate authorship is corporate authorship, and it has its own place on the hierarchy of trustworthiness. It's above "self-published" but below "published by a reputable, independent source".—S Marshall T/C 00:20, 27 May 2010 (UTC)

Crum's comments above indicate rather plainly that he believes that the author of a press release written as a work for hire is the employee rather than the employer. While this is true in a biological sense, it is absolutely untrue in a legal/copyright sense. WhatamIdoing (talk) 00:27, 27 May 2010 (UTC)

Well, I'm not concerned about that. What matters to me is how to evaluate the sources I see, and I don't pretend a corporate press release is "self-published" or "published". It's neither; in terms of trustworthiness, a corporate document doesn't fit in either of those two slots. It's a different thing that editors need to think about in a different way.—S Marshall T/C 00:37, 27 May 2010 (UTC)

A press release by a corporation is not SPS, because its author cannot send it to the public unless it undergoes professional scrutiny by multiple layers. It is generally a reliable source for WP purposes, but almost always a primary source, because it discusses the corporation in which the author is involved. And primary sources can only be used in limited ways and very carefully, per WP:PSTS. Crum375 (talk) 00:46, 27 May 2010 (UTC)

I think there are more shades of grey than that, and for me, it rather depends on the corporation. I'm sure a press release by MacDonalds undergoes professional scrutiny, but I wouldn't want to see a press release by Mrs Miggins' Pie Shop treated in quite the same way!—S Marshall T/C 00:52, 27 May 2010 (UTC)

It all depends on the topic. Mrs Miggins' press release announcing her latest Apple Pies would be a very reliable source for a description of her products, while MacDonald's would be OK for their latest menu addition. Neither one would be considered a reliable source for competing products. Crum375 (talk) 00:59, 27 May 2010 (UTC)

Well, maybe not always "very reliable" for a description of the products. "Mrs Miggins sells apple pies," fine. "Mrs Miggins sells the most popular apple pies in Oklahoma," I would not think is fine. But I'd believe MacDonalds if they said "MacDonalds sells the most popular burgers in the United States." Essentially, I think some press releases are more trustworthy than others.—S Marshall T/C 01:29, 27 May 2010 (UTC)

You may believe MacDonalds, but WP would require a better source for claims relating to competition. But we could always use in-text attribution and say, "According to MacDonald's, they sell more burgers than Burger King." The same for Miggins if she says she sells more than the store across the street. Bottom line: press releases by corporations are generally reliable primary sources, and generally to describe themselves. Crum375 (talk) 01:54, 27 May 2010 (UTC)

Again, I'm concerned that this is too simplistic and in fact I think it contradicts your earlier position, which was that verifiability is enhanced by the number of people paid to check the content. MacDonalds' press release would've passed before a lot of editorial eyes, but Mrs Miggins' one might've been written by the sales manager and sent to the newspaper. When it comes to verifiability, not all corporate sources are equal.—S Marshall T/C 09:46, 27 May 2010 (UTC)

Yes, you are correct that source "reliability" is not black and white, but it's also very much a function of the statement we are trying to support. There is no "absolute reliability", and we expect sources to be more knowledgeable than others about topics close them, although less objective. So corporations can tell us how very reliably how many types of widgets they sell, but not how good they are to your health, or how they stack up compared to the competition. And although a bigger corporation has more vetting layers of higher quality than the smaller ones, when it comes to their own products they would be likely be equally accurate, because the small corporation typically has a simpler product line and less information to process, so things may balance out. Crum375 (talk) 11:17, 27 May 2010 (UTC)

Let's all agree that the trustworthiness of a source is a continuum rather than a yes/no issue on whether it's self-published, and move on from that. On the specific example of the health effects of a product, the manufacturer can be very reliable. For example, here in the UK, thanks to COSHH, many manufacturers have to issue product safety data sheets, and a COSHH sheet for Portland cement is a highly reliable source for the health consequences of handling it even if it comes from the manufacturer. But a press release from the same maker would not be so reliable, because the press release doesn't have to adhere to the same exacting standards.—S Marshall T/C 12:04, 27 May 2010 (UTC)

Yes, source reliability is a continuum, as you say, and some are more reliable than others for the same information. This is where we have to use our editorial judgment and decide (by consensus) for each particular situation which source is best, among potentially many which are at least minimally "reliable". Crum375 (talk) 12:13, 27 May 2010 (UTC)

Specific proposal

I propose adding the following text to the top of WP:SPS:

Self-publication is the publication of a work by its author, without the involvement of an established, third-party publisher. This includes any and all individuals, small groups, and corporate authors who publish their own works on paper, electronically, or in any other media form, so long as the author is also the publisher.

Self-published and non-self-published sources may or may not be independent of the subject.^[1] Self-published and non-self-published sources can be primary, secondary, or tertiary sources.^[2]

1. ^ Examples: A book that is both written and published by a historian about the Roman Empire is an independent, self-published source. Memoirs written by a retired politician and published by a major publishing house is a non-independent, non-self-published source.
2. ^ Examples: A blog posting about a house fire, written by the person whose house burned down, is a primary, self-published source. A newspaper story about the same fire, written by a reporter on the scene, is a primary, non-self-published source.

I propose keeping all of the other text in the section the same (except, possibly, cleaning up the comma splice in the first sentence). This does not change the policy; it only provides a basic, verifiable (e.g., [16][17][18]) definition of what a self-published source is and dispels some unverifiable myths whose existence is amply proven by this discussion. WhatamIdoing (talk) 23:10, 26 May 2010 (UTC)

Your definition would rule out the New York Times as a source, since it is a corporation which publishes material on its own. Crum375 (talk) 23:33, 26 May 2010 (UTC)

No, it wouldn't. The New York Times is published by Arthur Ochs Sulzberger, Jr. — not by itself, not by its journalists, not by its owner, not by its editors — and that is practically the definition of an "established publisher", which the above directly names as a criterion for non-self-published materials. WhatamIdoing (talk) 23:36, 26 May 2010 (UTC)

The New York Times is a corporation which publishes its own material. It creates the content, vets it, and publishes it. According to your proposal, it would be a "self publisher" and therefore unreliable. Crum375 (talk) 23:45, 26 May 2010 (UTC)

You're missing some critical subtelties: The New York Times Company (the corporation) publishes its own materials — advertising rate cards, for example. It does not publish The New York Times (the paper). It owns the paper. The publisher is Arthur Ochs Sulzberger, Jr., and nothing in The New York Times (paper) except his rare publisher editorials is both written and published by him (and therefore self-published). WhatamIdoing (talk) 23:57, 26 May 2010 (UTC)

You're trying to introduce unnecessary confusion. Again I have to ask: what would be the benefit of the change in terms of how we approach sources in articles? SlimVirgin ^{talk contribs} 23:59, 26 May 2010 (UTC)

No, I'm trying to get editors to quit telling me that a source written by X and published by Y is self-published, and conversely that a source written and published by exactly the same entity is non-self-published if enough lawyers were involved. Editorial oversight is not the same as non-self-publication. WhatamIdoing (talk) 00:06, 27 May 2010 (UTC)

If a source written by X and published by Y is known or expected to have multiple vetting layers, or scrutiny by paid professionals who check it for accuracy and potential liability before it's able to go out to the public, then it's not self published. If author X is able to just press a button and send his material to the public, then it is self published. That's the basic framework. Use your common sense for in-between cases, where the guiding principle is how many vetting layers, including legal scrutiny, exist between the original author and the published document. Crum375 (talk) 00:14, 27 May 2010 (UTC)

Show me the reliable source that adds anything like your "guiding principle". Just show me a dictionary definition that says "Anything with editorial oversight or layers of vetting can't be self-published, even if the author is exactly the same as the publisher." Just one reliable source that says "Whether a publication is self-published is determined by how many vetting layers, including legal scutiny" are involved. Just one decent source -- that's all I need. WhatamIdoing (talk) 00:18, 27 May 2010 (UTC)

I quoted you a source above, which says "The simplest definition of self-publishing is when an author produces and publicizes his own book for public consumption." There are two separate issues here, self publishing and reliability, which are very related. The more vetting layers, the more we consider a source "reliable" for WP purposes. But to be not self published, you just need to have "some" editorial oversight, so you can't just "produce and publicize your own book for public consumption", per the above definition. Not being self published does not make you a reliable source, but it is a step in the right direction. Crum375 (talk) 00:31, 27 May 2010 (UTC)

Nope, I'm not finding the word "vetting" in your definition. Do you? Do you see "produces and publicizes his own book for public consumption without editorial oversight" in the definition? Or does the sentence stop without mentioning this concept? WhatamIdoing (talk) 00:49, 27 May 2010 (UTC)

Vetting means that other professionals are involved in scrutinizing the product for liability, accuracy, and other issues. If that exists, then by definition it won't be an "author [who] produces and publicizes his own book for public consumption". Professional vetting and legal scrutiny by others prior to publication is clearly the opposite of producing and publicizing your own material, i.e. not SPS. Crum375 (talk) 01:05, 27 May 2010 (UTC)

I said, "show me a source," not "draw your own, unsupported conclusions based on your assumption of the most common^{[according to whom?]} circumstance that results in self-publication". As far as I can tell, your source does not say a single word about editorial oversight. It does not make an exception for the author hired twenty lawyers to vet the manuscript, even if he took every single bit of their advice. I see nothing that says "produces and publicizes his own material except when the author is a forty-person committee in a large corporation with a formal structure that requires six vice presidents to sign off before publication".

