Wikipedia talk:Verifiability/Archive 33

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

This is an archive of past discussions about Wikipedia:Verifiability. 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.

Archive 30

→

Is detailed attribution good or bad practice?

Is giving specific details of a source in text ("According to NAME in PUBLICATION...") good practice? I think it is bad style, leading to weaseling, and inline citations are usually enough, but I cannot find a policy to cite to back this up.--_{Piotr Konieczny aka Prokonsul Piotrus| talk} 20:59, 31 March 2009 (UTC)

It depends on the info. It's absurd to have something like, "According to the New York Times, the Eiffel tower is in Paris." But sometimes it's necessary to attribute an opinion to a specific person, if that opinion is not widespread enough that it can be stated as a general fact. — Carl (CBM · talk) 21:13, 31 March 2009 (UTC)

My opinion, continuing to use Carl's example: Generally, if the information of interest is that the Eiffel Tower is in Paris, then mention of The NY Times belongs in the citation not the text. If the information of interest is what The NY Times says, then it would belong in the text.Ccrrccrr (talk) 21:52, 31 March 2009 (UTC)

What about something like list of topics characterized as pseudoscience? My belief is that since characterization is almost indistinguishable from opinion it would need to have in-text attribution of those characterization. But I am not sure about WP's actual policies regarding this. Thanks Unomi (talk) 06:21, 3 April 2009 (UTC)

The underlying dispute to Piotrus' question is my in-text attribution of sources for population data in Polish areas annexed by Nazi Germany. Since no definite population data exists, all these figures are estimates, and some differ significantly. The population data in this article is a very sensitive issue, dealing with expulsion, deportation, extermination, colonization, "racial evaluation" and similar matters. Thus I am convinced that these numbers need to be attributed to their respective source not only by inline refs, but by in-text attribution. Otherwise, numbers would most certainly be mistaken for scholary consensus or "facts" by readers who do not spend half an hour to check out the reflist. Not for all figures given in the article there is a range, some are singular specific estimates, but that does not make it a fact (and it will look like one if not in-text attributed). For the figures that could be given as a range, there is the problem that the reputation of the authors/sources is not equal, which would be suggested by giving an unattributed range. See Talk:Polish_areas_annexed_by_Nazi_Germany#Sourcing. Skäpperöd (talk) 10:01, 3 April 2009 (UTC)

IMO in general, if the fact is widely accepted by the relevant experts, one footnote reference, not attributed in the text, is enough. Where there are competing views or WP:COI issues it may be better to explicitly attribute refs in the text. It depends on the context though. It does seem like WP:CITE and WP:V both carefully avoid addressing this issue. Rd232 ^talk 11:05, 3 April 2009 (UTC)

I agree with the general consensus here... In cases where there are differences of opinion among scholars (or between scholars and amature authors) it is important to alert the reader the fact that there is disagreement, by using in text attribution. We need to tell the reader who holds which view. Blueboar (talk) 13:17, 3 April 2009 (UTC)

If the information is obvious and no one doubts it, does it need a reference to a published source?

Having a constant argument between editors at: http://en.wikipedia.org/wiki/Talk:Daybreak_(Battlestar_Galactica)#Robot_references right now. So I have some questions.

If you see a robot with the Sony name on it, and you can find pictures of the exact same robot on their site and thus find the official name of it, can you mention what type of robot it is in the article?
If you wrote in an episode summary that a character drove off in a certain type of car, would you need a reference from a published reviewer telling you that was the type of car they drove off in, or could you write down its name?
If there is absolutely no reasonable doubt that something is what it is said to be, no one disputing that fact at all, does it need a reference? How does everyone feel about changing the wording in the Verifiability article to reflect this, to avoid future never ending debates? The article was locked to prevent editing and the talk page is constant going, as some consider the literal interpretation of the rule, over what many of us consider common sense.

Dream Focus 10:17, 4 April 2009 (UTC)

Someone who has somewhat extended sysop rights here once said: "Many pictures do not require a reliable source for the simple reason that they are self-evidently what they purport to be. A zebra is a zebra." Skäpperöd (talk) 14:09, 4 April 2009 (UTC)

I am constantly amazed at what people will edit war over... No, it does not need referencing... but does mentioning the company or model really add anything to the article? Blueboar (talk) 15:18, 4 April 2009 (UTC)

Yes. It shows that it is in the modern time in the real world, and how far along we are to repeating the mistakes of the past, making robots even more humanoid. So, is it alright to edit the article page? How long should I wait for everyone to post their opinions? Or can I edit it now, adding in that sentence? Dream Focus 18:07, 4 April 2009 (UTC)

You misunderstand my comment... I agree that mentioning that a modern day humanoid looking robot is seen in a window adds something to the article... but does mentioning this robot by make and model add anything? To use the car analogy mentioned above... to say "John then got into a car and drove off into the sunset" is fine. To say "John then got into a Nisan Altima and drove off into the sunset" does not really add anything to the article that was not just as well conveyed by the non-product specific word "car" ... and specifying what kind of car boarders on using Wikipedia for product placement. Blueboar (talk) 12:55, 5 April 2009 (UTC)

I agree with Blueboar's point that while this can be determined to be the case but unless it's critical to the understanding of the article, there's no need to mention it. (For example, without even looking at the Battlestar Galactica thing, I know what part of the episode they are talking about, and there's absolutely no need to include specific references to what was shown that cannot be summarized by a link to, say, robotics or the like.

The other aspect is that there's a thin line that's not appropriate to include here between what's obvious to all and what's obvious to some. A scene in a work may be a shot-by-shot homage to another work, and an editor familiar with both may believe it to be obvious. However, this obviousness may not necessarily to the average reader or another editor, and thus without a source to establish that (a third-party RS that says that's the case, or comments from the creator themselves) the connection should not be mentioned; if it's critical to understanding the work, then likely the sourcing issue can be met. --MASEM (t) 13:06, 5 April 2009 (UTC)

Citing one article within another

There was a line in Friends which got reverted with reasoning that a Wikipedia article is not a reliable source. After a subsequent discussion (User talk:Cornucopia#Friends reverts, User talk:Jay#Re: Friends reverts), I added a summary at Wikipedia:Reliable sources#Primary.2C secondary.2C and tertiary sources which I based on WP:PSTS, where I added a clause for Wikipedia based on Wikipedia:Verifiability#Wikipedia and sources that mirror or source information from Wikipedia, which I modified to refer WP:SPS.

