Let me try again: It was inappropriate to move my comments out of the discussion and into this new section. My comments don't make any sense taken out of context, in my opinion. Dlabtot (talk) 18:16, 28 March 2008 (UTC)

Maybe I was wrong to move this discussion; however, I did link to the thread in which this started, so your meaning is not obscured. See my response to you on my talk page for further comments on this. PSWG1920 (talk) 20:49, 28 March 2008 (UTC)

I want to be clear: the comments I made in another section, don't belong in this section, they don't make sense in this section. The meaning of my comments was dependent on the context in which my comments were made. I'd like to respectfully ask that no one add anything to this section and sign my name to them. Thank you. Dlabtot (talk) 07:14, 29 March 2008 (UTC)

I am really sorry for the mess I have effectively caused here, and I have crossed out my responses to now non-existent comments. Admins, feel free to delete my crossed-out comments completely if you feel it is best to do so. PSWG1920 (talk) 07:52, 29 March 2008 (UTC)

Please can folk here look at the discussion at Wikipedia talk:When to cite#Citation in lists. The old "If you've got wikilinks, who needs sources" argument. Thanks. Colin°^Talk 17:56, 31 March 2008 (UTC)

Is it? Yes, please have a look, I'd like feedback too. - Dan Dank55 (talk) 22:52, 31 March 2008 (UTC)

I am faced with an interesting question relating to verifiability. The article on Leonard Neale (1746 — 1817), the first US Citizen to be ordained a Roman Catholic bishop, includes the following paragraph:

• "A story said to have been passed down by both George Washington's slaves and by the Jesuits is that several hours before Washington's death a Catholic priest was sent for from across the Piscatawney River. According to Jesuit tradition the priest was Father Leonard Neale. Both stories suggest that Washington converted to Catholicism on his deathbed."

Now, the idea that Washington became a death-bed Catholic certainly falls under the "Exceptional claims require exceptional sources" clause. It directly contradicts several eye-witness accounts which state that no clergy (of any denomination) were called to Washington's deathbed. So I and several other editors have been questioning this claim and carefully examining the sources given. At first the claim was (apparently by mistake) cited to an article from the 1950s that supposedly appeared in the Denver Catholic Register. This citation is given in several other more modern sources which repeat the claim. However, a volunteer editor checked this out in the Denver Library, and discovered that no such article ever existed in the DCR. When challenged on this, the editor who added the information explained that the citation to the DCR was an error and changed the citation to the National Catholic Register (which, according to its website, was the national edition of the DCR). Now, I suppose that it is possible that a story like this was printed in the national edition of the magazine, and not the local one... even though I think it unlikely. So... today, I went to the New York Public Library to verify the new source. It turns out that while the NYPL does have back editions of the NCR on file, they do not have any for the years in question (the earliest they have is in the 1960s) and neither do any of the other libraries and accademic institutions the NYPL is in partnership with. However, The editor who insists on keeping this legend in the article says he has a copy in front of him, and Assuming Good Faith, I should believe him.

This situation has all sorts of issues that relate to policy ... there are questions relating to WP:FRINGE, WP:UNDUE (the story takes up a fifth of the entire article) and probably several other policies and guidelines... but those can be discussed on the relevant talk pages and noticeboards for those policies and guidelines ... my question here is simply one relating to verification: At what point can we call a source unverifiable? If an editor claims to have a source in hand... but, after reasonable search in some of the largest public libraries in the US, it is impossible for anyone else to obtain a copy... can it be called verifiable? Blueboar (talk) 19:43, 31 March 2008 (UTC)

This gets even better... an editor has checked with the Library of Congress... they don't have it either. Blueboar (talk) 21:45, 31 March 2008 (UTC)

The archivist Karyl Klein at the Denver Catholic Register can be reached at archives@archden.org or Karyl.Klein@archden.org. Karyl e-mailed me the two articles from The National Catholic Register. Here is the publkic site where I got the contact information: http://www.archden.org/index.php?option=com_content&task=section&id=11&Itemid=421Dwain (talk) 22:29, 31 March 2008 (UTC)

Full disclosure from the get-go: I'm involved in what some might perceive as a citation dispute in eNom. I hope readers will consider the possibility that sometimes -- maybe even this time -- disputes might illuminate legitimate gaps in policies.

In this case, the disagreement is over citing a press release from an organization that grants annual industry awards, in the article about a company the award was granted to. No news media covered the award, so no citations exist that meet the strict guidelines of independent fact-checking, and yet within the context of the article, and the industry, the award is significant and it seems clear that citing the press release from the awarder is preferable to citing the press release from the awardee. I've seen similar questions raised about press releases about company acquisitions, and I think it's a safe guess that there are plenty of other examples from "long-tail" situations -- the enormous number of topics with relatively small followings that don't get coverage in mainstream media or peer-reviewed literature. So the question I'd raise is: WP:V is a beautiful treatise if read uncritically, but possibly suffers from leaving too much grey area between what's clearly acceptable and what's clearly not. Would it be reasonable to try to reduce the gap, at least in the limited case of press releases re: industry awards? For reference, previous discussion has suggested that such citations are "less than ideal", but have been allowed to remain (see summary sentence by Visviva at top of http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/GURPS_4e_Basic_Set_%282nd_nomination%29 ). Thirdbeach (talk) 21:40, 31 March 2008 (UTC)

The phrasing can be misleading. It implies that there's a distinction between truth and verifiability (which there is, and not just in a Popperian sense); and for a slogan, that's not a good thing there is. It's not that we exclude the truth, after all.

Not just truth, but verifiability.

You see how that "just" changes the connotations of the sentence? DS (talk) 19:51, 3 April 2008 (UTC)

Though the Popperian in me winces at the statement, I'm not actually sure we should disentangle truth and verifiability. "Truth through verifiability" comes very close to what we mean. Phil Sandifer (talk) 19:58, 3 April 2008 (UTC)

It may come very close to what you mean, but I don't believe your use of the word 'we' is really appropriate in this context. Dlabtot (talk) 20:04, 3 April 2008 (UTC)

The alternative seems to be the statement "We don't care if we get everything wrong as long as we can find other people who also get it wrong." If that is your belief, you ought not contribute to the project. Phil Sandifer (talk) 20:10, 3 April 2008 (UTC)

The policy says (or did when I last looked) that the threshold for inclusion is verifiability, not truth. That simply means that my believing something to be true — or even knowing it (e.g. if I saw it with my own two eyes) — isn't enough. Truth is neither a necessary nor a sufficient condition for inclusion. There is nothing controversial about that. No science journal, for example, will include a claim without evidence. Our requirement is simply the prior publication of the material by a reliable source. Again, this is not controversial. Lots of publications operate along the same lines, especially when dealing with unknown writers, as we do. SlimVirgin ^talk|edits 20:38, 3 April 2008 (UTC)

Truth, at least in the small-t form, is a necessary condition for inclusion. Phil Sandifer (talk) 20:57, 3 April 2008 (UTC)

Only in the sense that we have to know it is true that A said X, and that this is in a publication, or published by a publishing house, we would normally trust; or in some circumstances that A is a person we trust no matter who publishes his material. We don't have to know that X is true. SlimVirgin ^talk|edits 21:07, 3 April 2008 (UTC)

I suppose if you want to open a gray area between true and not-false you can, here, but for the most part I disagree - if our assessment of a piece of information is other than "it is true" than we should not put it in even with a reliable source. The lighthouse example above is a good one - Mangoe, in practice, is absolutely correct to figure out the correct location of the lighthouses and report that instead of presenting all of the various claims. Phil Sandifer (talk) 00:15, 4 April 2008 (UTC)

That's a simple exercise of editorial judgement, not a repudiation of our verifiability or neutral point of view policy. Dlabtot (talk) 02:34, 4 April 2008 (UTC)

Except that, as I've clearly stated the policy forbids what Mangoe did. Phil Sandifer (talk) 03:52, 4 April 2008 (UTC)

This policy doesn't forbid it. If someone were to challenge Mangoe, he'd need to find sources, yes, but otherwise he's okay. Having said that, I know that there are people who make unreasonable challenges, and that has started to concern me too in the last year, because I keep seeing ridiculous ones. I think it would be worth drafting a section to remind people that challenges have to be reasonable. SlimVirgin ^talk|edits 06:37, 4 April 2008 (UTC)

Well, let's try a nastier case that Blueboar and I have been involved in. There is apparently a Catholic legendarium wandering around that says that Leonard Neale came over from Maryland and converted George Washington to Catholicism on his deathbed. This thing was originally cited (through some website or another, as far as I can tell) to a pair of "Denver Register" articles from the 1950s, and to an inventory of Washington's personal effects. Well, the problem with the latter is that possession of a pair of holy cards doesn't prove much, even though it can be verified that he had them. That's a matter of judgement, and could be accused of being OR. The issue with the articles is more interesting, because they don't exist as cited. Someone graciously went the library and checked this out for us. However, it does seem possible that the articles were actually published by the National Catholic Register; we aren't sure because we haven't gotten our hands on them to assess them. Assuming the articles are real, there's then the problem of how to assess them against the testimony of the eyewitnesses on the scene. Right now we're stuck with the story in the article on Leonard Neale, albeit with the disclaimer that it probably isn't true, because the apparent fact of the articles' publication means that the story is verifiable, and therefore should be included. Personally, I'm inclined to leave it out entirely, because I think repeating apocrypha, even when disclaimed, lends merit to the dubious statements. But maybe that's just me. Mangoe (talk) 17:16, 4 April 2008 (UTC)

It's not just you - I think you're spot-on in this one. (And it's comforting to see that many of the editorial situations I'm familiar with from popular culture articles carry over to more traditional articles) The issue is that nobody actually writes an encyclopedia article with an eye towards verifiability over truth. (And nobody reads one that way either) Similarly, nobody writes an article "from sources" as such - they write it from their knowledge that has been informed by sources, and then go back and cite the sources to help make clear the links between the sources and the knowledge. Which is how all research works - one reads the sources, understands them, and then writes about the material with citations to establish the connection. The problem with WP:V right now (and to an equal extent WP:NOR, which I'm going to move on to after this discussion wraps up) is that it presumes a model where we somehow directly translate reliable independent sources into encyclopedia articles. The relationship between a researched document and its sources is nowhere near that straightforward, and simply does not work as described here. Phil Sandifer (talk) 18:21, 4 April 2008 (UTC)

I look forward to your comments about WP:NOR. Do you want to completely discard that policy as well? Dlabtot (talk) 18:36, 4 April 2008 (UTC)

I haven't looked at it closely enough to know yet - but it's certainly become something other than its original intention. But that's largely off-topic here (and I don't want to completely discard this policy - just its current wording.) Phil Sandifer (talk) 18:38, 4 April 2008 (UTC)

Apologies for backtracking on the conversation . . . DS's suggestion of "not just truth, but verifiability" does an exceptionally good job of both uniting and drawing the distinction between the concepts. Like others, I find the current text "verifiability, not truth" to be logically untenable. On first reading it expresses -- doesn't just imply -- a hostility toward truth; therefore it distracts and derails the reader. I think it fails to achieve what the writer intended. I think DS's alternative achieves the intent. Thirdbeach (talk) 01:49, 5 April 2008 (UTC)

There've been a lot of (often quite good) changes suggested since I made my detailed analysis, and while I still would prefer a more radical solution, I think it's certainly prudent to introduce some of them to the policy. Here are the ones, looking at the talk page, that seem to me to have consensus or, at the least, widespread support:

• Change the slogan to "Not just truth, but verifiability," thus stressing the goal of accuracy in Wikipedia.
• Slimvirgin's proposed change to the section that is currently handled via Jimbomancy.

On the latter, I still remain skeptical of "Any edit lacking a reliable source may be removed." For one thing, we ought think about content as information, not as edits. But more significantly, I don't think that's an appropriate standard - what we want, roughly, is that any information that is suspected of being inaccurate or unverifiable may be challenged, and in the event of a genuine dispute over the verifiability of the information, the burden of proof is that the information must be shown to be verifiable. Once a sensible challenge is made the burden of proof is on the people trying to add the information - but there's an initial and lesser burden of proof that needs to be met before the challenge is made - that is, a reasonable belief that you're identifying a problem as opposed to a hoop to jump through.

But there are a number of other issues I raised that I think need to be looked at and changed.

Two are relatively easy:

• De-academification. That is, while peer reviewed sources are clearly great and are essential for a number of topics, the academic bias of this policy doesn't reflect the sourcing issues involved in most of the material on Wikipedia.
• Fixing self-published sources. This is related to de-academification. Simply put, though self-published medical research is obviously not something we want, material like the Gray's Anatomy Writers Blog is self-published but can obviously be used as a reliable source without third party confirmation. J. Michael Straczynski's well-confirmed Usenet posts and writings in his self-published collections of his Babylon 5 scripts can obviously be used as reliable sources for statements about the actors and directors involved in Babylon 5 - even for negative material about them. In that case it should be clearly attributed, but his accounts of contentious interactions surrounding the show are clearly reliable sources. A formal review process is not nearly the be-all and end-all that it is treated as.

Two are harder:

• A realistic sense of what sources are available. Because commercial publishing is, well, commercial, decisions about what is and is not published are based on more than accuracy or even importance - they're based on marketability. The rise of the Internet and easy self-publication has made this more pressing - there's a lot of information that it's no longer worthwhile for a publisher to publish simply because there are people willing to compile and publish it for free on the Internet. We can't assume the information we need for subjects is stated straight-out in third party sources.
• Relatedly, we need to be realistic about how sources are used. No research happens "from sources" in the sense of just being taken directly from a text and dropped into Wikipedia rephrased and cited. Articles, like any research, are written by people who have read sources, learned stuff, and then write about the stuff. The point of citations in research is not to establish every fact, but to leave a paper trail that can be followed up on. Thus we need to both understand that synthesis is inevitable (because research is not just rephrasing sources) and that some information will need to be reflected in a general "references" section instead of line-by-line.

As I said, the first two, I think, can be accomplished with language rephrasing, albeit at times large scale rephrasing. The latter two are more fundamental shifts, but I think they are no less necessary. Thoughts? Phil Sandifer (talk) 18:25, 5 April 2008 (UTC)

I agree with the two changes that appear to have solid support. They're good ideas. Could you elaborate a bit more about "de-academification", both in terms of what you see as problematic and what you see as the solution?

I'd be a bit uncomfortable with such a liberalization of self-published source usage, both because of potential abuse concerns and because it runs contrary to common good practice on-wiki. However, much of your point could simply be covered by removing the tether from expert status to "reliable third-party publications". That portion of the policy could be simplified as simply requiring that the author be an "established expert" (leaving it, like much of policy, to local consensus regarding the specific meaning/application). That would move towards the goal of loosening the language of the policy appropriately, while maintaining the expert connection that is strongly supported. The use of the GA writers blog would already be permitted under the current policy (as they're established experts, published in a relevant field in a reliable publication), but such a change would further clarify the permissibility.

The use of such sources to comment about living people would need to be addressed at Wikipedia:Biographies of living people, as the reason for restriction on such articles arises from that policy. On a similar note, your concerns about synthesis would need to be addressed at Wikipedia:No original research, which covers such issues.

I am not convinced that the internet has had the effect on the print publishing industry that you indicate. I am an avid reader and involved in the treeware publishing industry. While the market has contracted to some degree, I have not seen the phenomena that you indicate is problematic. While traditional publishing markets have been somewhat slow to adapt, they have been (and are) adapting. Out-of-print books can be commonly purchased easily second-hand or in PDF format. Online libraries are thriving with business (largely bolstered by institutional subscriptions). Most periodicals now have online versions (and some have even made the move to new media versions being the version of record, including academic publications) and many have a thriving subscription business for archives or simply make archives freely available. That is my experience and perception of the matter, but I would be interested in hearing more about your concerns on this topic, including any examples you may have encountered. Vassyana (talk) 18:50, 5 April 2008 (UTC)

By de-academification, I mean that much of this policy is written for academic subjects in a way that is just not useful elsewhere, and, if followed elsewhere, could even be problematic. As far as I can tell I'm one of two scholars to write substantively on Calvin and Hobbes. I have a peer-reviewed article on it. But I'd be mortified if that article got used as the major source for the article due to its supposed reliability.

"Established expert" is a much better phrase - that should get added straightaway.

The issues certainly touch NOR and BLP, but BLP is ultimately just "Be super careful about V, NOR, and NPOV on these articles," and so I'm unconvinced that we can't deal with the self-published issue here.

As for the Internet, you're right inasmuch as I oversimplified the issue - certainly the "Internet is killing treeware" idea is wrong. What is perhaps more accurate to say is that commercial publishing is profounding influenced by, well, what's commercial. Even academic publishing is - I've routinely seen articles rejected from ImageTexT (the journal I work for) not because they're wrong or uninteresting but just because ImageTexT isn't quite the right place for them. Most of them are material that absolutely should go into an encyclopedia article on a topic, but aren't suitable for a theoretically-minded scholarly journal. And this happens in tons of other places - stuff isn't published in reviewed sources not because it's not true and important, but because it's not sufficiently innovative/interesting/marketable to make money off of. And a good amount of it falls into the gap between obvious and non-obvious synthesis - stuff that is obvious enough to be not worth publishing, but not necessarily obvious enough to be unsourced under our current policies. Again, we're approaching NOR territory here, but V and NOR are deeply intertwined policies. And certainly I intend to move on to NOR after this policy is fixed up a bit. Phil Sandifer (talk) 00:09, 6 April 2008 (UTC)

I knew it would be a good discussion, because I saw who was responding, and these people are very clued-in. You're doing a good job too, and some changes have already been made to WP:V, and more may come. Here are some responses to your points, and I hope you're starting to notice a pattern here: statements in WP:V that on the surface look like they're not good enough to get the job done do in fact represent quite a lot of blood, sweat and tears; it's just not always obvious.

First a general answer: lots of the things that you would like for people to talk about are talked about, every day, at WT:OR, WP:RSN and other places. For the sources they haven't been able to accept yet that you would want them to accept, my guess is that the software behind WP:Flagged revisions will eventually get us all the way to where we need to go. This is being tried first by the German Wikipedia, and maybe soon by the Russians and the English Wikibooks. Roughly speaking, it's intelligent data-mining that asks people to rate whether changes to articles ought to survive, and then makes assumptions about the reliability of the posters based on those ratings. Nothing too grand, but at least enough to sort out who the vandals are. In the same way, it could be and hopefully will be used to give us a rough idea of what we think about different sources. See the Wikiquality-l mailing list archives for detailed discussions.

Re: "Not just truth, but verifiability". I think we could be a little more pro-truth on this page. Truth is great, but in an imperfect world where we often haven't found the truth, and we often disagree, verifiability is quite often the best we can do and quite often good enough. Oppressed people come to us often and say, "Our people are dying because our governments can control the media; why aren't you telling the truth?" And all we can do is sadly reply that we'll do our best, and we want to help, but we have not found a satisfactory way to make Wikipedia any better than proven, reliable sources of information. Also, reaching for absolute truth has, at times, resulted in very bad results. Before you blame us for pushing verifiability over truth, look into the history of what has happened when we have tried it the other way around.

Re: "Any edit lacking a reliable source may be removed." I deleted the part about "unsourced information should be removed" a few days ago, and no one has reverted. It contradicted the part in bold, and it needed to go. The part you are quoting continues, "but editors may object if you remove material without giving them a chance...", and we've suggested some language to add to that in this discussion, but the bottom line here is that we've provided solid protection here for the editors trying to retain the unsourced information, because people take the "impose one's own view of 'standards to apply' rather than those of the community" clause from WP:POINT seriously around here. The sentence could be tweaked, but it's good enough to get the job done. - Dan Dank55 (talk) 00:32, 6 April 2008 (UTC)

Well, to be fair, it's not that I don't think that WP:V represents a lot of hard work. Nor do I think that it's up to doing the job as such - rather, I think that a lot of work has been put into actively damaging this page by pursuing an unrealistic and untenable vision of what research is.

Flagged versions will certainly be helpful, but I think the sorts of policy changes I'm indicating are necessary too.

You're misunderstanding my comments here - it's not that I don't think verifiability is essential. Unverifiable information should be removed. However untrue information should suffer the same fate. Hence "Not just truth, but verifiability," which makes clear that we do not want to publish untrue information, as opposed to the current version, which actively suggests that we are neutral on truth.

