Welcome to the MedLibrary.org Wikipedia Supplement on Wikipedia talk:Verifiability/Archive 16 -- Please Click to Return to Front Page

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Verifiability of television episodes

I am unsure of the specific impalementation of this policy in the case of television episodes. In an article about a specific television episode, is information about that episode not implicitly verifiable? If not, how is the episode supposed to be sourced?

The nature of this question is that Hell Comes to Quahog has been the location of a voracious edit war this past hour or so, dealing with the fact that one editor believes that content in the Notes and Cultural References sections of the article (pre edit-war version) is in violation of WP:V (among other policies). I am not a participant in the edit war, though I am trying to bring it to a resolution.

Thanks in advance for any help! –Dvandersluis 17:15, 15 November 2006 (UTC)

Any television episode, broadcast by say one of the top ten broadcasters, is also stored on video which may be purchased. Sometimes you can also get these on loans from libraries. So it is inherently verifiable. Wjhonson 17:35, 15 November 2006 (UTC)

True, but there may be a gap of several months between the first broadcast of a television show, and the availability of videotapes or DVDs for purchase. The show might be considered unverifiable during this time gap. --Gerry Ashton 20:04, 15 November 2006 (UTC)

Yeah, but that gap can be closed almost instantly at times with itunes and other websites offering episodes within minutes/hours after the initial airingGrande13 20:22, 15 November 2006 (UTC)

I think Gerry Ashton is also being rather USA centric there - if the programme is popular then it is likely to be transmitted in the UK, Canada, Australia, New Zealand, etc. within a few weeks to a few months of the first broadcast (e.g. Lost, series three started on 4 October in the US and will start on 19 November in the UK), allowing for verification. Also currently in the UK the most popular programmes (Doctor Who, Torchwood, Grey's Anatomy, Desperate Housewives) are often repeated within a week of transmission, again allowing one to verify claims. --Neo 20:26, 15 November 2006 (UTC)

Family Guy is one television shows page that should include all these cultural references and notes as some of the material used on the show is so abstract not everyone understands it ,so the episode page can be a point of reference and explanation. It is not a retelling of the joke, but more information the the reference in the joke. Since Family Guy is weak on plot, these cultural references are necessary. So basically the note and cultural references should remain 216.177.121.212 18:25, 15 November 2006 (UTC)

There is no problem using works of fiction as a source, provided that the edits conform to the established policies and guidelines. Some of the relevant sections are quoted below.

WP:WAF - a guideline, tells us "Wikipedia policy on verifiabilityrequires that articles "rely on credible, third-party sources with a reputation for fact-checking and accuracy." However, articles written from an in-universe perspective are overly reliant on the fiction itself as a primary source. Lacking as they are in any critical analysis of the subject, these articles may invite original research. In other words, lacking critical analysis from secondary sources, Wikipedia editors and fans of the subject often feel compelled to provide such analysis themselves."

WP:WAF-- "This often involves using the fiction to give plot summaries, character descriptions or biographies, or direct quotations. This is not inherently bad, provided that the fictional passages are short, are given the proper context, and do not constitute the main portion of the article. If such passages stray into the realm of interpretation, secondary sources must be provided to avoid original research. Note that when using the fictional work itself to write these descriptions the work of fiction must be cited as a source."

WP:OR a policy, states-- "Any facts, opinions, interpretations, definitions, and arguments published by Wikipedia must already have been published by a reliable publication in relation to the topic of the article."

WP:OR-- "An article or section of an article that relies on primary source should (1) only make descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) make no analytic, synthetic, interpretive, explanatory, or evaluative claims. Contributors drawing on entirely primary sources should be careful to comply with both conditions."

So as long as you are only describing something that happens in the episode and you follow the directions at WP:CITE for citing the source in the references section, then you may use the episode itself as a source. If however, you want to say that 1 work is an allusion/homage/parody of or to another work, then you must find a reliable 3rd party source that has already published that connection and cite that as a source. If you watch an episode and think that scene A looks like scene B in another show, you might be correct but to add that to the encyclopedia (absent a reliable 3rd party source that has already published that connection) is considered original research and is subject to removal. Please stop by the trivia cleanup project and check out what we are doing. Also, please check out Make Love, Not Warcraft which has been cleaned of cruft and is now being reviewed as a Good Article candidate. Cheers. L0b0t 21:02, 15 November 2006 (UTC)

Well the argument being made for family guy should have an exception, as it is consists of a much higher percentage of cultural references and callbacks and the format used for south park might not be produce the same benefits for an episode of family guy216.177.121.212 21:10, 15 November 2006 (UTC)

No, Family Guy should not have special treatment. We're not an episode guide, and we're not here to help people understand the jokes in Family Guy. -- Ned Scott 05:02, 16 November 2006 (UTC)

While family guy does need to be cleaned up quite a bit, I also believe that some sort of cultural reference section is needed so people have an easy and quick place to go to check upon a joke they that calls upon something they know nothing or little about. So while the note category should probably be fixed, the cultural references are a well used tool for deciphering some of the randomness of the show Grande13 22:24, 15 November 2006 (UTC)

I don't think there is a verifiability issue here for things that happen in an episode. Some fairly obvious references seem to fall into a grey area for verifiability. I think the real issue here is that we need to limit the amount of these notes. -- Ned Scott 05:02, 16 November 2006 (UTC)

• I think that in general we're overdoing the "cultural references". If something is an alleged reference (e.g. this sketch contains a fish so it might be a reference to that other sketch which also contains a fish) then it's original research. Using exact quotes is an obvious reference, but many inferred subtle references may be entirely coincidental, and it's not our job to speculate. (Radiant) 10:30, 16 November 2006 (UTC)

Gerry you're misunderstanding what I'm stating. Every episode, broadcast by a major station, is avaliable, immediately for purchase. You call the station, ask them for a copy, pay the fifteen bucks or whatever. There is no waiting. Wjhonson 18:37, 23 November 2006 (UTC)

Does this really apply to technical articles on facts that can be deduced logically?

I have a problem with the all-purposefulness of this "official policy"

I have a repeating problem with someone who feels he has the unquestioned right to undo any edits in a tech article where he sees no citations. And he comments his undos with the note "see WP:Verifiability".

An example: As someone who has years of experience on a specific topic, I have learned, partly from my own deductions, that certain things can be done. So I described them by adding them to an existing article. He, who has never heard (or thought) of this, undid my edit, asking me to provide a citation.

However, I can't provide a citation because it rather "new research", done by me or other with special knowledge, experience or understand. In this case, I then had to add explanations to the Talk page so others could follow my deductions to make this person realize that his undoing my edit was inappropriate.

Unfortunately, that same person has now again undone another's additions, although I fully agreed with those added statements, meaning that they were quite correct. Again, the said person removed them with the reasoning that there's no cititation to be found. Yet, more experienced people would understand that the newly added statements were correct.

On both cases it appears that the undoer is not able to comprehend the issue well, but still feels he must control it. Sure, there's also the need to educate this person that he should first ask, and only undo someone's edit after it has been discussed. But that's not my point here.

My point is: It should be made clear that this policy does not ALWAYS apply. In the cases I described, there is simply no citation to be found, yet they were correct. I do not like others to think they can question technical statements by a simple (and rather blind) reference to this policy.

Can something be done about this, please?