Now what I'm not convinced about is if article A makes a statement by backing it up with a reliable source, why can't article B cite article A for that statement? While Wikipedia as a whole is not a reliable source, all articles within Wikipedia follow the same rules, so why can't one article trust another? There is so much of reusability within Wikipedia, this can be extended to reliable source citations as well. Jay (talk) 10:10, 6 April 2009 (UTC)

Citing another Wikipedia article implicitly asserts that we are depending on the Wikipedia article itself as a source of information, rather than just the reliable sources cited in it (see Wikipedia:CITE#SAYWHEREYOUGOTIT). This gives a false impression that we deem our own articles as reliable sources in and of themselves, which they aren't. If a point needs to be repeated, the relevant citation(s) can be copied. Allowing direct citation to Wikipedia creates a maintenance nightmare. The article could be rewritten, split, merged, or changed in any number of ways that remove or relocate the cited information, and no one is going to want to manually check backlinks for dependencies, especially for frequently linked articles. Even if oldid permalinks were used, those become broken if an article is deleted. Moreover, oldid links are difficult to track because they don't show up in Special:WhatLinksHere. I'd say that it's far less work in the long run, and far better practice, to copy the references as needed. — TKD::{talk} 10:50, 6 April 2009 (UTC)

But only if you actually look up the original reference. If all you've looked at is the WP article, that's what you should cite, per Wikipedia:CITE#Cite_the_place_where_you_found_the_material. But to have an hoset, poor quality reference than a dishonest representation of the quality of the reference.Ccrrccrr (talk) 11:17, 6 April 2009 (UTC)

Of course; sorry for omitting that. It was my intent that the original source should be verified first. If it were inaccessible, I would avoid adding the information elsewhere on the basis of the Wikipedia article alone. I've seen enough cases where information in an article has been changed or updated without the corresponding reference replaced. — TKD::{talk} 11:36, 6 April 2009 (UTC)

Self published is confusing

If all the information in the articles comes from the novel, manga, move, or television series itself, and nowhere else is it mentioned, does that meet the requirement of the verifiability policy? Is it alright for List of weapons in Star Trek to reference the episodes they appeared in? What about list of weapons/spells/equipment/whatever in other series? If the only reference is what page and in what issue they were published in, does that count as valid? Is there a set rule for things, or does it depend on how popular the series is? If the series is already notable and has its own article, can it have a list of things in it, or does that violate the verifiability rule? Dream Focus 16:32, 7 April 2009 (UTC)

Are the novels, manga, movies, etc reliable sources for plot information? Yes, of course, so long as no personal interpretation is added. That isn't related to self-published. They are the primary sources, i.e. from the subject itself. Does it, however, count for notability. No. The other part of the question isn't related to verifiability, I'd say, but WP:NOT and other guidelines/policies. -- AnmaFinotera (talk · contribs) 17:10, 7 April 2009 (UTC)

Wait, alright. Self publishing is something you published yourself, not from a publishing company. Doesn't count if the manga is written by the people who own their own publishing company I suppose. I was thinking of something else. The two words didn't combine to one name in my mind for some reason. Self published, not information about one's self, which has been published in self. Alright, never mind that.

New question then. I was told that the verifiability rule means you need third party references, which makes no sense at all. I would think that was covered over in the nobility guideline/suggestion, which of course can be ignored by consensus(determined by whoever is around at the time to give their opinion). Thus the reason why popular series have the exactly same types of things that unpopular/undefended series do not, it a double-standard. We really should have a set standard, made into policy, about EXACTLY what is allowed, and what is not, without exceptions, to make it fair. Dream Focus 17:34, 7 April 2009 (UTC)

Hi Dream Focus, why does requiring independent sources make no sense, in your view? SlimVirgin ^{talk|contribs} 17:41, 7 April 2009 (UTC)

It makes no sense that some articles have to do it, and others do not. A double standard is a problem. Dream Focus 17:47, 7 April 2009 (UTC)

What kind of articles don't need independent sources? SlimVirgin ^{talk|contribs} 17:50, 7 April 2009 (UTC)

Episodes of popular series don't always need any media reference to survive, but only if enough fans are around to defend them. No reason not to include them all, but we need a set standard.
List of characters, villains, henchmen, weapons, equipment, etc. No reference necessary outside the series, even things that only appeared once in one episode, or were just briefly mentioned, if the series is popular enough. Look at the bottom of Doctor_who, and see the links to all the pages it has. A list of ships found in Star Wars or Star Trek will survive, but a list of ships found in a less popular series, shall not. And it doesn't matter if they were mentioned only in one television series or a single movie, or a single book. Some people want to delete every list out there, while others want to preserve them all, and it all comes down to consensus formed by how many people are around to protest one way or the other.

As for your previous question "Hi Dream Focus, why does requiring independent sources make no sense, in your view?" I'd like to point out that some bestselling novels don't get any third party mentions at all, while things that don't sell nearly as well, and have far fewer fans, do get reviewed, and thus get an article. Newspapers and magazines review certain types of things, more often than others. No matter how successful a manga is, you aren't likely to find it reviewed in any newspapers or magazines. On the AFD the past year or so, I've seen many manga series have their articles deleted, simply because of the bias system. Dream Focus 18:07, 7 April 2009 (UTC)

A few of us tried to merge WP:NOR and WP:V into a new policy, WP:ATT, a couple of years ago. One plan there was to introduce an exception for popular culture, where different kinds of sources could be used, for the reasons you outline. The proposal didn't succeed, so we're left with a situation where this policy doesn't describe what goes on, as it ought to. The result is inconsistency. Some editors get away with it in some areas. Others are called to task and have their articles deleted, citing this policy.