I like the language you're proposing, but it seems to me to remain the case - we do not mean to remove every edit without a reliable source, nor to consider that. We want to remove unverifiable information, and because you can't prove the negative that means that we remove unverified information. But fundamentally, the reason you remove should not be "I don't think it is verified," but rather "I don't think it can be." Phil Sandifer (talk) 14:36, 6 April 2008 (UTC)

←I wasn't proposing language, but we've had enough input now for a language proposal; give it a shot. I agree with your general point that WP:V has added language over time targeted at unsourced material, without spending equal time dealing with other problems, and that this may give WP:UNDUE weight to one point of view. I have some recommendations for how you can increase interest in your point of view:

• Read WP:RSN#Talking Points Memo Election Central (this link will go red soon when it's archived) and other conversations at WP:RSN, both to get a firm command of the details of sourcing, and so that you can discuss your points in terms of other people's arguments. People are much more interested in hearing their own ideas and words tweaked and reflected back to them.
• Keep track of the progress of WP:Flagged revisions, especially discussions about progress at de.wikipedia and en.wikibooks. Effort there could pay big dividends when it gets deployed, and it's not on the radar yet for many Wikipedians. (As always, there's a bias towards what's "hot" and "now".)
• WP:WHEN and its talk page are a very nice start on some of the issues you are interested in; perhaps more work could be done there and it could be promoted to a guideline.
• When trying to get people to pay attention to something they haven't paid much attention to before, there is no substitute for a large, careful, random survey of articles to demonstrate your point. If you leave any of those 3 adjectives out, it won't work. It also doesn't work to say "it's not just me saying this...", because that's not how Wikipedia works. Wikipedia is not a representative democracy; instead, a "wide" invitation (the width can vary) is issued, and if people don't show up, their voice isn't heard. - Dan Dank55 (talk) 17:16, 6 April 2008 (UTC)

I've gone ahead and made the two changes suggested at the top, and added a section below suggesting a change addressing de-academification. Phil Sandifer (talk) 18:42, 6 April 2008 (UTC)

How are religious texts or scriptures viewed in terms of verifiable material? Paulrach (talk) 08:24, 6 April 2008 (UTC)

They're only really good for verifying what they say themselves... any interpretation of that would need a secondary source, and in terms of trying to verify historical information, they're not really appropriate. SamBC(talk) 11:55, 6 April 2008 (UTC)

Absolutely. WP can and should report that source A says X and source B says Y, but should refrain from taking a position regarding the validity of either X or Y.-- Boracay Bill (talk) 12:31, 6 April 2008 (UTC)

The policy currently reads:

In general, the most reliable sources are peer-reviewed journals and books published in university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. As a rule of thumb, the greater the degree of scrutiny involved in checking facts, analyzing legal issues, and scrutinizing the evidence and arguments of a particular work, the more reliable it is. Academic and peer-reviewed publications are highly valued and usually the most reliable sources in areas where they are available, such as history, medicine and science. Material from reliable non-academic sources may also be used in these areas, particularly if they are respected mainstream publications. The appropriateness of any source always depends on the context. Where there is disagreement between sources, their views should be clearly attributed in the text.

The logic here is clear, but given the number of popular and non-academic subjects we cover, seems unhelpful - especially because academic sources can be, erm, problematic for some areas. I'm one of a few scholars to have published on Calvin and Hobbes in a peer-reviewed journal, but I will personally revert anybody who tries to rewrite the article to focus primarily on Lacanian interpretations of the strip.

Which isn't to say that academic perspectives aren't important, even on popular subjects. Here's the language I'd propose:

For areas of academic research like history, medicine, or science, peer-reviewed publications are highly valued and usually the most reliable sources. Material from reliable non-academic sources may also be used in these areas, particularly if they are respected mainstream publications. The appropriateness of any source always depends on the context. Where there is disagreement between sources, their views should be clearly attributed in the text. For non-academic areas, sources such as magazines, journals, mainstream newspapers, and books published by respected publishing houses are often the best sources. Academic sources may and often do exist for subjects that are not primarily academic, and should be carefully consulted and included, but may be inappropriate as the primary basis for an article. As a rule of thumb, the greater the degree of scrutiny involved in checking facts, analyzing legal issues, and scrutinizing the evidence and arguments of a particular work, the more reliable it is.

It could probably be cleaned up, as I am often needlessly wordy, but what do people think? I think it eliminates some advice that is not bad, as such, but is also not terribly helpful. Phil Sandifer (talk) 18:42, 6 April 2008 (UTC)

I think the intent was to keep this policy tightly focussed; as you noted, you were getting "needlessly wordy". Perhaps your clarification belongs in the Guideline WP:RS or even more appropriately in the essay WP:Reliable source examples. --SteveMcCluskey (talk) 18:54, 6 April 2008 (UTC)

To be fair, my text is only 21 words longer than the existing text. That comes from a lone sentence of clarification - "Academic sources may and often do exist for subjects that are not primarily academic, and should be carefully consulted and included, but may be inappropriate as the primary basis for an article." (And that's actually 32 words long, so I managed to trim 11 words from the existing text.) And so it's not like I'm trying to introduce a new wordy section. I was mostly just figuring a few words could be trimmed here and there.

If you want to move the whole discussion over to RS or RSE, I won't object, but right now this page is providing misleading advice about the use of academic sources. If that language is going to remain on this page, it should at least be made correct and useful. Phil Sandifer (talk) 19:07, 6 April 2008 (UTC)

I don't believe its misleading. I believe its where an encyclopaedia ought to go. I particularly disagree with the "primary basis" statement. --Relata refero (disp.) 10:14, 8 April 2008 (UTC)

You really think that Calvin and Hobbes should be rewritten with this [1] as its major source? (It's one of two published engagements with Calvin and Hobbes I'm aware of. There are probably a few others, to be fair, but it's one of a few, in a peer-reviewed academic journal). Phil Sandifer (talk) 14:10, 8 April 2008 (UTC)

Nice piece, Phil! Why not? --Relata refero (disp.) 21:25, 8 April 2008 (UTC)

More to the point, if that article focused less on recurring plotlines and more on critical analysis, it would indeed feel more encyclopaedic. Also, I would certainly think that rewriting the article around the piece would violate UNDUE as an unnecessary privileging of a Lacanian symbolic language.. so the "material from reliable non-academic sources" would be required for those areas that haven't had the beady eye of academia turned in their direction. --Relata refero (disp.) 21:33, 8 April 2008 (UTC)

Well, and to be fair, I think it's a good article too, and I wouldn't oppose in the slightest it being mentioned in the article. I mean to add notes about existing scholarship to comics articles where it exists. But as you point out, were the academic material on Calvin and Hobbes to be the primary basis for the article, it would be a problem of NPOV. Which is my main issue on areas where academic criticism is only a small aspect of the overall treatment of the subject - it's an NPOV problem. Where academic criticism exists I'm all for using it. Just... not always as the primary basis for an article. Phil Sandifer (talk) 02:59, 9 April 2008 (UTC)

I know images aren't typically considered under the "Verifiability" criteria, but the rule generally applies: Sourcing information needs to be complete so that it can be verified the license is correct. This would be a small part of the education of this planned project:

Help is needed to develop an "image cleanup month" (June). The goal is to "Educate, cleanup and move images here at Wikipedia". You don't need to be an expert or knowledgeable about images here to help. Need folks who can write well, copyedit, design connections/templates, organize, group, communicate, have connections to users to help advertise (once the month starts) or just want to help in any other capacity. Not knowing about images would be helpful as we can test our pages on you. Being knowledgeable you can help write the content. See the project central location at Wikipedia talk:WikiProject Image Monitoring Group#Wikipedia Image Cleanup Month (June) and dive right in to help. MECU≈talk 17:26, 8 April 2008 (UTC)

"Exceptional claims in Wikipedia require high-quality reliable sources; if such sources are not available, the material should not be included. Also be sure to adhere to other policies, such as the policy for biographies of living persons and the undue weight provision of WP:NPOV."

This is simply not true. Exceptional claims which are notable -have been covered in an RS- need Attribution, but may be included even from unreliable sources. This should read something like:

Exceptional claims must be notable in order to be included in Wikipedia. Claims and ideas concerning fringe topics which do not come from a reliable source must be carefully attributed. Also be sure to adhere to other policies, such as the policy for biographies of living persons and the undue weight provision of WP:NPOV.

As it is, this passage contradicts a lot- including BLP where it says that you can include someone's personal website. ——Martin^phi ☎ Ψ Φ—— 23:21, 8 April 2008 (UTC)

I think you're failing to make a sufficient distinction between including the claim, and including the claim of the claim. That is, X might be a controversial claim; John claims X is not controversial if we can reliably attribute it to John. The relevance of that statement is entirely a matter of significance and not verifiability. Someguy1221 (talk) 23:26, 8 April 2008 (UTC)

I thought of that, but if the claim is exceptional to begin with, we have no business claiming it as fact, no matter how reliable the source. ——Martin^phi ☎ Ψ Φ—— 23:54, 8 April 2008 (UTC)

It also says the material should simply not be included if not in an RS, rather than saying that the material should be well sourced. A distinction should also be drawn between exceptional on a topic, versus an article whose subject is the exceptional claim. Yes, this is really the problem. The current wording is aimed at exceptional claims in articles which are not about the claim. Yet, it is being used to apply to articles where the subject is the exceptional claim. Which is how I came to be here. ——Martin^phi ☎ Ψ Φ—— 23:58, 8 April 2008 (UTC)

If someone adds to the article Abraham Lincoln the claim that he was abducted by aliens, that needs a reliable source. If someone adds that he claimed to have been abducted by aliens, that still needs a reliable source. If someone adds Abraham Lincoln to a list of alien abductees, that also needs a reliable source. But not every sentence in the article Alien abduction needs a footnote to a reliable source. --FOo (talk) 02:40, 9 April 2008 (UTC)

I know. Here is how it is actually being used [2]. In other words, it is being used to exclude fring sources from fringe articles. That's the key: people are taking REDFLAG this way:

"Exceptional claims in Wikipedia require high-quality reliable sources; if such sources are not available, the material should not be included, even in articles on said exceptional claims."

In other words, you said it right, but that isn't really what the text says, and that's what it should say, because people are mis-using it. ——Martin^phi ☎ Ψ Φ—— 03:03, 9 April 2008 (UTC)

(I've also included this message on WT:NCGN, and I would welcome further opinions about the issues I raise.)

There is currently a discussion ongoing (its latest phase is on User talk:Jza84#English exonyms) about adding Welsh places names to places that are in England. Most of these places are close to the English-Welsh border, but not all of them, and so in some cases, it is reasonable to consider adding the Welsh names. However, there is some disagreement between myself and an editor about verification of the facts added when one wishes to add a Welsh name to an English place. I maintain that to add a Welsh name to an English place one needs two kinds of verification (a) an appropriate reference (say, to a suitable dictionary) that can confirm that the added Welsh name is accurate, because we have had disputes or incorrect names added previously (e.g., Hereford, as can be seen from its edit histories and discussion page, though there is at least one other place which I cannot find just now); and (b) some justification, appropriately verified, as to why the addition of the Welsh name to a particular English place is reasonable (for example, if a place in England was frequently used as a stopping point on an old droving road by Welsh farmers that it gained a Welsh name that it was used and by which it was known by in Wales, and if this can be verified by means of a suitable citation, then it should be used to justify the addituion of the Welsh name in a suitable place within the article for the English place.) Of course, sometimes one single source could serve both purposes. I maintain that these requirements are merely a working through of the normal wikipedia requirements for verification and citing sources of any non-obvious factual information that is added to an article. The one person who is objecting to this has said in a previous discussion that this is an unfair extra burden to be placed on the addition of Welsh names. I would like some advice about this from others. The present discussion can be viewed for a bit more information. If previous discussions are needed, I can search them out if requested. Many thanks.  DDStretch  (talk) 23:53, 4 April 2008 (UTC)

You certainly need a reputable reference, not just several people's opinions. It may be a burden, but then getting a reference for anything may be a burden, so that argument doesn't impress me. You also need some sort of justification I'd say.Doug Weller (talk) 13:08, 5 April 2008 (UTC)

It strikes me that adding Welsh place names for places not in Wales, given this is the english language wikipedia, would be a case of WP:UNDUE unless there's a source giving a good reason for it. Do the articles for London and Dover give the French-language names for those places? Okay, Dover does, but only because it's an indication of the importance of Dover in international terms, which is rather the point I'm making. SamBC(talk) 19:01, 5 April 2008 (UTC)

See Wikipedia:Writing better articles#Use other languages sparingly. -- Boracay Bill (talk) 23:47, 5 April 2008 (UTC)

I have made some revised suggestions at User:Ghmyrtle/Sandbox 2. I am conscious that this may not give as much rigour as some might like - though personally I tend towards flexibility rather than rigidity of approach - and also that it tends to focus on the Wales/England issue, which has been my main concern and has generated many words on many talk pages. I haven't changed Jza84's suggested usage table, simply because I'm undecided how useful it would be (although I'm very grateful for the stimulus it has offered). All comments and thoughts welcome. I'm copying this message to various pages, but I suggest that further discussion should be coordinated at the WP:UKCITIES talk page. Ghmyrtle (talk) 20:26, 9 April 2008 (UTC)

Article on Yelena Tregubova heavily relies on her book, which is heavily sensationalist (evident from name "Kremlin Mutants"). I have a feeling that it crosses borders described in WP:SELFPUB (especially 3, 4 and 7), but I was involved in a number of disputes with wikipedian who inserted various claims in the article using said book. I would like somebody uninvolved and more experienced in Verifiability analysis to take a look. Thank you in advance, RJ CG (talk) 15:26, 8 April 2008 (UTC)

This user tells about a book that was published, several times reprinted, and sold in a huge number of copies by an independent publishing house in Russia: Tregubova, Elena (2003). The Tales of a Kremlin Digger (Байки кремлевского диггера). Moscow: Ad Marginem. ISBN 978-5-93321-073-3. {{cite book}}: Cite has empty unknown parameter: |coauthors= (help)

It was translated to Italian and German and published again. The Italian translation of "Tales of a Kremlin Digger" ("I mutanti del Cremlino"), came out in 2005 at the publishing house Piemme.^[1] The German publishing house Tropen Verlag released the book in German ("Die Mutanten des Kremls") in October, 2006.^[2]

This is not self-publishing by any means, and this is not a source about herself. She writes about many notable Russian politicians. She is a reputable and famous journalist.Biophys (talk) 16:06, 8 April 2008 (UTC)

Statement above contains several unsourced claims. There's no proof that book was either several times reprinted or sold in a huge number of copies (Ad Marginem is pretty small publishing house, that does not even have decent webpage). Claim about her being a reputable and famous journalist is POV again. She is a journalist. And I still kindly ask someone uninvolved to take a look. RJ CG (talk) 16:15, 8 April 2008 (UTC)

These books are obviously not self-published. So, the discussion is out of place.Biophys (talk) 16:26, 8 April 2008 (UTC)

These books are obviously not self-published. This statement begs for proof too. RJ CG (talk) 16:34, 8 April 2008 (UTC)

Proof has been offered; the name of the publisher and year of publicaton are given. If you lack access to a library that lets you find the book in question, or find other books by the same publisher, tough. It is not the responsibility of Wikipedia to give you a library. --Gerry Ashton (talk) 17:16, 8 April 2008 (UTC)

"the name of the publisher and year of publicaton are given" To me that proves fact that book was published indeed (which I did not deny). It does not cover any of my concerns. RJ CG (talk) 19:54, 8 April 2008 (UTC)

More than 300,000 copies of the book were sold in Russia [3].Biophys (talk) 02:51, 9 April 2008 (UTC)

Oh yes, and now you can try to convince anyone neutral that this number was based on independent circulation verification and not on Tregubova's statement. Did you notice that circulation numbers for Germany (which can be easily verified) are gloriously absent in German article? Although article does contain one interesting bit. Book had been published by publishing house notorious for it's work with Russian Radical Nationalists. That automatically makes it WP:RS for Eastern European section of Wikipedia, isn't it? RJ CG (talk) 13:51, 9 April 2008 (UTC)

A fairly long article is wholly unreferenced. Should it:

• a) Have a {{fact}} tag at the end of every single sentence
• or b), be tagged with {{unreferenced}}?

Thanks. Neıl ☎ 23:00, 9 April 2008 (UTC)

I would say it should be tagged with {{unreferenced}}, and maybe a few specific statements that really need sources could be pointed out with fact tags as well. But putting a fact tag after every single sentence would just be overkill. PSWG1920 (talk) 23:34, 9 April 2008 (UTC)

These are the alternatives I've seen mentioned to the phrase "not truth, but verifiability," which seems to me, at this point, to clearly lack consensus in its current wording (though not in its intended principle).

1. "The threshold for inclusion in Wikipedia is verifiability"
2. "The threshold for inclusion in Wikipedia is not just truth, but verifiability"
3. "The threshhold for inclusion in Wikipedia is not just accuracy, but verifiability"
4. "Wikipedia strives to be a complete and accurate encyclopedia. However, the threshhold for inclusion in Wikipedia is not truth, but verifiability." (i.e. reinsert the old opening sentence of the policy)

Are there other suggestions? Does anybody, reading the talk page, have a sense that one of these options is closer to consensus than others? Are there any we can definitely rule out? Phil Sandifer (talk) 01:30, 9 April 2008 (UTC)

Sounds very close to the dispute on SPOV on the NPOV talk page. ——Martin^phi ☎ Ψ Φ—— 01:43, 9 April 2008 (UTC)

The only one that may work for me would be #4. ≈ jossi ≈ (talk) 01:44, 9 April 2008 (UTC)

• I absolutely disagree with your assertion that "not truth, but verifiability" lacks consensus, which is why I opened an RfC on the subject, which so far has garnered only 6 responses, including yours. I don't see any clear consensus among the seven editors, including myself, who have expressed an opinion on the topic. Even if there were a clear consensus among these seven editors, the opinion of seven editors is not sufficient to change long-standing core policies, when we have a community of thousands, very few of whom are likely even aware of this discussion. I understand that some may think it important to change the policy before the upcoming deadline, but I believe it is more important that changes to core policies have wide community support. Dlabtot (talk) 01:56, 9 April 2008 (UTC)

Adding the phrase never had the sort of wide support you describe. The discussion here has been on RfC, and I announced broad discussion on VPP. If these are the people who show up, these are the people who form a consensus. Phil Sandifer (talk) 02:02, 9 April 2008 (UTC)

I don't agree, and I don't understand why you believe that repeating your points so many times makes them more persuasive or adds something to the discussion. It doesn't. Making your assertions twice, three times, or a hundred times is the same as making them once. Dlabtot (talk) 02:19, 9 April 2008 (UTC)

I am not sure how one goes about disagreeing with simple and demonstrable fact, but if you have managed it, I congratulate you. I am not sure, however, how your accomplishment adds to the quality of this discussion. Phil Sandifer (talk) 02:22, 9 April 2008 (UTC)

That "simple and demonstrable fact" that you can't imagine anyone disagreeing with... does it also happen to be the truth? lol Dlabtot (talk) 02:31, 9 April 2008 (UTC)

Please, then - show where wide support for the phrase was demonstrated. Phil Sandifer (talk) 02:32, 9 April 2008 (UTC)

It has been demonstrated by thousands of editors following the policy and invoking phrase in talk pages discussions in the years since it was first put in place. It has been demonstrated by remaining as policy after myriad discussions and challenges on this talk page. Being challenged to point out the obvious in this manner bores me; I have no interest in engaging in further pointless arguments with you. I look forward to hearing the opinions of the community of editors. Dlabtot (talk) 02:47, 9 April 2008 (UTC)

Duration is not evidence for anything. As I said below, a search term for child pornography remained in the article on child pornography for over a year and through hundreds of edits. The phrase was inserted from a draft of NOR that had few people looking at it, and it was barely, if at all, discussed at the time. Since then it has had virtually all of its original context stripped away, and far more people have objected to it than ever initially discussed it. And the reasons for those objections - that it is blatantly silly and that it depends on an esoteric usage of the word "truth" that is dramatically removed from day-to-day usage - are things that you and many of the other people opposing a change have simply declined to respond to. For you to then complain that the objections and arguments you are so flagrantly ignoring are being restated is deeply unfair. If you wish to leave the discussion, please feel free - it is not as though your posts have been engaging with it thus far. Phil Sandifer (talk) 02:52, 9 April 2008 (UTC)

I also fail to see how "not truth, but verifiability" lacks consensus, a few people have expressed disagreement with it, but that alone is no basis for saying long standing policy lacks consensus. I for one think it describes our position rather well. But then, so does suggested wording #4. _(1 == 2)^Until 02:26, 9 April 2008 (UTC)

It's the combination of that and the way that it has, over time, drifted away from its relatively clear original explanation, which, notably, never had that much input gathered either. Phil Sandifer (talk) 02:29, 9 April 2008 (UTC)

The thing is, all the alternatives you have proposed are both less clear, and actually encourage people to do things that is beyond their (or anyone's competence). If we encourage people to write "The TRUTH", then we lose any hope of making the encyclopedia NPOV. Jayjg ^(talk) 02:40, 9 April 2008 (UTC)

I have answered this twice already, though I'm happy to do it again - I see no reason to believe that a normal reader, upon reading "not truth, but verifiability," will make the leap to the esoteric and philosophical concept of absolute TRUTH as opposed to the more conventional usage of the word to mean "stuff that is not wrong." We do, in fact, encourage people to add material that is not wrong, and discourage them from adding material that is wrong. Phil Sandifer (talk) 02:44, 9 April 2008 (UTC)

My experience, however, is the exact opposite; editors insisting that reliable sources should be rejected, because the aren't telling the truth, or that the truth must be told, sources be damned. At this very moment I am involved in a debate with an editor who insists that he can insert original research into an article, because what he is saying is the truth, and the alternative is false. Jayjg ^(talk) 01:03, 10 April 2008 (UTC)

Just chiming in to say that my experience echoes that of Jayjig. I've been down that road a number of times, once with an editor who had over 5,000 edits and who scolded me: "It's ridiculous that you're treating me to a lecture on Wikipedia policy." and said: "When I made the edits I clearly marked them as unverified but true." -- Boracay Bill (talk) 03:08, 10 April 2008 (UTC)

I am not sure how these examples show the usefulness of the current wording. Nor do I see how "not just truth/accuracy, but verifiability" alters the nature of the argument you're making against these users. Phil Sandifer (talk) 03:50, 10 April 2008 (UTC)

If Phil were right, there would be no Jesus article, there would be two articles, both of which would violate NPOV. Slrubenstein | Talk 17:31, 9 April 2008 (UTC)

Ummm... no, actually. That's completely untrue and a needless bit of hyperbole that does not so much advance a reasonable argument or engage with issues as engage in hysterical denunciation. Please try again, this time with reading comprehension and/or sanity. Thanks! Phil Sandifer (talk) 03:50, 10 April 2008 (UTC)

April 11, 2006 "The threshold for inclusion in Wikipedia is verifiability, not truth", April 9th, 2007 "The threshold for inclusion in Wikipedia is verifiability, not truth.", April 9th, 2008 "The threshold for inclusion in Wikipedia is verifiability, not truth."