-- Tempel 14:14, 16 November 2006 (UTC)

• It really depends on the complexity of your deduction; could you show us an example please? Basic math doesn't require a source (e.g. stating that "77 is the natural number following 76 and preceding 78, and the smallest positive integer requiring five syllables in English." is self-evident). But complex deduction that may be easy to an expert in the field could be incomprehensibly arcane to an outsider. (Radiant) 14:23, 16 November 2006 (UTC)

Well, since I've now brought up the issue with the undo edit on the affected page, I might as well refer to it now from here. It's File_Allocation_Table. The two cases I meant were deductions that could be made from working with the FAT system. To someone who works with this enough, he'll simply agree that the things are what they are. In detail, these were:

• I claimed that while a common belief is that a FAT16 formatted disk can only have certain maximum number of files (that number was 512) in the root directory, it would be possible to enlarge this number at time of formatting the disk. The person did not believe this. I explained to him that the spec provides for this without explicitly stating it (but nothing disallowed it either). He found an article saying that disk always have only up to 512 files in the root dir and took this as a spec - but it's not a spec, it's only a description of what is currently usual or common. Later others added comments telling of little-known programs or options that actually made possible what I had claimed. But those were hard to come by and I had not known of them - was that reason enough to deny my statements?
• Recently, someone added a long section about how file fragmentation happens on a FAT formatted disk. What the person wrote was technically correct, I knew this from experience. Yet, there's no article explaining this. But "reading" a FAT disk would just explain it. Of course, not many people can do that. But the fact exists, and anyone with enough understanding would be able to see this. It's just not to be found in an article on the web.

I just realized that the person who undid my first edit referred also to Wikipedia:No_original_research. This he did when I argued that from reading the specification on the FAT file system, one could deduce the limitations I had described. Somehow, I had the impression he's not able to know the difference between a specification and a interpretation. He did quote the interpretation the the truth and the deduction from the spec incorrect. So, what would you suggest here? There was not quote to find to support what I had written, yet it could be deduced from a spec or by "reading" things readily available, even if they're not present as an article on the web or in another known form of publication. And the person I fought this over with simply said "you did original research and that does not belong in WP". Is he right? --Tempel 14:53, 16 November 2006 (UTC)

• Well, in the first case I would simply link to the specs (I'm sure they're on the 'net somewhere?). In the second case, I would add a link to e.g. the description of Norton Speed Disk, or a different defragging tool, or if you find none, link to the spec again. The issue in the second case may also be that the language used in the explanation is overly technical; maybe the explanation needs to explain things further? HTH. (Radiant) 15:00, 16 November 2006 (UTC)

Wording of nutshell

Can we add either "question" or "challenge" to the third point of the nutshell so it reads: "The obligation to provide a reliable source lies with the editors wishing to include the material, not on those seeking to question and/or remove it." ? Harryboyles 09:51, 19 November 2006 (UTC)

Done Martinp23 10:59, 19 November 2006 (UTC)

It's not the nutshell, it's the policy (the rest of the page being merely discussion). But the change is an improvement, jguk 15:05, 19 November 2006 (UTC)

I would posit that it is not an improvment. Unsourced info should be removed on sight, not tagged {{fact}} for months waiting to see if the editor that added it will eventually come back and follow policy. We have been, I think, far too tolerant with unsourced claims, especially in articles about fictional subjects. If one can't be bothered to provide a source then one's edits should be excised. The encyclopedia is filling up with original Research, speculation, and inference of connections based on facts not in evidence. Absent reliable sources, Wikipedia ceases to be a general purpose encyclopedia and becomes a blog. L0b0t 16:12, 19 November 2006 (UTC)

it should be removed eventually, but not instantly unless it is pure nonesense, maybe the period for which the tag exists should be a week/month so there is ample time to find proper sources or to integrate the material in some other format. If the item cant be sourced within that time period then it can be removed no questions asked. 67.184.160.211 18:48, 19 November 2006 (UTC)

I disagree, you either come correct with a sourced edit, or you don't edit the encyclopedia. The editorial standards at Wikipedia are VERY lax and if an editor is unwilling to abide by the few simple policies we do have, then their contributions are not welcome. Cheers. L0b0t 20:43, 19 November 2006 (UTC)

that rule could be fine for new info to be added, but for information that is currently already listed, then there should be a period to find sourcing after its tagged 67.184.160.211 20:58, 19 November 2006 (UTC)

Wikipedia:Speedy deletion criterion for unsourced articles

Regulars to this page may be interested in this proposal. Please comment on the talk page there. jguk 21:45, 19 November 2006 (UTC)

The first sentence

I'd like to amend the first sentence from

Wikipedia:Verifiability is one of Wikipedia's three content policies. The other two are Wikipedia:No original research and Wikipedia:Neutral point of view.

to

Wikipedia:Verifiability is one of Wikipedia's content policies. It is complemented by Wikipedia:No original research and Wikipedia:Neutral point of view.

as I believe it's more important to emphasise that these policies work hand-in-hand rather than the number of them. jguk 13:27, 24 November 2006 (UTC)

It's particularly worth noting that WP:NOT is also a content policy, but isn't on this list, yet the current phrasing implies that the list is complete. JulesH 09:54, 25 November 2006 (UTC)

These three policies are our three content policies, so I do not see the need to change the wording. ≈ jossi ≈ (talk) 22:59, 25 November 2006 (UTC)

WP:NOT is one of our fundamental policies, but it is not really a content policy. This formulation has had the support of the community for quite a while, and I do not see what is prompting the need for that change. ≈ jossi ≈ (talk) 23:02, 25 November 2006 (UTC)

It's more a case of stressing which policies are complementary. WP:NOT, WP:Copyrights and WP:GFDL are all key content policies - indeed any content not complying with these should be removed. And any discussion of what constitutes a content policy would need to consider them. I don't see anyone in the community disagreeing with that - realistically. They're just different types of policy to WP:V, WP:NPOV and WP:NOR, which are complementary. It is the complementary nature of the policies that is key here. jguk 00:05, 26 November 2006 (UTC)

The understanding that there are three main content policies V, NOR, and V is a long-standing one. That is all I am saying. If you want to change that, you'll need to do a lot of convincing around. ≈ jossi ≈ (talk) 00:54, 26 November 2006 (UTC)

I'm suggesting that the important thing to note is that V, NOR and NPOV complement each other. They work hand-in-hand. Hence the suggested revised wording. jguk 12:15, 26 November 2006 (UTC)

The classification of all policies, at Wikipedia:List of policies, has the three main content policies in the "Content and Style" category, along with four other (lesser) policies. "Legal and copyright" is a separate category.

Perhaps it would suffice to change this policy so that the first sentence reads three main content policies rather than three content policies? John Broughton | Talk 22:26, 26 November 2006 (UTC)

Isn't the key point that the policies work hand-in-hand? If so, we should stress that, not how many similar policies there are. jguk 12:46, 28 November 2006 (UTC)

As the suggested rewording does not dispute the number of content policies (eg by suggesting a different number), which appears to be the concerns discussed I above, I'm tempted to introduce it into the text (absent further comments here to the contrary). jguk 12:35, 29 November 2006 (UTC)

As said above, these three policies are our three main content policies, so I do not see the need to change the wording besides adding "main" ≈ jossi ≈ (talk) 17:44, 29 November 2006 (UTC)

Why do you see the number as so important? Indeed, more important than saying that they need to be considered together rather than in isolation? That's what's puzzling me. jguk 07:21, 30 November 2006 (UTC)

I support the change. The number isn't important and it's accuracy is debatable (it doesn't matter if it is accurate or not, this debate proves it's debatable), so it might as well be removed. --Tango 23:46, 3 December 2006 (UTC)

If you want to change this, you will need to argue the case also at WP:NOR and WP:NPOV as these use exactly the same wording. ≈ jossi ≈ (talk) 00:03, 4 December 2006 (UTC)

It should only be argued once. I'll put a note on those talk pages pointing to this discussion. --Tango 00:05, 4 December 2006 (UTC)

The three are so tied together they should be made into a single policy. Efforts to join at least two of them is ingoing. The point is that they are tightly interconnected; not that they are "main". As they are not more important than the policies which say to stay legal by not invading the privacy of nonpublic persons (BLP), not breaking IP laws (mainly copyright), and not defaming (people and corporations). WAS 4.250 00:15, 4 December 2006 (UTC)

I agree. Hopefuly we can soon get enough support for WP:ATT that does already an admirable job of combining V, NOR and RS into one sensible policy. ≈ jossi ≈ (talk) 00:23, 4 December 2006 (UTC)

I don't think this is such a big deal, but I note that in general our policy (and guideline) pages don't start with "this is one of our 5 legal policies" or "this page is one of our 7 spam guidelines", so I don't quite see why this one should. An important reason for having policy pages is educating novice users. Thus, listing related policies is very useful, and stating exactly how many policies we have doesn't really help them. (Radiant) 15:44, 4 December 2006 (UTC)

Prohibiting "self-published" sources?