What I see is "an author [who] produces and publicizes his own book for public consumption" -- end of sentence, no further conditions, no exceptions.

Let me be clearer: You apparently believe that "A source is not self-published, even if the author and the publisher are the same individual human, so long as enough paid professionals have vetted the manuscript". Show me a reliable source that says this plainly and directly enough to include that statement, in Self-publishing -- without violating WP:NOR.

(I believe that no such source exists, because nobody except you defines self-publishing this way, but I'm willing to change my mind if you can produce a source.) WhatamIdoing (talk) 01:17, 27 May 2010 (UTC)

You said, "You apparently believe that 'A source is not self-published, even if the author and the publisher are the same individual human, so long as enough paid professionals have vetted the manuscript'": Correct. If the publisher of the NYT writes an editorial in his paper, he will have lots of paid professionals vetting his manuscript, so he will not be publishing his own material by himself. And because he doesn't do it himself, he is not "self published", and neither is any other reporter or columnist on his payroll. They must all have their manuscripts undergo professional scrutiny by others before publication, so they are not "self published". Again, we use the term "self published" on WP in its most common^{[according to whom?]} sense, which corresponds to the common definition I linked to above, i.e. material published by the author without being screened by professionals prior to publication. Crum375 (talk) 01:33, 27 May 2010 (UTC)

(edit conflict) I'm sorry, but that makes no sense, to me. The amount of vetting affects the reliability of a source, as you've said. It has no bearing on whether the source is self-published.  Chickenmonkey  01:48, 27 May 2010 (UTC)

The amount of vetting affects the reliability, but essentially any vetting by paid professionals means the author is no longer doing the entire publication process by himself, and according to the definition I provided above, the source is no longer an SPS. Crum375 (talk) 02:01, 27 May 2010 (UTC)

So where's your source? If your additional restriction really is "the most common^{[according to whom?]} sense", then why is it not defined that way in any reliable source? If your made-up insertion of editorial oversight is "the most common^{[according to whom?]} sense", then why doesn't even one single reliable source mention that issue? Why do you keep claiming that the definition linked above includes language like "without being screened by professionals prior to publication", when nobody can actually find words like that in the source you're supposedly citing? WhatamIdoing (talk) 01:43, 27 May 2010 (UTC)

As I mentioned above, we use the term SPS in its most common^{[according to whom?]} meaning, which is an author doing the entire publication process on his own. I provided a link to that definition above. Therefore, once you introduce additional scrutiny and vetting layers, that author is no longer doing the entire publication process on his own, hence that is no longer an SPS. Thus the NYT is not an SPS, as expected. Not sure what else you want. Crum375 (talk) 01:59, 27 May 2010 (UTC)

I want a source that says your bit about "the entire publication process on his own". Your source doesn't say that. WhatamIdoing (talk) 02:10, 27 May 2010 (UTC)

My source says the "author produces and publicizes his own book for public consumption". You seem to read into it that it can include an entire professional organization which helps produce that material. But by your reading, it would make the NYT "self published". So clearly your reading is incorrect, and "author produces" means the author does it, not others. That is also the common understanding of every person: "self published" book means you write it up, and print it, without letting professionals screen it for accuracy or legal issues. Crum375 (talk) 02:58, 27 May 2010 (UTC)

---

For one, Self-published does not equal unreliable. For two, The New York Times is a publication which publishes the work of reputable journalists reporting on third-party information. If The New York Times published a piece on how great itself is, that would be of questionable reliability, right?  Chickenmonkey  23:58, 26 May 2010 (UTC)

No. On WP "self published" means an individual author who can press a button at will and distribute his material to the public. The New York Times is not "self published" by this definition, because we focus on the individual, not the corporation. And when the NYT writes about itself it becomes a primary source for that material; still reliable, but must be used very cautiously and in a restricted fashion. Crum375 (talk) 00:04, 27 May 2010 (UTC)

You keep saying "on Wikipedia". Why should Wikipedia redefine words? I agree that The New York Times is not "self published", but for apparently different reasons. I understand primary, secondary, and tertiary sources.  Chickenmonkey  00:09, 27 May 2010 (UTC)

Wikipedia does not redefine words. But some common words, which can have multiple meaning, are defined on WP so we can all be on the same page. Crum375 (talk) 00:16, 27 May 2010 (UTC)

It makes sense, if there are multiple meanings to a word, for Wikipedia to dictate which definition it is going to adhere to. That's true. However, the definition you're offering is one entirely created by some unknown entity, because WP:V does not contain it. Reading WP:V, I am entirely fine with how it is currently written. Where is this definition written that you keep quoting? I feel like I'm arguing, and I don't want it to seem like I am. So, if it does seem that way, I apologize ahead of time. In my opinion, I am on the same page as Wikipedia policy and you're quoting something I've never heard before.  Chickenmonkey  00:31, 27 May 2010 (UTC)

We define reliable sources by saying in WP:V, "In general, the best sources have a professional structure in place for checking or analyzing facts, legal issues, evidence, and arguments; as a rule of thumb, the greater the degree of scrutiny given to these issues, the more reliable the source." This tells us that reliable sources have in place multiple layers of vetting, for legal issues, technical accuracy, etc. The WP:SPS section tells us "Anyone can create a website or pay to have a book published, then claim to be an expert in a certain field. For that reason self-published media—including but not limited to books, newsletters, personal websites, open wikis, personal or group blogs, Internet forum postings, and tweets—are largely not acceptable." Clearly it focuses on "self published" as material published directly by the authors without professional vetting layers. SPS and RS are related, though not one-to-one, in that SPS is generally not RS, except about the author. But being non SPS does not make a source automatically reliable. Crum375 (talk) 00:40, 27 May 2010 (UTC)

That is, exactly, what it says. "The best sources" not "The only sources we can use". Also, "including but not limited to", means that is not the complete listing of materials that fall under "self-published". Yes it focuses on these materials, but it states that there are other materials that also fall under "self-published". I'm not concerned with your definition of "self-published" because I feel WP:V states it fine. Can we at least agree that just because a source is "self-published" (under either definition) doesn't necessarily mean it isn't reliable?  Chickenmonkey  01:01, 27 May 2010 (UTC)

As long as we agree that SPS is a source published by an author with minimal or no editorial oversight or legal screening, yes, it can be used as a reliable source to describe the author himself, per WP:SPS. Crum375 (talk) 01:16, 27 May 2010 (UTC)

A self-published source is a source published by the author of said source. We can agree to disagree on that. Also, per WP:SPS, a self-published source can be used as a reliable source on a subject upon which the author is proven to be knowledgeable, except with WP:BLP. I believe this discussion, or at least my part in it, as reached an end. Good day.  Chickenmonkey  01:30, 27 May 2010 (UTC)

Importantly, the quotation that Crum provides here is not in WP:SPS. It's in WP:SOURCES, and it is equally true for both self- and non-self-published sources. SOURCES does not say, "Non-self-published sources have a professional structure in place for checking or analyzing facts...": It applies to both self-published and non-self-published sources; to primary, secondary, and tertiary sources; to independent and non-independent sources. This is a description of editorial oversight. It is not a statement about whether the author and publisher are the same. (And again: I fully support the use of sources with editorial oversight. The issues here is whether Wikipedia is going to make up a definition of "self-published" as meaning "without editorial oversight" instead of following the reliable sources. WhatamIdoing (talk) 01:08, 27 May 2010 (UTC)

There are no outside sources which tell WP how to formulate its policies. We use "self published" in the most common^{[according to whom?]} sense: a source which is typically an author publishing his own material, which is an individual pressing a 'Save' button to publish his stuff, with no additional layers of legal scrutiny or accuracy checking by professionals. And as such, an SPS is unreliable, per WP:SOURCES and WP:SPS, except as a possible source about the author himself. Crum375 (talk) 01:16, 27 May 2010 (UTC)

My definition includes your scenario; it also includes the scenario in which an individual presses the 'Save' button to publish his stuff, with dozens of layers of legal scrutiny, accuracy checking by professionals, an imprimatur from the Pope himself, and an endorsement from the Dalai Lama. My definition includes this scenario because all of the reliable sources include this scenario, because the common-sense definition includes this scenario, and because conflating issues self-publication with editorial oversight (or primary/secondary/tertiary status; or independence from the subject matter) creates confusion among editors who are told "press releases aren't self-published" when they most certainly are (and thus, for example, are ineligible under WP:N). WhatamIdoing (talk) 01:23, 27 May 2010 (UTC)

A press release by a corporation is very definitely not "self published" by the common definition, which I linked to above, and if your definition makes it so, then your definition is plainly wrong. A press release by a corporation undergoes serious vetting by multiple paid individuals, as well as legal counsel, before it goes out. This is definitely not "self publishing" by the common or WP definition, which says that "self publishing" is an author publishing his material "by himself". In the press release of a corporation, there is no single individual who makes up a press release and presses a button to publish it, any more than the NYT publisher can write his editorial that way. Corporations screen their output through multiple layers and therefore that's not "self publishing", not the NYT, and not Coca Cola. Crum375 (talk) 02:12, 27 May 2010 (UTC)

Are you aware that the definition you link above is entirely silent on the issue of "serious vetting by multiple paid individuals, as well as legal counsel"? That those words really, truly do not appear in the definition you cite? That there is, in fact, absolutely no prohibition in that definition against the self-published author hiring large team of paid individuals and legal counsel? WhatamIdoing (talk) 02:27, 27 May 2010 (UTC)

The definition says "The simplest definition of self-publishing is when an author produces and publicizes his own book for public consumption." If there is a professional organization which vets the material, checking it for accuracy and screening it for legal issues, then the author is not producing his own book. Other professional people are doing it. Those other professionals are what makes the NYT not a "self published source". Crum375 (talk) 02:34, 27 May 2010 (UTC)

Your cited definition says the self-published "author produces and publicizes his own book". It does not say that the "author produces and publicizes his own book all by himself".