I really don't think it had drifted all that much over time. And something that has remained unchanged for that long clearly has some sort of consensus behind it. _(1 == 2)^Until 02:34, 9 April 2008 (UTC)

Have a look at its original formulation here: [4]. Yes, the slogan (being the catchy part) has not changed, but the context that made clear what we mean has largely vanished. And, when that was added to WP:V, we still had language stressing "The goal of Wikipedia is to become a complete and reliable encyclopedia. Verifiability is the key to becoming a reliable resource" in the intro. As for lack of change, the article on child pornography had, for over a year, explicit mention of a search term used to find child pornography on the web. It was edited hundreds of times over that year, and nobody removed it. Duration is not evidence of consensus - it's evidence that nobody looked at it, went "Wait a moment..." and then did something about it. Phil Sandifer (talk) 02:41, 9 April 2008 (UTC)

Phil: I figure you're looking for Consensus can change, which is not just a good idea, it's actually policy. (Or for more in-depth, you could point people to its expansion at silence and consensus) --Kim Bruning (talk) 00:10, 10 April 2008 (UTC)

The question is, is there consensus to remove the long standing text? _(1 == 2)^Until 02:43, 9 April 2008 (UTC)

"not just truth, but verifiability"? Again, we cannot decide what the "truth" is; instead, you should encourage people to report what reliable sources have said. Jayjg ^(talk) 02:37, 9 April 2008 (UTC)

I agree that the word "just" is misleading in that it implies our criteria may be partly based on truth, when in fact it is a non-issue because it cannot be objectively determined. _(1 == 2)^Until 02:40, 9 April 2008 (UTC)

I do not think that a normal reader, upon seeing the word "truth," gets into questions of epistemology and the objective grounding of all knowledge. I think most people see the word "truth" and take it to mean accurate, correct, not wrong. Which is where our problem arises - because accuracy, correctness, and not-wrongness are things we strive for and require - inaccurate information ought to be removed. Phil Sandifer (talk) 02:46, 9 April 2008 (UTC)

• How about The goal of Wikipedia is to become a complete and accurate encyclopedia. Because of this, it is important to make it easy to verify the accuracy and neutrality of your content through citations to your sources. Adapted from text in the 2005 version. The introduction of the term truth has always bothered me a bit, to be honest. We seem to be placing too much emphasis on the fact that something is verifiable, to the exclusion of whether it is neutral or reliable. I think we need to drop the line that "the threshold for inclusion in Wikipedia is", because it isn't true. There are other thresholds as well. Hiding T 10:49, 9 April 2008 (UTC)

Perhaps "objectively verifiable fact" might dodge the whole "what is (T/t)ruth" problem. LeadSongDog (talk) 13:41, 9 April 2008 (UTC)

I must disagree with Phil, people do see the word truth and think it means "what they know to be true" and insist on it, while this is noble our goal is not to be correct, it is to reflect what reliable sources believe to be correct. Someone will say "That reliable source is wrong, I was actually 3 minutes faster, I know I was there", they are so sure it is the truth they insist on it. We need to show them that truth is not enough and that it needs to be verified.

To extend this, in the 60's we would have posted that Cannabis is 3 different species, this is not true but is was the verifiable fact at the time, and the scientific consensus. With modern genetic research we know they are 3 strains of the same species, this is likely the truth, and it is verifiable. Even if something is flat out contrary to the truth, that is what we must record if that is what the existing verifiable sources state. The reason is that it is our job not to be "right", but to reflect the knowledge of the world as it is now.

Anything else would be original research, and that I think is why we explicitly mention that truth is not our threshold. Our goal is not to be "right" or "not wrong", our goal is to accurately reflect what has been said by reliable sources(even if the whole world is wrong about something). _(1 == 2)^Until 15:02, 9 April 2008 (UTC)

The issue is, though, that this isn't really a rejection of accuracy, but rather a careful limiting of the sorts of things we say. In general we would avoid saying "There are three species of Cannabis" in favor of "Scientific consensus is that there are three..." And in that regard, we ensure accuracy. The real issue, in other words, is that we avoid making claims about the world, and limit ourselves to claims about what is said and generally thought about the world. Phil Sandifer (talk) 15:06, 9 April 2008 (UTC)

My concern is more that, if it were easier to get to the old claim that there were three species rather than the new claim that there is only one, we would have a fight to prefer the outdated information. Indeed, it might be hard to document which view is preferred, at least in a manner which would convince the stray layman. I'm not as concerned about the "scientific consensus" disclaimer. Mangoe (talk) 16:23, 9 April 2008 (UTC)

Precisely, we should avoid making claims about what is true, and stick to claims about what can be verified. With that in mind, I think we have had the right wording over the last three years. _(1 == 2)^Until 15:07, 9 April 2008 (UTC)

Well, no - if we say "Source X says Y," it had better be true that Source X does, in fact, say Y. We don't abandon truth - we limit the claims we make to ones that can be verified. Phil Sandifer (talk) 15:15, 9 April 2008 (UTC)

Well what if they guy was misquoted in truth, but several publications stated "he" said something. The point is that truth is not really something we can base things off of at any level, and that needs to be made clear. We verified that he said something, that is verifiable but we can't tell if it is truth. Someone may show up saying "I was there, that is not what he said", but we have to take the source over the truth. _(1 == 2)^Until 15:58, 9 April 2008 (UTC)

Find me an example in which such a contorted explanation for why we would say "Source X says Y (Citation to source X)" when, in fact, it is not true that Source X says Y - that is, an actual article out of our two million. Otherwise, this seems more than a little abstracted from reality. Phil Sandifer (talk) 03:50, 10 April 2008 (UTC)

• I prefer option #1. Everything else seems wishy-washy obfuscation. Why not direct POV pushers to WP:NOT instead? No reason to keep a statement that, while it may be useful for a small minority, those with POV issues, confounds others as an obvious contradiction, on a page that is otherwise focused on explaining why we cite sources? This debate will be continually revisited as long as the "not truth" statement is there, or even "not there," it seems. But really, WP:NOT provides an obvious alternative to direct POV pushers and other big-T truth believers to, in lieu of a controversial opening phrase "here" that the rest of the policy doesn't even try to explain, leading to ceaselessly recurring arguments from those of a positivist bent. Amerique^dialectics 16:22, 9 April 2008 (UTC)

Sorry to jump in in the middle, but where's that RFC? The link just took me back to the article. "Verifiability, not truth" has been in the policy at least since 2003 when I first became active and I think it is extremely important language.

One reason why it's important is that editors frequently disagree on what "the truth" is, but quite often editors that disagree on what the truth is can agree on what verifiability is. Editors who are committed to verifiability can produce useful and stable articles. Editors who are committed to "the truth" tend to edit-war.

Verifiability is an appeal to external authority. If external authority can't be accepted as the touchstone, then who is the authority? "Wikipedia editors in general... and me in particular?" Dpbsmith (talk) 16:25, 9 April 2008 (UTC)

The RfC is above, entitled 15 RfC: Should the threshold for inclusion in Wikipedia be changed from verifiability, not truth to not just truth, but verifiability? Dlabtot (talk) 16:31, 9 April 2008 (UTC)

Dpbsmith has managed to express my opinion on this matter very well. _(1 == 2)^Until 16:39, 9 April 2008 (UTC)

My position is, leave "truth" out, and we eliminate this source of conflict on this page. We can't have our cake and eat it too, we can't say "v≠t" and then pretend that this doesn't amount to a definite position on the concept of truth. The meaning of this statement has been the subject of constant debate, that has been widely off track of the purpose of this page, which is to explain sourcing guidelines. I am all for keeping this page focused on verifiability, but to do that we have to get rid of the concept of truth, which leads us into these debates that serve no purpose re: the improvement of this policy. Amerique^dialectics 16:59, 9 April 2008 (UTC)

Are you saying that verification does equal truth? Or are you saying we should not take that position? This seems to me the be logically sound, you can verify something through reliable sources and it may or may not be true. The whole point of that sentence is to make it clear that the concept of truth is not part of the verifiability policy, removing the comment would make that less clear.

I also don't think avoiding debates is a good reason to remove it, I think these debates do serve to improve the policy. Several people here have made it clear that they think it has value to the policy. _(1 == 2)^Until 17:23, 9 April 2008 (UTC)

We should leave it as it is. It is fundamental to Wikipedia. It does not require us to take a specific stance on "truth;" the whole point is that different editors may have their own ideas of truth and criteria for truth. "Verifiability not truth" means that no matter what you think is true, or what your standards are for truthfulness, it doesn't matter, rather, attribution to a verifiable source is what matters." Most of us (meaning, not those of us with this page on our watchlist, but most editors who do not bother with these talk pages and instead focus on articles) understand this quite well and accept it as a bedrock principle for Wikipedia. Slrubenstein | Talk 17:28, 9 April 2008 (UTC)

I have found that phrase to be a great tool to bypass POV pushers insistence that they are right despite the sources disagreeing with him/her. I would really hate to lose that tool, which really does accurately describe our best practices. _(1 == 2)^Until 17:33, 9 April 2008 (UTC)

My sense is that all of the people who like the current wording hear the same thing from "not truth": they hear the words coming from someone else's mouth, not their own. I think the general principle is that when we say a word, it's our word, and we're responsible for the meaning, unless we attribute it to someone else or put scare-quotes around it. There are good reasons to trust Blueboar's judgment, because he deals with these kinds of objections at WP:RSN and elsewhere on a minute-by-minute basis ... although maybe I'm just saying that because I really like his suggestion above, quote: The threshold for inclusion in Wikipedia is verifiability, not truth. While Wikipedia strives to provide accurate information on every topic it covers, editors should not add material simply because they know or think it is "true". The information also needs to be verifiable. "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. (end quote) I think the scare-quotes on "true" are enough to get the idea across that we are attributing truth to someone else. - Dan Dank55 (talk) 18:10, 9 April 2008 (UTC)

The thing reads quite differently if we replace "truth" with "correctness" or similar wording: The threshold for inclusion in Wikipedia is verifiability, not factual correctness. The implication is that we can knowingly insert false material as long as we can find a half-decent source for it (such as an ill-informed newspaper reporter). In a policy as important as this one we really ought to distinguish "truth" in terms of factual correctness from The Truth^TM. Raymond Arritt (talk) 18:58, 9 April 2008 (UTC)n

Reliable sources may disagree about what is 'factually correct' as much as they may disagree about what is 'true'. Examined out of context, the phrase "verifiability, not truth" seems nonsensical, but when examined as part of an interlocking set of policies, one see how WP:V works together with WP:NPOV. Until and unless we have a Content Committee that will rule on questions of Truth, editors should simply present all significant viewpoints that have been published by reliable sources, and leave determination of what is true or not true to our readers. Dlabtot (talk) 19:13, 9 April 2008 (UTC)

A fact is a piece of information presented as having objective reality, whereas truth is the property of being in accord with reality. One is something that can be objectively demonstrated, the other is a reflection of how reality is. That is to say, a fact while being objectively demonstrated can change, but the truth is immutable because it represent how things really are. These terms are not interchangeable. _(1 == 2)^Until 19:12, 9 April 2008 (UTC)

People quite often disagree on what the "facts" are; it's a term that's as personal as "truth". Jayjg ^(talk) 01:03, 10 April 2008 (UTC)

Truth often depends on belief. Let us take two contradictory statements: 1) "Jesus is the son of God", and 2) "Jesus is not the son of God". There are millions of people who will tell you that statement #1 is absolutly True and represents "how things really are". There are also millions of people who will tell you that statement #2 is True and represents "how things really are". So which should we say is the immutable truth? Blueboar (talk) 20:45, 9 April 2008 (UTC)

This discussion is precisely why we need to say front and centre that we're not in the Truth business. Or the truth business. If the whole world is lying, we repeat it quietly. --Relata refero (disp.) 20:48, 9 April 2008 (UTC)

Yes. Let's leave as it is now, which makes that point quite unambiguously. ≈ jossi ≈ (talk) 20:54, 9 April 2008 (UTC)

Tell me if I'm getting pushy, but I want to repeat myself, a little more strongly (even though I could live with the current wording; as I've said, we're doing just fine like we are). Every person who has voted to keep the current wording (does that mean the whole paragraph, or just the first sentence?) has said something that suggests that you feel that "not truth" is referring to what's in someone else's head, a misimpression that we're correcting, it's not a statement that you feel represents your own take. All I'm asking for is wording somewhere in the paragraph, such as Blueboar's, which makes sure that we're all on the same page, that we all take that comment the same way that you do. - Dan Dank55 (talk) 21:08, 9 April 2008 (UTC)

Well, I also mean in my own head. When I edited happily as an anon back in 2003-4 and wrote large chunks of the ancient Rome articles, I was writing what was true in my own head, without checking each point. This was what made Wikipedia fun back in the day, along with the belief that if one was wrong, someone else's recollection would correct it. We aren't in that place any more - quality has improved to the point where my unaided recollection may not actually be more accurate than the current text, or the expected level of accuracy of new material - so yes, I restrict myself to what I can check is correct. I hope this answers what you were answering, and not that I've misunderstood it. --Relata refero (disp.) 21:31, 9 April 2008 (UTC)

"Every person who has voted to keep the current wording has said something that suggests that you feel that "not truth" is referring to what's in someone else's head, a misimpression that we're correcting, it's not a statement that you feel represents your own take." Huh? You've lost me. I don't think the phrase "Verifiability, not truth" requires any fancy parsing, nor do I believe the word "truth" is in any way ambiguous to an 'average reader'. Dlabtot (talk) 21:47, 9 April 2008 (UTC)

Dlabtot, I may have gotten it wrong, but when you said "The truth according to whom? ... Should WP tell people the answer to this question?", I took that to mean that you're referring to other people's notion of "the truth". That is, you were not saying "I know what truth is, and I know what verifiability is, and Wikipedia is interested in the former and not the latter," you were saying, "Hey dude, we have no obligation to accept something because you assert that it is true." Relata: I stand corrected, I think your username makes it clear that that paragraph looks good to you just like it is :) I am personally very attracted to your position (if I understand it correctly) that "Truth is irrelevant, everyone might be wrong, I don't care, I just record the words." I have lived long enough to be just that skeptical. But it's a decidedly minority opinion, even in Wikipedia but especially outside, and when people read the sentence that way, it seems to cause argumentation (as Amerique's links show). - Dan Dank55 (talk) 22:21, 9 April 2008 (UTC)

If this discussion has made anything clear it is that "truth" is not an objective criteria. It is important to right at the beginning of the page make it clear that truth is not our criteria for inclusion. _(1 == 2)^Until 22:32, 9 April 2008 (UTC)

Then we should be upfront: Wikipedia editors are allowed to knowingly include false information, as long as it is attributable to a source. I wonder how many of our readers are aware of this? Raymond Arritt (talk) 22:35, 9 April 2008 (UTC)

What are you talking about? Nobody has said that. _(1 == 2)^Until 22:40, 9 April 2008 (UTC)

Knowingly adding false information would clearly be disruptive editing. I would argue that deliberately misconstruing the arguments of fellow editors qualifies as disruptive editing as well. Dlabtot (talk) 22:44, 9 April 2008 (UTC)

To Raymond: if you think about it, trying to WP:WFTE has some elements of that. Also consider this: there's a particular movie actress' article I watch (under a sock since the person in question's linked to my RWI). Reliable sources have her birthday slightly wrong. There's also a chap who regularly comes along and changes the date to one off by three years, based on a thoroughly unreliable source. I revert this dastardly individual with equal regularity, keeping information I know is wrong in. In a nice cartoon-like manner, that expresses a reason for your (equally caricatured?) objection. --Relata refero (disp.) 00:02, 10 April 2008 (UTC)

What's being misconstrued? We've already had several people endorse the statement "If the whole world is lying, we repeat it quietly." In other words, we knowingly insert false information. Sounds like you're trying to have it both ways: we don't concern ourselves with truth, except when we do concern ourselves with truth. Raymond Arritt (talk) 00:02, 10 April 2008 (UTC)

Raymond, if you don't believe that a proper balance of reliable sources are reflecting the world as it actually is, that's its own problem. By not concerning ourselves with truth on an individual basis, we expect the project as a whole to asymptote towards truth. --Relata refero (disp.) 00:05, 10 April 2008 (UTC)

To humor the suggestion by Raymond arritt, then if the scientific consensus is that the moon is X kilometers wide, but you know for a fact that that is wrong because you went there a measured it, but your findings have not been accepted by reliable sources on the subject for some reason(like they don't believe you have a spaceship) then that truth is not welcome. We would have to go with the existing findings rather than accept this "truth" that contradicts them. We are not a creator of original research, we are not seeking the truth, we are reflecting existing knowledge as it is presenting by existing reliable sources on the subject. _(1 == 2)^Until 22:48, 9 April 2008 (UTC)

If you want to see how the general public views the truth, this Google search may be enlightening. _(1 == 2)^Until 22:53, 9 April 2008 (UTC)

I think the nature of Google's search is unlikely to give representative results. I can get comparable data in the other direction by using "true that" for the search, thus getting more casual phrasings. Phil Sandifer (talk) 13:48, 10 April 2008 (UTC)

Dlabtot, I'm guessing you're referring to the previous edit and not mine as disruptive, but tell me if I'm wrong. I guess I've said all I have to say here, let me just hit my topic sentence one more time, which is: if I am wrong, if we are not "talking past" each other, then I have no problem with keeping the wording just like it is. I think it will lead to more arguments, and when it does, we'll just just repeat the summaries and clear things up. (Hey! What a great use for summary boxes.) My point is: "Until", when you just said that "truth is not objective", I understand you to say that we can't be sure what truth is. But without Blueboar's explanation and scare-quotes on "true", the sentence as it stands contradicts you. It asserts that we do know the meaning of the word "truth", because we wouldn't use a word that we don't know the meaning of. If we say it, then it's our word, and we're responsible for assigning meaning. Again, I'm not trying to be argumentative; I genuinely believe that half the people here hear one thing when they read the sentence and half hear something else. If I'm wrong, then I have no problem leaving it like it is. - Dan Dank55 (talk) 22:55, 9 April 2008 (UTC)

How does saying that our criteria for inclusion is not truth make any sort of statement about the meaning of truth? We are only saying we don't use this method of decision making, that does not mean we are taking a stance on that method of decision making. Some web sites do use truth as a measure, and they have their own meaning, we are side stepping the whole sticky affair by making it clear we don't use truth, regardless of the meaning. _(1 == 2)^Until 23:10, 9 April 2008 (UTC)

I came to this page with concerns much like Phil's above. The problem I'm having is that the "wiki way" is not easily reducible to a sentence or two about verifiability, citations and truth. It is really a much more complex process that involves discussions about the nature of truth and consensus reality. The value of a wiki is that it models consensus reality and thus is a microcosm and/or a model of consensus reality.

A similar conversation happened at Wikipedia talk:Consensus. On that page, the emphasis for quite some time was on what consensus was or should be. It only improved when the discussion shifted to trying to understand the process of how consensus happens on a wiki. That same discussion shift would benefit this page as well.