Hi.

Does "self-published" sources include publishing your research in an academic journal? Also, why the prohibition, anyway? Is it because there is no way to check the accuracy of such things, at least not in a Wikipedia-compatible way (because Wikipedia is NOT a peer reviewer for reference sources, Wikipedia is just an ENCYCLOPEDIA)? 70.101.147.74 03:21, 26 November 2006 (UTC)

If you publish a paper in an academic journal, then the owner of the journal is the publisher, so unless you own the journal, that isn't self-publishing. The prohibition is meant to keep people from using Wikipedia to publish crackpot science, fringe historical analysis, and other such material. Since most professional publishers have some level of fact checking or other quality review, this limitation provides some minimal level of quality assurance for Wikipedia. dryguy 15:26, 26 November 2006 (UTC)

It also limits the use of advertisements, press releases, company web sites and so on. While these things may be valid supplementary material, the core of an article about a person or organization should be third-party reports, not the unchecked claims of the subject or affiliated persons (first party) nor of anyone with a potential adversarial relationship to the subject (second party). The reasoning is exactly the same: we rely on others who are better qualified and better equipped to check the facts. Robert A.West (Talk) 18:23, 26 November 2006 (UTC)

That is correct. Self-published sources can be used alongside other third party sources, with proper attribution of the self-published sources. Press releases if properly attributed, can present the organization's viewpoints on certain aspects that may be needed for NPOV. Remember that NPOV asks editors to describe significant viewpoints, and these include the viewpoints of the subject of the article. On the other hand, if there are no third party sources on a subject, then most probably the subject does not deserve an article in WP. ≈ jossi ≈ (talk) 19:41, 26 November 2006 (UTC)

Precisely why I want to make sure that phrasing remains clear in the proposed WP:ATT. Robert A.West (Talk) 19:52, 26 November 2006 (UTC)

And yet WP:ATT still includes that absurd (and thankfully oft-ignored) requirement that if we use a primary source, that source has to be intelligible to someone knowing absolutely nothing about the subject. If we're going for clarity, and common sense, we need to make clear that any reliable source can be used to substantiate claims. The nature of the source is a factor taken into consideration when deciding if it is reliable, not a determinant of whether we should use it. jguk 20:43, 26 November 2006 (UTC)

From what I have seen, 99% of the time, the primary/secondary advice is right on target. Professional historians are taught to use primary sources because they are supposed to conduct original research, which Wikipedia eschues. OR is something that should be done by trained professionals or under professional guidance: Don't try this on Wikipedia. If you can use a primary source in a way that does not constitute OR, then you are either using it to support a reliable secondary source's interpretation (acceptable) or its interpretation does not require special expertise. There's no third option. Robert A.West (Talk) 20:53, 26 November 2006 (UTC)

Two points, if I may. First, it is perfectly reasonable for a specialist writing on Wikipedia to deduce from a primary source information that any specialist in that field would deduce. That, of itself, is not original research since there is nothing original about it. Second, although the approach stated on WP:ATT and WP:NOR has some validity when talking about contentious issues where there is significant research and debate being conducted, it is far too rigorous for non-contentious points, particularly in non-academic areas. It's reasonable, for example, to use a first-hand account of a sports match written in a newspaper as a reliable source for what happened in that game. It's also reasonable to use an official communiqué from a summit as evidence of what was concluded at that summit, even if it is couched in technical jargon. Rigour has its place - sometimes we need to be as rigorous as &*!@ - particularly in our more academic articles - but at other times we can accept that for low-impact, non-contentious information, lower quality sources are ok. jguk 21:03, 26 November 2006 (UTC)

I agree that all Wikipedia-vetted specialists can do as you suggest. There's just one problem: Wikipedia does not vet specialists, so the set of editors who can validly do that is empty. If the specialist can make his reasoning apparent to enough editors to establish consensus that the reasoning is straightforward and obviously correct, then specialist knowledge is not absolutely required. It's helpful, but not essential. As a case in point, the legal articles benefit greatly from the input of legal specialists to distinguish rationes decidendi from obiter dicta, but lots of persons other than legal technicians are sufficiently skilled in close reading to appreciate and evaluate the results. Moreover, the conclusions can be checked in secondary authority, which answer to secondary sources in historiography. This is not a matter of disresepecting experts, but of not accepting unverifiable claims of expertise and going with the genuine, published experts. Robert A.West (Talk) 21:41, 26 November 2006 (UTC)

The newspaper account of a sports match is, for Wikipedia purposes, sufficiently separated from the event that it is not what we mean by a primary source, unless there is a serious issue of bias or culture. Consider some of the unfair and unreasonable things that were written about Jackie Robinson at the time. I agree that the terminology breaks down, because an historian would consider it a primary source, but I have lost that point repeatedly, and think that it would be better to explain in a FAQ than argue further. As for the official communiqué, using them for the bare facts of who was in attendance is clearly permitted. On the other hand, any official description of what was accomplished, unless a signed protocol, needs interpretation. What do we do with, "A frank exchange of views and productive dialogue on core values?" We all know that probably means a verbal row, yet we can't rely on "common knowledge" to do that. Robert A.West (Talk) 21:51, 26 November 2006 (UTC)

What if your self-founded, self-owned publisher becomes a very big, trustworthy and reputable company? 170.215.83.4 08:17, 28 November 2006 (UTC)

Robert West wrote that is OK to use a primary source in WP, provided " its interpretation does not require special expertise." This is not quite true. It's perfectly OK if the reader will need special expertise to interpret a passage that is a direct quote or close paraphrase of a primary source; it is only if the editor is using specialist expertise to interpret the source that there is a problem. Also note that when editors discuss a passage on a talk page, it is not unusual for the editors to have specialist knowledge, and the conversation may very well be unintelligible to a reader chosen at random. --Gerry Ashton 23:13, 26 November 2006 (UTC)

• Talk pages are not encyclopedia space, so are rather beside the point: one cannot discuss whether to use a source and be fully within NOR/V/NPOV. In any case, I thought there was no ambiguity that we are talking about editor-supplied interpretations. Readers are always free to interpret anything. Of course, in some cases inclusion of the uninterpreted primary-source content would violate NPOV, but that is another matter. Robert A.West (Talk) 23:29, 26 November 2006 (UTC)

My point is that when discussing whether an article passage correctly represents the primary source it is drawn from, the editors discussing the passage are likely to use their specialist knowledge. It would seem Robert West is telling us that a consensus of the editors that the passage agrees with the source is insufficient; there should be a consensus of the editors that any randomly chosen reader could see that the passage agrees with the source. --Gerry Ashton 23:44, 26 November 2006 (UTC)

There are four basic cases. I consider the minutes of the Wannsee Conference as the source in each case.