There's nothing in your cited definition that prohibits the author from hiring a proofreader to read the book, an artist to design the cover, a printer to print the book, a binder to bind the book, a magazine to tell readers about the book — not one word about these things. Your definition says, "author produces and publicizes his own book" and stops. WhatamIdoing (talk) 03:03, 27 May 2010 (UTC)

As I tried to explain to you, if you use your interpretation that "author produces" means "author plus a professional organization which screen the material produce", it would include the NYT as "self published source". Ergo, your interpretation is wrong, and "author produces" means what it says, and nothing more. Crum375 (talk) 03:18, 27 May 2010 (UTC)

If you're right, then I'm sure you'll have no trouble showing me a reliable source that says self-publication is "author produces and publicizes his own book, but not when they author plus a professional organization which screen the material produces and publicizes his own book". Any reliable source. One reliable source. All I need from you is exactly one, single, solitary source that doesn't stop dead when it gets to the end of the plain statement that "author publishes his own work," but keeps going on with your caveats about professionals or layers or vettings.

And if you cannot possibly find one single reliable source that says this, then perhaps you'll admit that your personal, private definition is neither verifiable nor the "common" definition. WhatamIdoing (talk) 03:48, 27 May 2010 (UTC)

There is no "personal private definition". There is the most common^{[according to whom?]} definition which I have linked to above, which says that SPS means "the author produces his own book." It is you who is trying to interpret "the author produces" to include an organization of professionals screening the book for accuracy and legal liability issues. This interpretation would make the NYT a "self published source", so it's clearly wrong. If you think it's not wrong, then you have to explain why. Crum375 (talk) 04:00, 27 May 2010 (UTC)

"The author produces his own book" does not preclude the author from hiring help. Is it your theory that if the author doesn't physically print the pages himself, or hires a proofreader, or asks his lawyer to look it over, that it quits being "his own book"? WhatamIdoing (talk) 04:27, 27 May 2010 (UTC)

If the author has a paid professional organization in place, then that combined mechanism, the author as the producer of the raw data and the other staff as the means for vetting it for accuracy and potential liability, stops being "the author produces his own book." It might be the author's company produces his own book, if he is the owner, but that's not the common definition of "self" as in "self published". And again, if we accept your interpretation that the definition of "self published" does allow for a professional vetting organization to be there along with the author, it would include the NYT as "self published", so clearly your interpretation defies comon sense. Crum375 (talk) 11:28, 27 May 2010 (UTC)

Is it your theory that the existence of a paid, professional staff makes the author not be the author, or that it makes the author not be the publisher?

Or perhaps I've misunderstood, and you think that the definition's use of the singular is the critical point? That is, that "the author publishes his own work" is self-publishing, and "the authors publish their own work" is non-self-publishing? WhatamIdoing (talk) 17:32, 27 May 2010 (UTC)

See my reply just below from 11:45, 27 May 2010 (UTC). Crum375 (talk) 17:42, 27 May 2010 (UTC)

I have not been finding myself to be lucky in interpreting your responses. Your response below, for example, says to me, "I have conflated self-publication with the absence of editorial oversight, and I think that WP:SPS says that reliability is sometimes low if the Library of Congress record has the same name in the "author" and "publisher" fields, but that it means that reliability is sometimes low if there is no editorial oversight".

(IMO, both are true on Wikipedia: self-publication is less reliable than non-self-publication, and publication (whether self- or non-self-) without editorial oversight is less reliable than publication with editorial oversight. IMO this page should have a "layers of vetting" section very similar to the "self-publication" section, and also an "independent from the subject matter" section -- but NB that they would be separate sections.)

Since I don't seem to be understanding your response, please consider answering exactly the questions I've asked:

Given that the author, if acting alone, would unquestionably be considered the publisher for my hypothetical source:

1. If a paid, professional staff is involved, does the author-publisher quit being the author?
2. If a paid, professional staff is involved, does the author-publisher quit being the publisher?
3. If multiple humans write a source, and exactly the same multiple humans publish it, is this maybe "selves-publication" or "non-self-publication" in your mind? WhatamIdoing (talk) 18:12, 27 May 2010 (UTC)

Here's a practical example: Kelly Link self-published the chapbook 4 Stories by founding Small Beer Press. Small Beer Press has an editor. Does that mean Kelly Link did not self-publish her chapbook?

Source: about.com  Chickenmonkey  04:40, 27 May 2010 (UTC)

When an author owns his own publishing company, we need to make a judgment call as to whether it is "self published". If it's a large company which publishes works other than those by the same author, then it's more likely to be not "self published". So the NYT is clearly not self published, even when the chairman writes an opinion piece in it. If it's a cover for a vanity press, then it's still self-published. What we are looking for are paid professional vetters, who can tell the author, "this fact is wrong," or "if we publish this, we'll get sued." If it's basically the author pushing all buttons and publishing whatever he wants, he is self published. Crum375 (talk) 11:45, 27 May 2010 (UTC)

The source I provided from about.com regards Kelly Link as "self-published". The New York Times is published by The New York Times Company. Therefore, no, it is not self-published. For that matter, the journalists working for The New York Times are not self-published either; they are published by Arthur Ochs Sulzberger, Jr.. It's a gray area, for sure. We, on Wikipedia, have to judge the journalistic integrity of such organizations as The New York Times, USA Today, and The National Enquirer, separately. None of them are "self-published", but have varying degrees of reliability. The same applies to materials that are "self-published", such as: corporate press releases, personal websites, and certain books. It doesn't matter if an editor regards a source as "self-published" or not, the same judgment of integrity is made. Therefore, you can continue believing your definition of "self-published" and I will continue believing my definition "self-published".  Chickenmonkey  18:39, 27 May 2010 (UTC)

The NYT and its various related companies are effectively one single entity: there is no one company which writes the material and another which vets it. There is one combined entity which writes, vets and publishes the material. This is true for most newspapers, and of course it's not "self published" in the WP sense. When we say "self published", we mean it in the "vanity press" sense, that an individual author can write a document, press a few buttons and send it out to the world, perhaps after paying some fee. If he has to go through an organized system of professional fact checking and legal scrutiny, including people who can tell him "no, this fact is wrong", or "if you publish this, we'll do under", it's no longer a "vanity press" or a "self publishing" operation. Corporate press releases, although normally primary sources, are never "self published", since they normally undergo professional vetting and legal scrutiny before they are released. The only definition of "self published" that clearly distinguishes a small newspaper and a vanity press operation is that "self published" means you as individual write the material, press some buttons and release the material yourself, with minimal or no interference. Crum375 (talk) 19:00, 27 May 2010 (UTC)

The New York Times is not "self-published" in any sense, WP or otherwise. The New York Times is published by The New York Times Company, just like The Boston Globe. Journalism is filled with gray areas. Yes, The New York Times and The Boston Globe are part of the same entity -- The New York Time Company --, but in terms of journalism, they are separate. This is especially true in how Wikipedia should treat them and the journalists they employ. In terms of reliability, they must all be judged separately based on their integrity. Corporate press releases should be treated differently from journalism, whether you think it's "self-published" or not. If The New York Times Company puts out a press release, it is "self-published" even though it probably went through internal vetting layers, unlike The New York Times Company publishing The New York Times or The Boston Globe.  Chickenmonkey  19:42, 27 May 2010 (UTC)

As I tried to explain above, that the NYT consists of several related corporations makes no difference, because you don't have one company writing and another company vetting. And in any case, most smaller newspapers are just one company, which does the writing and the vetting, and they are still not "self published" in the WP sense, any more than any other corporation. Crum375 (talk) 19:48, 27 May 2010 (UTC)

The New York Times doesn't consist of several related corporations. The New York Times Company does, is that what you mean? I don't believe newspapers are corporations and they should not be treated as such. I'm not saying any newspaper is self-published. Newspapers have writers, editors, and publishers. If a company releases a press release, like the one I linked to above, the company writes it and releases it and it doesn't have to worry about journalistic integrity.  Chickenmonkey  20:24, 27 May 2010 (UTC)