So here is my big point: Reality is not some thing that exists out in the world, that we discover and write about on Wikipedia. Consensus reality is the fruit of an article maturing. When dozens, hundreds or thousands of people can read an article and think that it is basically correct, and the article undergoes no major changes, then it has succeeded. The reason to insist on verifiability is to help in resolving conflicts, not to give editors a reason to remove everything that is uncited and thus create conflict. The criteria is for verifiability, the criteria is not that everything is verified. This is an important distinction. I've seen far too many verifiable things deleted because they cannot be verified with academic or "respectable" sources. -- ☑ SamuelWantman 08:26, 10 April 2008 (UTC)

More pernicious is the idea that if it's not verifiable "on the web" that it is somehow invalid as a citation. Get thee to a library, folks. Dlabtot (talk) 08:46, 10 April 2008 (UTC)

kCoupled with complaints that a source is in an academic journal, therefore elitist, impossible to find, etc.Doug Weller (talk) 10:03, 10 April 2008 (UTC)

Very much agree with these last two points, though note that wikipedia asks for secondary sources in preference for primary sources.  DDStretch  (talk) 10:35, 10 April 2008 (UTC)

Strongly agree with everything above, with just a little caveat. We trust the mob instead of "big ideas", and, despite the fact that that's a recipe for disaster outside Wikipedia, it works well here, especially if you're willing to be patient. I propose a large, random, careful study of articles at roughly the maturity of WP:GAN to see when we feel the need to cite and when we don't. (And btw, the recent changes to WP:V will be very helpful, because I agree with Sam and Phil, I have also been a little disappointed by the apparent position of "WP:V gives me the right not to have to think about the article content; all I have to do is insist on cites everywhere, and that covers all questions." Hopefully that's fixed now.) WP:GAU is currently doing a style survey at the WP:GAN level, and it looks like just the right level: it's after people have given serious thought to their own preferences, but before they've been heavily influenced by WP consensus, at least for the ones who write more outside of WP than inside. - Dan Dank55 (talk) 15:37, 10 April 2008 (UTC)

P.S. On a personal note, I'd appreciate feedback about saying "agree with ... Phil" and other encouraging things. That's just how I roll; if I agree with someone, I'll say it. But I'm very aware that he's been a bit of an angerball, and that many editors react with exasperation if you encourage people whose position is that Wikipedia sucks in some way. I think if someone has something useful to offer, I want to err on the side of being encouraging, and only cut them off if that approach doesn't work, repeatedly. I don't usually ask for advice on "how to conduct a conversation", that's just too "meta", but it keeps on sticking in my brain, so maybe it's useful to ask. - Dan Dank55 (talk) 15:50, 10 April 2008 (UTC)

The sense I get from the discussions is that the sky will not fall if we remove that clause from the page, and just leave the part about BLP. Are we agreed that that's wrong ... that information that isn't about a living person, which hasn't been challenged and is unlikely to be challenged doesn't need to be removed? - Dan Dank55 (talk) 03:32, 4 April 2008 (UTC)

Would this change help? (I've reverted myself.) My suggestion is to remove the quote from Jimbo, and replace it with something like:

Any edit lacking a reliable source may be removed, but editors may object if you remove material without giving them a chance to provide references. It is important to strike a balance between being quick to removed unsourced material that is clearly wrong or in some way damaging, while at the same time making sure that challenges are reasonable, and that editors are given enough time to find supporting sources. Before you challenge unsourced material, ask yourself whether you really do doubt that the material is accurate. Unsourced material should not be removed simply because of a difference of opinion, or in an effort to make life difficult for other editors. As with all policies, this one must be applied with common sense.

The exception to this is with unsourced biographical material about living persons, which should be removed immediately if it is in any way contentious.

Thoughts? SlimVirgin ^talk|edits 07:09, 4 April 2008 (UTC)

A definite improvement. Maybe change the first sentence to "Any information that is suspected of being unverifiable may be removed." That both widens and narrows the net a little. Phil Sandifer (talk) 13:51, 4 April 2008 (UTC)

I think that "Unsourced material should not be removed simply because of a difference of opinion" could be misinterpreted. If I had a difference of opinion on the validity of an unsourced claim, then that would be legitimate cause to removed the unsourced claim. I am not sure, but I think the intention of the sentence is to make it clear that one should not game this rule to gain advantage in other disputes. _(1 == 2)^Until 14:08, 4 April 2008 (UTC)

Also, I cannot agree with Phils suggestion of replacing "Any edit lacking a reliable source may be removed" with "Any information that is suspected of being unverifiable may be remove". It is the burden of those seeking to include information that need to provide the sources. What is more, proving something cannot be verified is a bit like proving there are no blue chickens. It is an example of Negative proof. _(1 == 2)^Until 14:10, 4 April 2008 (UTC)

Hm. Perhaps my wording could be clarified. What I envision is this - unsourced material is of uncertain verifiability. If you look at something and suspect that it is, in fact, unverifiable, you may challenge it. At which point the burden of proof is on the person arguing for its inclusion to show that it is verifiable. So no negative proof - just suspicion of the negative. Phil Sandifer (talk) 14:21, 4 April 2008 (UTC)

There may be more work to be done, I agree, but the one clause I pointed out is the only one that is directly contradicted in its own paragraph (it's contradicted by the part in bold, no less). We've had a lot of active participation in the last few days, so this is the right time I think to make any needed changes, and no support that I can see for keeping that clause even if it weren't contradictory, so I'm going to be bold-ish and remove it ... WP:BRD advocates small steps at a time. - Dan Dank55 (talk) 14:25, 4 April 2008 (UTC)

The only worthwhile way to show that something is verifiable is to verify it. Any citation, no matter how spurious it is, is potentially verifiable, at least on the level of showing that the material exists. With the mixture of verification and reliability, presumably it could also be put to the test of accuracy, but that's even further out there. Mangoe (talk) 15:22, 4 April 2008 (UTC)

I think the instruction above is a significant departure from the Jimbo version ("It should be removed, aggressively, unless it can be sourced."). It is also long and unclear. This should be formulated shorter. We want to tell only two things:

(1) Any edit lacking a reliable source may be removed, and must be removed immediately if potentially damaging for anyone's reputation, especially in BLP articles, and

(2) prior to removing the unsourced materials (except in BLP), please notify other editors by placing the "unsourced" and other labels in the article and wait for at least two days prior to deletion of the unsourced material if the sources were not provided. Biophys (talk) 15:52, 4 April 2008 (UTC)

I would not put a definitive time limit on the wait... that is a judgement call that has to be made on an article by article basis. For example, what if an editor contacts you and says he can provide sources, but is busy right now and will not be free to attend to sourcing for a few weeks. Good faith requires that we give him that grace period. If you are going to say "tag, notify and wait"... then simply say "a reasonable amount of time". Blueboar (talk) 16:10, 4 April 2008 (UTC)

I think you're misreading the Jimbo quote - note that he is specifically targetting " random speculative 'I heard it somewhere' pseudo information." That is not equivalent to "all unsourced information." I sincerely doubt that Jimbo's intention is to introduce Cartesian doubt to Wikipedia - rather, I think he's talking about a particular class of information - the sort of plausible but mildly outlandish claim that raises an eyebrow. Which we've all seen in articles. This is a statement about unsourced and kind of dodgy-looking information - not about run of the mill unsourced information. Phil Sandifer (talk) 16:18, 4 April 2008 (UTC)

Fine, but let's make this text shorter and more clear. I would also put a recommended waiting timeframe before deletion. Let it be, for example, two weeks if two days are not enough.Biophys (talk) 16:33, 4 April 2008 (UTC)

I wouldn't want to see a recommended waiting time, because it will very much depend on context. The aim of introducing a new paragraph about this is to lighten the burden of rules a little, not to add to them. SlimVirgin ^talk|edits 17:23, 4 April 2008 (UTC)

←Can we clarify this idea of how long to wait? If we aren't really sure, then new editors aren't likely to be sure, either. WP:CfD allows 5 days discussion; some processes allow more time. Is deleting random text in a random article likely to need more time for discussion than the typical debate over deleting a category or an article? Why or why not? I'm not disagreeing with Blueboar that it should be "a reasonable amount of time"; I'm just saying that it won't hurt to contrast and compare with other more well-known processes. - Dan Dank55 (talk) 18:41, 4 April 2008 (UTC)

I don't think any firm timeline can be set. Many unsourced statements are just fine, despite not having an inline citation, and there's no reason to remove them at all. For example, I occasionally put a cn tag on my own writing to remind myself that I want to look up a source someday. On the other hand, if I think something is actually false, I will probably remove it pretty quickly. — Carl (CBM · talk) 18:46, 4 April 2008 (UTC)

I typically wait at least a few months after a ^{[citation needed]} tag has been added to remove the material, especially in articles that don't seem to have a lot of watchers. I don't believe this should be codified, however, but left as a matter of editorial judgement. Since we have the page history, nothing is really deleted anyway. Dlabtot (talk) 18:48, 4 April 2008 (UTC)

Argument from Jimbo is a useful guideline, but often a poor way to write policy. The suggested phrasing seems to describe the way sources are in fact requested, and probably should be requested. I agree with Phil and Slim on this point - David Gerard (talk) 20:46, 6 April 2008 (UTC)

I think that the new wording is not an improvement. It seems that the Jimbo quote is being sidelined to only cover "negative information about living persons". What he said was "I can NOT emphasize this enough. There seems to be a terrible bias among some editors that some sort of random speculative 'I heard it somewhere' pseudo information is to be tagged with a 'needs a cite' tag. Wrong. It should be removed, aggressively, unless it can be sourced. This is true of all information, but it is particularly true of negative information about living persons" I agree with the statement. The new wording takes us back to the bad old days were contentious stuff hung about for months or even years. Common sense is needed, but better that the the guideline errs on the side of deletion than retention. See below #This is not wrapped up --Philip Baird Shearer (talk) 09:06, 11 April 2008 (UTC)

I'm disturbed by the magnitude of the changes taking place here recently. The suggested use of {{fact}} etc tags now conflicts with Wikipedia:Biographies_of_living_persons#Remove_unsourced_or_poorly_sourced_contentious_material which suggests that they should never be used. -- John (Daytona2 · Talk · Contribs) 23:02, 8 April 2008 (UTC)

From the current wording of this policy, The exception to this is that any unsourced material may be removed from biographical material about living persons at any time, and should be removed immediately if it is or even might be contentious. What is conflicting here? Someguy1221 (talk) 23:05, 8 April 2008 (UTC)

It's the quote -

It also conflicts with the item listed in Wikipedia:Verifiability#Further_reading "it is better to have no information, than to have information like this, with no sources.".

To make changes of this magnitude requires community wide consensus afaic, until it has been gained, I will be ignoring it. -- John (Daytona2 · Talk · Contribs) 17:45, 10 April 2008 (UTC)

First, WP:VPP was invited into the discussion and showed up, there was an RfC, and people who care tend to keep an eye on WP:V. There have been maybe 50 (I'm guessing) contributors, and probably an order of magnitude more people who read along and presumably were fine with what was happening. It doesn't get any more "community wide" than that. Second, this conversation has not been archived ... if you want wider discussion, tell us where to issue an invitation. Third, what's not to understand about the quote? If someone says some kind of "random speculative 'I heard it somewhere' pseudo information", remove it. As you point out, the quote lives on in WP:BLP. Fourth, the information from Further reading was BLP information, so that quote doesn't disagree at all with what WP:V now says. Now ... what are you saying? - Dan Dank55 (talk) 19:11, 10 April 2008 (UTC)

50 people (0.1%) do not a mandate for change make. When, using the latest figures, in September 2006, there were 43,000 active editors on the English language Wikipedia. When no effort has been made to contact every active editor, in spite of such an excercise being relatively easy to accomplish - it regularly occurs on other issues. Given this, your comment that "It doesn't get any more "community wide" than that." is in my opinion arrogant and an attitude which is damaging to Wikipedia.

The comments, although highlighted on and referring to WP:BLP also refer to non WP:BLP unreferenced edits. They conflict with WP:Vs - "Alternatively, you may tag a sentence by adding the {{fact}} template" -- John (Daytona2 · Talk · Contribs) 09:23, 12 April 2008 (UTC)

It was more like the kind of issues discussed in WP:OWN than arrogance, which is why I decided to back off. I felt like you were threatening my personal accomplishment; that's what makes it time to leave :) In my mind, WP:V hasn't changed one bit on the support it gives to insist on sources when community consensus is that sources are needed, so it seems like not a big deal, and it seemed to me that an RfC and invitation to WP:VPP was enough. - Dan Dank55 (talk) 12:25, 12 April 2008 (UTC)

The changes I tried to make to this section seemed to have good acceptance when they were proposed by SlimVirgin above, and none of the people who have been reverting them have actually contributed to the existing discussion or registered their objection. Obviously, it is rather difficult to generate consensus over unstated objections.

To my mind, the logic of the change is straightforward - disruptive challenges of information are a genuine problem, and it is not our goal (nor has it ever been) to footnote every sentence - that's not a viable or realistic research practice. Challenges are for information that is thought to be unverifiable. Furthermore, the phrasing as it stands depends on Jimbomancy, and furthermore, as I've explained, seems to me to badly misread Jimbo's statement. Certainly spurious challenges and fact tagging has been considered disruptive and unacceptable in the past, and anybody who ran a bot to remove all unsourced information would be blocked on sight, so the clarification that challenges should be substantive, rather than a formal hoop-jumping procedure seem to be consistent with existing practice.

Would any of the people who disagree with this change be so kind as to elucidate their objections? Phil Sandifer (talk) 21:50, 6 April 2008 (UTC)

Dude, I'm trying to help.

• Read back over my suggestions; that is, follow those links and look around for things you care about that seem to already have support from many people. Then pick one small focus that you think is the most important, and work just on that.
• Your key point, as I see it, is that the arguments concerning when things should be cited got restated many times in many ways because that's what the people wanted, and in the process, WP:UNDUE attention was focused on just one side of what WP:V was designed to handle. Re-frame the debate, and do it fast before you're sunk. Say that you don't want to change what WP:V says about the importance and need for verifiability one bit; instead, you want to restore balance by paying attention to things that are just as important, such as why to cite, and when not to cite.
• Use fewer words. - Dan Dank55 (talk) 23:44, 6 April 2008 (UTC)

So no objections to the proposed wording then? Phil Sandifer (talk) 23:51, 6 April 2008 (UTC)

Challenges are not just for unverifiable content, they are also for content that is dubious to someone. I certainly agree the bad faith or disruptive challenges can be discarded, but we already do that. It is not reasonable that one to have to determine if something is unverifiable to challenge its presence, that would put the homework for the person who sought to include it on the shoulders of someone else who is simply trying to ensure accuracy. _(1 == 2)^Until 00:27, 7 April 2008 (UTC)

All that being said, the changes do seem to reflect my opinions adequately. Once again though I must object to the wording "Unsourced material should not be removed simply because of a difference of opinion" which makes it appear that a difference of opinion about the challenged fact is not grounds to challenge it, which it of course is. I assume this is meant to state that an unrelated difference of opinion should not result in a pointy challenge to a fact? Perhaps the wording "Unsourced material should not be removed in response to unrelated disputes" would correct this ambiguity, though I really do think this idea goes without saying. _(1 == 2)^Until 00:31, 7 April 2008 (UTC)

Well, I'd prefer wording that suggests that challenges are for content that is suspected of unverifiability - that is, we don't have to determine that it's unverifiable - we just need to be skeptical that it is. So the standard of proof for a challenge is doubt, and the standard of proof for responding to a challenge is a reliable source. Is there a wording you can think of that captures this (which is a point I think we're in agreement on, though I could be wrong.) Phil Sandifer (talk) 02:29, 7 April 2008 (UTC)

"You're not allowed to challenge that, because you haven't demonstrated doubt" is, I'm guessing, too big a change, too quickly. The most likely result of making big changes to one of the 3 core content policies pages is that Wikipedians will continue to insist that what's in their heads is the real policy, have a few protracted wars over it, and eventually revert this page, after wasting a lot of everyone's time. An approach which is scientific, gentle, and grounded in respect for community standards has a better chance of sticking. Simply do a large, careful, and (reasonably) random study to determine when uncited material tends to remain in articles, make and test some hypotheses about what the commonalities are, and add the results of the study to WP:V. - Dan Dank55 (talk) 02:48, 7 April 2008 (UTC)

If our goal it to write an accurate encyclopedia I think we should keep the burden fully on the shoulders of those who seek to include facts. Perhaps I am misunderstanding, what exactly is the difference between being dubious of a claim, and suspecting that you cannot verify a claim? If the challenge is to provide verification, the why should the challenge require an assumption that it cannot be verified? Why not just state that a challenge can be done if you are dubious of the claim? _(1 == 2)^Until 03:17, 7 April 2008 (UTC)

Fair enough - what about simply wording it as "any unsourced information may be removed if an editor finds it dubious?" Phil Sandifer (talk) 04:45, 7 April 2008 (UTC)

I'm getting sleepy so I may not be thinking straight, but there are two new sentences that I think probably won't survive: "Be sure challenges are reasonable" (because it's either redundant to the following sentence, or if it is stronger than the following sentence, then it's too big a change), and "Unsourced material should not be removed simply because of a difference of opinion, or in an effort to make life difficult for other editors." The three core content pages try to have as few words as possible, because we don't want to run the risk of constant tweaking. Policy pages that constantly change are even worse than bad policy. Making "life difficult" probably has the wrong tone; it seems to violate WP:AGF. - Dan Dank55 (talk) 04:16, 7 April 2008 (UTC)

To be clear, do you actually object to either piece of wording, or do you merely suspect others might? For my part, I think this is an important clarification - spurious challenges are a genuine problem we have in the article namespace, and we need to make clear that challenges are to be done because you think something is wrong, not because you feel obliged to challenge statements that lack sources. Phil Sandifer (talk) 04:45, 7 April 2008 (UTC)

I like the way "any unsourced information may be removed if an editor finds it dubious" sounds, short, simple, and to the point. _(1 == 2)^Until 04:54, 7 April 2008 (UTC)

Until, do you think someone could read your proposed sentence as support for "You're not allowed to challenge that, because you haven't demonstrated doubt"? Phil, you're quite right to ask for clarification. What I mean is that I could see myself reverting just those two sentences, if further discussion doesn't seem to support accepting them, and if no one else removes them, but I'm not in a hurry. I'm listening. Understand that there is a larger presumption against adding any new material to core content policy pages than any other page I can think of, so new sentences really have to be bullet-proof and have community-wide support; it really is completely unacceptable for them to change this month, change back the next month, get tweaked the next month, etc. Unstable policy creates FUD. Even if the comments from the current RfC support the change in viewpoint (and there's no consensus of that yet), that wouldn't be enough; we would need a notice on both WP:VPP and WP:COUNCIL#Current notices, if there is to be any solid shift in intent on this page, which is, of course, what you're after. And that might be a lot of trouble for nothing: I encourage everyone to look at WP:Flagged revisions, particularly the part most relevant to this discussion, WP:Flagged revisions/Quality versions. There's no definite timeline for this stuff; people are looking to see what happens on the German Wikipedia first. But it could easily come within 3 months, and then we'd have to completely refactor the discussion. As I said: if you reframe the debate to make it clear you're not attempting to affect the role of verifiability or prohibit people from being able to challenge material if they want to, and instead focus on supporting and describing community standards for when it is already considered unacceptable to challenge, that's not so big a change, and in that case, IMO, we wouldn't have to do a big WP:VPP thing to get support. (P.S. I am very nearly violating the "use fewer words" proscription, but I can't see which words could be left out.) - Dan Dank55 (talk) 13:14, 7 April 2008 (UTC)

I do, notably, think that the new proposed language (which is definitely fewer words, as it cuts the Jimbo quote) is not so much a change to WP:V's substance as an acknowledgement of actual practice. Flagged revisions may necessitate changes to policy, but I doubt it will change to a system where every statement really does need a citation, simply because that system is fundamentally untenable. I am thus inclined to make the changes needed to have WP:V reflect an actually employable practice. As for the wording, I don't think the misreading is likely - "an editor finds it dubious" clearly imposes a subjective and individual standard whereby only one person needs to find it dubious to challenge it. Phil Sandifer (talk) 13:46, 7 April 2008 (UTC)

There is a possibility that User:MikeGodwin, the WMF's attorney, will have an issue with removing that quote from Jimbo from a core policy page; both its presence and its effect may provide protection in lawsuits, such as this one. I'll go ask. - Dan Dank55 (talk) 14:29, 7 April 2008 (UTC) P.S. He answered, I'm asking for permission to quote him here. - Dan Dank55 (talk) 15:48, 7 April 2008 (UTC)

I am deeply skeptical of that reasoning, but if Mike says to reinsert it I will, of course, yield. Until then, though, I really don't see any objection to the wording as being discussed, so I'm going to reinsert it with the "if an editor finds it dubious" clause. Phil Sandifer (talk) 15:20, 7 April 2008 (UTC)

←I'm reasonably clued-in about legal things; my concern was a good-faith objection, but Mike Godwin isn't insisting that we keep either Jimbo's quote or the exact content of it, so we're fine. I think we would do everyone a favor if we come up with language that has a reasonable chance of not offending anyone too much either here or at WP:VPP, which I think would be more or less a waste of everyone's time, since we're potentially a very short time away from having to refactor the whole discussion in light of WP:Flagged revisions. Amerique made a wonderful suggestion above:

Why not just eliminate the unnecessary mention of "truth" in the first place, and open with "The threshold for inclusion in Wikipedia is verifiability"? This would save us the problem of trying to develop an overly complicated or simplistic explanation of what WP's relationship to, or stance on, "truth" is, and then we could keep this page focused on discussions of verifiability in terms of accuracy.