1. The editors make no interpretive claims, and none are required. Example: The minutes are cited as the source for attendees, etc. This is not a problem, since minutes as such are within common experience. If we have another source that says Heydrich was late for one meeting, we can expect most readers to understand that minutes often fail to reflect tardyness.
2. The editors make no interpretive claims, and some are required. Example: The minutes are quoted to the effect that "forced emigration" was the goal. This is improper: a non-specialist might not realize this is euphemism.
3. The editors make interpretive claims on their own. Example: The euphemistic language is exposed without secondary source. This is improper: a reliable source is needed for such a major claim.
4. The editors cite a number of secondary sources on the nature of the euphemisms involved, and use these to interpret the primary source. This is not a problem, as the sources for the claim are reliable and secondary.

With which of the above is Gerry Ashton disagreeing? Robert A.West (Talk) 01:47, 27 November 2006 (UTC)

The question in my mind is when does an allowable paraphrase become a disallowed interpretation (assuming the paraphrase is created by a WP editor and is not taken from a secondary source). I don't see anything in the minutes of the Wannsee Conference that is apt to generate grey-area paraphrases, but the problem could easily occur in technical articles. For example, text from an old published lab notebook is paraphrased, but units are converted into modern units. The editors discuss it on the talk page, and a concensus forms that the conversions are correct, so the paraphrase is OK. The editors who frequent the talk page can do the most complex unit conversions in their sleep. But, the conversion is beyond the skill of many Wikipedia readers, so perhaps the conversion is contrary to the Verifiability policy. --Gerry Ashton 02:32, 27 November 2006 (UTC)

One does not need to be a specialist to do unit conversions. Any reasonably bright eighth-grader science student should be able to do them. Now, some Wikipedians lack that level of skill, but enough have it that we have a reasonable expectation that there are hordes of editors and readers who remember dimensional analysis and can check the work. Now, I understand that there is a gray area -- there always is. What if the reasoning involved an integration by parts? To how many Wikipedians is it obvious how Bayes theorem applies to Democratic peace theory? The last is not common enough to salvage such an observation from being OR. I am not sure which side of the line I would place the integral: it would probably depend on whether the math project could be expected to take a look. Robert A.West (Talk) 02:49, 27 November 2006 (UTC)

• Oh, becuase that might allow bias on the part of the "expert interpreter", pontentially leading to a breach of WP:NPOV? 170.215.83.4 08:17, 28 November 2006 (UTC)

• • Or, he might not do the work correctly and just be flat-out wrong. A published source will have been double-checked by someone competent. Robert A.West (Talk) 08:33, 28 November 2006 (UTC)

• • • What do you mean that he is "flat-out wrong"? I thought that statements need not be true, it even says so on WP:V. Any clarification here? 170.215.83.4 03:43, 29 November 2006 (UTC)

By "double-checked" we can only say "it says what it says". We cannot repeat the analysis or experiment that led to it saying what it says. Verification means, going to that source and checking that it actually says what it says, nothing more. "Flat-out wrong" is not something we do here. We *can* however use conflicting sources to show that some fact is in conflict. We cannot present results of our own experiments to show a conflict. We *can* use sources, not experiments, to show that there is a conflict. Wjhonson 19:45, 2 December 2006 (UTC)

JBKramer's edit

In the subsection WP:V#Self-published sources (online and paper) JBKramer added the following sentence:

However, for articles that are about content that was published first and even exclusively on blogs, linking to the specified blogs is not only acceptable, but practically a requirement.

I feel this is overly broad. For the most part, content that has been exclusively published in blogs should not have a Wikipedia article about it. --Gerry Ashton 02:41, 27 November 2006 (UTC)

• I agree wholeheartedly. Robert A.West (Talk) 02:51, 27 November 2006 (UTC)
• I reverted. Apparently, the concern was over the Rather/Bush papers affair, where blogs played an acknowledged role. Once secondary sources start analyzing blog reports, I can't see any reason to treat them less well than other primary sources. The only risk is that the blogger will modify his blog to emphasize his role, and that is not such a huge risk. Robert A.West (Talk) 03:07, 27 November 2006 (UTC)
• I edited the policy to what appeared to be strong consensus, to which I deferred. JBKramer 03:09, 27 November 2006 (UTC)
  • I saw no discussion here of those issues. What consensus? Robert A.West (Talk) 03:12, 27 November 2006 (UTC)
    • It was clear from the results of the RFC on killian documents that I misunderstood policy to prevent blogs as being used as sources. If it turns out that I accurately understood policy, then I suggest there are a number of individuals responding to RFCs that are not remotely versed in policy. JBKramer 03:38, 27 November 2006 (UTC)

That article is a very specific case. There is no need to change policy to accommodate exceptions. ≈ jossi ≈ (talk) 03:55, 27 November 2006 (UTC)

And not much of an exception, either. The blog claims formed the basis for research by journalists, and were verifiably cited, IIRC. Wikipedia is relying on the confirming research moreso than the blogs. Robert A.West (Talk) 06:07, 27 November 2006 (UTC)

Generally, I don't think 6 or 7 posts over a holiday weekend constitute a sufficient concensus to change the policy. Should we discuss whether "stories that broke on blogs, then moved to MSM" (the Killian documents and Mark Foley scandal pages are fine examples) need a specific discussion, and whether it should be on WP:V, WP:RS, or WP:ATT? TheronJ 14:33, 27 November 2006 (UTC)

I don't see what the big issue is. As soon as MSM picked it up, the blog became a primary source supporting a secondary source, and we have rules for using those. If something can come in under either of two theories, treat it less restrictively, not more. Robert A.West (Talk) 14:52, 27 November 2006 (UTC)

• Yes, WP:NOT a blog and a directory. If you want to comment on blog entries, comment on the blog's comment page, not on the Wikipedia (and opinionated comments are just about all you can make if all you have as a reference ("exclusively") is the blog, so these comments would be "by definition" totally and irredemably unverifiable and POV, so also unsuitable for inclusion by those as well.). 170.215.83.4 19:42, 29 November 2006 (UTC)

Challenged

ElectricEye, not everything that is unreferenced is likely to be challenged. That 2 plus 2 is 4 is not likely to be challenged. The phrase implies that a degree of common sense is needed. SlimVirgin ^(talk) 03:14, 27 November 2006 (UTC)

Yes, that is correct, but common sense without presenting a reliable source is original research. ^_^ In my opinion, everything can be sourced and if things arent they should be sourced or simply erased. --ElectricEye (talk) 03:23, 27 November 2006 (UTC)

That is your opinion, but it is neither consensus, nor sensible. By the very nature of human discourse, a few tidbits will always slip through uncited. Mostly, they will not even be noticed. It isn't original research to condense a list of offspring into "Five sons and four daughters." Wikipedians can count, add, subtract and note that if Paris is the capital of France, it is certainly in France. Robert A.West (Talk) 03:36, 27 November 2006 (UTC)

LOL. No need to tell me this, I already know it: it's common sense. I'm talking about something else. --ElectricEye (talk) 04:56, 1 December 2006 (UTC)

EE, if we were going to delete every uncited sentence, we'd have to delete most of every article. By all means start the process and we'll see how you get on. :-) SlimVirgin ^(talk) 04:22, 27 November 2006 (UTC) (that's a joke by the way, just in case ...) ;-D

Hehe, I've seen that joke before! --ElectricEye (talk) 04:58, 1 December 2006 (UTC)

Fast forward: User gets banned for vandalism. —Centrx→talk • 05:57, 27 November 2006 (UTC)

I would WP:AGF if any editor removed all unsourced listings of Prominent figures in some the Somali clan articles, for example. Which is part of the basis for my coming here and bringing this up. ^_^ --ElectricEye (talk) 04:59, 1 December 2006 (UTC)

Awwww...my money was on WP:POINT. Rats! Robert A.West (Talk) 06:01, 27 November 2006 (UTC)

How much money? =) --ElectricEye (talk) 05:00, 1 December 2006 (UTC)

Take a look at the Somali clans where they start listing Prominent figures. You'd think common sense says to remove the unverified information to the talk page for discussion, verification and the results written into the article. Not the other way around. --ElectricEye (talk) 05:01, 1 December 2006 (UTC)

It's a judgment call as to whether disputed items are removed from the text and taken to the talk page, or kept in the text until after someone is given an opportunity to provide their citation. The only exception is if the claim relates to a living person or is otherwise highly controversial (ie damage could occur because of a mistake). Different WPians prefer different approaches. Pick whichever one suits you best.