"I don't believe newspapers are corporations": If you can find one major newspaper which is not a corporation I'd be surprised. Can you please tell me which major newspaper (or other news organization) is not a corporation? Crum375 (talk) 20:35, 27 May 2010 (UTC)

I may be wrong, but I would say newspapers are publications which are owned and published by corporations. If you are correct, that newspapers are corporations, then that would mean they are recognized as separate legal entities from their owners. I don't believe this to be the case.  Chickenmonkey  20:56, 27 May 2010 (UTC)

I have no idea what you mean, sorry. You say, "newspapers are publications which are owned and published by corporations". What legal entity is a "publication"? Crum375 (talk) 21:10, 27 May 2010 (UTC)

Publication#Legal definition and copyright I'm certainly not an expert on this (or anything, really heh). My understanding is, newspapers are the publication of collected works of which the copyright holders have granted permission to be published, through their employment with the newspaper -- much like the writer of a book grants a publishing company permission to distribute said book.  Chickenmonkey  21:25, 27 May 2010 (UTC)

I think you may be confused. A publication is a piece of work, it's not an organization. A newspaper or magazine, as a legal entity or organization, is normally a corporation, as far as I know, though I am willing to learn more. Crum375 (talk) 21:54, 27 May 2010 (UTC)

No, really: A newspaper is a publication. See the first sentence of Newspaper: "A newspaper is a regularly scheduled publication..." See wikt:newspaper, "A publication, usually published daily or weekly..." Using "newspaper" to refer to "the entity that prints the publication" is just a figure of speech (a metonym). WhatamIdoing (talk) 22:00, 27 May 2010 (UTC)

Yes, exactly what WhatamIdoing has said here.  Chickenmonkey  22:47, 27 May 2010 (UTC)

Not at all. A newspaper is also a pile of paper, so it all depends on context. I was asking what kind of a legal entity is the company or organization which produces a newspaper, i.e. the legal entity that pays the various individuals to produce it and is liable for their output. As far as I know, it is generally a corporation, but I am willing to learn otherwise. Crum375 (talk) 23:27, 27 May 2010 (UTC)

Yes: A newspaper is a publication — a pile of (printed) paper (with a certain type of content). Its legal status is essentially the same as a book — a thing (with intellectual property implications), not a group of people or an activity. A newspaper is not the business organization that writes, produces, and distributes it.

It is possible for the business that owns the publication to be any form of business, e.g., sole proprietorship, partnership, LLC, corporation — anything. In the US, where incorporation is a trivial act and provides asset protection and tax benefits, most, but certainly not all, newspapers are owned by a corporation. The New York Observer and the The Philadelphia Inquirer, for example, are owned by limited liability companies. The companies (whatever their form) fairly often own more than one newspaper; for example, in my local area, the daily paper is one of about 50 owned by a large, privately held corporation, and then there's a small-town weekly that is owned by a company that also happens to own the weekly in the next town over. Most of the employees work on both publications; altogether, the business has one publisher, one editor, and one reporter (and five non-editorial employees) writing, producing, and distributing the two newspapers.

In those parts of the world where incorporation is more of a bother, the smaller businesses might be less likely to choose incorporation. WhatamIdoing (talk) 01:35, 28 May 2010 (UTC)

Crum, can you look at these two sentences, and tell me if they are identical?

1. "Self-publishing is when an author produces and publicizes his own book for public consumption."
2. "Self-publishing is when an author produces and publicizes his own book for public consumption, with minimal or no interference." WhatamIdoing (talk) 19:16, 27 May 2010 (UTC)

No, they are not "identical", but they are equivalent, and I can accept them both as defining "self publishing". Self publishing on WP is a guy and his dog pushing some buttons and getting their stuff out. When professionals stand in the way, who are paid to say "no" when the material doesn't pass muster, it's not self publishing in the common—and WP—sense. Crum375 (talk) 19:43, 27 May 2010 (UTC)

Okay: You personally believe that the common definition of self-publication includes "with minimal or no interference", right?

Now: Can you produce an actual, reliable source that includes any words even remotely like "with minimal or no interference"? Pretend that I'm trying to put sentence #2 into the lead of Self-publishing, and I need a source that directly says something like "with minimal or no interference" to overcome a WP:BURDEN challenge from an editor who says that the words "with minimal or no interference" are not supportable. Can you identify such a source for me -- one that says something like "with minimal or no interference" in a plain, direct, NOR-compliant fashion? WhatamIdoing (talk) 19:57, 27 May 2010 (UTC)

Yes, I do believe that "self publishing" is when you can distribute your stuff to the public "with minimal or no interference", and this is very compatible with the definition from the source I gave you, as well as the wiki article. But we are not discussing here the wiki article — that discussion belongs on the article's talk page. The topic here is the verifiability policy, and for policies we need working definitions, which are based on or related to the corresponding wiki articles, but not directly tied to them. Therefore, you won't normally see us providing sources or references inside policy pages, because they are not articles. Content policies need to clarify what we expect editors to do when writing articles, not explain to them how the world works. But in any case, I don't see a big difference between our working definition and the article, or the definition I linked to above. Crum375 (talk) 20:13, 27 May 2010 (UTC)

Another break

Do you agree that your additional criterion, while compatible with the definition, is not actually part of the definition you provided? WhatamIdoing (talk) 20:17, 27 May 2010 (UTC)

If you were to require a mathematically exact definition, none of this would work, and we'd have no encyclopedia. For our own working definition, and within the boundaries of common sense, the WP definition and the definition I linked to, as well as the current self publishing wiki article, are all the same or very similar. Crum375 (talk) 20:30, 27 May 2010 (UTC)

I'm not looking for a mathematically exact definition. You have asserted that the common definition includes words to the effect of "with minimal or no interference". If that sort of language really is part of the common definition, then you should have no trouble at all showing me a source that contains language at least remotely like that.

If, on the other hand, that criterion really isn't part of the common definition, then can you and I acknowledge that apparent fact -- a bit like adults who, having looked into the cupboard and discovered there's absolutely no food in it, don't keep saying that there really is a whole lot of food in the cupboard, even thought we can't see any? WhatamIdoing (talk) 20:38, 27 May 2010 (UTC)

I am sorry for not following your cupboard analogy. In fact, I am not really sure what your point is. We have a working definition on WP which is consistent with those commonly used, such as the one I gave you and the wiki article. That definition is essentially that "self publishing" is a guy publishing his own stuff, with no paid professionals to stop him when he is wrong or when there are legal issues. It seems to me we are going around in circles. Crum375 (talk) 20:45, 27 May 2010 (UTC)

I look into the sources, and I find no words like "with minimal or no oversight".

You look into the sources, and you also find no words like "with minimal or no oversight", right?

But you tell me that words like "with minimal or no oversight" are actually part of the common definition -- despite the fact that neither of us can actually find words like this in any reliable source, right?

If this criterion is actually part of the common definition, why can't you show me one source that actually includes this criterion? Is this the kind of "common" like "so popular that nobody goes there any more", or "so common that none of the sources mention it", or "so full of food, that the cupboard is completely empty", i.e., "common" meaning "exactly the opposite of common"?

Is it, in fact, possible that this allegedly common criterion is not actually part of any source's definition -- that the common definition actually doesn't include a condition like "with minimal or no oversight"? WhatamIdoing (talk) 20:55, 27 May 2010 (UTC)

Again, for WP policies working definitions we don't need to follow some external definition verbatim, it's enough to be close, esp. since the external ones may vary. In this case, the wiki article says, "Self-publishing is the publishing of books, micropublishing on-line works and other media by the authors of those works", the definition I linked to says "self-publishing is when an author produces and publicizes his own book for public consumption", and WP effectively defines it as a the guy who can press some buttons, and with no paid professionals or a vetting mechanism to stop him, can send his stuff out to the world. Those are all consistent with each other, and as I said, I think we are going around in circles. Crum375 (talk) 21:19, 27 May 2010 (UTC)

Are we agreed, then, that "with minimal or no oversight", or words to that effect, don't actually appear in any reliable definition? That, in fact, it is possible for an author to subject his work to an enormous level of oversight, and still publish his own work himself, and -- no matter how reliable the source might be as a result of that vetting -- the resulting publication is actually, technically, according to every definition you've found, still technically self-published, solely on the grounds that the name of the author is exactly the same as the name of the publisher?

Because that's what I want to add to this policy: a plain, bald statement that "[wikt:self-publishing|Self-publication]] is the publication of a work by its author, without the involvement of an established, third-party publisher" -- and to stop there, without adding any unverifiable statements about oversight.