There's a problem with the sentence "The exception to this is with unsourced biographical material about living persons, which should be removed immediately if it is in any way contentious", namely, there are many humorous examples of things you couldn't imagine people cared about that they care very much about. So in BLP material, it should be absolutely clear that anyone can remove material at any time that is not sourced, for any reason or for no reason. So, let's try: "The exception to this is that any unsourced material may be removed from biographical material about living persons at any time, and should be removed immediately if it is or even might be contentious." There's also a problem with "Unsourced material should not be removed simply because of a difference of opinion, or in an effort to make life difficult for other editors." The tone does not come close to meeting community standards for a core content page; it assumes that someone is taking some spurious action merely out of spite. It's got to go. I see that the other short sentence I had an objection to is gone, via your reversion. I have the same problem the other editor had with the "balance" bit. Does anyone have a problem with: "Unsourced material that is clearly wrong or contentious should be removed; challenges to other unsourced material should be reasonable, and editors should be given reasonable time [I'm following Blueboar's advice here to say "reasonable"] to provide references." After saying "reasonable" (twice!), I think it would be both redundant and too wordy for a core content page to add "As with all policies, this one must be applied with common sense." I am agnostic on "Before you challenge unsourced material, ask yourself whether you really do doubt that the material is accurate." There's certainly a hint of lack of AGF there. - Dan Dank55 (talk) 16:48, 7 April 2008 (UTC)

I think that mentioning that our threshold is not truth is important, because many new users come here thinking just that and can't understand why we won't accept their truth. _(1 == 2)^Until 18:07, 7 April 2008 (UTC)

What attracted me to it is the need to be sparse on core content pages, not to say anything that's even a little bit hard to parse, so that the page tends not to change over time; the idea of avoiding metaphysical arguments over what truth is on WT:V, so that we just focus on verifiabilily and accuracy here, and argue truth on some other page, sounds great to me. But it's true that it has said "not truth" for a long time, and if people are counting on finding that phrase here, if it's annoying if they don't find it, I have no problem with leaving it. - Dan Dank55 (talk) 17:03, 7 April 2008 (UTC)

I'm reminded of a Depeche Mode song:

You'll see your problems multiplied

If you continually decide

To faithfully pursue

The policy of truth

Words to live by. Developing an actual policy on truth could be fun for all the wiki-drama it would cause, but i don't think it could ever be practical. Amerique^dialectics 18:27, 7 April 2008 (UTC)

←Okay, here's my conversation with Mike Godwin (General Counsel for WMF), word for word (after the formalities). This is useful to prevent our deletion of Jimbo's quote from prompting uninformed "legal opinions" in future discussions:

I'm asking for a legal opinion here; if you you feel that either the actual presence of that statement in a core content policy, or its effect, is useful to WMF legally, then I feel pretty sure that we'll all take your word for it and leave it. Thanks, Dan.

I am guessing Jimbo was trying to address the risk of harmful defamation here more than simply the risk of legal liability.

Certainly any controversial or negative statement should be sourced, especially if it concerns the biography of a living person. Otherwise, controversial statements should be removed.

--Mike

Thanks for the quick reply ... man, WMF folks are on the ball. With your permission, I will reproduce my last sentence and your reply verbatim at WT:V. Dan

Sure, but keep in mind that this is an off-the-top-of-my-head opinion.

--Mike

- Dan Dank55 (talk) 18:30, 7 April 2008 (UTC)

Another note from Mike Godwin, verbatim: "My feeling is that you don't need to cite a source for uncontroversial claims such as "George W. Bush has served as president of the United States from 2001 to the present, and his term in office is expected to end in January 2009. --Mike" Note that we still have the "impose one's own view of 'standards to apply' rather than those of the community" clause from WP:POINT if someone starts insisting that things like this need a cite; and as soon as we're done nailing down the language, I'm going to suggest a careful random survey to help nail down when it may be actually inappropriate to ask for cites. Still, does anyone want to suggest a tweak to the BLP language? - Dan Dank55 (talk) 23:29, 7 April 2008 (UTC)

What kind of tweak do you feel is needed, Dan? SlimVirgin ^talk|edits 23:50, 7 April 2008 (UTC)

Well, my feeling is that what Mike said gives some support to what I suggested above, which is "The exception to this is that any unsourced material may be removed from biographical material about living persons at any time, and should be removed immediately if it is or even might be contentious." Note that up until recently, this part read, "Do not leave unsourced information in articles about living persons." But many BLP articles and sections have unsourced information that no one has removed; as Mike points out, some things don't need sourcing. So saying that it "may be removed ... at any time" seems to me to better reflect actual practice. Of course, a core content policy is like a constitution; make one little change and it can have unintended effects on the "lower court rulings". So if anyone can think of an unintended consequence, please speak up. If Mike was suggesting that we should further weaken it, I can't support that, that would be too big a change from what it was, too quickly. Let's give changes a chance to percolate before we even consider that. - Dan Dank55 (talk) 01:46, 8 April 2008 (UTC)

Argumentum ad Godwin is not a sounder approach than Argumentum ad Jimbo. Let's stick to thinking about the issues. What are the cases where it is appropriate to challenge a statement? Phil Sandifer (talk) 02:05, 8 April 2008 (UTC)

Dan, the policy currently says, "The exception to the above is with unsourced biographical material about living persons, which should be removed immediately if it is in any way contentious." Are you saying that's too strong, or not strong enough? SlimVirgin ^talk|edits 02:06, 8 April 2008 (UTC)

Phil, don't put words in my mouth. I didn't say that the argument was over because Mike weighed in, I said that his position seems to support mine. He's one contributor among many. Your question is premature, because a clever answer isn't good enough; the right answer, the one that respects decisions made on over 2 million pages, can only be achieved by a large, careful, random survey. We also don't need to answer that question to nail down the language here. Slim, I'm saying the current wording is not strong enough, because it's too far from what was here for a very long time: "Do not leave unsourced information in articles about living persons." I'm suggesting a fairly strong change of wording: "may be removed ... at any time", which changes a requirement (which wasn't always honored) to permission. I'm suggesting that because that obviously reflects better what editors do and don't do. Many BLP articles and sections have unremoved, unsourced information, and the sky has not fallen; WP:BLPN doesn't have any trouble with these issues. I'm also suggesting it because it seems to have support in this discussion. The current wording goes farther; it doesn't say anything at all about uncontentious BLP material. Whether that would be a good change or not is, I think, irrelevant; it's too much, too fast. Let's make just the change I suggested (unless someone wants to suggest better language), and let the effects percolate, and clean up the unintended consequences; then later we can see if there's support for making a further change. - Dan Dank55 (talk) 02:32, 8 April 2008 (UTC)

Unsourced contentious information in BLPs should be removed on sight. Unsourced uncontentious information is less of a problem, but there is in fact an "ought" involved for contentious unsourced information on BLPs. Phil Sandifer (talk) 03:05, 8 April 2008 (UTC)

That said, and unindenting, we've gone a few rounds on the "any unsourced material may be removed" line. I'd really like to see this clarified - we don't mean it. Anybody who made a bot to remove all unsourced material would be banned in seconds. There are *clearly* cases where removing unsourced material is simply the wrong thing to do. What are the circumstances where unsourced material should be removed? That is, how can we better delineate the domain here? Preferably in a few marvelously elegant words. My best phrasing so far is "that an editor finds dubious," which requires only that somebody actually look at it and go "Really?" It's a very low bar to clear, but it does stress from the beginning that this is only to be done after a modicum of thought. Phil Sandifer (talk) 03:05, 8 April 2008 (UTC)

Phil, like I said, I was hoping that we could put some language on the page that has a chance of not offending anyone on this page or in WP:VPP. If, 24 hours from now, we haven't reached a rough consensus on the language, then we're kind of in a pickle, because everybody here has decided that the old language had problems, so we can't just revert, but we can't represent to all of Wikipedia that the opinions of a few people are now a core content policy, either. At that point, we'll have no choice but to go to WP:VPP. I have explained why several of the sentences currently on the page won't be accepted at WP:VPP, but if you need to see for yourself, then that's what we'll do. I was hoping not to make this a big deal. - Dan Dank55 (talk) 03:23, 8 April 2008 (UTC)

Baloney. This page didn't become the mess that it is because of widespread and careful community input. If it had, it would actually be followed in its prior form, which it wasn't. And the changes were certainly not made based on testing across two million articles - they were made based on a desire to try to legislate our most pathologically messy articles into submission. Should any objections of fact (as opposed to insisted objections that will some day arise) we can discuss them. But I am hard-pressed to address concerns that have not been raised.

I ask my question again, however - a question that is designed to get actual concerns as opposed to your hypothetically-existent ones. When are challenges appropriate? Phil Sandifer (talk) 03:27, 8 April 2008 (UTC)

Challenges are appropriate for contentious or exceptional material or material that an editor believes in good faith to be inaccurate or mischaracterized, to give a few examples. Spurious or bad-faith challenges should be dealt with as disruption, but it's not a problem as far as I know. The assumption of good faith appears to be working. Dlabtot (talk) 21:58, 8 April 2008 (UTC)

Dlabtot, you frame the issue well here (and I apologize again for our formerly starting out on the wrong foot). However, in my experience, (presumably) bad-faith challenges are a problem unresolved by disruption-handling methods, particularly when any of several editors I know deletes sourced content as "unreliable" (or "nonnotable"!) without initiating debate upon reliability or properly going to the RS noticeboard. I will address this in a separate section. JJB 01:08, 14 April 2008 (UTC)

I just put an "inuse-section" template on the Burden of Proof section, because the current conversation that talks about "current language" won't make any sense if people wander by and change the language around. I also wanted to warn other editors not to take the current language too seriously while we're trying to figure out what to do with it. I encourage everyone to weigh in so that we can at least put down some wording that we are semi-confident about within 24 hours; this is, after all, a core content page, and much-used. - Dan Dank55 (talk) 02:57, 8 April 2008 (UTC)

I'm very puzzled by this. Right now we have what seems a mostly uncontentious version, and no rephrasing is gaining any particular traction. I would be surprised if, in 24 hours, there are any consensus changes. Phil Sandifer (talk) 03:05, 8 April 2008 (UTC)

Then read what I wrote above; I gave you arguments for why some of those sentences do not meet community standards. - Dan Dank55 (talk) 03:17, 8 April 2008 (UTC)

Both of the sentences you identify are clearly not changes to practice - spurious challenges and overzealous challenges are already clearly frowned upon. Things like Ashburn, Chicago where virtually every sentence was, until I just removed them, fact-tagged are not OK. There may be phrasing tweaks to be had, but the message that challenges need to be based on more than the desire to remove unsourced information for the sake of doing so is non-controversial, as evidenced by nobody disputing it here. Phil Sandifer (talk) 03:23, 8 April 2008 (UTC)

I'm continuing the conversation with Phil by email for the moment; we don't seem to be connecting, and we seem to have lost other contributors. Phil has suggested that since his wording is on the page at the moment, and no one is speaking up, that proves that everyone agrees with him and not with me. To test this theory, I may change the wording to the things I've suggested tomorrow morning, just to gauge the response. - Dan Dank55 (talk) 04:10, 8 April 2008 (UTC)

Okay, since this is a core content page and we can't simply take a break from having a stated policy, I went ahead and made the edits that I felt absolutely had to happen in WP:V#Burden of evidence, with a short but hopefully sufficient reason for each in the edit summary. We will survive WP:VPP much better if there is wide discussion; please share your point of view, everyone. - Dan Dank55 (talk) 09:33, 8 April 2008 (UTC)

Jayjg made the edit just before mine, way above, and I answered up there, regarding "not truth". Other than that, the only edit that might be unclear is the one that says: "Until Apr 6, this was 'Do not leave unsourced information in articles about living persons.' Too big a change to imply it's not ok to remove unsourced noncontentious material)". Actually it was stronger than that (I was at the character limit), it said "or at all about living persons". That was changed to the language that was there yesterday because we all knew that that rule was not generally followed, or even regarded as the right rule. But the language yesterday was too big a jump all at once: it went from saying that you must remove it to implying that, for noncontentious material, you can't remove it without a very good reason. Regardless of whether that's the eventual language we want, the norm on any policy or even guidelines page is to make changes methodically, wait for reaction from the community, and fix the unintended consequences, before moving on to the next step. - Dan Dank55 (talk) 09:58, 8 April 2008 (UTC)

I think that the new wording is not an improvement. It seems that the Jimbo quote is being sidelined to only cover "negative information about living persons". What he said was "I can NOT emphasize this enough. There seems to be a terrible bias among some editors that some sort of random speculative 'I heard it somewhere' pseudo information is to be tagged with a 'needs a cite' tag. Wrong. It should be removed, aggressively, unless it can be sourced. This is true of all information, but it is particularly true of negative information about living persons" I agree with the statement. The new wording takes us back to the bad old days were contentious stuff hung about for months or even years. Common sense is needed, but better that the the guideline err on the side of deletion than retention. --Philip Baird Shearer (talk) 09:14, 11 April 2008 (UTC)

Talking about what effects this might have is the natural next step; could we perhaps move this down towards the end and the active conversation? Or give it its own section at the end? - Dan Dank55 (talk) 12:42, 11 April 2008 (UTC)

I think it is better that the wording in the section should be reverted to the version that was here until the recently proposed changes and then we can start a new section about changing it at the end of the page. --Philip Baird Shearer (talk) 15:55, 11 April 2008 (UTC)

Okay. I'm very happy with the level of discourse, and what we've done. I've had quite a lot of input into the discussion, so I'm going to back off for a while from participating because of the general principle behind WP:OWN; or maybe "If you love something, let it go." - Dan Dank55 (talk) 16:19, 11 April 2008 (UTC)

See below #Changes to "burden of evidence" section 2 --Philip Baird Shearer (talk) 12:16, 12 April 2008 (UTC)

I have several problems with the recent changes to the burden of evidence section that have recently been made. So I have reverted the changes, but left a couple in the first sentence that I think improve it.

I have also rearranged {{fact}} and moving it to the talk page as I think that the emphasis should be that way around. However I am not wedded to either of those changes and if anyone wishes to revert I will not worry.

The problems as I see it are with these changes: Before you challenge unsourced material, ask yourself whether you really do doubt that the material is accurate [sentence 1]. Unsourced material should not be removed simply because of a difference of opinion [sentence 2]. and the removal of the Jimmy Wales quote.

The first sentence is a breach of the assumption of good faith the information may well be accurate but as JW says in the quote "'I heard it somewhere' pseudo information", just because one has heard it does not mean it is accurate. For example I do a lot of work on battles, people often add numbers to men involved and casualties to the info boxes, partly I think because the info boxes look empty without them and they are crying out for those factoids. Now when a person adds a number to the box it may be accurate or it may be a guess, without a source there is no way to tell, so demanding a source for the numbers should be done whether or not the "you really do doubt that the material is accurate".

To give an example of such a thing. See the Battle of La Suffel and this edit [5]. I reverted the change not because I thought it was wrong but because I did not know it was right. As as can be seen on the talk page although the addition had been in good faith it was a guess. Later after more research I found a source that confirmed the guess and I was able to readd the information. But the wording of sentence 1 could have made this course of action difficult if the person making the guess was not a reasonable editor. In cases like this which is better to include information that might be right or not to include it at all? --Philip Baird Shearer (talk) 12:14, 12 April 2008 (UTC)

The second sentence is the wrong way around "Unsourced material should not be readded simply because of a difference of opinion", otherwise it nullifies the lead sentence of the section "The burden of evidence lies with the editor who adds or restores material.".

The Jimmy Wales quote: "I can NOT emphasize this enough. There seems to be a terrible bias among some editors that some sort of random speculative 'I heard it somewhere' pseudo information is to be tagged with a 'needs a cite' tag. Wrong. It should be removed, aggressively, unless it can be sourced. This is true of all information, but it is particularly true of negative information about living persons." is about more than just negative information about living persons and I think should remain in this section, (and not relegated to the obscurity of the ghetto of biographies on living people), as over the last year many many contentious articles have been improved out of all recognition thanks to the structure the PROVEIT section and the quote helps assert the point of view that all contentious material should be cited particularly when the editor who is defending the insertion of uncited material is passionate and knows that it is neither controversial or wrong. --Philip Baird Shearer (talk) 12:06, 12 April 2008 (UTC)

My problem is that, reading that quote, I think it's serially misinterpreted in a way that is unfortunate, especially given that Jimbomancy is problematic to begin with. Simply put, that quote does not appear to me to be supporting the policy as it was written. The claim does not seem to me to be "all statements must be sourced" or "removal is preferable to the citation tag" (both of which are, frankly, idiotic positions that, if applied across the encyclopedia would gut our coverage and fork the project). The key phrase is the sort of information Jimbo describes - it's a particular type of bad information that we're all familiar with - the maybe-sorta plausible but unsourced and kind of rumory pseudo-information that most plagues BLPs, but is a problem for all articles. To expand that statement to the sort of citational fascism that the burden of evidence section previously suggested was unwarranted and mandated a practice that was controversial at best and rejected wholesale by practical consensus at worst. Phil Sandifer (talk) 14:02, 12 April 2008 (UTC)

No one is suggesting that all comments must be sourced just "All quotations and any material challenged or likely to be challenged". No one is interpreting the quote in the article, one is free to draw ones one conclusions having read it (something someone can not do if it is not in the section). Also you have not addressed the my criticisms of the two new additions sentences that have been added to the section. Please address these issues before altering the text. --Philip Baird Shearer (talk) 15:19, 12 April 2008 (UTC)

As it stands your version puts the bar at likelihood of challenge, but contains no provision discouraging unreasonable or pedantic challenges - in other words, it puts all unsourced material into a capricious danger zone. With no check on when one ought to challenge, given the mechanical way in which policies apply, all material is "likely to be challenged." In effect your reversion removes all meaningful checks on disruptive source demands, to say nothing of the simply idiotic ones. A clear consensus exists that the wording you advocate is problematic. If you have issues with the attempts to improve it, look at the wording and see how it can be improved. But outright reversion is unhelpful in an attempt to garner consensus - it is clear that the old wording is unacceptable. As for your specific concerns, I quite strongly feel that removing and challenging material without a genuine suspicion of error or unverifiability is bad practice. If, looking at the information, you do not even have a doubt of its accuracy you should not challenge or remove it. Doubt is a very low bar to clear. I am less attached to the "difference of opinion" wording, due largely to the fact that I think "difference of opinion" is weak wording that has no real content. However I think your objection, as it stands, is deeply destructive. To be perfectly blunt, challenging and removal of unsourced material is not an inalienable right that can be exercised without judgment. Material challenging must be done with a modicum of care and sense. As for the Jimbo quote, no policy should depend on a Jimbo quote for its justification. If Jimbo's point cannot be justified on its own merits and without an appeal to Jimbo than it is not suitable for a policy statement. Phil Sandifer (talk) 16:40, 12 April 2008 (UTC)

My version is the version (with some minor modifications) that has been here for some considerable time. It is you who wishes to change the wording of the section so I think it would help, if before those changes are made, that we first to concentrate on the two additional sentences that I have highlighted above, as those are the two major additions to the text. The differences we are discussing are of emphasis and as such there is no reason why we can not talk about them first and agree on changes before they are made. I have given you an example of why I think new sentences are not appropriate and I think perhapse we need to talk through the ramafications of those sentences before they are inserted. I would appreciate it if you would consinder the example I have given above about Battle of La Suffel and if you see how the wording may effect such issues. --Philip Baird Shearer (talk) 16:53, 12 April 2008 (UTC)

I don't, honestly - it seems to me that your revert (though I think a fact tag would have been more appropriate) was based on exactly what such challenges should be based on - a sense that something isn't quite right from somebody who knows stuff about the issue. It would, on the other hand, be wholly inappropriate for me to go and remove such information - I've not even heard of the battle, and have no business challenging anything but the most clearly unusual information in such an article. Phil Sandifer (talk) 17:00, 12 April 2008 (UTC)

If I had placed a fact tag there and it had turned out to be another chap, then for the time it was there the Wikipedia article would have been inaccurate. In such cases it is better not to have the commanders name in the text than to put in one that might be wrong. It is a matter of which does the less harm. No information is better than incorrect information. And as to the specific formular of words it is not that I "really do doubt[ed] that the material [was] accurate" it is just that I did not know it was correct, after all I could have added that information myself when I wrote the article from the unreliable sources I had found on the net, but I do not think that is how Wikipedia articles outgh to be written and hence my objection to the wording in that sentence. --Philip Baird Shearer (talk) 17:34, 12 April 2008 (UTC)