Just remember that the aim is to keep valid information in WP, and this means encouraging the citing of reliable sources. The way that encouragement is given is not so important. jguk 09:14, 1 December 2006 (UTC)

Thanks, Jguk. I'll bring this up with some other editors who seem to be watching these articles and see what they think we should do. --ElectricEye (talk) 11:55, 1 December 2006 (UTC)

Confused by change

I haven't been following this closely, but I see that the following change has been made: "The obligation to provide a reliable source lies with the editors wishing to include the material, not on those seeking to challenge or remove it." Can someone give a short, clear summary of why this change was made? It seems quite odd to me. - Jmabel | Talk 07:03, 29 November 2006 (UTC)

I agree that it seems odd. Some Wikipedians, particularly for less controversial claims, prefer to leave information in the text but challenge it on the talk page (perhaps with adding a "citation needed" message in the text too). My understanding is that action, provided it is done in good faith, gives the obligation on editors wishing to retain material to justify it. The old text covers this situation. The revision doesn't. jguk 12:28, 29 November 2006 (UTC)

seeking to challenge is an odd way to put it, I think. More proper would be not on those challenging it or seeking to remove it. Putting a {{fact}} tag into an article IS a challenge to whatever material is tagged. (And I support reversing the change, with slightly clearly language.) John Broughton | Talk 01:41, 30 November 2006 (UTC)

Avoiding confusion by making language clearer

There's one section of this page that has long been a bugbear of mine - largely because it is so confusing. Those used to teaching people or taking people through new topics will know that once you have confused someone on a key concept, it takes a lot to get them unconfused. Much better to lay down what you mean in a clear, unambiguous logical format first.

The passage I don't like is:

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

The confusing bit is the "not truth". It seems weird to state upfront that WP does not care about truth. Of course, that's not what it means, but then there's nothing other than the "not truth" bit, in bold font, that a new reader has to work on to understand what we mean. It's no surprise that even those who defend keeping a reference to the point in the policy freely admit that many WPians are confused on the point, and no mistake that it is listed as the most frequently asked question on WP:ATT/FAQ.

We really should address this confusion, and after much thought, I think I have a formula that has a fair chance of succeeding. The revised text (absent bolding) would be:

The threshold for inclusion in Wikipedia is verifiability. It is not sufficient for something to be true, you need to demonstrate that it is true by providing a reliable source.

"Verifiable" in this context means that any reader should be able to check that material added to Wikipedia has already been published by a reliable source. Editors should provide a reliable source for material that is challenged or likely to be challenged, or it may be removed.

I reckon that this covers the key point of the "not truth" bit, and yet, by explaining it rather than stating it as a stark statement, means that there would be little confusion. jguk 14:50, 1 December 2006 (UTC)

This has been discussed already many a times. The formulation of "Verifiability, not truth" has been there for a long time and it is designed to make the point across. ≈ jossi ≈ (talk) 16:27, 1 December 2006 (UTC)

I'm not disputing what it has been designed to do. I'm pointing out the quite obvious issue that it patently fails to get the right point across, and suggesting a way to put that point across that is clear. jguk 16:58, 1 December 2006 (UTC)

Maybe something like this, as a compromise? (I've tinkered with the last sentence, too.)

The threshold for inclusion in Wikipedia is not truth but verifiability. "Verifiable" in this context means that any reader should be able to check that material added to Wikipedia has already been published by a reliable source. It is not sufficient for something to be true; you need to demonstrate that it is true by providing a reliable source. This is particularly true for material that is challenged or likely to be challenged; without a source, it is likely to be removed.

-- John Broughton | Talk 17:21, 1 December 2006 (UTC)

The rationale has to come as soon as the concept of "not truth" is introduced. Otherwise, you start by introducing confusion. A slight reordering of what you suggest (and either removing the bold font or putting the first two clauses in bold too) could achieve this (although I think my initial suggestion is better still):

It is not sufficient for something to be true, you need to demonstrate that it is true by providing a reliable source: the threshold for inclusion in Wikipedia is not truth but verifiability. "Verifiable" in this context means that any reader should be able to check that material added to Wikipedia has already been published by a reliable source. This is particularly true for material that is challenged or likely to be challenged; without a source, it is likely to be removed.

jguk 17:33, 1 December 2006 (UTC)

I'd reword the first two sentences, for better flow/grammar, and change a word in the penultimate sentence and punctuation in the final sentence -- all tweaks, admittedly:

It is not sufficient for something to be true -- you need to demonstrate that it is true by providing a reliable source. The threshold for inclusion in Wikipedia is not truth but verifiability. "Verifiable" in this context means that any reader should be able to confirm that material added to Wikipedia has already been published by a reliable source. This is particularly true for material that is challenged or likely to be challenged: without a source, it is likely to be removed.

But I think you're going to run into trouble with this because the core concept - threshold - is in the second sentence, and I think a lot of poeple will argue that it should remain in/as the first sentence -- John Broughton | Talk 15:35, 2 December 2006 (UTC)

My original wording addresses this point:

The threshold for inclusion in Wikipedia is verifiability. It is not sufficient for something to be true, you need to demonstrate that it is true by providing a reliable source.

"Verifiable" in this context means that any reader should be able to check that material added to Wikipedia has already been published by a reliable source. Editors should provide a reliable source for material that is challenged or likely to be challenged, or it may be removed.

jguk 18:04, 3 December 2006 (UTC)

Internal links to third party sources

See also previous discussion Wikipedia talk:Verifiability/archive8#Explicitly say Wikipedia articles cannot serve as references for other articles and Wikipedia talk:Cite sources/archive10#"Note: Wikipedia articles can't be used as sources"

I think this issue needs to be discussed further because it has been raised in the debate taking place at Wikipedia:Speedy deletion criterion for unsourced articles.

Using internal links to another article which contain the sources for a statment is very common in Wikipedia articles. An example given in the earlier discussion was let us suppose that there is an article called Ten highest mountains in Africa and that, that list had been derived from 100 sources, with a detailed analysis of why those particular peaks have been chosen for the Wikipedia list. The current wording of this plolicy implies that all 100 sources need to be included in every article which mentions the list of 10 highest mountains in Africa. Placing every reference in every article for every Wikipedia link used as a source becomes a real maintenance issue: suppose that the sources for the list changes, then it would be necessary to edit all the subsidiary articles which use the list as well as the list itself. In day to day editing of Wikipedia articles, people are willing to accept statements like "mountain xyz is the fourth highest mountain in Africa", because they are willing to follow the link to the parent article for the sources. Only if the link article (parent article) does not have adequate sources are they likely to demand under this policy that the statement in the child article contain a third party source for verification. I think that this policy needs changing to reflect this because many/most articles rely to a lesser or greater extent on this premise.

There is of course a limit to how many nested links one should have to follow to find third party sources. I think that only a single link should be allowed, otherwise there is a danger of recursion where a circular link is created with no third party source cited for a disputed fact.