But while I agree with your point, I do wonder which best-selling novels you're thinking of that don't get any independent reviews. SlimVirgin ^{talk|contribs} 18:46, 7 April 2009 (UTC)

http://en.wikipedia.org/wiki/Dragons_of_Summer_Flame This one had no sources, but we elected to keep it anyway in consensus. wp:Common sense outweighs the nobility guidelines, as long as there are enough reasonable people around to use it. Dream Focus 19:16, 7 April 2009 (UTC)

Amazon lists a review of it, just as one example. SlimVirgin ^{talk|contribs} 19:19, 7 April 2009 (UTC)

Let's see...has that article been AfDed? No, then you can't argue it was "kept" by consensus or has been claimed as notable despite having no references. -- AnmaFinotera (talk · contribs) 19:22, 7 April 2009 (UTC)

And of course, the delitionist sees a chance to delete something else, so nominates it. It was prodded for deletion before, not sent to the AFD. My mistake. I believe common sense will prevail, and it'll be kept, however it depends on who goes over to form the consensus. The notability guidelines should be used as a suggestion/example of how things can be done, not an excuse to delete anything you can. Dream Focus 19:32, 7 April 2009 (UTC)

"Best selling" novels without significant third-party coverage get deleted the same as anything else. Notable manga does have plenty of third-party coverage, including reviews. You've already been told, repeatedly, that that argument that it is not is blatantly false as is shown BY the articles on those series. That is why manga is covered under WP:BK. Alas, I thought this was a legitimate question, but it now appears to be yet another forum shopping attempt to get around established consensus and try to enforce ILIKEIT over notability standards. -- AnmaFinotera (talk · contribs) 18:12, 7 April 2009 (UTC)

No they don't. We saved a recent Dragonlance novel, and I've seen Goosebumps articles saved, that had no reviews anywhere at all. And there was no established consensus for any of the policies, since only a very small percentage of people who use the wikipedia ever had any say in it at all. It comes down to people enforcing what they want, by camping out at various policy and guideline policies, and arguing nonstop the same illogical nonsense, because you don't like something, and want an excuse to get rid of it. And I am not forum shopping, I am answering a legitimate question. I came here to find out if you absolutely need a third party reference outside the series to pass the verifiability policy, since there was a debate over that in an current AFD. Dream Focus 18:17, 7 April 2009 (UTC)

Short answer, for what you seem to be wanting, is yes, a third party reference outside of the series is necessary to verify anything beyond pure plot summary (including the existence/release/publication of the work). -- AnmaFinotera (talk · contribs) 18:26, 7 April 2009 (UTC)

Howdy Dream Focus, sorry to butt in everybody. DF, I fully understand your frustration. I would posit, however, that this not a problem of different sourcing standards but, rather, that some fictional works have a more vocal fan-base involved in editing Wikipedia. Try deleting unreferenced cultural references from a The Simpsons or Family Guy article and see how long it stays excised. Our coverage of fictional topics also suffers from systemic bias through a sort of prism of obscurantist modernity. That is, while we cover the most inane modern media in excruciating detail, we devote very little space to earlier works. We have articles for every single The Simpsons episode, the entire 20 year run while Amos 'n' Andy or Captain Video and His Video Rangers get short shrift. Our coverage of topics is directly related to the number of people willing to work on that topic and with a collaborative process that is unlikely to change. Cheers. L0b0t (talk) 18:37, 7 April 2009 (UTC)

Refusal for inline citations.

What is the appropriate Wikipedia page to discuss verifiability and citation issues regarding inline citations? I have made a request for specific inline citations to Classification of Japanese, which contains a general list of references with no indication which parts of the article where generated from which references or which pages. A editor reverted the request with reasons given at Talk:Classification of Japanese#Inline citations. I would appreciate suggestions or a review from others. If there is a more appropriate forum, please let me know. Regards, 124.214.131.55 (talk) 02:27, 9 April 2009 (UTC)

Foreign language sources

I recently quick failed a GA nomination for Follo Line for two reasons; WP:CRYSTAL and the fact that 10 out of eleven cited references were in Norwegian - with no English translation provided. My reasoning was that I could not verify the references. To me this would appear to be valid. The nominator disputes the neccessity for providing such translation and has gone to WP:GAR as is perfectly proper.

My question is how can one verify the reliability of such sources and whether or not they support statements in an article. This is especially so in the case of a languge for which no reliable online translation resources are freely available. I would suggest that the current guideline at Wikipedia:Verifiability#Non-English_sources is somewhat unclear in that it omits any mentions of cases where the cited material is not presented in the article as either a footnote or in-line. Jezhotwells (talk) 14:36, 12 April 2009 (UTC)

I don't know if there is precedent for this specific case, but here is how I would approach it.

When all references for an article are in Norwegian, then very likely quite a few Norwegians are interested in it. If there aren't enough Norwegian speakers around at the article, it would make sense to leave a note at WT:WikiProject Norway (and perhaps also at WT:WikiProject Sweden and WT:WikiProject Denmark, since the languages are mutually comprehensible). If this seems to be too much effort, then you should leave handling the nomination to someone else.

I think failing a nomination only because you can't read the language of the sources is normally not acceptable, just like it wouldn't normally be acceptable to fail a nomination because all its references are printed books that you can't easily obtain. Verifiability means verifiability in principle, not instant verifiability for every reader of the article. There are obvious exceptions; such as when an implausible negative claim about Obama is sourced to a book in the Maltese language. But this doesn't seem to come close to any of these red flags. --Hans Adler (talk) 15:03, 12 April 2009 (UTC)

Thanks for your comment. I shall leave the article at WP:GAR as I don't feel I could conduct an honest Good Article review when I can not satisfactorily tick the boxes for verifiability of sources or even establish if they are reliable. The other reason for failure was WP:CRYSTAL as the subject is apparently not slated to begin construction until 2013. Jezhotwells (talk) 18:20, 12 April 2009 (UTC)

Contradictions in reliable sources

An issue brought up by current debate in Talk:Australia (see the RFC there) - what do we do when it becomes clear that there is an error in something that would usually satisfy WP:RS, either because two RSes contradict one another, or (as in this case) because a RS contradicts itself?

IMHO, common sense says that unless there are very clear grounds for ignoring one source (e.g. general consensus of error on one side of things), neither side of the claim should be considered verified, and neither version of events belongs in the article unless the disagreement itself is noteworthy. But WP:V and WP:RS don't seem to be explicit on this point. --GenericBob (talk) 08:24, 14 April 2009 (UTC)

If another source is available, my advice would be to not use the source with the error. If no other source is available, I think you are on the right track with saying you should simply not discuss it at all. 17:09, 14 April 2009 (UTC) —Preceding unsigned comment added by Blueboar (talk • contribs)

Affirmed. JJB 17:24, 14 April 2009 (UTC)

Help amending this policy

I need to include an addendum to the Verifiability policy, but am not sure how to integrate and word it.