Does that work for you? Based on what you've learned, do you agree that the common definition of self-published does not actually include any statements about layers of vetting or editorial oversight? Do you think that the above definition is wrong? WhatamIdoing (talk) 22:07, 27 May 2010 (UTC)

No, you are missing or ignoring the key point. The issue is not that the author has "subjected his work to enormous oversight". That could still be self published. What will make it not self published is if there is a structure in place, of paid professionals, whose job it is to vet the author's raw output, for factual accuracy and legal liability, and to actually stop such output from being published if it fails to meet their standards. Even if the author happens to be the "big boss", like "Chief" in the Daily Planet, they can still tell him, "this is wrong, we can't say that", or "this is libelous, we'll get sued if we print it", etc. Just having the author himself vet his output, or subject it to vetting as you call it, when he can at any point decide, "this is enough", and press the "Save" button, is still "self publish". So again, "Chief" is not self published, even if his "Daily Planet" is very small, as long as there is a professional organization in place to control his output before it goes out. Crum375 (talk) 22:43, 27 May 2010 (UTC)

Okay, I think that "enormous oversight" is the same as "a structure in place, of paid professionals, whose job it is to vet the author's raw output, for factual accuracy and legal liability". Do you?

I think that the ability "to actually stop such output from being published if it fails to meet their standards" is a defining duty of the publisher. (The publisher is the person who decides whether it goes to the public.) Thus, if someone other than the author can "actually stop such output from being published if it fails to meet their standards", then the author is not the publisher. Do you agree? WhatamIdoing (talk) 22:50, 27 May 2010 (UTC)

No, "enormous oversight" is not the same as "a structure in place, of paid professionals, whose job it is to vet the author's raw output, for factual accuracy and legal liability." The question is only whether there is such structure, not whether there is "oversight" of any kind. And the key point is that those professionals who vet the output before it's published are paid to stop it if it's libelous or factually incorrect, etc. You are forcing the position of a "publisher", which may or may not exist. That publisher could be just a figurehead, or someone who does actual vetting, or completely nonexistent as an individual, in which case it is simply the corporate "person". But there is no need for any such "publisher" for the professional structure to be in place and strictly enforced, which will make the publication not "self published". And no, I don't agree with "if someone other than the author can 'actually stop such output from being published if it fails to meet their standards', then the author is not the publisher." Take my "Chief" example from above. Chief could be the publisher, not the editor in chief, but he could still be stopped from releasing junk or libel. He could potentially override the objections, but not likely, because all those objecting pros would make excellent plaintiff witnesses if he is subsequently sued for libel. Crum375 (talk) 23:40, 27 May 2010 (UTC)

Just what is your structure supposed to be doing, if not overseeing things? Twiddling its thumbs?

Yes, I really am forcing us to confront the idea of a publisher. I'm doing that because every single relevant reliable source believes that self-publication requires this idea that publishers exist.

In my world (and the world of the reliable sources) a publisher can be an individual human, a group of humans, or a corporation. Whatever entity decides to produce and distribute a publication is the publisher. Every single publication was published by someone. Publications do not spring forth fully formed from the forehead of Zeus: they are deliberately published by their publishers.

In my world (and the world of the reliable sources), an author can be an individual human, a group of humans, or a corporation. Whatever entity writes the publication is the author. Every single publication was authored by someone. Publications do not spring forth fully formed from the forehead of Zeus: they are deliberately written by their authors.

Publications are written by people we call "authors" and published by people we call "publishers".

If Chief can be stopped by someone — really, truly stopped, against Chief's own will, not merely advised in the strongest possible terms that this is a disastrous idea that will land them all in court and told that if Chief doesn't come to his senses right now, then the lawyer is going to call for the nice men in the white coats to bring a padded truck ASAP, because Chief really, really, really needs to make a different decision — then Chief is not the publisher, because Chief is not the entity that actually decides to produce and distribute the publication.

Do you see how this works? "Person who writes = Author". "Person who publishes = Publisher". "Person who follows other people's directions about publication = Not the publisher. WhatamIdoing (talk) 23:57, 27 May 2010 (UTC)

The simplistic division of labor you specify, into an author (or authors) vs. a publisher (or publishers), is incorrect, even in large publishing houses, because there are intermediaries between the ones who do the writing and the ones who approve the release for publication. In fact, those intermediaries, who do much of the factual and legal legwork, contribute significantly to the overall vetting structure. But you don't need to have this polarized structure at all. You can have an author, a vetting organization, perhaps some final proof reading, and a final mechanical publishing step. The "big boss" could be uninterested in the daily details, only that the material is well vetted. And he himself may end up writing some stuff, which he expects to be vetted no less than that of his underlings. In such cases there is no "authors" vs. "publishers" dichotomy, simply one or more authors, a professional vetting structure, and a purely mechanical final publishing step. The "publisher" concept is not needed to make the output reliable, and not "self published". Again, on WP by "self published" we specifically refer to the concept of individuals creating and distributing their own content, with minimal if any professional vetting. This is the most common^{[according to whom?]} definition of the term, and this is what we mean by it in our content policies. Crum375 (talk) 00:21, 28 May 2010 (UTC)

I agree that it can, on occasion, be difficult to identify exactly who the authors are, and who the publishers are. The fact that it's hard to figure out exactly which names to put in the "author" blank or the "publisher" blank on a form — and the fact that far more people might be involved than those who do authoring and those who do publishing — does not change the fact that every single publication was authored by someone and was published by someone, or the fact that when these "someones" are the same, then the publication is self-published according to every single reliable source on the planet.

Your definition is NOT "the most common^{[according to whom?]} definition". It is your own, made-up definition. You have failed to find a single source that shares your opinion. To quote Jimbo, if it is true that "with minimal if any professional vetting" is part of the common definition, then it should be easy to supply a reference, shouldn't it?

I'm really not sure how to get this through to you, without resorting to something that sounds a lot like I'm saying, "Liar, liar pants on fire": There is not one single reliable source on this planet that restricts self-publication to situations in which the author is the same as the publisher "with minimal if any professional vetting." Not one. For you to keep saying that this criterion, which cannot be found anywhere, is "the most common^{[according to whom?]}", is verging on intellectual dishonesty. If your definition cannot be found anywhere, it is NOT "the most common^{[according to whom?]} definition". Definitions that cannot be found in any dictionary anywhere in the world are NOT "common" definitions. WhatamIdoing (talk) 00:46, 28 May 2010 (UTC)

My definition is the same as the common one that I supplied, and consistent with the current wiki article. The point about the vetting is simply extra explanation, since many editors could see a typical newspaper as "self published", i.e. when it clearly isn't so under our definition. So what distinguishes a small newspaper from the guy and his dog who just push a button to publish their stuff? The professional structure that the newspaper has in place to vet the material. This is our working definition, so editors will be able to know when an operation is self published, like the button-pushing guy, vs. not self published, like the small newspaper or equivalent. Crum375 (talk) 00:57, 28 May 2010 (UTC)

That's a good explanation of "the specific kinds of self-published sources that we are particularly worried about" (and IMO with good cause); it is not, however, an accurate definition of "self-published". It is self-published-plus, not just plain self-published.

I think we need to address self-published ("plain", not "plus") in this section. I think that we need another, separate section that addresses failure to vet things. I believe that "Is the author the publisher?" is a completely separate question from "Did that source use reasonable controls?".

For example, Dressed to Kill (book) is self-published: The authors are the only people involved in the tiny business that nominally published it. It is also badly vetted and talks absolute nonsense about the cause of breast cancer.

By contrast, Coca-Cola's website was written by the business, and published by the business -- which makes it "self-published" -- but it is well vetted, with excellent controls (within reasonable limits, that are apparent to you). The fact that it's well reviewed doesn't change the identity of the author or the publisher (and therefore its status as a self-publication), but that fact does change the reliability.

Finally, consider the case of the ravings of a crackpot, or a hoax, that is published by another, completely separate entity: Consider, e.g., The Archko Volume, written by W. D. Mahan (now dead, and mercy on his soul) and currently published by the major publishing house McGraw-Hill (among others). This is not self-published, but it is badly vetted. No matter how much distance is between the author and the publisher, it is still a badly vetted hoax -- but it is (no longer) self-published.

I proposed the addition of a four-sentence definition/description above; you basically seem to agree with it at this point, but you seem to want it to talk about more than just self-publication. Here's the text; would you read it again, and let me know if this works for you, strictly as a definition of the single consideration of self-publication, not as a catalog of every important consideration of reliability?

Self-publication is the publication of a work by its author, without the involvement of an established, third-party publisher. This includes any and all individuals, small groups, and corporate authors who publish their own works on paper, electronically, or in any other media form, so long as the author is also the publisher. Self-published and non-self-published sources may or may not be independent of the subject.^[1] Self-published and non-self-published sources can be primary, secondary, or tertiary sources.^[2]

^ Examples: A book that is both written and published by a historian about the Roman Empire is an independent, self-published source. Memoirs written by a retired politician and published by a major publishing house is a non-independent, non-self-published source.

^ Examples: A blog posting about a house fire, written by the person whose house burned down, is a primary, self-published source. A newspaper story about the same fire, written by a reporter on the scene, is a primary, non-self-published source.

Thanks, WhatamIdoing (talk) 02:00, 28 May 2010 (UTC)

I disagree with the premise that "an established, third-party publisher" is required, or even a key criterion, for sources not to be "self published". There are many reliable sources which have no "third party publisher", but have a good vetting structure in place and are therefore not "self published" within the wiki policy meaning. The insistence on a "third party publisher" would turn these reliable sources into "self published sources", which on WP are considered by default unreliable, so all existing articles which depend on such sources would be in limbo, for no good reason. Clearly this won't fly. Crum375 (talk) 02:08, 28 May 2010 (UTC)

I took that from a source, but I think other sources would say "established or third-party" rather than "established and third-party".