Some article types which rely very heavily on the use of links providing sources are:

• Wikipedia:Summary style (this includes many small stubs)
• Wikipedia:Timeline
• Wikipedia:List
• Wikipedia:Disambiguation.

If this policy is taken at face value then all disimbaguation pages need third party sources, unless one is willing to interpret the phrases: 2 Editors adding new material to an article should cite a reliable source, or it may be challenged or removed by any editor.. 3 The obligation to provide a reliable source lies with the editors wishing to include the material, not on those seeking to challenge and/or remove it. to mean, as they do in practice, that an internal link can be used to provide reliable sources.

To stop pedants using WP:V as a club (and wasting a lot of time on reverts and talk page discussions), I think that it is necessary to include in this policy a statement, that reflects how Wikipedia links are already used, that third party sources can be provided by in-line links to other Wikipedia articles that already contain adequate third party sources.

--Philip Baird Shearer 12:23, 22 November 2006 (UTC)

I don't think this is the intention of the statement that Wikipedia can't be used as a source. I always consider this kind of thing as "including references by reference", so to speak. But what we can't do is take one unsourced article and use it to back up another one; that just wouldn't work, and is what (I believe) the wording in the policy was intended to prevent. JulesH 12:01, 24 November 2006 (UTC)

I think JulesH is correct. WP:V has to be applied using common sense. A Wikipedia article cannot, of itself, be a source for another Wikipedia article - but it can be a useful reference for a list, say, provided the Wikipedia article being linked to clearly references the information being used in the article being linked from. If Philip, or someone else, can think of revised wording that makes this clear, then I'd support adding it to the policy page, if not, I'm inclined to leave it. jguk 13:16, 24 November 2006 (UTC)

DAB pages, lists and redirects are commonly held verifiable if they point to properly-sourced articles. Redirects never have sources. This is less a matter of WP sourcing WP than of a distinction between entities in article space and actual articles. Articles written in summary style are articles and tend to rely on the constituent articles, but again, they are not intended as stand-alone entities. Robert A.West (Talk) 19:33, 24 November 2006 (UTC)

I'm afraid I don't agree with the addition of "Reliable sources can be provided by an in-line link to another Wikipedia article that already contain adequate third-party sources" into the policy box.

Although true, I don't think it is as essential as the other three points already in the box. Plus, some will inevitably misunderstand it to mean that it is uniformly desirable to have internal links as sources.

I wouldn't be so concerned, however, if the text were removed from the policy box and placed later down in the discussion of the policy. jguk 13:54, 1 December 2006 (UTC)

I removed a numbered item that encouraged providing a reliable source by linking to another Wikipedia article that has the source. With the specific exceptions of dab pages and articles written in WP:Summary style, this is a generally bad practice, because there is no guarantee that the source will remain in the target article. It also encourages laziness, because the target article may not use the source accurately. Robert A.West (Talk) 17:22, 1 December 2006 (UTC)

• I should add that I would be flabbergasted to learn that there was consensus for the addition. Robert A.West (Talk) 17:23, 1 December 2006 (UTC)

Right... "This statement is referenced somewhere in the history of some other article" is a terrible way to reference something. Jkelly 17:24, 1 December 2006 (UTC)

There's discussion of this up above at Wikipedia_talk:Verifiability#Internal_links_to_third_party_sources. Can we somehow merge the discussions? jguk 17:35, 1 December 2006 (UTC)

• I have boldly moved the section referenced to become the parent heading of this discussion. Robert A.West (Talk) 04:09, 2 December 2006 (UTC)

To say "this is a generally bad practice, because there is no guarantee that the source will remain in the target article." is no more valid than saying that linking to on line sources is a bad idea because there is no guarantee that the source will remain on line. Are you really saying that all the articles on Wikipedia which rely on in-line links need to be source. The place to challenge for more sources is in the in-line link page not on the sentence that contains the in-line link. After all that will fix the source problem on two or more pages simultaneously.

The claim that in-line links to other Wikipedia articles are not sufficient is a massive move away from what is currently accepted in the day to day editing of Wikipedia articles. See for example the list I put higher up the page, are you claiming that all those type of articles should be removed if they do not contain reliable sources? Also what about the practical issue of the overheads on maintenance? --Philip Baird Shearer 15:53, 2 December 2006 (UTC)

You previously asked for a detailed discussion, and I gave it to you in the section below. I am not sure what more you want.

As for policy. you have it backwards. Wikipedia policy is and has been that a Wiki (any Wiki) is not a reliable source. The main reason has always been as I stated -- in contrast to most sites, a Wiki is subject to arbitrary change without notice. The second reason is that there is no guarantee that any Wikipedia article has accurately represented its references. If the editor of article A goes to the trouble of checking the references in article B, he may as well go ahead and include them in A. The only reason to cite article A is to save the trouble of reference-checking in the first place.

As for the four special cases you cite, and which I discuss below, the deleted phrasing did not make me think about them -- it looked to me like blanket permission to bypass verifiability for any article by reference to any other article. Followed to its logical conclusion, Wikipedia could become a daisy-chain of bad information: article A relies on article B which relies on article C which once had a dubious source but now relies on article A. Yes, there are undoubtedly thousands of articles that do this. There are also undoubtedly thousands of copyvios, instances of original research, unsourced defamatory statements about living persons and so on. The purpose of a policy page is not to describe what we do when we are at our worst and then enshrine it as acceptable practice. Robert A.West (Talk) 18:33, 2 December 2006 (UTC)

I am aware of the daisy-chain problem if you look at the previous discussions you will see an example I gave (Bundesland), which is why I phrased the addition the way I did: "Reliable sources can be provided by an in-line link to another Wikipedia article that already contain adequate third-party sources." as this removes the daisy-chain problem. It is also phrased in such a way that it is not the tertiary information in the Wikipedia in-line link article that is being cited, but the adequate third-party sources in the inline-link article. --Philip Baird Shearer 12:35, 3 December 2006 (UTC)

The five cases

There are five cases where one article might rely on another for verifiability. Some are valid, and others are bad practice. Two of the valid cases I would class as "navigational aids" rather than "articles." If the intention of the addition I reverted was to provide for the valid cases, then I think it was unfortunately phrased. In that event, it could be mentioned as a set of exception farther on down.

1. Pure navigational aids: redirects and categories. There is simply no practical way to include references, so anyone demanding we do so is being a dick.
2. Disambiguation pages. So long as they only contain enough information to find the correct article, they are really another species of navigational aid -- they certainly do not stand on their own. I would commend including references for the redlinked items, since one can hardly rely on an article that hasn't been written yet, and the reference can serve as a starting point for the article.
3. Lists and timelines. Some of these are similar to disambiguation pages; some make substantial assertions. If the information given is brief, central to the article linked and so indisputable that it is inconceivable that it would ever be superseded, then no purpose would be served by a separate reference other than to make work. Otherwise, we should insist on sources.
4. Articles written in Wikipedia:Summary style. Again, these are articles that are not intended to stand on their own. Substantial assertions, and any assertion that could conceivably be removed from the subarticle should be referenced. This is already covered.
5. Articles that should stand on their own. Such articles should never use other articles as a verifiability shortcut. I grant that we have thousands of examples of articles that do just this, but we shouldn't endorse it. There are two risks: the target article might be wrong, and the statement referenced might vanish from the target article. If it is obvious that the source cited in the target article is a good one, importing the correct source should be the work of a few moments.