I am with PBS here, for two reasons. First, the sentences "Before you challenge unsourced material, ask yourself whether you really do doubt that the material is accurate. Unsourced material should not be removed simply because of a difference of opinion" are a non-sequitor. The first deals with challenging, the second with removal. I happen to believe that if someone really really is of the opinion that material is inaccurate, to the point where, in their opinion, it does damage to the article, it is perfectly reasonable to remove it to the talk section until it is verified (although I have no objection to someone instead just adding the fact tag). Second, I think "It should be removed, aggressively, unless it can be sourced" gets to the heart of verifiability. What is essential here is the meaning of the word "verifiable," which is different from "verified." Verifiable means that one could find a source for it, not that one has found a source for it. And I think it is essential to the integrity of the encyclopedia that substantive claims made by the encyclopedia (by which I mean: not the claim that the sky is blue, but the claim that the sky is blue because the molecules that make up the atmosphere absorb and eventually radiate high-frequency radiation from the sun mor than low-frequency radiation) be verifiable. They may not yet be verified, but it must at least be possible to trace them to some source. This is all Jimbo is saying, "...it can be sourced..." He is not saying "unless it is sourced." If we think there could be a source for a claim in the article, it is reasonable to ask for one and depending on how contentious one thinks the claim, it is reasonable to add a fact tag, or remove it to the talk section, until someone provides a source. If a proposition central to the topic of the article simply cannot be verified, if we know that no one will be able to find a source for it, well, then it should indeed be removed. Slrubenstein | Talk 17:54, 12 April 2008 (UTC)

If you actually removed it because you were wholly neutral on whether it was accurate or not then I am inclined to say you should not have removed it, simply because, were that standard applied across the project, we would rapidly gut our accumulated coverage. Anyone who spent their time loading random articles and removing all unsourced information they did not know to be accurate would be rightly banned. Ergo it is clearly not accepted practice to do so. Phil Sandifer (talk) 21:08, 12 April 2008 (UTC)

I am confused which "you" are you referring to? Do you mean "one"? --Philip Baird Shearer (talk) 21:23, 12 April 2008 (UTC)

Actually, I meant you and screwed up the indentation - it was a reply to your comment about the battle article intended to note that if you really did not have some sense that the information might be inaccurate (as opposed to merely a lack of sense that it is accurate) then it was inappropriate to remove it. Phil Sandifer (talk) 21:34, 12 April 2008 (UTC)

I think it depends on the page and the age of the information which is being reviewed -- see below. --Philip Baird Shearer (talk) 05:58, 13 April 2008 (UTC)

PS writes, "Anyone who spent their time loading random articles and removing all unsourced information they did not know to be accurate would be rightly banned." There is a slippage here from unsourced to inaccurate. If we stick to one thing, it simplifies the matter. People shoud not remove something solely because it has not been verified. But people should remove something because they think it is unverifiable. There is a third possibiity - people think it is verifiable but it has not yet been verified. I think PBS's edits clarify what to do in this situation. Slrubenstein | Talk 21:52, 12 April 2008 (UTC)

If I understand correctly, though, we basically agree on all three of these circumstances - simply unverified, leave it be or make a note on the talk page. Believed unverifiable, remove. Think it needs verification, add a fact tag. Phil Sandifer (talk) 22:07, 12 April 2008 (UTC)

Yes. The only thing I would add is that in some cases i think something - something I (assuming I am a well-informed editor) consider especially unlikelyu and controversial that sould be removed to the talk page pending verification. This way it is not lost - it stays on the radar and people hopefully wil find the source - but it does not risk compromizing the quality of the article. Slrubenstein | Talk 22:20, 12 April 2008 (UTC)

OK. Then since we both seem to agree that challenging material (i.e. cases 2 and 3) is a thoughtful, not automatic process, how can we change the wording of the policy to clearly reflect this fact? Phil Sandifer (talk) 22:28, 12 April 2008 (UTC)

I think in practice how one deals with information it depends on how long the information is on the page:

• New additions to an article (typically on one's watch page): How credible is information and how well cited is the other information already on the page. In the example of my edit to the Battle of La Suffel, an article that was fully cited and on my watch list so deleting the information and discussing it on the talk page is the right approach. If it is new information to an article that does not carry many citations then use the template {{fact}} because it is much easier and more reasonable to ask an editor who has just added information to provide the source, than an editor who added stuff months or years ago and may not have the original source to hand.
• Old information on a page which is being reviewed for accuracy: How credible is the information? Unless the information is obviously wrong (in which case remove it), check what one can with easily available sources and "consider taging a sentence by adding the {{fact}} template, a section with {{unreferencedsection}}, or the article with {{refimprove}} or {{unreferenced}}", put comments on the talk page asking for citations with a link to WP:PROVEIT, then come back in about a month and start to prune the article if citations have not been provided (As I did recently to the Jolly Rodger article -- an article that still needs lots of work!).

Phil Sandifer, I if I understand your position then I disagree with you on the level of citations needed. If you look at the article the Battle of Waterloo as it was at the start of 2007 and how it was at the start of 2008, the information provided by the text on the page has not changed much, but the level of citation has improved greatly. Much of the information on the page had been there for a number of years, long before citations had been seen as desirable so there weren't many. Adding the citations was done by a number of editors working on the old information and retrofitting citations to the article using the {{fact}} template and discussing problems on the talk page. Fully citing an article makes the article much easier to maintain for interested editors, and it makes the article much more reliable/trustworthy for passing readers of the article. If Wikipedia article is to be considered a reliable source then fully citing all factoids is essential. --Philip Baird Shearer (talk) 05:58, 13 April 2008 (UTC)

I am reading a consensus that challenging material is a thoughtful, not automatic, process, and agree with Sandifer that this needs to be clarified. However my own prior proposal is different enough from this talk topic that I will give it a new section below. If it happily subsumes this discussion, well and good. JJB 03:16, 14 April 2008 (UTC)

The sentence "Do not leave unsourced information in articles about living persons." suggests that every sentence that does not have a source should be removed. I suggest it be rewritten to something like "Do not leave unsourced quotations or any material challenged or likely to be challenged in articles about living persons." --SaskatchewanSenator (talk) 05:32, 13 April 2008 (UTC)

I think this should be "Do not leave unsourced information about living persons in articles". --Philip Baird Shearer (talk) 06:06, 13 April 2008 (UTC)

Do you mean that every bit of information without a reference in a Wikipedia:BLP should be removed?

Perhaps referring to the Sources section of BLP would be better than trying to include specifics on BLP's here.--SaskatchewanSenator (talk) 07:43, 13 April 2008 (UTC)

It is not just biographies where one has to be careful of with publishing information about living people, it also applies to any article where information about a living person or an organisation published. Is the average editor of a Wikipedia article able to judge if the information is libellous under American law? I do not know the situation under US law but "Almost uniquely in English law, in libel cases the burden of proof lies with the author / publisher and not the complainant. In other words, you have to prove that what you write is true. The person you’ve targeted does not have to prove that you’re wrong."[6] So perhaps include in the sentence something along the line of "Do not leave unsourced information about living persons or organisations in articles that damages their reputation.". --Philip Baird Shearer (talk) 10:07, 13 April 2008 (UTC)

Good points. I can see you've laboured over the wording, but I suggest "Do not leave unsourced information that may damage the reputation of living persons or organisations in articles.". Editors should err on the side of caution.--SaskatchewanSenator (talk) 17:19, 13 April 2008 (UTC)

Better still. Does anyone want to suggest any other wording or make a comment? --Philip Baird Shearer (talk) 19:09, 13 April 2008 (UTC)

Support. Aggressive removal pertains to material potentially damaging to potential litigants and other very clear-cut cases like vandalism. JJB 03:16, 14 April 2008 (UTC)

Could you answer two questions, please? 1. Do scientific biological databases qualify as reliable sources? 2. Do the recommendations about "primary" and "secondary" sources apply to "primary databases" (original depositary of data) and the "secondary" (derivative) databases? Should we prefer the "secondary" sources if appropriate?

As a particular example, one can compare a "primary" database PDB: 2axt​ and a "secondary" database PDBsum 2axt. What link would you recommend to use? I believe that the second (a derivative) database is better and more convenient because it makes a reference/link to the primary PDB source and other primary sources, but the primary source does not - see discussion of Template_talk:PDB. Thank you.Biophys (talk) 20:05, 12 April 2008 (UTC)

A tentative response: it depends on what the database is being used for. If it is used to make purely descriptive claims about something, maybe, if it is "published" in some form that makes it likely to find in a public library or on-line. But if a Wikipedia editor uses the database as a source of data to support an analysis proposed by the Wikipedia editor, that would violate WP:NOR. Slrubenstein | Talk 10:31, 13 April 2008 (UTC)

Would it be possible to clarify the policy to say that (most) blogs are lousy secondary sources? Quite frequently, when discussing current events, a blog is the primary source for a fact. In particular, when one of the participants in a notable event makes a public statement via a personal blog (or a comment on another blog), that's the definitive primary source for the fact that so-and-so said such-and-such.

For example, when talking about Alberto Gonzales, it is notable and newsworthy that certain accusations were made on certain dates in the Talking Points Memo blog. The truth of those accusations was not established by the postings, but because they were verified and led to his resignation, the fact of the accusations themselves is notable, and a reference to the original blog posts is the best possible citation.

If I'm talking about the history of DragonflyBSD, Matt Dillon's announcement is a primary source document for the fact that Matt Dillon announced the project on that date.

Or when talking about the recent Expelled brouhaha, PZ Myers' post to Pharyngula (blog) is central to the story.

I think WP editors are generally acting sensibly in this regard, but I think the policy wording in WP:SELFPUB needs updating to reflect reality. These days, blog postings are sometimes noteworthy events in and of themselves.

71.41.210.146 (talk) 12:08, 13 April 2008 (UTC)

This is a good point.

Like all other primary sources, to be used with caution. The Pharyngula post, which I agree is absolutely central to the Expelled article, is definitely a primary source. As it is frequently cited as central to the Expelled controversy, there shouldn't be any problem with citing it even as things stand now.

In other cases, it might be the case that the date isn't important. If the blog post in question has been referenced as relevant, a couple of uncontroversial details from it are hardly going to be objected to unless the objector is being disruptive. --Relata refero (disp.) 13:24, 13 April 2008 (UTC)

For background, see here. I have realized that to get anywhere with this, we've got to focus on one small issue at a time. Therefore, I suggest for now we discuss whether the first sentence of WP:SELFPUB should be reworded. I suggest we change it from Material from self-published and questionable sources may be used as sources in articles about themselves (an unclear and grammatically questionable statement) to the following:

Self-published and questionable sources may be used in articles about themselves to reference relevant information, so long as:

1. it is relevant to their notability;
2. it is not contentious;
3. it is not unduly self-serving;
4. it does not involve claims about third parties;
5. it does not involve claims about events not directly related to the subject;
6. there is no reasonable doubt as to who authored it;
7. the article is not based primarily on such sources.

I don't see how this would be an improvement, or even a real difference. When they are used in articles about themselves, then it's obviously going to be to reference relevant information, but then again, almost everything about a person or group is relevant information in an article about that person or group. We try to keep out vanity material, but the lines are a little blurry. The qualification is anyway repeated in the list of restrictions: not unduly self-serving etc. SlimVirgin ^{(talk)(contribs)} 07:21, 24 March 2008 (UTC)

Well, I'm trying to take this one tiny step at a time, because I now realize that's the only way consensus will ever be even remotely achieved. Part of my reasoning is that the current first sentence is unclear and grammatically questionable. And this rewording would show that it's the information being referenced which is subject to these restrictions, not the sources per se. PSWG1920 (talk) 07:34, 24 March 2008 (UTC)

I've fixed the grammar. Thanks for pointing it out. I think "relevant to their notability" needs to be removed, and in fact I can't remember why or when that was added. It's precisely for material that isn't relevant to notability that we most often use self-published sources in articles about themselves e.g. where they were born, pivotal childhood experiences etc. For material that's relevant to their notability, we usually have the normal reliable sources. SlimVirgin ^{(talk)(contribs)} 07:43, 24 March 2008 (UTC)

Yes, I had thought that as well. I would also say that the material could still be interpreted as referring to the entire source rather than just the specific information referenced, so perhaps that can be further clarified. PSWG1920 (talk) 07:49, 24 March 2008 (UTC)

Well, the material used still seems a bit vague. What about the information cited? PSWG1920 (talk) 08:10, 24 March 2008 (UTC)

I can't see how "material used" is vague. :-) The problem with using the word "information" is that it implies it's accurate, whereas material is neutral. I know people talk about "false information," but that's actually an oxymoron. SlimVirgin ^{(talk)(contribs)} 08:56, 24 March 2008 (UTC)

My concern was that "material used" could still be seen as referring to the whole page or article cited. I've been trying to think of other ways to put it, but everything seems to have potential pitfalls. Well, how about statements referenced? PSWG1920 (talk) 09:08, 24 March 2008 (UTC)

The material might not consist of statements, as such. I really do think "material used" is clear enough. SlimVirgin ^{(talk)(contribs)} 09:18, 24 March 2008 (UTC)

I think #7 is important though. It's just as true that Wikipedia should not contain articles on subjects for which very little independent material may be found as that it shouldn't contain articles for which none can be found. The latter is already here, but it should indicate the former as well. Seraphimblade ^{Talk to me} 09:21, 24 March 2008 (UTC)

We do mention the notability issue elsewhere in the policy. The problem with talking about "very little" is that the meaning of that will always boil down to editorial judgment, so really we have to leave it up to editors to work out whether they have enough sources to start an article. SlimVirgin ^{(talk)(contribs)} 09:25, 24 March 2008 (UTC)

Well, everything boils down to editorial judgment, unless you're wrong, in which case it boils down to AfD participant judgment. I like the idea of having clear guidance though, that articles should be based mainly on independent, secondary, reliable sources, with primary sources used only for supplementary, incidental material. Seraphimblade ^{Talk to me} 09:28, 24 March 2008 (UTC)

I would say that the next logical step is to distinguish #6 and #7 from the rest, since the first five apply specifically to the material used (thanks SlimVirgin ;), whereas the last two are more general. How about something like this:

Self-published and questionable sources may be used as sources in articles about themselves, so long as:

• there is no reasonable doubt as to who the author is;
• the article is not based primarily on such sources.

In addition, the material used in wikipedia:

1. should be relevant to their notability;
2. should not be contentious;
3. should not be unduly self-serving;
4. should not involve claims about third parties;
5. should not involve claims about events not directly related to the subject.

I don't agree that the existing wording is in any way unclear or misleading, and I don't believe you've made a convincing case that there is any reason to change WP:SELFPUB. Dlabtot (talk) 07:04, 29 March 2008 (UTC)

What I am currently suggesting is a simple reorganization, which would clearly distinguish the source in general and the material cited specifically in wikipedia. If one believes that the current first five restrictions apply to the entire source, then you could never cite Mein Kampf (using an oft-cited example) even in an article about itself, Hitler, Nazism, or anti-semitism. Or, in a more benign example, you couldn't cite Young Earth Creationist literature even in the Young Earth Creationism article! In other words, there would be little point in ever allowing any self-published or otherwise questionable sources to be cited, in which case you wouldn't need WP:SELFPUB. PSWG1920 (talk) 21:58, 31 March 2008 (UTC)

Hitler etc can be cited in articles about themselves, just not elsewhere. That's the issue that sometimes causes problems, but changing it might cause even more, which is why it is as it is. SlimVirgin ^talk|edits 22:10, 31 March 2008 (UTC)

I understand that, but I'm not sure others do. Which is why I have suggested the reorganization above; to unmistakably distinguish restrictions on cited sources in general from restrictions on material cited specifically in wikipedia. This particular reorganization has nothing to do with what sources are appropriate for what articles, as "articles about themselves" remains (and could perhaps be dealt with later.) PSWG1920 (talk) 22:22, 31 March 2008 (UTC)

I'm not really getting the distiction between "restrictions on cited sources in general" and "restrictions on material cited specifically in wikipedia." SlimVirgin ^talk|edits 18:38, 1 April 2008 (UTC)

"There is no reasonable doubt as to who the author is" and "The article is not based primarily on such sources" would seem to be general restrictions regarding the use of questionable sources, while the other five regard the material used specifically in wikipedia. PSWG1920 (talk) 19:44, 1 April 2008 (UTC)

<---It says:

Self-published and questionable sources may be used as sources in articles about themselves, so long as:

1. the material used is relevant to their notability;
2. it is not contentious;
3. it is not unduly self-serving;
4. it does not involve claims about third parties;
5. it does not involve claims about events not directly related to the subject;
6. there is no reasonable doubt as to who authored it;
7. the article is not based primarily on such sources.

All seven are about when this kind of source may be used. I'm still not seeing the distinction you're making. A source can be a person or the material that a person writes; the meaning changes on WP depending on context. SlimVirgin ^talk|edits 23:50, 1 April 2008 (UTC)

Well, #7, "the article is not based primarily on such sources", doesn't directly concern the sources themselves, just how they are used in the article overall. #6 excludes certain written sources from being used at all. The first five deal with what specific material from questionable sources is acceptable for wikipedia. I grant that source can mean different things, but can you at least see the categorical difference between #1-5, #6, and #7? PSWG1920 (talk) 01:12, 2 April 2008 (UTC)

I can see that the sentences are structured differently, but all refer to the same material. It must not be contentious, it must not be self-serving, the authorship of it must be clear, it should not be the primary support for articles etc. It = material from self-published and questionable sources. SlimVirgin ^talk|edits 15:58, 2 April 2008 (UTC)

In #6, it applies to the entire article/publication/web site, unlike in #1-5. That is what I think may confuse some people. PSWG1920 (talk) 17:37, 2 April 2008 (UTC)

No, it refers throughout to the material being used as a source. Each point refers to the same thing. SlimVirgin ^talk|edits 17:42, 2 April 2008 (UTC)

Yes, it technically refers to that, but in practice it applies to the entire article/publication/web page. PSWG1920 (talk) 17:45, 2 April 2008 (UTC)

It refers to the same thing technically and in practice. It = the material being used as a source. If that is an article, it refers to the whole article. If a book, it refers to the whole book. If a website, it refers to the whole website. It means that there should be no doubt as to who wrote the book/article/website; that it should not be unduly self-serving etc etc. Really, I think you have now gone beyond even splitting hairs. You are now splitting sub-atomic particles. :) SlimVirgin ^talk|edits 18:09, 2 April 2008 (UTC)

I got the impression from what you said previously, and from the recent edit you made as a result of this thread, that it only refers to the specific part of the book/article/website which is referred to in wikipedia. That was my whole reason for starting this discussion. With #6, that still could be technically true, but in practice it applies to a larger scope, and my concern is that that will mislead people into thinking #1-5 apply to that larger scope as well. I guess it's about time for me to quit (at least with this particular point) as we don't seem to be getting anywhere, and you've been very kind to take the time to discuss this with me. But do you see at all where I'm coming from here? PSWG1920 (talk) 18:23, 2 April 2008 (UTC)

"It" refers to the source material, whatever that is. If the authorship of a whole book is unclear, then the authorship of each part of it is similarly unclear.

I'm sorry if I'm being dense, but I really don't see where you're coming from with this, and I don't know what you mean by a "larger scope." The source material ("it") should not be unduly self-serving, it should be of clear authorship, it should not involve certain types of claims etc. I can't see any grammatical issues or lack of clarity. You might disagree with the substance of it, of course, but the syntax is clear. SlimVirgin ^talk|edits 18:30, 2 April 2008 (UTC)

Since you don't understand where I'm coming from, I'll try to explain as clearly as I can. The "larger scope" I refer to is the whole book/article/website which is cited. In practice, that's what it in #6 usually applies to, because if the authorship of any part of a source is unknown, chances are that is the case for the entire book/article/website which is the source in question. Someone reading this policy may thus see #6 and intuitively associate "it" with the entire book/article/website, and thus conclude that's what "it" refers to in #1-5 as well, and by that logic exclude almost any questionable source from any article. That is the reason I suggested separating (what is currently) #6 from #1-5. PSWG1920 (talk) 19:06, 2 April 2008 (UTC)

Do you mean that people would think if there is, say, a claim about a third-party on page 300 of a self-published book, it means that material from page 100 may not be used as a source? I don't think anyone would think these words implied that, and especially not in the order they're currently in. But at least I did get what you were saying! (I hope.) My apologies for being so slow! SlimVirgin ^talk|edits 20:04, 2 April 2008 (UTC)

Yes, that is basically what I meant. I think I have seen it interpreted in essentially that way, but with web pages. In that case, sources were labeled as promotional, and deleted, even though they were only used to reference material that seemed (to me) very appropriate for the article. Perhaps I should have discussed this further with the editors in question, but since I wasn't sure what the rule actually meant, I came here, and noticing discussions about how WP:SELFPUB could be changed, started offering my own ideas.