Would the insertion of the word "or" change reconcile you to this definition? WhatamIdoing (talk) 02:53, 28 May 2010 (UTC)

I don't see how it would help anything. If you take all existing reliable sources which have a good reputation for accuracy and fact checking, with a professional vetting structure, and have no identifiable individual or corporate "publisher" in place other than the organization or group proper, they would be instantly deemed "self published" and hence unreliable by this change. Crum375 (talk) 03:01, 28 May 2010 (UTC)

"Self-published" does not equal "unreliable", and I don't see how this change would alter any source, anyway.  Chickenmonkey  03:05, 28 May 2010 (UTC)

"self published" = "unreliable" on WP. In general, such sources can only tell us about themselves, if they are not unduly self serving. Crum375 (talk) 03:14, 28 May 2010 (UTC)

That's simply not what it says. It says self-published sources are "largely not acceptable". Self-published sources should be used with caution. We use our best discretion to determine if the self-publisher is reliable. Self-published sources should never be used as third-party sources on living people.  Chickenmonkey  03:23, 28 May 2010 (UTC)

"Largely not acceptable" means that by default they are unacceptable. You have to make a special case to show that they are acceptable, and normally it would be only when they talk about themselves, per WP:SELFPUB. Crum375 (talk) 03:31, 28 May 2010 (UTC)

What? I'm not an expert on the publishing field; so, I can accept that I may not be one hundred percent clear on what exactly "self-published" means (though, I think I do), but I know what "largely" means and it's not "completely".  Chickenmonkey  03:35, 28 May 2010 (UTC)

If you read the entire WP:SPS section you'll see what "largely" means. It refers to two special SPS exceptions: one is the established expert allowance, and the other is any SPS describing itself. There are no other listed exclusions, and "largely" refers to those two only. Crum375 (talk) 03:41, 28 May 2010 (UTC)

That's what I said. We use our discretion to determine if the self-publisher is reliable (either is an expert, or is talking about the self-publisher itself). The fact that there are exceptions means "self-published" does not equal "unreliable".  Chickenmonkey  03:49, 28 May 2010 (UTC)

The vast majority of sources don't meet those exceptions, which means that for all of them, being downgraded from reliable (today) to unreliable (because of being reclassified as SPS) would cause all articles currently relying on them to disappear, or be drastically chopped. This is for sources that have an internal structure run by paid professionals for vetting their published material, and do a good job of fact checking and liability screening. I don't think you'd get this kind of decimation of previously-reliable sources to be accepted by the community. Crum375 (talk) 04:50, 28 May 2010 (UTC)

Confusion

Crum, I'm confused by your response. You don't think that the publisher of a large newspaper would generally be considered an "established publisher"? Or is your concern that editors whose answer to the question, "Who is the publisher?" is "I haven't the foggiest idea" would declare that "unknown publisher" is "same name as the author"? WhatamIdoing (talk) 03:09, 28 May 2010 (UTC)

Where did I say that the publisher of a newspaper is not considered "established"? And yes, if you have a group where no one in particular is identified as "the publisher", its publisher would be by default the group itself, which is a tautology, since every source is a publisher in a sense, because it published the material in question. Crum375 (talk) 03:19, 28 May 2010 (UTC)

You said, "if they...have no identifiable individual or corporate "publisher" in place other than the organization or group proper, they would be instantly deemed "self published"..."

On what grounds to you claim that anything published by an organization or group would be deemed self-published? On the grounds that the editor can't figure out who the publisher is, and sloppily equates "I don't know" with "Author's name"? On the grounds that the editor thinks that a newspaper publisher isn't "an established publisher"?

This really isn't that hard:

• This source was written by ("author") Clifford Krauss and John M. Broder. It was published by ("publisher") Arthur Ochs Sulzberger, Jr. Are these the same? No? Then it is not self-published.
• This source was written by Giles Whittell. It was published by a member of the News International Group (very probably Times Newspapers Limited). Is "Giles Whittell" the same as "a member of the News International Group"? No? Then it is not self-published -- even though there is "no identifiable individual or corporate "publisher" in place other than the organization or group proper."

It's exactly the same analysis: "Do the author and publisher match?" is the question, not "Who is the publisher?". So where's the problem? WhatamIdoing (talk) 03:50, 28 May 2010 (UTC)

The problem is that you say "as long as the author is also the publisher", and in a different place above you said that both "author" and "publisher" are amorphous terms that could cover groups of individuals. Therefore, for any smaller organization, where the exact composition of the "author" and "publisher" groups is vague or has significant overlap (even by their own members), you'd end up with a situation where author (group) = publisher (group), therefore, "self published", therefore, unreliable. So according to your proposal, such an organization would be deemed unreliable and not usable as a general source even if they had a structure in place, consisting of paid professionals, for vetting their output. This is unacceptable, because it conflicts with WP's concept of "self published", and would render many currently well sourced articles unsourced. Clearly this does not make sense. Crum375 (talk) 04:03, 28 May 2010 (UTC)

Yes: Where author (group) = publisher (group), it is self-published.

No: Where author (group) = publisher (group), it is not (necessarily) unreliable.

Self-publication is not a fancy way of spelling unreliable. Many self-published sources are highly reliable sources. WhatamIdoing (talk) 04:46, 28 May 2010 (UTC)

No, groups can be so amorphous that you won't know for sure who is the "publisher" and who is the "author", yet they can have a high quality structure in place for fact checking and legal screening. They can have a reputation for accuracy and fact checking, and yet they'd become "self published" overnight and hence unreliable if your version becomes policy. That's clearly unacceptable. Much WP content is based on such sources, and we can't downgrade them to "unreliable" and "self published" just because you introduce a new requirement for "third party" publisher. Crum375 (talk) 04:59, 28 May 2010 (UTC)

If you do not know who the author is, or you do not know who the publisher is, then you cannot determine whether a work is self-published.

Consequently, in those situations in which "you won't know for sure who is the "publisher" and who is the "author"", the only answer you can give to "Is this self-published?" is I DON'T KNOW. Do you understand that?

Would you be happier if the first proposed sentence were shortened to "Self-publication is the publication of a work by its author"? I had thought that an explicit exception for the involvement of an established publisher would have pleased you, by making it perfectly clear that The Times is not self-published. WhatamIdoing (talk) 15:57, 28 May 2010 (UTC)

The problem is that your proposal leaves the critical terms vague. Who is "author"? One human? Two? Twenty? The entire organization, or a significant subset? And same for "publisher", is it a human person? A corporate entity? What happens in the case of a smaller organization where many people participate in the writing and publishing process, who are the author and publisher in such cases? Is the material "self published" if both are amorphous?

To summarize, as I see it, your proposal will at best add confusion, at worst knock down a big fraction of article content which is currently based on reliable sources which have a good reputation for accuracy and fact checking with good structure in place for vetting their content by professionals. I just don't see any gain. Crum375 (talk) 16:07, 28 May 2010 (UTC)

You don't think that we can use basic, plain-English, dictionary definitions for "author" and "publisher", e.g., wikt:author#Noun and wikt:publisher#Noun, and leave the application to the editors' best judgment?

In defining self-published, I see a substantial gain: No editor will be able to look at this definition and say, "Ah, self-published means 'without lots of layers of vetting, no matter who the author and publisher are'" or "I see, on Wikipedia, self-published means 'a secondary source written and published by the same person, but not a primary source written and published by the same person'". And since both of those absolutely erroneous definitions have been put forward on this and other pages, surely defining the term as meaning "not these errors" would reduce the confusion that is being caused by these erroneous, unverifiable assertions.

Do you think that it is desirable to have this demonstrably false definitions circulating around Wikipedia? WhatamIdoing (talk) 16:29, 28 May 2010 (UTC)

(ec) Trying to reduce confusion by adding more is pouring gasoline on a fire. There is no necessary connection between "self published" and the primacy of sources — self published sources can be primary, secondary or tertiary, but they are almost always considered unreliable as WP sources, per WP:SPS. And SPS does mean that it is a source where the author and publisher are essentially the same, typically one human, without any layers of vetting such as fact checking and legal scrutiny. So it seems to me that all your proposal would achieve is more confusion, and not help in any way I can see. I believe SV asked you this before: can you provide an example of an article where the present policy language fails, in your view? Crum375 (talk) 16:56, 28 May 2010 (UTC)

Gavin provided an example below, and since WP:N directly references WP:SPS on this point, notability is a relevant example.

The absence of any definition at all on this page permitted an editor to make up his own definition that conflated primacy with self-publication. Do you actually think that it is desirable to have these demonstrably false definitions circulating around Wikipedia, uncontradicted by this page? WhatamIdoing (talk) 17:16, 28 May 2010 (UTC)

There are lots of people around with misconceptions about our policies. In most cases, I think it's because they don't read them carefully. Do you have your own example of an article you were personally involved in, where in your view a change in WP:SPS wording would have helped? Crum375 (talk) 19:31, 28 May 2010 (UTC)

My personal experience is being told by SlimVirgin at WP:Notability that Coca-Cola, Inc.'s website is not written and published by the company.