Cases 1-4 are valid exceptions, although 3 and 4 are not complete exceptions. Case 5 is what I meant by "very bad practice." I have been guilty of that particular sin myself, but that does not mean that doing so improves Wikipedia. Robert A.West (Talk) 05:08, 2 December 2006 (UTC)

I think there is a case where articles in category 5 can do this validly, and that is where the information being sourced is tangential to the subject of the article. If we have an article about a mountaineer, say, and want to say "he has climbed both Everest and K2, the two tallest mountains in the world", I think it would be acceptable to point to List of highest mountains, a well-sourced article on the subject, as a reference for this. It's highly unlikely that those sources will be removed, and even if they are it will almost certainly be because something is wrong with them. So I can't say I really understand your concerns. JulesH 11:34, 2 December 2006 (UTC)

In the example cited, we would need a reliable source for the fact that the mountaineer climed both mountains. (Its just the sort of claim we shouldn't accept from a self-published source.) The tangential fact that Everest and K2 are the two highest mountains in the world could, IMO, pass without a citation, because the fact is trivially checked using obvious works of general reference that nearly everyone has. Its the sort of fact that is not "likely to be challenged." Using a wikilink to reinforce the point is harmless.

Now, contrast this with the case where the mountaineer helped identify the bodies from a disaster. This fact is significant to both the biography of the climber and the account of the disaster and should be well sourced in both cases. Neither article should lean on the other, because it is uncertain how each will change. As noted in the edit summary of the paragraph I reverted, there are thousands of articles that lean on other articles in this way. It is preferable to no sources, but we should not give blanket permission for it.

I have seen lots of articles that lean on other articles for sources and that should not. I have never seen an editor demand a source for a redirect, a true dab page or a list that was just a set of article pointers. The proposed paragraph seems to be solving a problem we have not seen, at the risk of exacerbating a problem we have. Robert A.West (Talk) 12:57, 2 December 2006 (UTC)

Robert, clearly you have not been involved in the debate about Wikipedia:Speedy deletion criterion for unsourced articles, because it is a problem, believe me if it is not spelt out in this policy page then if that idea is implemented expect all the type of pages you have mentioned above to be flagged for deletion as they do not meet the WP:V policy. It is no good arguing as you have above that custom and guidelines (eg the above example Wikipedia:Summary style#Citations and external links | already covered) cover it, because custom and guidelines are not policy. The policy needs adjustment to reflect the real Wikipedia world and as it stands at the moment almost all of what you say above fail the policy test of "1. Articles should contain only material that has been published by reliable sources. 2. Editors adding new material to an article should cite a reliable source, or it may be challenged or removed by any editor. 3. The obligation to provide a reliable source lies with the editors wishing to include the material, not on those seeking to remove it."

Many articles are both articles in their own right, but also contain lists or are in part summary pages. Above you write "IMO, pass without a citation, because the fact is trivially checked using obvious works of general reference that nearly everyone has. Its the sort of fact that is not "likely to be challenged." Using a wikilink to reinforce the point is harmless." The trouble with this is who defines what is trivially checked out. For example the an article might state "A mercenary is an unlawful combatant." or "Napoleon Bonaparte lost the Battle of Waterloo." or "There is no positive international law banning the use of depleted uranium sabot rounds" or "Charles I was executed on January 30 1649". What is an obvious fact to someone and easy to look up in a reference that they have, is to another an obscure fact which is why they are reading the article on Wikipedia and the link provides an easy way to check the fact in a reference that they posses. This policy should reflect the everyday fact that this is how pages in wikipedia are structured using in-line links to other more detailed accounts. This article "meta:wiki is not paper" has an interesting section (Style and functionality) which say:

It is unclear if IBM's code page 437 (a character code set)
was based on the VT-220 terminal (a computer input/output
device) of DEC (Digital Equipment Corporation), or if the
reverse was the case.

It is unclear if IBM's code page 437 was based on the VT220 terminal
of DEC, or if the reverse was the case. 

I think that this argument also applies to references. --Philip Baird Shearer 19:54, 2 December 2006 (UTC)

ROFL. Obviously, you are talking about a different Wikipedia talk:Speedy deletion criterion for unsourced articles, because I helped start that thread, the proposal to limit it to Biographies of Living Persons is mine and the UfD is mine. Since 18 November, the proposal has contained (Revision as of 05:59, 18 November 2006 ) and still contains (Proposed_text_.231) an explicit exception (variously phrased) for the sorts of pages I covered in cases 1-3. Few summary style pages can go completely without references, so it is more a standard verifiability issue -- how many references does such an article need? Robert A.West (Talk) 18:01, 3 December 2006 (UTC)

Thank you for pointing out that you are aware of the problem, in which case I fail to see why you write "The proposed paragraph seems to be solving a problem we have not seen, at the risk of exacerbating a problem we have". The proposals at (Speedy deletion criterion for unsourced articles) seemed to have been added after I raised this issue on the talk page (88155683). But I notice that in the latest proposal #4 (23:37, 21 November 2006 208.20.251.27) that the wording is now "The process should not be applied to disambiguation pages or those written insummary style" The trouble with this wording is that an article may be written in summary style, but the in-line references may be to wikipedia links which do not contain any third party sources! (see for example this version of List of War Crimes (Revision as of 13:44, 21 March 2006) this is clearly unnaccptable and it does not cover summary style articles which link in-line to summary style artiles (the diasy-chain problem). The proposed wording I have suggested for this policy covers text in all article, (removing the potential problem of a policy relying on guidlines for policy statments eg "what is a summary page?") and does away with the problem of an in-line Wikipedia link pages not containin adequate third party sources. --Philip Baird Shearer 21:32, 3 December 2006 (UTC)

• OK, we have a number of misunderstandings here. The first and foremost is that a proposal should be responded to by major changes in existing policy. If there are defects in the proposals, or exceptions that should be made clearer, then the relevant project page is the place to address them. The second misunderstanding is confusing process (XfD or a CSD criterion) with the policy that it is meant to enforce (Verifiability). The vast majority of articles that should be deleted are not valid speedy deletions: speedies are only for the clearest cases where no discussion is required. I'll avoid going on, because the proposal is not germane to this page, and is unlikely to become policy, except maybe in the form of BLPs only. Robert A.West (Talk) 22:08, 3 December 2006 (UTC)

The "SDcdfua" is only tangential to this policy page. However it highlights a problem with this policy. The experienced editors who edit this page tend to be practical experienced editors who are more than capable of making sensible judgements about when a piece of text needs a third party citation. But as experienced editors they are also immersed in the current culture of editing Wikipedia pages, and, often without realising it, they interpret Wikipedia policies and guidelines with a lens which includes custom. However when it becomes clear that the wording of a policy is far removed from what is done on a day to day basis when editing pages, then I think the policy should be altered to reflect best practice. Most experienced editors accept a sentence or paragraph which uses in-line links to well sourced Wikipedia article, as reliably sourced. They are only likely to object if the sentence draws a "novel narrative or historical interpretation." or "Surprising or apparently important claims that are not widely known" that are not directly supported by the in-line links. In such a case they will demand a citation from reliable sources even if the in-line links are well sourced e.g. "The moon is a satellite of the earth and it is made of cheddar cheese". My suggested changes will not protect such sentences. But they will save the need to put in citations for such statements as "At the end of the Potsdam Conference, the Allies issued an ultimatum to Japan". Adding a sentence reflecting the wide spread acceptance of such in-line Wikipedia references into this policy will stop the need for guidlines like "SDcdfua" to add a complicated clause to put into words what is the customary interpretation by experienced editors of WP:V. This change, to bring into line the Verifiability policy with its customary interpretation, will help stop trolls and well intentioned inexperienced editors wasting everyone's time. --Philip Baird Shearer 07:33, 4 December 2006 (UTC)

Well, err, no: I'd stick to the rationale explained in Wikipedia:Don't use internal sources for verification.