If we're talking about the order the restrictions are in, it seems to me the way to make this most clear is to place (what are now) #6 and #7 above the rest and categorize the other five as applying to the "material used in wikipedia". PSWG1920 (talk) 20:19, 2 April 2008 (UTC)

It may be that people objected to other material on a webpage and felt it undermined everything on it, in terms of whether it was a reliable source overall. That's sometimes valid, sometimes not. That's where it has to boil down to editorial judgment, because we can't produce rules that cover every eventuality. SlimVirgin ^talk|edits 23:10, 2 April 2008 (UTC)

The bottom line is, websites promoting the Bates Method are not going to pass muster as reliable sources for the encyclopedia article about the Bates Method, because such promotional material is by its very nature unduly self-serving. Dlabtot (talk) 23:50, 2 April 2008 (UTC)

WP:SELFPUB deals with sources which are already considered essentially questionable. If SELFPUB comes into play, we know the source is not considered "reliable". That doesn't mean it's completely excluded from wikipedia, which is why we have SELFPUB.

As for being "unduly self-serving", that applies only to the specific material used in wikipedia, not to the entire source. PSWG1920 (talk) 00:02, 3 April 2008 (UTC)

Let me ask this: Is there any downside to the reorganization which I have suggested above? PSWG1920 (talk) 21:49, 3 April 2008 (UTC)

The downside, in practical terms, is that you haven't been able to convince anyone that this change should be made. WP operates by consensus - not by that editor who happens to be the most tenacious or the most verbose or the editor who gets in WP:THELASTWORD. Dlabtot (talk) 19:52, 4 April 2008 (UTC)

Please explain why you think, per your revert of my edit, that the existing text is "perfectly clear". In this thread I have highlighted a potential ambiguity. Also, do you think my edit in any way changed the meaning? PSWG1920 (talk) 19:59, 4 April 2008 (UTC)

I don't agree that you have demonstrated any ambiguity. I don't feel that I have an obligation to respond ad nauseum to attempt to explain why I understand the existing policy. I will admit that I don't understand your explanation of what you say you find ambiguous. Of course your edit changes the meaning; it creates an unnecessary and improper distinction. I don't intend to continue this discussion unless some new point, argument, or information is brought forth, however, any silence on my part should in no way be construed as agreement. Dlabtot (talk) 01:50, 6 April 2008 (UTC)

The current ambiguity I see is this: when SELFPUB says "Self-published and questionable sources may be used as sources in articles about themselves, so long as: the material used is relevant to their notability, it is not contentious, it is not unduly self-serving, it does not involve claims about third parties, it does not involve claims about events not directly related to the subject, there is no reasonable doubt as to who authored it", what does it refer to? Authorship is not limited to a single statement, so for practical purposes, it can be seen, in restriction #6, as referring to the whole article, book, or website which is cited. Which could make it seem as though it means the same thing in the first five restrictions; if so, a source that contains any statement which is contentious, self-serving, or involves third parties would be completely excluded from wikipedia. But every indication I've seen is that the first five restrictions are only on the specific material referred to in wikipedia which originates from such a source, not on the entire source. If so, there is a categorical difference between #6 and #1-5. Do you now understand why I think the current version is ambiguous (a simple yes or no will suffice if you don't feel like saying anything else, and indicating that you understand will not be seen as implying that you agree)? PSWG1920 (talk) 02:28, 6 April 2008 (UTC)

Self-published and questionable sources may be used as sources in articles about themselves, so long as:

Self-published and questionable sources may be used to reference material which is appropriate for the article, so long as:

Could anyone tell please if this qualifies as an "educational blog" and therefore can not be considered a "reliable source". Obviously, one person asks and another replies, without any editorial oversight. I am asking because this "source" contradicts several published books. Thank you.Biophys (talk) 20:29, 3 April 2008 (UTC)

I would regard that society as a reliable source within its field, which it says is the support of its members in the practice of their professions in occupational and environmental radiation safety. Bear in mind that they will have a POV, because they seem to be a lobby group of some kind. Are you encountering what seem to be errors, or just differences of opinion? SlimVirgin ^talk|edits 20:34, 3 April 2008 (UTC)

I think this is an opinion of a person (and without any visible oversight) rather than an official statement by society. This is a case of Nikolai Khokhlov who was poisoned by radiactive thallium, according to almost every publication, including several books; and now this has been disputed [7], based on something I believe is a private opinion of an expert posted in his society blog.Biophys (talk) 21:28, 3 April 2008 (UTC)

The opinion is signed by someone called George Chabot. There is a radiation specialist by that name. [8] What I would do in your shoes is e-mail that person to ask whether that is his post. If yes (or if you find it's another suitably qualified person), then it's a reliable source, but it needs to be written with in-text attribution — something like, "George Chabot, emeritus professor of X at University Y, argues that ..." SlimVirgin ^talk|edits 06:45, 4 April 2008 (UTC)

Thank you for explanation. I think this is his post. In the future, I will treat such sites as reliable sources. That was not obvious for me.Biophys (talk) 15:26, 4 April 2008 (UTC)

I note the following from the top of the page "The following question was answered by an expert in the appropriate field:" selected by the society. DGG (talk) 00:17, 15 April 2008 (UTC)

I just did a line by line reading of this page, and was shocked at the number of egregious problems it displayed.

Standard note of qualifications: I am a PhD student in English who fairly regularly teaches writing courses including ones that are expected to cover basic research skills. I am not a composition specialist, but I am at least tangentally aware of the trends there and know pretty well the general sense of "what makes good research-based writing."

The full list of severe problems is long, and I've posted it in my userspace at User:Phil Sandifer/Verifiability.

In short: I do not see a way in which application of this page could be used to improve most articles beyond what any reasonably informed editor would instinctively see to improve. For the most part, application of this page on anything but the barest of stubs would actively diminish the article. Furthermore, the page is increasingly not even a good page with some questionable advice. It has, at this point, rotted to its core. It would not meaningfully harm the project to delete it and start over, or to tag it as "not policy," replacing it with the single sentence "Editors should use their best judgment and ensure that information is reliably sourced, and that readers can follow up on those sources."

It is, in short, a train wreck. Phil Sandifer (talk) 18:30, 1 April 2008 (UTC)

There may be some good points in your critique, but I would argue that to make changes to an established policy that is one of the three core content policies and widely referred to, you will need more than a critique to demote it. Wikipedia:Consensus can change, of course, but you need to establish it first for your proposal to have any kind of traction. ≈ jossi ≈ (talk) 01:31, 2 April 2008 (UTC)

I am skeptical that the changes that took this page from a cornerstone of policy to a pile of festering nonsense gathered proper consensus. Changing it back should be equally straightforward. Phil Sandifer (talk) 01:43, 2 April 2008 (UTC)

Can you provide a diff of the version that was a crnerstone of policy? ≈ jossi ≈ (talk)

I haven't poured over the history of changes to figure out the point where it went wrong, and I'm sure the changes were incremental. [9] is a good start, though. Phil Sandifer (talk) 02:04, 2 April 2008 (UTC)

I think April 1 was yesterday, so I may be missing something here. :) ≈ jossi ≈ (talk) 04:05, 2 April 2008 (UTC)

The page has drifted through a series of small cuts and instruction creeps from a pretty good page to a disaster. I am hard pressed to find a diff of just before it went wrong, so I went to the first good diff, i.e. the original version of the page. If you want a more thorough but still pretty good version, [10] is also still good. It looks like the foundations for what went wrong got laid somewhere in 2005, as the first version of 2006 starts to pick up the language that became inflexible and problematic, but it's still pretty good too: [11]. You can definitely see the problems starting to creep in, though - "Any edit lacking a source may be removed. If you doubt the truthfulness of an unsourced statement, remove it to the talk page. Otherwise, just request a source." has shown up by then, which is overly pedantic advice - truth be told, we don't want footnotes at the end of every sentence and we know it. By 2007, the page was definitely starting to be a bit of a mess: [12]. By this point the page has become a manual for dismantling articles instead of building them. Unsurprisingly, this drift towards problem becomes more pronounced the more frequently the page gets edited - from 2003 to 2005 the page had fewer than 50 edits. 2005 you can get through in about 250 edits. 2006 is about 1000 edits, as is 2007. Phil Sandifer (talk) 04:30, 2 April 2008 (UTC)

That only shows that this page has evolved as many others have. I see no reason for a dismantle and rebuild approach as you suggested, as the current page is the result of vigorous debates by hundreds of participants and reflects, IMO, current application of the spirit of this policy. If there are specific aspects or formulations that you believe are superfluous, bring these to the attention of editors here for discussion. ≈ jossi ≈ (talk) 15:12, 2 April 2008 (UTC)

The idea that more editors being involved necessarily equates to quality or accuracy is not a logical one. As for your desire for specific complaints, follow my link above - I identified several passages that are deeply flawed and provide actively bad or insane advice. I'd be happy to discuss any of them with you.

As it stands, this page should not be considered policy. It does not reflect the reality of article editing, it does not reflect the consensus that forms on individual articles, it does not reflect our best practices, and it does not even offer anything resembling a plausible description of how to write an encyclopedia. Regardless of what tag it wears and the process through which this monstrosity developed, it is not accurate to say that this page reflects Wikipedia policy. The page is wrong. It should not be followed, and any editor applying the page as written is being a deeply irresponsible editor. Phil Sandifer (talk) 15:45, 2 April 2008 (UTC)

I must just note, with a chuckle, that I disagree, not only with everything you've said here, but apparently with all of the underlying premises that you've left unsaid as well. Dlabtot (talk) 19:28, 2 April 2008 (UTC)

I think there is little that I have left unsaid. What do you think you disagree with? Phil Sandifer (talk) 20:27, 2 April 2008 (UTC)

I've already stated that I disagree with every single statement you've made, as well as the all the unstated premises that underly the statements you've made. Have I really been in any way unclear? Please explain. Dlabtot (talk) 06:27, 3 April 2008 (UTC)

Inasmuch as you have not stated what you believe those underlying, unsaid premises to be, yes, you've been quite unclear. Phil Sandifer (talk) 13:06, 3 April 2008 (UTC)

Phil... I don't think you can say the page does not reflect Policy, when by definition it is Policy. I suspect what you intended to say is that (in your opinion) it does not reflect consensus and practice. I disagree with that but, for the sake of discussion, could you give us some examples of where you think the current wording has gone wrong? Blueboar (talk) 16:19, 2 April 2008 (UTC)

No - policy is not defined as "what people have put a policy tag on." It is defined as "how one ought to go about contributing to Wikipedia." This page is not that. As for specific objections, I already have - there's a link above to a detailed critique. Phil Sandifer (talk) 16:26, 2 April 2008 (UTC)

The place to be specific in discussing this article is here, at the discussion page for this article, not on some sub-page in your userspace. Do you have any real commentary to make beyond calling it a 'monstrosity'? Dlabtot (talk) 19:19, 2 April 2008 (UTC)

I'm discussing it here. I offloaded a lengthy line-by-line reading to my userspace, then linked to it to discuss it here. You're welcome, I suppose, to ignore that reading because you dislike what namespace I put it in, but it says little good about the quality of consideration on this policy. Phil Sandifer (talk) 20:27, 2 April 2008 (UTC)

I'm not ignoring your rant, I'm simply pointing out to you that if you want to discuss the verifiability policy of Wikipedia, you'll need to do so, at the talk page of the verifiability policy of Wikipedia. Dlabtot (talk) 06:32, 3 April 2008 (UTC)

Actually, you're dismissing my critique as a "rant" and insisting (on the talk page of the verifiability policy of Wikipedia no less) that I am in the wrong venue. I off-loaded the full critique because it was long and would needlessly overwhelm the page, and because I thought surely nobody would mind a single mouse click to get to my full comments and point. Was this wrong of me? Phil Sandifer (talk) 13:05, 3 April 2008 (UTC)

Yes. Dlabtot (talk) 16:39, 3 April 2008 (UTC)

If your investment in careful discussion of this policy page is such that you are unwilling to exercise your index finger to even the minutest degree to read a lengthy analysis that I did not feel like cluttering a talk page with then I have to say, I am hard pressed to treat your opinions as sufficiently thought out to be worth engaging. Phil Sandifer (talk) 17:00, 3 April 2008 (UTC)

You don't feel like 'cluttering' this talk page? The unintentional irony of that statement, considering the current volume of your verbiage here, is rather amusing. Do you have any specific suggestions for improving the Verifiability policy? Dlabtot (talk) 17:38, 3 April 2008 (UTC)

Yes. You've been refusing to read them. Either read them and respond to them or go find somewhere else to be an unhelpful twit. Phil Sandifer (talk) 18:42, 3 April 2008 (UTC)

I have read the sub-page in your userspace to which you are referring. Twice. I didn't see in that subpage, any specific suggestions for improving the Verifiability policy - though no doubt the fault is mine. Rather than specific suggestions for improving the policy, it looks like a complete and total rejection of the policy. If I've misunderstood you, perhaps you could help me out by enunciating one of the specific suggestions that you have. Thanks in advance. Dlabtot (talk) 18:59, 3 April 2008 (UTC)

1. Return the page to a short statement of first principles.

1. Build it up organically with attention to practice not only on our most pathological limit cases but with attention to what responsible research practices are.
2. Avoid trying to craft a page that can be applied to write or edit an article about a topic you know nothing about - this is not a case we should be encouraging, and indeed, it is one we should actively discourage. Phil Sandifer (talk) 19:37, 3 April 2008 (UTC)

There doesn't seem to be any consensus for your suggestion to "Return the page to a short statement of first principles" - in fact you appear to be a minority of one. Dlabtot (talk) 19:51, 3 April 2008 (UTC)

However, and this is a very important point, I am right, inasmuch as right now the page offers guidance for how to conduct research that is laughable by any professional standards whatsoever. Phil Sandifer (talk) 19:53, 3 April 2008 (UTC)

Suggest detailing more specific examples of requirements that you believe should be changed. You've suggested one, a requirement that unverified statements be removed, with a statement that we don't want footnotes at the end of every sentence. On our most contentious and sensitive topics, we really DO basically want a footnote at the end of every sentence. We've found from experience that under the realities of our system of group editing, this approach is the only way to prevent both constant insertion of false information and constant challenging of existing information when the subject-matter is highly contentious. You might not find such an outcome ideal from an English-prose-style point of view, but it is an outcome that works for us -- it beaks deadlocks and enables things to get done. It's certainly true that most articles and subjects are not highly contentious or sensitive. WP:CITE#When to cite provides more detailed guidance about problem areas requiring hightened sourcing. There is a proposal, WP:When to cite, that would make when to be strict and when to be lenient a little clearer. Best, --Shirahadasha (talk) 16:42, 2 April 2008 (UTC)

I think, actually, our most contentious articles are where we can least simply cite every sentence, simply because that's where the most careful judgment is needed to properly organize and contextualize the information. But I am deeply skeptical that more programmatic and rigid rules break deadlocks and lead to progress. Note that I am not trying to lessen the importance of citations. Rather, I'm asking for citations to be used in a manner similar to how they're used elsewhere - to provide the body of sources that one can go back and check the work from not as a sort of admission certificate for every line. The belief that we can transparently take information from sources and present it straightforwardly and without interpretation is, by any respected standard in the field of literary studies, utter nonsense. The page is thoroughly based on a false premise about how language, sources, and interpretation work. The correct first step is to return to a page about principles instead of about dogmatic and false advice. "Editors should use their best judgment and ensure that information is reliably sourced, and that readers can follow up on those sources" would be a nice start. Phil Sandifer (talk) 20:27, 2 April 2008 (UTC)

Phil, the problem here is that you are trying to explain the real academic purpose of citation to people who believe it can, and therefore should, be done mechanically and deterministically. This is the version of citation that people get taught in high school and the first year or two of college, just to get them citing at all. It's training wheels for future academics.

But they're taking that to a ridiculous extreme, and want to fix the existence of disputes at all by mechanical rules. Of course, this doesn't work, and is nothing like the purpose of citation in real work, and will in fact provably be counterproductive and the opposite of knowledge, as you demonstrated (and I must say I'm impressed by counterarguments like "it's on your talk page so I don't have to think about it, nyah nyah").

But you have to convince them of this. That is, you have to give them cluefulness about citation that people only just manage to get after three or four years of college, if then. So lotsa luck managing it on a talk page! - David Gerard (talk) 22:55, 2 April 2008 (UTC)

Some of the people involved in developing and maintaining this have PhDs, so level of education isn't the issue. It's a question of how to find a good balance. I also dislike the citation-after-every-sentence approach, but there's no doubt that it's sometimes the only way to stop disputes in contentious areas. In non-contentious ones, it doesn't happen and it doesn't need to happen — there's certainly nothing in this policy that says every single claim needs a source, and the page stresses the importance of editorial judgment. But it also makes clear that, if you want to add something and you're challenged (reasonably), you do have to produce a source. I can't see how challenges would ever be satisfied otherwise.

People may sometimes take the words on this page too far — I've definitely seen unnecessary challenges — but that's true of all the policies and guidelines. They all have to be applied with common sense, and there's no legislating for that. SlimVirgin ^talk|edits 23:06, 2 April 2008 (UTC)

Actually, the page never says anything about the reasonableness of challenges, instructs the swift removal of all unsourced information, and does not use the word "judgment" anywhere. One cannot legislate common sense, but this page seems intent on legislating against it. Phil Sandifer (talk) 23:50, 2 April 2008 (UTC)

We used to have some words advising editors not to remove things without good reason. I forget what it said exactly, but I remember adding it twice, and I see it's gone again, or watered down, so it's clearly not the kind of thing that people want to see. I suppose some editors feel it opens loopholes.

I just did a search for the word "judgment," and you're right, it's not there. I must have been thinking of NOR. We could always bring something like that sentence over here e.g. "Appropriate sourcing can be a complicated issue, and these are general rules. Deciding what kind of source is appropriate on any given occasion is a matter of common sense and good editorial judgment." SlimVirgin ^talk|edits 05:54, 3 April 2008 (UTC)

It's a sad commentary upon human nature that the idea that we should exercise judgement is in some way controversial. Dlabtot (talk) 06:23, 3 April 2008 (UTC)

←Phil, you could make a useful contribution to Wikipedia if you wanted to, so please stop shooting yourself in the foot :) How likely would it be that I could help your English department by announcing on my arrival that everyone reads Wikipedia, very few people read anything produced by English departments, and it would be better for all of you to give up? This page, in fact, is performing its function exquisitely. I know you're dubious, so let me explain:

• There are many regular readers of this page who have encyclopedic knowledge of past policy discussions and have excellent judgment. The main reason they don't simply enlighten us all and solve all disputes is WP:OWN: it's a mistake for any one person to "run the show", it interferes with the evolutionary pool of ideas and with respect for all people.
• You are correct that this page is unbalanced in the sense that it gives more support to people who want to add citations than to people who want to subtract them, but this is the Wikipedia way. We tend to respond, in a slow and methodical way, to the problems we have, to people as they actually are. There is a constant problem of people asserting things that are not verifiable; there is less of a problem of people being disruptive by adding "citation needed" tags against community consensus. And when that happens, it can be handled by the "impose one's own view of 'standards to apply' rather than those of the community" clause from WP:POINT. These things most certainly are discussed, frequently, but it has been thought (so far) that working these discussions into WP:V would give them undue weight. For instance, see the current discussion at WT:CITE#When is it okay to remove a "citation needed" tag?.
• It's a common criticism that core content pages don't quite seem to nail down the policy; but just as in the courts, the interpretations do nail down the policy. People rely on policy pages to change very slowly or not at all. Policy pages are kind of like a national constitution; when they change, then your interpretation of all the laws and all the lower court rulings has to change, and that can be a hassle.

Having said that, it couldn't hurt to share knowledge about what citations are, and when they're not needed according to current Wikipedia standards, which (rightly) vary from page to page. It's true that this kind of information could be discussed in a more scientific and public way. I just finished one such conversation over at WP:WHEN. - Dan Dank55 (talk) 14:00, 3 April 2008 (UTC)

It's difficult to know where to begin, given that your bulleted description seems to me to utterly not describe the practical evolution of policy on Wikipedia or how this page evolved. Certainly I have no desire to own the page - if I did I would have been a more substantial contributor along the way (my edits to the page are minimal, in practice) and would involve myself on the talk page more than I do. The issue of how citations are (mis)used on Wikipedia is one I have an odd sort of interest in - it is, from my perspective, one of the most interesting issues surrounding Wikipedia. But it is also one where progress is difficult at best, and it is generally deeply frustrating to try to work with it due to the entrenched disinterest in thought displayed by many people on this issue. So I tend to involve myself only when I am reasonably persuaded that I will enjoy, or, at the very least, not hate the experience. All of which is a long-winded way of saying that, truth be told, I'd prefer to avoid the page and see it evolve. When, however, it is evolved along a route that is based on deeply deluded premises about the nature of information and citation there is, I think, good reason to step in.