I am hoping to not have to bloat WP:N with an apparently endless list of statements like, "Corporate websites are not evidence of notability, no matter how many 'layers of vetting' you believe they have. Slick brochures are not evidence of notability, even if they're secondary sources..." (and so forth).

I am hoping, in fact, that we can all agree that a company's own advertisements are both written and published by the company, and therefore self-published, and so we can stick with our existing, brief, simple "no self-published stuff" statement.

In my experience, having an admin like SlimVirgin assert that Coca-Cola's website isn't self-published as far as Wikipedia is concerned is exactly the kind of nonsense that multiplies into endless problems.

But never mind my motivation: I ask again -- do you, or don't you, choose to go on record as saying that in your opinion Wikipedia benefits from not contradicting these demonstrably false definitions of self-publication? WhatamIdoing (talk) 19:59, 28 May 2010 (UTC)

WAID, forget "self publishing" for the moment. In the case of most companies, their website is typically very reliable, but is also a primary source. This is because it generally describes the company itself, its products and its activities, hence it is involved in the content being described, and is considered a primary source for it, per WP:PSTS. As you know, we are not allowed to use primary sources to establish notability, and we are also not allowed to base an article solely on them. So the issue in your case is not if a company's website is "self published" (generally no), or "reliable" (generally yes), but whether it is primary, which it most often is. And for notability purposes, or to create an article without other secondary sources, it can't be used. Hopefully this helps. Crum375 (talk) 21:12, 28 May 2010 (UTC)

No, you forget the circumstances that prompted my proposal, and answer my direct and repeated question.

• You agree that these definitions are false.
• You agree these false definitions exist on Wikipedia.

Are you actually prepared to go on record as saying that Wikipedia is best served by not contradicting these demonstrably false definitions? WhatamIdoing (talk) 21:31, 28 May 2010 (UTC)

You say that these are "demonstrably false", but you have yet to demonstrate an article you were involved in where there was any problem. The Coca Cola issue is hopefully resolved, since that was a primary source, and I don't see anything "false" anywhere, nor anything "contradictory" that needs fixing. If you can point to something specific, please do. And again, please show a real article you were involved in where the policies as currently written were problematic, in your view. Crum375 (talk) 22:06, 28 May 2010 (UTC)

Am I to understand that you now believe that a source can only be self-published if it is a secondary source? I thought we had agreed (1) that such a definition was given [as Wikipedia's definition, not merely somewhere in the world] and (2) that such a definition was actually false. Have you changed your mind on that point? WhatamIdoing (talk) 22:43, 28 May 2010 (UTC)

Sorry, not following. Can you show me where I said or implied that "a source can only be self-published if it is a secondary source"? I thought I made it quite clear before that self-published sources can come in all flavors: primary, secondary or tertiary. And I am still waiting for you to demonstrate a problem in the existing policy which you encountered while working on an article. Crum375 (talk) 01:37, 29 May 2010 (UTC)

You haven't made that particular claim, but other editors have: '"self-published" refers to secondary sources'. Do we once again agree

1. that "self-published" does not refer to secondary sources AND
2. that this false definition was put forth by a mistaken, but well-intentioned editor? WhatamIdoing (talk) 02:14, 29 May 2010 (UTC)

I am sorry but I don't have any special knowledge why other people do or say things. As I tried to explain several times now, SPS can come in all flavors: primary, secondary and tertiary. All of them are normally unreliable, except in limited circumstances, as described in WP:SPS. I am still waiting for you to show me the example from an article you have worked on, where you consider the existing policy problematic. Crum375 (talk) 02:31, 29 May 2010 (UTC)

I'm going to take the above as an admission that you are aware that these demonstrably false definitions exist, and that you, personally, believe these other editors to be wrong.

Now: You are requesting evidence that these false definitions have caused problems in the main namespace. Please show me the policy that says the community can't, or shouldn't, address known misconceptions until after the problem's blown up in the main namespace. WhatamIdoing (talk) 02:56, 29 May 2010 (UTC)

I am sorry, but I haven't made any "admissions" about any "demonstrably false" anything — I am simply describing to you what the current policies say. It sounds from the above that you have not encountered any specific problem with the policies, but are trying to prevent future ones. But the way we design policies is to handle known cases, otherwise they become unrealistic or impractical. This is why it's important for you to provide actual examples where the current policy runs into problems, because then we can try to find a way to cover the problem case, without causing breakage elsewhere. In most situations, at least in my experience, the problem is not the policy itself but the way it is applied (or misapplied). Crum375 (talk) 03:07, 29 May 2010 (UTC)

So you think that the linked statement didn't happen? Or you think that primary sources can't be self-published? Or, perhaps while you object to the term "admission", you really do think that an experienced editor really did espouse that definition, and you really do think that definition really is wrong? WhatamIdoing (talk) 03:25, 29 May 2010 (UTC)

This is a talk page for a policy, and our goal is to improve the policy. I try to avoid guessing why editor X says Y, and I normally just assume good faith. I don't object to the word "admission" on principle, but I have not made any "admissions" that I am aware of. As far primary sources and SPS, I tried to explain this several times now, but I'll try one last time: SPS can come in all flavors: primary, secondary and tertiary. Which logically also means that each of those three flavors can be either SPS or not SPS. Not sure what more can be said about it. Crum375 (talk) 03:53, 29 May 2010 (UTC)

Yes: SPS can come in all flavors: primary, secondary and tertiary.

Since we have agreed on this point about a dozen separate times now, I have added it to the policy page. WhatamIdoing (talk) 19:25, 2 June 2010 (UTC)

update

Well, it seems that SlimVirgin has removed it as a "can of worms. She'd be one of the editors who has made up her own, unverifiable definitions, of course. WhatamIdoing (talk) 20:37, 2 June 2010 (UTC)

FWIW, I didn't see the utility of your addition to WP:V, although I admit I haven't been following this discussion. Jclemens (talk) 22:29, 2 June 2010 (UTC)

I'd like to get a plain definition of "self-published" into the policy, so that a small number of editors (SlimVirgin is one) will find it much harder to assert their own made-up definitions in discussions on Wikipedia.

As Crum and I have agreed that the content of that particular sentence is True™, verifiable, and relevant (e.g., examples of people falsely asserting the exactly opposite have been documented with diffs), it seemed uncontroversial. Crum's main sticking point seems to be with saying what self-published is (author=publisher), not with saying what it isn't (primary/secondary/tertiary). WhatamIdoing (talk) 01:14, 3 June 2010 (UTC)

---

It should be clear to all by now that reliability and self publication are independent characteristics. As for "who is author" and "who is publisher" the answers are the same as for anything else here. They are who reliable sources say they are. If the publication has any real credence it will be on WorldCat and in at least one depositary library such as the British Library or Library of Congress. It will have been catalogued to indicate the publisher. For major newspapers there will often be books about the newspaper, but they will not always reflect current information about the publisher. For that purpose official sources such as EDGAR are often useful, wherein legally required filings identify publishers that are publicly traded corporations. For a normally trustworthy publication or publisher, we extend that trust to the assumption that the byline correctly attributes the article to its authentic author (if for no other reason than that writers don't like seeing their work attributed to others and tend to make a stink when it happens). In short, "I don't know" is irrelevant. "I can't cite" matters. LeadSongDog come howl! 16:44, 28 May 2010 (UTC)

Grandma is an author

'Self-published sources' is an odd term... it seems to be more appropriate to an era where it would be very uncommon for on to have the ability or means to 'publish' (i.e. make available to a material amount of people) their opinions or reporting. Although I believe the term Vanity press was used to refer to on one hand, those with the means who could publish whatever fit their 'vanity' to a large amount of people, and on the other hand some hardscrablle authors who believed so much in there work they paid the way for their works to get published - some known today, likely many unsuccessfully and unknown today. Self published sources would not always be primary sources, but when being secondary sources, when do they become reliable? To continue the analogies:

1. Grandma blogs about her house burning down
2. Gma blogs about her neighbor's house burning down
3. Gma blogs about her neighbor's house burning down on a web site she has ran for her retirement community for years, which has been mentioned and linked to multiple times by a local news paper
4. Gma blogs about her neighbor's house burning down on the local newspaper web site using some type of user content area
5. Gma blogs about her neighbor's house burning down as a paid contributor to the local newspaper
6. Gma blogs about her neighbor's house burning down and her blog post is converted to article format and published in the paper version of the local newspaper
7. Gma wirte about her neighbor's house burning down for the new york times

Maybe she stops being 'self-published' around step 4 or 5... when does she become a reliable source? I don't think the two necessarily are tied together.Cander0000 (talk) 03:10, 27 May 2010 (UTC)

The distinction being proposed here appears to only add confusion, not relieve it. I'm not even sure if it's accurate or true, and I certainly cannot understand what "real-life" editing issues it could possibly hope to solve. Jayjg ^(talk) 06:45, 27 May 2010 (UTC)

The Grandma analogy is clever - maybe we can kick her around a bit more to develop a shared understanding of this issue. Going back to the start of analogy, lets speculate that "Grandma started the fire in order to make a bogus insurance claim". From this perspective, Grandma may not be a self-published source in all cases, but she has strong connection with the subject matter, to the point where she is suspected of not being a disintersted source of information. Thus, whether the source is self-published or not, if the source is not independent of the subject, the reliablity of the source must be brought into question regadless of whether it is a primary, secondary or tertiary source. --Gavin Collins (talk|contribs) 09:26, 27 May 2010 (UTC)

I think that #5 is usually when Grandma is normally considered non-self-published, although the details of #4's set up matter here. (I assume that Grandma isn't the newspaper's publisher of record.)