As a reminder, one of the issues that led to me writing that essay was the birth date that used to be mentioned in the Jimmy Wales article. What happened? In short, some time ago Jimbo (himself) had (obviously copy-pasted) some autobiography in the Foundation wiki. Then, later, that content of the Foundation wiki was used as a reference for Jimbo's birthday in the Jimmy Wales article. It cost Jimbo a whole lot of trouble to get the erroneous date deleted from the Jimmy Wales article.

So, no, internal links should not play a role in the Verification mechanism of main namespace content, apart from the obvious exceptions mentioned in Wikipedia:Don't use internal sources for verification ("obvious exceptions", e.g. copyrights-related attribution in edit history when translating/importing content from another WikiMedia project; attribution of media via "image" namespace, and that's about it). --Francis Schonken 10:51, 8 December 2006 (UTC)

The wording I am proposing does not suggest that the content of the Wikipedia in-line article that is being used as a source, but the third party sources in that article. There is nothing exceptional about this, it is what is done in many (most?) articles to a greater or lesser extent. --Philip Baird Shearer 13:19, 8 December 2006 (UTC)

I don't understand why, in cases of conflict over this, you could not, just simply cut-and-paste the citation from one article to the other? I really doubt anyone is going to be such a dick as to request references for trivially verifiable content. We do not however, want to get into a position where we have circular references. So in any case of conflict, the least confrontational approach is simply to repeat the citations in each article where they are requested. Where is the harm? Wjhonson 18:40, 8 December 2006 (UTC)

My objection to relying on a wikilinked article for sources is it makes the passage in the source that backs up the Wikipedia article harder to find, because when you go to the linked article, it may have a broader range of information than the linking article. If, for example, an article about Aberdeen relied on a source from the Scotland article, the poor reader would have to look through 61 different sources, many of which have nothing to do with Aberdeen. Even if the article to be linked to does not have a long list of references now, it may in the future. (Clarified based on Wjonson's comment immediately below.) --Gerry Ashton 23:08, 8 December 2006 (UTC)

You object to my idea? I don't understand your objection. It appears you think I'm saying "cut and paste all sources". I'm not. I'm saying copy the one requested citation from one article to the other. If one article says "the tallest Mountain in Scotland is Mount Googleblat (see Hitchins, p20)" then you simple copy that one citation to the Mount Googleblat article. Not all citations in the Scotland article. Hope my position is clearer. Wjhonson 23:02, 8 December 2006 (UTC)

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Third-party sources

I have removed the word "third-party" from the sentence: "If an article topic has no reliable, third-party source Wikipedia should not have an article on it." I'm not sure how it got to be added, but I think it must have been in error. WP:V is not a notability criterion - notability is dealt with on other page. The verifiability requirement is, and should be, for a reliable source. If a source is not 3P there may (or may not) be some question marks over its reliability. Indeed, we are likely to be more rigorous in assessing its reliability. But if the source is reliable, it is good enough to be used. jguk 17:45, 1 December 2006 (UTC)

Re. the latest removal by Jguk. Would it not be better to say reliable secondary sources, as an article with only primary sources is a very bad idea and just inviting original research? L0b0t 17:43, 1 December 2006 (UTC)

• (edit conflict) I could not disagree with this novel change more. The words "third-party" have been in that sentence since it was added to the policy, and that was done after significant discussion and consensus was established. The requirement not to include topics based on self-claims is essential to maintaining any sort of credibility for Wikipedia. Robert A.West (Talk) 17:55, 1 December 2006 (UTC)

I disagree with the assertion that an article "with only primary sources ...[invites] original research". Original research is not the same as source-based research. I agree with removing "third-party". Wjhonson 18:00, 1 December 2006 (UTC)

I would respectfully invite you take a look at some of the television show episode articles (The Simpsons, Family Guy, American Dad, etc) where sole reliance on the episode itself immediatly leads to fanboy speculation and interpretation (This scene with a fish in show X is a parody of this other scene in show Y that has a fish.) Cheers. L0b0t 18:06, 1 December 2006 (UTC)

"Third-party" is not the same as "secondary" and we should not confuse them. A source can be secondary in form, and apparently reliable, but still be controlled by the subject of the article. This is often true with corporations that sponsor research related to their goals. Further, a source can be primary and yet unrelated to any of the participants. Thus, a newspaper account of an accident is a primary source, but is a third-party look at what happened. If the topic has no coverage that is independent of the subject, then Wikipedia will never be able to construct an NPOV article on that topic. Robert A.West (Talk) 18:13, 1 December 2006 (UTC)

See also Wikipedia:Don't use internal sources for verification --Francis Schonken 18:56, 1 December 2006 (UTC)

I agree that articles that rely only on self-published sources by people or organizations with a vested interest in the subject of the article are bad, and sources that are not technicallly self-published but are under the control of people or organizations with a vested interest are just as bad. I am concerned, however, that some readers of the policy might see the words third-party and think that sources written by a person with a vested interest, but published by an independent reputable publisher, are ruled out. It suffices that the publisher be a third-party; the author need not be. --Gerry Ashton 20:17, 1 December 2006 (UTC)

• Wikipedia's verifiability policy is already phrased to regard the publisher moreso than to the author. Professor Dryasdust's blog is a questionable source. If the exact same content by the same author is published in The Journal of Xerology, we regard it as a reliable source. With a newspaper, we regard the paper, not the reporter. The exceptions (advertisements, letters to the editor, family-written obituaries, the gossip column, etc.) are based on merely cursory editorial review for those sections, and affect all content in those sections, regardless of author. Robert A.West (Talk) 03:07, 2 December 2006 (UTC)

"Third-party" only serves to confuse not clarify the issue. We have lived with primary and secondary for a long time. Adding another layer does not help the situation. Wjhonson 19:49, 2 December 2006 (UTC)

I hadn't expected this amendment to be controversial. I really did see it as a mistake that had somewhere slipped into the text. Had I thought different, I would have raised the point here on the talk page first.

The main point of this page is that claims should be supported by reliable sources. Other issues are discussed in addition to that, but they, important as some are, are all either ancilliary or subsidiary.

WP does not have a definitive definition of "reliable sources" - and it is probably impossible to have one - but we all have ideas of which sorts of sources tend to be more reliable than others. I don't think that, using the words in their normal English meaning, anyone would agree that non-third-party sources may be reliable. Accordingly, where a non-third-party source is assessed as reliable, we allow it to support a claim in Wikipeida.

The statement that where no reliable sources exist, we should not have an article on them, is something that follows automatically from the main point that claims should be supported by reliable sources. The statement adds no new requirement: it is not possible to agree with the main point of the policy and disagree with the statement.

Adding a "third-party" to the statement introduces a new requirement. And I think it is an unreasonable requirement. This is because the policy applies to everything in the article namespace - including stubs - including new, not very developed, articles - including fairly unimportant topics. It is reasonable to ask these articles to provide a reliable source. But what is gained by requiring a third-party source? If we have an article on a Simpsons episode that quotes the episode and the Simpsons website, what really is to be gained by adding a third-party source for some of the text (which would probably be in the nature of a brief synopsis in TV Guide or a TV review in a local newspaper)?

I just think we should be realistic here. Long, developed articles without third-party sources are most certainly lacking something in the way of credibility. I agree we should be worried there. But a stub - the beginnings of an article that may develop further? If the content can be referenced to a reliable source, surely we should let that article be, and hope it does develop, rather than argue that we should kill it at birth for want of a third-party source? jguk 18:18, 3 December 2006 (UTC)

That is not what the policy says. An article can be killed only if no reliable third-party sources exist at all for the topic. Under those conditions, we can never have an NPOV article on the topic, so we shouldn't have one at all. If sources can be found, but are not in the article, the article should be improved. That is clearly stated in Wikipedia:Deletion policy. Robert A.West (Talk) 19:48, 3 December 2006 (UTC)