Per an administrator-determined consensus at Talk:33 (Battlestar Galactica), "trivial" unverified information falls outside the scope of the burden of evidence section. Specifically, the information having been tagged, challenged, and finally removed many times, it is an administrator's determination that consensus has determined that as it is of a "trivial nature" it doesn't fall under the purview of this policy and may remain in the article untagged and unverified. Since policy is derived from consensus, that discussion leads me to needing to update this one.

I'm unsure whether this addendum should be integrated into the preexisting prose, or if a separate section is warranted. Should the nature of "trivial" be specifically quantified by this policy, or should "triviality" be determined by administrators on a case-by-case basis (as in the exemplar)? Should the nutshell address the unnecessity of verifying trivial information? Should trivial information that doesn't need to be verified be tagged in some fashion so as to show future editors that X information has been determined by an administrator to be "trivial" and that it falls under this special provision? Currently WP:TRIVIAL redirects to Wikipedia:Trivia sections; could it be usurped so that administrators could have a shorthand to point newbies to this new provision?

I've never directly edited a policy page before with such changes/additions to implement and would really appreciate input and assistance in my efforts here so as not to botch the whole thing. If any elabouration or clarification is needed, either refer to the aforementioned talk page for a more in-depth discussion as to this new consensus, or feel free to ask me here! Thank you! — pd_THOR ^{| =/\=} | 21:54, 11 April 2009 (UTC)

Please do not disrupt wikipedia to make a point. If you want, go start an RfC over including '101' as a production code, don't go changing WP:V over a trivial scifi tv series, consider their actions an invocation of WP:IAR and move on. Sincerely, Unomi (talk) 00:26, 12 April 2009 (UTC)

I'm discussing changing this Verifiability policy because apparently it's consensus that this policy need not be applied universally as written. Since an administrator has determined consensus to be in contravention of this policy (which should normally be followed by all editors), this policy needs adjustment to reflect consensus; as the Consensus policy says, "Policies and guidelines document communal consensus rather than creating it." I'm not trying to make a point; something has to apply across the board with regard to verifiability, an aspect one of our five pillars, and if community consensus has been determined to be different that what we're espousing it is, we should change what we're claiming.

For what it's worth, including specifically unverifiable information isn't an improvement to the English Wikipedia, and as such falls outside the purview of the Ignore all rules policy. In fact, it sounds more like vandalism; and since it has to either be one or the other (a change in Consensus as opposed to Vandalism), I assume administrator intervention makes it the former. Please assume good faith here; I just want a wholly factual, verifiable, and accurate article to be the end result, as determined by the policies herein. — pd_THOR ^{| =/\=} | 00:47, 12 April 2009 (UTC)

I appreciate and applaud your desire for clarity and coherence, as you point out yourself it is not an improvement to allow unverifiable information so why change the policies to encourage it? What constitutes trivial? My advice regarding opening an RfC was sincere. I do not have the stomach to get involved with your discussion on '33' but clearly something is amiss if you believe it is not verified and they do. Another question might be to what degree the production code is significant, especially if it is unverifiable. A workaround might be 'according to blah blah' in which case it is at least attributed to the website they mention. An admin is just an editor with some tools, they are not generally considered an oracle or, necessarily, arbiter of consensus outside of situations where they are needed to close discussions, such as AfD etc. Do not count the discussion as over and don't expect the outcome to be an amendment of WP:V. If you feel you are in the right there are many options for dispute resolution open to you. Unomi (talk) 01:07, 12 April 2009 (UTC)

Actually, the discussion has clarified that no source explicitly verifies the information, but it's just too "trivial" to worry about needing a reliable source. I'm unfamiliar with the RfC process, but am wary of doing so because (a) I don't know what input I'll receive vs. what I want to receive, and (b) I don't want to risk the ire of the administrator who's steadfast in the absolute correctness of the article as it currently stands. Thanks for the suggestion though, I may look into it. — pd_THOR ^{| =/\=} | 18:45, 14 April 2009 (UTC)

I understand the desire to stick to policy, but I would think that many people will view this as a "Since I lost the game, I'll change the rules" type of direction. Why not just accept that too much instruction creep is not always a good thing? — Ched : ? 01:14, 12 April 2009 (UTC)

It's not so much that I want the rules changed, but that I've been told the rules are XYZ, just not yet documented. I firmly appreciate and implement the current Verifiability policy to the best of my ability. But the administrator involved actually made plain that Consensus = XYZ; since Consensus = the Rules, the Rules (policies and guidelines) should be changed to reflect such.

My rub is that either (a) I'm way outside the bounds of the rules as they stand, and therefore the rules need to reflect the actual bounds so I and other editors know what they are, or (b) I'm right in my implementation of the rules, and have been edit warring with an administrator in the wrong. Ultimately, the former seems more likely, and here I am. — pd_THOR ^{| =/\=} | 01:29, 12 April 2009 (UTC)

I can understand where you're coming from, and I applaud your efforts to be accurate in what you do. I usually try to go with the flow of the people I'm working with on articles. If I'm editing in a section like movies, and a fair number of folks that are regular editors to those articles prefer XYZ, then I'll try to do XYZ. Then if I see that over in a BLP, that folks don't use XYZ, but rather adhere to ABC, then I'll try to go with ABC on those articles. Wikipedia is a global community, and we can't have a written "rule" for every single contingency. An example would be that in the US, what we consider "football" .. is completely different from what people in the UK consider "football". Sometimes you just need to accept a group consensus to avoid confusion and disruption. It's fine, and even preferred that you ask about conventions you're not sure of; but, once they've been explained - to continue to belabor a point, and say "but bla, bla, bla says this" is only going to have the affect of editors viewing your editing habits as disruptive. Now I'm not telling you what you should do, or how you should do it - but I will say that once an editor has gotten a reputation for being disruptive, it's a difficult task to remove that albatross. Sometimes it's good to just step back from the computer, and ask yourself - how important is this individual item in the whole grand scheme of things? Is it important that I push an issue like episode numbers to the point that other people don't want to work with me? You may have a valid point, and you may even be able to wiki-lawyer your way into forcing changes - but is it worth the trouble in the long run to you? Personally, I'm content to accept that episode 105 means Season 1, Episode 5 when I see it in an article. I don't have a need to have it proved to me. As you may have noticed, I tried to provide a link that showed the correlation between the number and the actual episode. I noticed that you went to RS/N and requested some sort of "can I remove this reference" question. One editor suggested that it wasn't a reliable source; and while I may agree with that if I were using that reference to attempt to establish WP:N for a BLP, I can accept some leeway for minor things. When you removed the reference, I didn't say a word, I just left the article as quietly as I came - I'd rather edit in areas that achieve the most bang for my buck so to speak. Again, i can't tell you what you should do here - that choice is yours. I'm just suggesting that there are some things that can be accepted in order to keep friction to minimal levels. I wish you the best of luck in what ever direction you chose to go ;) — Ched : ? 02:40, 12 April 2009 (UTC)