As for the bias of the page, my concern here is actually less based on a pattern of actual abuses (though I do think there are some, most notably in the realm of deletion) than on a destructive paradigm shift that has been happening where policy pages increasingly seem intended to be executable by a machine that will somehow automatically write and edit an encyclopedia. Certainly some of this will always exist, but on the other hand, this page all but encourages it. I think there is a good page to be written on the subject of verifiability - a stern one that gives good editors all the tools they need to remove crap from articles. But this isn't it. It may incidentally serve as a tool for that purpose, but it's also serving as a fundamentally erroneous description of how to write an encyclopedia - one that encourages writing that is not merely bad in a mechanics/presentational sense (as someone suggested) but in a perniciously way that is antithetical to our mission. I am loathe, however, to write such a page because it would come unfortunately close to WP:OWN. Better, I think, to cut the page down to a lone statement and try to be more responsible in rebuilding it. Phil Sandifer (talk) 16:19, 3 April 2008 (UTC)

Regarding "your bulleted description seems to me to utterly not describe the practical evolution of policy on Wikipedia or how this page evolved", in order to respond, I'll need to know if you're right. Is he right? - Dan Dank55 (talk) 16:37, 3 April 2008 (UTC)

To be more specific, I think we are painfully unmethodical in our development of policy - we tend to formulate policies around a handful of our most deeply pathological articles (which represent a tiny percentage of the overall encyclopedia) and then to watch as they seep outwards, being applied to articles that simply do not need that level of dogmatism to keep functioning. (Not that the level of dogmatism seems to help our pathological articles all that much either) Phil Sandifer (talk) 17:00, 3 April 2008 (UTC)

With apologies for backtracking, I'd like to express support for re-adding the concept of "editorial judgment" to the main WP:V article. I know it provides wiggle room and fodder for contentiousness, but it also gives guidance regarding sources that fall into the gap between reliable and questionable sources. Thirdbeach (talk) 01:49, 5 April 2008 (UTC)

"This page (Wikipedia:Verifiability) documents an official English Wikipedia policy, a widely accepted standard that all users should follow." Where exactly is the reference backing this up? There is none, and it's certainly not accepted by me. Phil, you have diagnosed a disease plagueing Wikipedia.

In an article about a philosophy it was claimed that: the explanations of the philosophy's author could not be used (extensively) because it was not third party material. This is sick to the extreme. Who can be trusted best? Someone (clearly) explaining his/her own idea, or the interpretation of a third party? (Who's idea the editor now has to interpret.) When I brought this absurdity to the surface on the current discussionpage, it lead to the final version of: "If no reliable, third-party (in relation to the subject) sources can be found for an article topic, Wikipedia should not have an article on it." This is the second bullet point at Phil's page User:Phil Sandifer/Verifiability. (I clearly failed in resolving the underlying problem.)

Thanks Phil, for bringing it up again! I hope others who agree with us will also speak up, as to make clear the current policy is no longer widely accepted. ErikvdL (talk) 11:11, 5 April 2008 (UTC)

Several posts in the "Detailed analysis of this page" equate providing a source for a claim with providing an inline citation for a claim. If the people equating these ideas sincerely believe that is what the policy requires, and are not just being argumentative, then the policy must be revised to emphasize that any claim that can be looked up in any of the general references without undue difficulty is sourced. --Gerry Ashton (talk) 15:24, 3 April 2008 (UTC)

Yes. This would help tremendously - if we could have a meaningful sense whereby sources were treated as being related to the overall page instead of treating the page like a set of independent and discrete facts. We would still have the rather large problem of this idea of transparent interpretation, but that is one problem instead of two. Phil Sandifer (talk) 16:19, 3 April 2008 (UTC)

This is a "sometimes yes and sometimes no". If I wrote an article with a bibliography with a thousand books at the end, any one sentence or claim in the article would be practically unverifiable. As a practical matter, no individual could look up all of the thousand books and see if one of them contained the claim. And what is needed here is practical verification - it has to actually be possible for a human being to check a claim made against the references given. So once an article and its references reach a certain size and complexity, inline citations become essentially required to achieve the purposes of the verification policy. One difference between Wikipedia and an academic publication is trust -- an academic publication generally trusts that claims made are true and doesn't generally expect detailed, low-level verification to actually occur. But Wikipedia, as an encyclopedia editable by anyone, operates with a lesser level of trust than one editable only by pre-screened experts, and detailed, low-level verification has to occur and the policy has to support verification being done in such a way that it can be supported. It seems to me that part of what you are saying is that Wikipedia's basic model is self-contradictory and impossible, because a well-written work has to rely on a higher trust level than Wikipedia is in a position to have. You may be right. Best, --Shirahadasha (talk) 18:07, 3 April 2008 (UTC)

To back this up, I have at times tried to verify claims made by the author Zacharia Sitchin. Sitchin has huge and impressive bibliographies, with many good references (which contradict his claims). It is rare that I can find a page where Sitchin even makes it clear what book he is using for a claim, let alone what page. So his books are for all intents and purposes largely unsourced. Doug Weller (talk) 18:47, 3 April 2008 (UTC)

That doesn't seem to me to be a problem with the style of presentation of sourcing, or the level on which he sources (i.e. line by line or more holistically), but rather with him simply misusing and abusing his sources. That is not something that the in-line vs. holistic debate covers - any number of sources can be misused (and, indeed, often are - it's one of the biggest problems with source citing on Wikipedia - often things are cited to sources that just don't support the claims being made. This is far worse than uncited information.) Phil Sandifer (talk) 18:55, 3 April 2008 (UTC)

In-line citations are without a doubt important as well. But the problem is that in-line referencing was never intended to be used as a sort of step-by-step verificationism. That's just not how citations evolved as an academic institution. They don't work to do that. And inline citations don't really help either in an article with a thousand sources, because fundamentally the problem you're facing there is one of information presentation - a thousand references just isn't manageable as a dynamic document.

As for the possibility of Wikipedia's basic model being impossible, that seems to me true only if its basic model is that articles will be written by non-experts. That doesn't seem to me to actually be our model, though. Our model, I think, is more accurately described as "World-wide peer review coupled with dynamic authorship is an adequate substitution for credentialism." The difference there is huge. Phil Sandifer (talk) 18:53, 3 April 2008 (UTC)

(responding to Doug) The lack of sourcing in some academic work is a serious problem, and I'm thinking particularly in history. There are historians who are a delight to read because they footnote every claim, so you can follow exactly where they got their information from and what they made of it, and others who throw out opinions and expect you to take on board whatever they say. The latter is increasingly no longer acceptable, because "experts" are no longer regarded as the font of all wisdom, and that's a good thing, a revolutionary thing, which Wikipedia is part of. That doesn't mean we go to the opposite extreme, but it does mean the question, "What is your evidence for making this claim?" is always valid — of all experts, all politicians, all lobbyists, everyone who is trying to persuade us of something, and yes, of all Wikipedians too. SlimVirgin ^talk|edits 18:58, 3 April 2008 (UTC)

Even if there are a substantial number of general references for an article, and it would not be practical to read all of them to verify a claim, it is still not necessary to provide an inline citation for every claim. A decent reader should be able to discern which of the general references provide comprehensive coverage of the topic, and look up the claim in index or table of contents of that reference. Claims that can be verified in this fashion don't need an inline citation. --Gerry Ashton (talk) 19:00, 3 April 2008 (UTC)

I suppose the question is why someone wouldn't want to give an inline citation if asked for one. If I've added material to Wikpedia, and I've read the books and know my stuff, and if someone then asks me for a page number, why would I refuse to give it? SlimVirgin ^talk|edits 19:04, 3 April 2008 (UTC)

We can't assume that the author of a given line of an article is still contributing to the project, watching the page, etc. Phil Sandifer (talk) 19:08, 3 April 2008 (UTC)

There's no reason to refuse to give a page number if you have one at hand, but that doesn't mean it should be included in the article. It might be easier to simply tell the person who asked and leave it at that.

This is especially true for the introductory parts of subjects like chemistry, physics, mathematics, etc. For example, if someone asks for a page-number reference to the definition of a group, I'll be glad to provide one. But, by asking for a page number, they reveal they have not consulted any of the general sources. — Carl (CBM · talk) 20:03, 3 April 2008 (UTC)

Nice, Carl. Well, people, there are a lot of good suggestions here, and anyone who's interested in putting this stuff into practice is welcome to come join us over at WP:GAN, where we have a backlog of around 200 articles. - Dan Dank55 (talk) 21:08, 3 April 2008 (UTC) P.S. I'm particularly into "well-written" issues these days, so anyone who's particularly interested in citation issues who wants to pair up with me on Good Article reviews is more than welcome. - Dan Dank55 (talk) 21:21, 3 April 2008 (UTC)

In reply to SlimVirgin, the editor who contributed a claim to an article might have borrowed the source from a library, and no longer have it at hand. Also, the original article can be written much faster if it isn't necessary to write a citation for every claim where there is a remote chance the claim might be challenged. --Gerry Ashton (talk) 02:57, 4 April 2008 (UTC)

Indeed - excessive demands for sourcing seem to me to go against the whole "anybody can edit" philosophy. Or, in practice, they deepen our bias towards online sources over print ones. With the exception of BLPs, an attitude of eventualism seems not only more prudent but more likely to produce good results. Phil Sandifer (talk) 04:00, 4 April 2008 (UTC)

In practice, I don't follow this policy. Which is not to say that I do things that conflict with verifiability in the abstract, but that I don't find it helpful to consult the policy as written, and that its definition of reliability is problematic. Indeed, I think the biggest issue (and one that lies behind many of Phil's criticisms) is that this policy is being made to do too much work.

I think the attempt to combine verifiability and reliability into a single policy doesn't work. They are really two different and independent issues. We want our articles to have verifiable sources, which is to say, we want to be able to find out where the material is coming from. But we also want those sources, when produced, to be reliable.

Verifiability could be addressed succinctly, as long as the actual matter of verification were addressed. Reliability is a much thornier problem, and I don't think we are well served either by having two policies covering the issue, or by the actual content of those policies. We keep trying to describe reliability in terms of source form, when real reliability is determined by source content. Sources that publish accurate material are reliable; those that do not, are not. Those that cannot be assessed on this basis can be generally judged untrustworthy, and thus unreliable.

"Verifiability, not truth" devalues fact-checking. I've written a bunch of lighthouse articles, using a standard set of about four sources. I have to do a lot of fact-checking to put these together, because all of them have mistakes. I particularly find a lot of location errors, so those I check against charts and aerial photography. It's easy enough to verify the content of these articles as far as finding the sources is concerned, though I have eschewed the blow-by-blow citation style as absurd in context. But people want more than this; they want the articles to be accurate. Mangoe (talk) 18:30, 3 April 2008 (UTC)

In practice this policy, like others, is useful if you combine it with common sense. We do want accuracy, but only as a derivative of proper sourcing. Our goal is not find out where a lighthouse stands — it is to tell our readers where the reliable and verifiable sources say it stands. If we have conflicting sources, we do our best to sort them out, and present them in the best possible way. That eventually boils down to, after following all the policies and guidelines, common sense. Crum375 (talk) 18:38, 3 April 2008 (UTC)

No. We, ideally, would like to tell people where the lighthouse is. I mean, given the choice between giving correct and incorrect information, we would like to give correct information. Anyone who does not want to give correct information probably shouldn't edit an encyclopedia. "Verifiability, not truth" is a deeply problematic statement, if only because it suggests we don't like truth very much. "Truth through verification" is really closer to what we do. Phil Sandifer (talk) 18:45, 3 April 2008 (UTC)

Not really. Our goal as a tertiary source is to summarize existing sources in the best possible way. If a WP editor's knowledge conflicts with the reliable sources, adding the editor's information would violate WP:OR. There are many cases where there is no one "correct" answer, like in political or ideological conflicts. And even in scientific and technical matters, the so called "exact" sciences, things change over time. Our goal is not to provide absolute truth, but to act as librarians and point users to the best possible sources, properly balanced and presented. Crum375 (talk) 18:54, 3 April 2008 (UTC)

That's a different statement than the "Verification, not truth" as well, though. I do not think, in practice, that most people, when they speak about true information, mean it in some sort of absolute and ontological sense. Indeed, the project of Wikipedia is not to provide truths that exist independently of the existing cultural context. But that's not what we usually mean when we say "truth." Usually we mean "what is generally accepted and regarded as true at the moment." Which is what we do try to present. Verification is picked because it's a good way of ensuring truth that can be checked by most people and thus encourages peer review. Phil Sandifer (talk) 19:18, 3 April 2008 (UTC)

The truth according to whom? The location of a lighthouse is not a good example because it is not likely to be contentious. But what about the question of whether Liancourt Rocks is Japanese or Korean? Should WP tell people the answer to this question? Or just present the significant viewpoints that have been published by reliable sources? Dlabtot (talk) 19:26, 3 April 2008 (UTC)

Wikipedia should absolutely provide the correct and true answer to that question. Of course, it should do so from a NPOV perspective. You're trying to collapse a distinction and make the word "true" do things that, while it does do in some limited philosophical contexts, it does not generally do. While it is no doubt true that Wikipedia does not pursue some Platonic ideal of Truth, it clearly does pursue small-t truth - that is, accuracy, correctness, rightness. Specifically, it pursues the small-t truth about the question "what are the important aspects of this topic?" And it is worth noting that the answer we do give on the matter of the Liancourt Rocks - that they are the subject of a dispute between Japan and South Korea - is unquestionably true. The most ardent partisans on either side would not argue that no dispute exists - they would merely argue what the correct resolution to that dispute is. Wikipedia pursues truth - it's just selective about what questions it asks. Phil Sandifer (talk) 19:34, 3 April 2008 (UTC)

Umm, no. Dlabtot (talk) 19:46, 3 April 2008 (UTC)

I am not entirely sure how to respond here. This is because of one major problem - you are wrong. If you are claiming that the phrase "verifiability, not truth" is interpreted as a statement about Platonic big-T truth, you are wrong. If you are claiming that there is no practical distinction between big-T and small-t truth, you are, again, wrong. If you are claiming that, in fact, Wikipedia is unconcerned with small-t truth, which can basically be taken as a synonym for "accuracy," you are, once again, wrong. If you have some other meaning, I suppose, it is possible that you are not wrong. However that meaning is not clear to me, and I would appreciate its elucidation, since, as it stands, you are standing about advocating a position that holds no credibility whatsoever. Phil Sandifer (talk) 19:56, 3 April 2008 (UTC)

I'm not saying any of the things that you are claiming I'm saying - I'm saying what I actually said. For example, concerning whether Liancourt Rocks is Japanese or Korean, I asked the rhetorical question: "Should WP tell people the answer to this question?" I did not ask: "Should WP tell people that a dispute exists?", as your comment falsely implies. Dlabtot (talk) 20:01, 3 April 2008 (UTC)

And I answered that question - we are selective in the questions we ask. That is not a question that we should ask or answer, no. (Or, if we answer, it should be a not-quite answer along the lines of "it's complicated." In general, we should shy away from questions to which there is no small-t true answer. (And, in practice, the main problem with your question is that it begs the question. We shouldn't answer the question "Who owns Liancourt Rocks, the Japanese or the Koreans" because the question assumes a false premise - that one of the two does clearly own it. If the question is opened to one that has a clear small-t true answer - "What is the current situation regarding the ownership of Liancourt Rocks" then the problem you're identifying quickly vanishes.) Phil Sandifer (talk) 20:08, 3 April 2008 (UTC)

(big undent) I would like to point out that any question is as controversial as anyone would like to make it. Information about lighthouses ought not to be controversial, but if someone decides to make a point of a particular fact, then there is controversy.

As far as the Liancourt Rocks issue is concerned, the discussion is completely missing the point of my main criticism. It is probably pretty easy to come up with a citation from the Japanese government that say the rocks belong to them, and a similar citation can most likely be obtained from the Korean government. In that wise, verifiability is not a problem at all. The problem is that, under the circumstances, neither of them is a reliable source, at least in and of themselves. And if one is being a WP:NOR stickler, one would have to come up with a third source documenting the conflicting claims. And that third source would be reliable not because of what it is, but because it would be right! Mangoe (talk) 20:58, 3 April 2008 (UTC)

Of course those wouldn't be reliable sources, and unsurprisingly, no such citations to government pronouncements appear in the article. Apparently I am completely missing the point of your main criticism. Would you be so kind as to re-state it? Dlabtot (talk) 21:28, 3 April 2008 (UTC)

I have a hard time figuring out why we wouldn't, in general, consider statements from the two governments to be reliable sources. Certainly, in many cases, we consider government publications reliable sources. None of our sourcing policies particularly indicate a problem here. Which is, of course, not to say that we should use either of those sources for a claim about the ownership of Liancourt Rocks. The issue, however, is not one of reliability -at least not as we generally use the term. (It approaches an issue of reliability in the more general use of the concept in research, but even there the issue is one of appropriate use more than reliability.) Phil Sandifer (talk) 00:25, 4 April 2008 (UTC)

Would it help to differentiate between "Fact" from "Truth" in this discussion... To me, Truth is subjective while Fact is not. We definitely want to strive for Factual Accuracy in our articles. Blueboar (talk) 00:51, 4 April 2008 (UTC)

While we could surely make such a distinction, my main objection here is that I do not think that "truth' is generally taken to have the sort of subjective or Platonic meaning that we seem to be using it to mean. That is - as long as we use the word "truth" in the policy page, people will quite reasonably read it as a synonym for "factual." Phil Sandifer (talk) 03:53, 4 April 2008 (UTC)

I'm not sure I understand the difference between true and factual. If a statement is true, it just means it describes an actual state of affairs, which is what a fact is.

The issue here boils down to common sense. It's going to be quite rare that every single reliable source we find for something is wrong. In fact, can we even think of such an example? The area where people get frustrated is when there's a gap in published knowledge (e.g. the location of the lighthouse), in which case it's fine to add the information, even though it's OR, because no one is likely to object. Bear in mind that material that isn't challenged or likely to be challenged doesn't need a source, and all the policies have to be applied with common sense. SlimVirgin ^talk|edits 06:33, 4 April 2008 (UTC)

Well, when a source is wrong, it isn't reliable, at least to that degree. But that leads to the problem point: how do we determine that some statement is wrong? There are actually three level of verification here: first, that the cited material exists; second, that it says what it is claimed to say; and third, that it presents accurate information. The first two are pretty mechanical and can be written down succinctly. The last is hard.

And I don't see where appeal to common sense gets us. I'm someone who has been pushed through higher education to the degree that my common sense on sources is pretty good. The thing is that the elaboration of policy on reliable sources is essentially a substitute for that common sense. Mangoe (talk) 15:43, 4 April 2008 (UTC)

Well put. We should attempt to ensure all three levels of verification. The first is easy. The second is often easy (though harder with primary sources and specialized sources). The third is tricky. But it's also probably the most important. Phil Sandifer (talk) 15:45, 4 April 2008 (UTC)

There is nothing wrong with applying verifiability throughout the whole of Wikipedia. What is wrong is the concept used to define it. If someone wrote "On a clear day, the sky is blue" in an article, then the most logical and straightforward way for users to verify such statement would be to step outside and look at the sky. As simple as that. Instead, what Wikipedia policy makers want us to do is step outside, don't look up (that is, ignore the colors of the sky), and go to the local library, fetch a Geography or a Meteorology book referenced in the article to find out if its authors wrote something about the color of the sky. As if the eyes of us common Wikipedia editors and users were not as reliable as the eyes of book authors. I find such policy to be absurd, to say the least. --AVM (talk) 21:25, 15 April 2008 (UTC)

I was thinking about how we describe WP:V. What do people think of the following proposed nutshell:

Wikipedia aims to collate what is known about a wide range of topics. To do this, it reports the views and statements of sources that are believed to speak with some degree of authority on a point. Statements where doubt may reasonably exist, should be attributed to a reliable published source that readers can use for source verification and for their own assessment. Statements that cannot be so attributed, even if believed true, may be removed.

FT2 ^{(Talk | email)} 20:28, 3 April 2008 (UTC)

This would be a good start for the page, I think. Phil Sandifer (talk) 20:35, 3 April 2008 (UTC)

It's too long for a nutshell, and essentially says what the current nutshell does, which by way of comparison is: "Material challenged or likely to be challenged, and all quotations, must be attributed to a reliable, published source." SlimVirgin ^talk|edits 20:49, 3 April 2008 (UTC)

Yesno.. what about as an intro explanation though? And this as the nutshell:

All material where doubt may reasonably exist, and all quotations, should be attributed to a reliable published source that readers can use for source verification and for their own assessment.

FT2 ^{(Talk | email)} 21:13, 3 April 2008 (UTC)