The fact that Grandma in #5 is the author of material published by someone else does not tell us whether Grandma has a massive conflict of interest/that she's independent of the subject matter, or whether the story was properly vetted (e.g., if the editors were out sick that week because Grandma poisoned them all with her famous fruitcake so that she could publish her story without any interference). These other factors definitely affect reliability, but they don't tell us whether Grandma is both the author and the publisher. WhatamIdoing (talk) 20:14, 27 May 2010 (UTC)

You forgot one senario... Grandma worked as a fire Marshal for 40 years and has written several well critiqued books about firefighting... and then blogs about her neighbor's house burning down. Her blog is completely self-published... but due to the "expert exception" her blog might be considered highly reliable. Blueboar (talk) 20:20, 27 May 2010 (UTC)

I agree: In that scenario, Retired Fire Marshall Grandma is both a highly reliable source and a self-published source. WhatamIdoing (talk) 20:40, 27 May 2010 (UTC)

Example?

It's hard to follow what the aim of the above is. Could someone give a real example of a problem that's being caused by our current use of "self-published"? SlimVirgin ^{talk contribs} 21:41, 27 May 2010 (UTC)

Yes: The current lack of an explicit definition has resulted in three editors (including yourself) asserting that "self-published" means something different from what every single reliable source says it means. You and Crum have asserted that it means "self-published without editorial oversight", and George has asserted that it means "self-published secondary source".

If we provide a plain statement of what self-published means, according to every single reliable source, then editors will not be able to make up their own unverifiable definitions to suit themselves, and they will not be able to falsely tell other editors that "Wikipedia defines self-published this way", when it does no such thing. WhatamIdoing (talk) 22:10, 27 May 2010 (UTC)

What I'm requesting is a real example from an article showing that the way the policy uses "self-published" is causing a problem. SlimVirgin ^{talk contribs} 22:45, 27 May 2010 (UTC)

I could give the example of a dispute about self-published sources regarding the author Dan Willis, for whom notability was claimed using only some forum postings and book listings. There was clearly a difference of opinion between the two administrators and myself about what is self-bublished, but the ambiguity of current definition may be the source of the disagreement. Although Dan Willis may or may not have written the coverage cited in the article himself, the fact that he could have done so, or his publisher or agent could have done so suggest to me that a self-published source is one that cannot be classed as being reliable because it has not been the subject of editorial oversight (aka fact checking), but also because the source is not independent of the subject matter. WP:SPS as it is written is a combination of these two missing characteristics: editorial oversight and independence. I think we need to seperate the two to make this policy clear. --Gavin Collins (talk|contribs) 14:05, 28 May 2010 (UTC)

You're talking about notability now, Gavin, and this policy isn't about that. But using Dan Willis, can you give me an example from that dispute of a source where calling it a primary source wasn't enough? That is, an example of a source that needed the words "self-published" to be added to the description before it could be disqualified. SlimVirgin ^{talk contribs} 16:00, 28 May 2010 (UTC)

I am late to the discussion but want to add a historical note. From the inception of WP:CORP (one of the first pages to address the trustworthiness of corporate publications), press releases and corporate webpages were considered prohibited or at least highly suspicious self-publications. Even though multiple employees may have been involved in the publication, they are not truly independent of each other - they all share the same conflict of interest with their employer. Rossami (talk) 21:36, 2 June 2010 (UTC)

Thanks for your note. I agree with everything you say here. Unfortunately, we seem to have a very small number of editors who prefer their made-up definitions of self-publication, which exclude corporate websites. I would like to address this plainly in the policy, so that every editor is on the same page, and so that any editor who is given a false definition can become better educated (and you know how "some apparently experienced editor once told me" turns into irrefutable facts), but I don't know if we'll be able to. Perhaps the community really does want to have these false claims go uncontradicted. WhatamIdoing (talk) 16:57, 3 June 2010 (UTC)

As I stated above, I don't see what real-life problem is being "solved" here. I found the material inserted by WhatamIdoing today to be very confusing to the reader, and not particularly relevant to the policy. Jayjg ^(talk) 01:52, 3 June 2010 (UTC)

I think the problem with the way WP:SPS is written is confusing, because it mixes three (but admittedly not unconnected) characteristics of a reliable source:

1. editorial oversight;
2. independence;
3. what consitutes a reliable "published" source.

Going back to the Dan Willis disucussion, there was a length discussion about whether the various sources cited in the article could be used as evidence of notability. There were three types of source which were disputed: a book review on an educational website; an interview with the author himself and a biography on his publisher's website. The reliablity of these sources were brought into question for the following reasons:

1. forums and fansites that host third party content do not exercise any editorial oversight;
2. content published for promotional purposes are not independent;
3. website content that can be changed or altered at short noticed is "on display" rather than being "published".

I think that items 1 and 3 are covered quite well in WP:RS, but it is item 2 (independence) that is not spelled out in any policy, which is suprising, because in the real world, independent sources are highly prized. If WP:SPS is confusing, it is because it focuses on specific types of self-published source, rather than on their general characteristics. If we were to deconstruct this section, it would be possible to make it more simple and clearer without the need for an exaustive description of what constiutes a self-sublished source. --Gavin Collins (talk|contribs) 08:39, 3 June 2010 (UTC)

Jayjg, I think that it would have made more sense if all of the pieces had been there. The pieces are:

• Self-publishing means (just) that author=publisher
• Self-published does not apply solely to primary/secondary/tertiary sources (e.g., George's assertion that only secondary sources can be self-published)
• Self-published does not apply solely to sources that are independent/non-independent of the subject matter
• Self-published does not tell us anything about how many lawyers vetted the source (e.g., SV's assertion that Coca-Cola's website isn't self-published, that is, isn't both written and published by the corporation)

But Crum seemed skittish about providing a complete, verifiable definition, and I didn't want to go any further than what we had agreed. I assume that he's also pretty tired of vehemently agreeing with me about self-publication being possible in non-secondary sources. So my effort was a baby step in the direction of providing the information that we definitely agree upon. SV has not said why she decided that this a factually accurate statement is a "can of worms"; I assume, though, that she prefers having her made-up definition continue to be uncontradicted and therefore silently endorsed by this policy. WhatamIdoing (talk) 16:57, 3 June 2010 (UTC)

• I was asked to comment by WhatamIdoing. What I understand SV to mean, is that among self published sources, the official web sites of large organizations have a certain degree of reliability. In my opinion, sheis correct. This is just a more detailed statement of the accepted policy that self published sources can be used for the description of the basic facts about the organization if uncontroversial, and in any case about its official position about things and events. More generally, there is no sharp distinction between reliable and non-reliable sources; sources have a position on a spectrum of reliability. There is no source whatsoever that it totally reliable for all purpose, and almost no source that is totally worthless for at least some purposes. The most reliable newspapers and academic journals have been known to publish not merely errors, but material which turned out to be totally fraudulent; the web site of the loonies crank organization is good evidence for what they believe. Every reference book known to me contains some errors.

Whether a source is self-published is determined not by the ostensible facts of publication, but the editorial control.Coca-cola's website is under the control for the company and says such things as the company wishes it to say. But the company in this case has a certain amount of reputation, and is not likely to publish recklessly. The NYT is under the controls of its editors. It is reliable even for news stories about itself (or a direct competitor), because these are prepared by the editorial staff, not the publisher. But any such story is not quite as free from doubt as what it may publish about a company in an unrelated industry. If a company says how much it profit it is making in a press release, it is not as reliable as in an audited statement. But even such audited statements have been known to be deliberately counter-factual.

What is a promotional purpose is not a sharp consideration either. Everything any organization, whether a company or a government publishes in a sense has a direct or indirect promotional purpose. The accurate plain authenticated statement of favorable facts is promotional; the publication of unfavorable facts is promotional also, in showing a reputation for honesty--we properly tend to distrust material from a country that reports only the good things about itself. Everything a person publishes has at least a secondary purpose of promoting their views, their reputation, their career, or their financial-well being. We ourselves write at WP both to help humanity, and to show both that we wish to do so--and also that we are capable of writing competently.

I think it wise to maintain here the pre-existing traditional language, because so many other statements are based on them. The details of sources is a matter for WP:RS, and the two pages should be kept separate, this as policy, that as guideline. The interpretation of sources in any particular instance if disputed is a question for the RSN, the Reliable sources noticeboard, and the actual statement of our prior practice is its archives. DGG ( talk ) 22:31, 4 June 2010 (UTC)