pd_THOR, I certainly see your point and understand your frustration. I can only presume the other editors are not intending to take the article to GA or FA level, as such unreferenced (or invalidly referenced) production codes would have to be stripped anyway (as has been seen in other episode lists and articles - you can't just proclaim a production code). And while, some may say its "obvious" the truth is, many episodes do NOT use that kind of production code, and without a real source no one can claim it is the real one. Case in point List of The Real Ghostbusters episodes, which has the actual, real production codes as published by the producers. The "guessing" 105 is the code is purely WP:OR, and goes against WP:V, no matter who is saying otherwise. The problem is not that consensus disagrees with the policy, but is simply choosing to ignore it (i.e invoking IAR) because they think it looks better for whatever reason. So I disagree with the need to change this policy, but I do think you are right, though Ched, unfortunately does make a valid point. People tend to rather just attack the editor than accept something that might seem so basic, so to save editorial reputation, it may be better to let it drop for now. -- AnmaFinotera (talk · contribs) 02:58, 12 April 2009 (UTC)

Potentiality for GA/FA is one of my concerns; I raised it at WT:FA? as was told that such contested and unreferenced information is an immediate no-go for such candidacy. At a random sampling, The Simpsons, Arrested Development, Duckman, Little Britain, Gomer Pyle, U.S.M.C., and Prison Break all use differing formats for their episodes' production codes. And for that matter, there are no reliable sources that Battlestar Galactica even uses such codes.

I haven't edited the article for 2+ weeks, but what exactly is the "proper" amount of time I should wait before again trying to apply this policy and remove plainly unverifiable information? Should I mayhap take the article to GAN as an impetus to remove the information? I think it warrants such, or at least the process to see where improvement could be made; I've been holding off because of the bad info, so would I be attacked by the administrator again for being pointy by nominating it? — pd_THOR ^{| =/\=} | 18:45, 14 April 2009 (UTC)

Hasn't this thread drifted from the general to the specific? pd_Thor originally raised the question of amending this policy to allow for a "trivial fact" exception to the Verifiability policy. I think this would be codifying the actual practice in Wikipeida. Most sentences do not have references but any editor, arguing out of ignorance or malice, can place a fact tag on a sentence and then argue strongly that the tag must remain because that is policy. The trivial fact exception needs to be codified and I would support such a change to this policy. Like pd_Thor, I'd appreciate some help in how to go about doing this. Tom94022 (talk) 21:41, 22 April 2009 (UTC)

I very much disagree with the idea that trivial facts should be given a blanket exemption from being referneced. If something is challenged, then it should be cited. And doing so is not that difficult... A lot of trivia is self-referenced (ie to the book, TV show, movie it comes from). Blueboar (talk) 23:45, 22 April 2009 (UTC)

Exceptional claims require exceptional sources

This section in a nutshell:

Is the claim, supported by source A, that a self-proclaimed or known adherent of an exceptional claim has expressed his belief in such a claim in source B, itself an exceptional claim?
Does either source A or source B need to be of exceptional quality to warrant the inclusion of this information in an article?

Note: This section does not discuss the notability or relevance of the information with regard to the article.

Related archived discussions on this page: "Exceptional claims require exceptional sources" - Exceptional claims: Is a change of heart exceptional?

I support the guideline that exceptional claims require exceptional sources. This guideline, however, often seems to be misunderstood. The main reason is that people either differ about what the actual substance of the claim is, or do not think about determining this at all.

An exposition of the logic here might be too philosophical or too long. With the hope of starting a discussion on the issue, I am instead giving some examples that might illustrate the point.

"The earth is flat." -- Here, the claim is that the earth would be flat. There are probably no exceptional sources to support this exceptional claim.

"My neighbor recently said that U.S. President Obama believes the world is flat." -- Here, the claim is that Obama would believe that the world is flat, and the neighbor is given as the source. Given that the claim is exceptional, the source is clearly lacking in reliability.

"A recent study found that some indigenous people in the Amazonas region believe that the world is flat." -- In this case, the claim is that some people believe in something that is false. However, the claim itself might be true, so the source should be assessed with regard to its validity.

When the claim is not correctly identified, this also negatively affects the discussion on the question of notability and relevance. While the claim "indigenous people in the Amazonas region believe that the world is flat" is not relevant with respect to the Wikipedia article on the planet Earth, it would be relevant in the article Flat Earth. (Any claim that "the earth is flat" would not have sufficient support to be included in either article, of course, even if it could be verified that some people believe in it.)

Yet, in a number of cases comparable to the one described above, editors have started to argue that the earth is obviously not flat, that the source would not be exceptional, and that the information should not be included in the article. This clearly misses the point, and in addition, it does not follow the basic policy that the "threshold for inclusion in Wikipedia is verifiability, not truth".

These misunderstandings could, of course, be addressed in on a case-by-case basis, between the editors involved. However, this very often does not seem to produce any results other than disruptions, personal conflict, and additional administrative work. Often, both sides involved in a particular conflict have a tendency to argue about the truth instead of presenting arguments on the basis of established WP policies.

A clarification of these policies is, in my view, urgently needed as a basis for resolving the recurrent disputes around these issues. --Cs32en (talk) 22:55, 12 April 2009 (UTC)

New comments

Please leave comments here, or in the respective subsection.

Hmm, well, I may be spoiling your commenting format by inserting interruptions below then, but when one proposes a different format, one is prepared for others not to go along. JJB 17:29, 13 April 2009 (UTC)

Comment by JJB, 12 April 2009

Thanks for starting the ball rolling. Permit me to play devil's advocate, as one who's been through this section before. I think there are a few improvements to be made.

First, editors should avoid prejudgment during discussion: consider that most such cases (even the flat-earth ones) involve editors who have experience with the minority viewpoint and who require good faith, and so equating the minority view with "something that is false", as you happened to, is just the sort of slip we wish to avoid.
Second, I agree editors should distinguish claims: let me annotate because there are multiple claims involved in such sentences, not just single claims:
- 1: The earth is flat.
- 2: The Flat Earth Society says that 1: the earth is flat.
- 3: My neighbor recently said that 4: Obama believes 1: the world is flat.
- 5: A recent study found that 6: some indigenous people in the Amazonas region believe that 1: the world is flat.
Your point, then, is that 1 and 4 are obviously exceptional and 6 is not; and that people mistakenly argue that the exceptionality of 1 forbids the inclusion of 6 (or 5) from WP. Well, I agree that such an argument is invalid, but do not confuse 5 with 6 in the process of delineating 1; 6 is just as includible as 5, or 2.
Third, I agree editors should distinguish relevance: relevance means that a statement is somewhere in the top 500 or 1000 different statements that an article should contain about its topic, and editors often miscategorize data into the highest-level article without realizing that more specialized articles exist to handle data not in the top tiers of relevance. As an inclusionist, I have for 2 years emphasized that such insertions should be moved to the specialized topics, and not deleted, in most cases.
Fourth, exceptional claims can still confer notability: most widespread exceptional claims, such as the flat earth, are notable due to coverage, not due to who agrees with the claim. Notability is relatively irrelevant.
Fifth, we should not blur in-line versus footnoted attribution. In-line attribution softens one claim by extending it into another claim, that is, WP does not include 1: "the earth is flat" (exceptional), but does include 2, 6: "people say it is" (wholly unexceptional). Claims 1 and 4, being exceptional, are not includible even with a footnote. Claims 2 and 6 are includible with footnotes (e.g., sourced to "Flat Earth Society" and a published indigenous Amazon), and without further attribution. If the Flat Earth Society left it ambiguous whether they believe the earth is flat (as would be the case if they objected to terminology foisted on them by a majority outgroup), then claim 2 becomes disputable (but not exceptional) and needs further attribution, such as 7: "Sociologist X says that the Flat Earth Society's claims amount to the earth being flat." Claims 3 and 5 would be footnoted to "my neighbor" and "recent study", but claim 3 would then be dismissed as an unreliable (as well as unexceptional) source. Claim 5 is technically not necessary (redundant with 6), but typical, because the more direct source (the indigenous Amazon) is usually less available.
Sixth, many people who show up here (like me, originally) have been involved in content disputes. I haven't checked, but my general view on a content dispute over something like this is that editors should be able to "get it" and build consensus without a policy rewrite, and if they don't the forum is dispute resolution not policy. But what change were you going for, anyway? Perhaps: A statement to the effect that an adherent believes some exceptional claim is not itself an exceptional claim. JJB 23:59, 12 April 2009 (UTC)

First, editors should avoid prejudgment during discussion: [...] equating the minority view with "something that is false", as you happened to, is just the sort of slip we wish to avoid.

The correct characterization in this context would probably be to say that it is an "exceptional claim". I specifically said that it is "false", stating my personal opinion instead of staying within the logic of the presuppositions, in order to avoid giving certain people an opportunity to start a distractive discussion on whether I would contemplate the possibility that the earth was flat or not.

With regard to multiple claims, you are correct. I just focused on one of the claims in each statement, because you can treat each claim separately. I will continue to do so, in order to keep the conversation as simple as possible.

Well, I agree that such an argument is invalid, but do not confuse 5 with 6 in the process of delineating 1; 6 is just as includible as 5, or 2.

For 2 and 5, you would need appropriate sources. Therefore, I only referred to 6.

Extended content

Well, WP:V is about the idea that every WP claim eventually needs sources, but that we don't source everything immediately. For 6 you need a source also, which can either be self-published (an Amazon) or third-party (a study). JJB 17:29, 13 April 2009 (UTC)

In the article in question, not a single claim would escape a revert if it would not be instantly supported by a source, probably. Anyway, for claim 6, there is source 5, which could be used, if found to be reliable. I did not give a fake source as a footnote, but I assumed that source 5 would actually be a fully valid source, if found to be WP:RS. --Cs32en (talk) 13:48, 21 April 2009 (UTC)

Fourth, exceptional claims can still confer notability

This is correct. However, in this case, the discussion should not be about whether the claim is exceptional or not, but whether the claim that the exceptional claim would be notable is exceptional or not. The question of notability and relevance should be addressed after consensus has been reached on the issue of verifiability. Strictly speaking, a claim cannot be notable, but only the expression of such a claim.

Extended content

Getting closer to the issue. But claims that "X is WP-notable" are not settled by exceptional sources but by ordinary notability criteria. And you're talking about whether X (claim or expression) is notable for inclusion in an article, not whether X is notable enough to be the topic of an article. Generally that is decided by whether the source is reliable, and not by whether we can verify whether the source said it, which is usually academic. JJB 17:29, 13 April 2009 (UTC)

I generally agree with you, although the issue of notability would, in the end, very like boil down to the question of whether WP:RS sources have attributed such notability to the issue. In the context, it has been argued that something about which four major newspapers in Denmark, and a government-owned TV-station has reported, with two interviews with the person who makes the "exceptional claim", would not be notable. A further line of defence seems to be "we should really avoid foreign sources". I'll set these issues aside for the moment, as they are not relevant to WP:Verifiability. --Cs32en (talk) 13:48, 21 April 2009 (UTC)

we should not blur in-line versus footnoted attribution

Inline vs. footnote is, of course, a technical difference. The issue here is whether you need a source or not. Otherwise, your reasoning in this section seems to be correct.

Extended content

I'd hold that it's more than a technical difference. Inline in these cases usually indicates that the source is not quite highest-quality enough that its claim stands on its own. Also inline changes the claim WP is making. If we say "indigenous people believe[1]", we say WP agrees wholly with study X; if we say "study X says[1]", we say only that WP agrees that the study says it. JJB 17:29, 13 April 2009 (UTC)

Sorry, I misunderstood that. I have been using the terms "supported by the source" (WP makes the claim), and "attributed to the source" (WP makes the claim that the source says so) to distinguish the "footnote" vs. the "inline" citation. (Both versions, of course, use footnotes.) I have also seen the term "inline citation" used to distinguish articles with "inline" footnotes vs. articles in which just some sources were given at the end of the text; that usage is probably wrong. --Cs32en (talk) 13:48, 21 April 2009 (UTC)

But what change were you going for, anyway?

Most important would be to have a policy on how to determine the claim that is actually relevant to WP:Verifiability, and to make people aware of the fact that it is really important to identify this claim. Everyone who uses "exceptional claims require exceptional sources" as an argument should, as a policy guideline, also clearly define the claim he is talking about. There are a number of editors who refer to the "wrong" claim, i.e. a claim that is not relevant with regard to WP:Verifiability, use the sentence "exceptional claims require exceptional sources" because it sounds convincing to the average WP user, and directly proceed to initiate administrative measures when challenged with regard to their logic. Because their reasoning sounds convincing at first glance, and they can refer to the section on WP:Verifiability, they can easily convince other editors and administrators that their position is correct, and that other people would be soapboxers, POV-pushers, distractors etc. Therefore, it's really important to amend WP:Verifiability, after a consensus on the policy is reached. --Cs32en (talk) 02:24, 13 April 2009 (UTC)

Pretty simple, and not likely to need new policy. Having taken only a peek now at the debate, but continuing the flat-earth illustration: it is perfectly appropriate to say, in the flat-earth article, that "indigenous people believe[1]" if no one disputes the reliability of the study, or "study X says[1]" if they do. But if someone disputes whether the statement is worthy of inclusion at all, that person is IMHO playing deletionist tactics with includible data, because a proper argument against inclusion would only be along the lines that too many such claims might constitute undue weight. It appears that you are dealing with someone who thinks "it's exceptional to say that an indigenous people [or else an otherwise reputable scientist] believes a minority theory", which I think is a total non-starter. (And if this is just one dispute with another editor or three, then go to WP:DR, not here.) Now you could allow a little WP:BLP and say that, if the adherent of the exceptional claim (e.g. the indigenous people) might be hurt having the claim attributed to them, then you should stick with the way that party self-identifies, or (in the case of unlikely publishers like the indigenous people) the most reliable descriptors of that party's beliefs, or (in the case of disputes) all views about what that party's views are. Particularly, if it were exceptional for the party to believe what it is said they believe, that fact should be verifiable, such as to balance claim 5 above by saying, "Sociologist X is a member of Society Y,[2] which has taken the official position that no indigenous tribes believe in a flat earth[3]"; anything short of that is not fairly arguing the position your fellow editors appear to be taking. But, note, the original view under consideration is still wholly includible. In this case I will repeat my clarification as an appropriate addition to policy: A statement to the effect that an adherent believes some exceptional claim is not itself an exceptional claim, perhaps amending an adherent to a self-identified adherent. (P.S. I must respectfully decline your offer to hammer this out on user talk, and preferably this talk can also be kept to a minimum if it is determined the correct forum is elsewhere.) JJB 17:29, 13 April 2009 (UTC)

I agree with your argument, and I share your assumptions that I am probably dealing with deletionist tactics here. I have described a similar issue below. The problem is that there are some established editors who are supporting each other, or at least, are ignorant to the flawed reasoning. Without making the policy more explicit, there is just not enough pressure on everyone to justify the arguments put forward, and the dispute resolution would very quickly not revolve around the actual policy issues, but along all kinds of diversionary tactics and ad hominem insinuations. (Some editors advocating the inclusion of certain pieces of information are not always using the best ways to promote their arguments as well.) --Cs32en (talk) 13:48, 21 April 2009 (UTC)

Comment by Jehochman, 13 April 2009

Note that Cs32en has been attempting to insert unverified info into 9/11 articles. They appear to be trying an end run around consensus by asking for changes in policy. Jehochman ^Talk 05:14, 13 April 2009 (UTC)

Asking for changes is the appropriate way to contribute to the development of an existing consensus, wouldn't you agree, Jehochman? I should add that this is not about changing policy, but about amending the wording of a guideline to make it more useful to actually support an established consensus on policy. --Cs32en (talk) 05:28, 13 April 2009 (UTC)

Your contributions here have been lawyerish and disruptive, in my opinion. You are starting threads on multiple pages to discuss the same issue: you want to insert poorly verified or unverified info into 9/11 articles, in violation of WP:ARB9/11. You're very new to English Wikipedia. It is probably unwise to dive headlong into changing policies and customs before you even understand how things work here. Jehochman ^Talk 06:11, 13 April 2009 (UTC)

I would like to point out that Jehochman is not new to threatening users and systematically using loaded language. He always find "disruptive" everybody who does not share his opinions and always think they are wikilawyering and deserve some heavy sanction. I could cite plenty of diffs showing this pattern in his behaviour which is btw very uncommon between wikipedia editors.--pokipsy76 (talk) 11:01, 13 April 2009 (UTC)

Wholly agree, as Jehochman did exactly the same to me my first time on this page. JJB 17:29, 13 April 2009 (UTC)

Actually, Cs32en, if you're not proposing policy changes, this branch of the conversation is technically over, and we can continue somewhere like WT:RS. JJB 17:29, 13 April 2009 (UTC)

In my view, the wording of the respective section of the project page associated with this talk page reflects the policy in a way that leads to misunderstandings. This is why I propose to change this wording, and therefore I think this talk page is the best place for this discussion. --Cs32en (talk) 20:17, 13 April 2009 (UTC)

To Jehochman: I knew it would be something like that. a: Let me go on the record as saying that b: I understand that c: a number of namable notable engineers believe d: it would scientifically be an exceptional claim to say that e: certain pancake collapses are not due to controlled demolition. Boy am I in trouble now, no matter how far I remove myself from the claim! JJB 17:29, 13 April 2009 (UTC)

Comment by Blueboar, 13 April 2009

Let's address the issue and not the editors please... regardless of what Cs32en's reasons for posting here may or may not be, we should still address his concerns and questions... He presents three examples of "extraordinary claims":

"The earth is flat." -- I think we would all agree that this is an extraordinary claim. It would certainly need a very strong source (and that is unlikely).
"My neighbor recently said that U.S. President Obama believes the world is flat." -- Cs32en states that "the claim is that Obama would believe that the world is flat"... but that is not quite right. What he is talking about would be the simple statement: "Obama believes the world is flat"... with that we clearly have an extraordinary claim that would need an extraordinary source... but as written things are not quite that simple. There are actually two claims being made... a) that his neighbor said something about Obama, and b) that Obama believes something. A lot depends on what the subject of the article is. In an article about the neighbor (assuming the neighbor is notable enough for one) then the claim that she said something like this may not be all that extraordinary, and might be worth discussing. In an article about the neighbor, the focus is more on the fact that the neighbor said something, and less on whether what she said was true or not. However, in some other article, expecially one about Obama or about the Earth, then the focus shifts to the second claim. In this case, the statement is extraordinary and we need a better source.
"A recent study found that some indigenous people in the Amazonas region believe that the world is flat." -- Here again we have a dual claim... a) that a study exists and that it found something to be true, and b) that people in the Amazon believe something about the earth. In this case neither claim is all that extraordinary. The only question is the reliability of the study.

In other words... sometimes it is difficult to determine whether a statement is extraordinary or not. It often depends on the exact subject of the article and the exact wording of the statement. All we can do is issue a general statement warning people to be alert for extraordinary claims, and to demand high quality sources when we find them. Blueboar (talk) 17:51, 13 April 2009 (UTC)

I agree with Blueboar that the statements that I provided in the example contain more than one claim, except for the first one. I did not think about the possibility of a Wiki article about the neighbor with regard to the second statement of the example, but Blueboar is correct to point out that the relevant claim in this case would be that the neighbor believes something about Obama. Much of the confusion about which of the assertions that are contained in a given statement would be the relevant claim with regard to WP:Verifiability is due to misunderstandings, but there also seem to be instances where editors, inluding established editors who are aware of WP policy, deliberately exploit this confusion in order to support their arguments. We need specific advice in the guideline that it is necessary to determine which is the relevant claim with regard to the subject of the article. --Cs32en (talk) 18:37, 13 April 2009 (UTC)

I disagree with the idea that you need "specific advice" in the policy. The more specific you get, the more wikilawyers will find loopholes and ways around it. Our polices should not be viewed as law codes that try to cover every consevable specific circumstance. Instead they should be viewed as general statements of princilple that can be applied with some flexibility in any circumstance (For a very good essay on this, read: Wikipedia:The rules are principles). editors need to stop trying to parse each policy statement for subtle shades of meaning. Instead, they need to think about the intent of the policy. I think the intent is clear in this case ... The more extraordinary a claim seems to be, the more we should demand very high quality sources to support it's inclusion. Whether a given claim is or is not extraordinary can (and probably will) be argued at the article level. Blueboar (talk) 19:51, 13 April 2009 (UTC)

Yeah, I fully agree that we need a principles-based, not a rules-based policy. Now, the problem here seems to be the other way round: editors who take the wording "exceptional claims require exceptional circumstances" by the letter, and then just finding any claim in a given statement that they consider to be false. Specific advice (my words) is maybe not the correct term, rather it's a clarification of the intent of the policy, which - in an interpretation that does not reflect the underlying reasoning of the policy - is currently used as if it was a rule. --Cs32en (talk) 20:17, 13 April 2009 (UTC)

It seems fairly clear to me. If someone finds a claim to be extraordinary, and the source for it not of sufficent quality and reliability, then they should question it. What you seem to be discussing are the situations where editors disagree as to whether the claim is extraordinary or not... resolving such disputes is not what a Policy is for. Resolving such disputes is a judgement call that needs to be made by those who know the subject, and should be resolved by attempting to reach a consensus among the editors working on the article. If a consensus cann't be achieved, and no compromise can be made, then you just have go through the normal dispute resolution process... filing an RfC, asking for third party opinions, requesting mediation, etc. Blueboar (talk) 00:30, 14 April 2009 (UTC)

No, it's really not about people disagreeing on the nature of a particular single claim. It's about cases where there are multiple claims in a given statement, and instead of identifying the claim that is actually relevant with regard to WP:Verifiability, people are obscuring, whether purposefully or not, the fact that there are multiple claims. This would be not really serious if it happened in an overall friendly atmosphere. But these arguments, based on misconceptions of policy, are also used in circumstances in which words (and arguments) should be chosen more carefully, e.g. in the context of administrative measures. --Cs32en (talk) 00:56, 14 April 2009 (UTC)

I think a lot of the intricate discussion on this point can be short-circuited by recongnizing that ordinarily, Wikipedia articles should not contain statements that are likely to be false. There can be a few exceptions, such as articles that address false or unlikely beliefs held by a significant minority ("the Earth is closer to the Sun during northern hemisphere summer"), but most articles should not contain such claims at all. This concept goes beyond the narrow statement "exceptional claims require exceptional proof". --Jc3s5h (talk) 13:49, 21 April 2009 (UTC)

Discussion on specific proposals

"A report to the effect that a self-identified adherent believes some exceptional claim is not itself an exceptional claim."

I reverted the recent addition about this because it wasn't clear what it meant. [1] SlimVirgin ^{talk|contribs} 01:25, 14 April 2009 (UTC)

Could someone explain in a nutshell what the issue is here? There's quite a lot to read, and a quick glance hasn't helped. SlimVirgin ^{talk|contribs} 01:44, 14 April 2009 (UTC)

I am not exactly sure what the issue here is either. Blueboar (talk) 02:31, 14 April 2009 (UTC)

Hi Slim and Blue! At risk of horrendous misstatement: This appears to be a question from Cs32 (de.wikipedia) who would like certain 9/11 alternate theory info included but does not know what specific policy proposal to make. I gather that the argument is (Cs32) "this is a reliably sourced peer-reviewed foreign journal" vs. (other(s)) "it can't be so or English sources would notice". The dispute is thus whether the claim "peer-reviewed science journal favors controlled demolition" is itself exceptional and requires another source. Seems to me that, if (1) it passes WP:RSN and (2) editors still invoke WP:V, then it should be includible as stating its own POV significant to that article, and such a footnote, which means exactly what it says, is advisable as it relates to a recurring policy question (what is an exceptional claim). Your thoughts under WP:BRD? JJB 15:57, 14 April 2009 (UTC) Need I add that naysayers essentially create an infinite loop that can exclude any "nonmainstream" source: because if peer-review 2 affirms peer-review 1's conclusion, that too would be "exceptional" for us to believe and requires peer-review 3's affirmation (until we reach "mainstream", whatever that means)? This is an article about minority POV, after all. JJB 16:02, 14 April 2009 (UTC)

Thanks John... this helps. A peer reviewed journal is usually considered reliable... but it may or may not qualify as an "exceptional source"... it depends on the reputation of the journal. The fact that a journal is foreign language does not really impact that reputation significantly. However, given the topic, I do think we need to be cautious... first what specific article are we talking about: the 9/11 conspiracy theories article, the main September 11 attacks article, the World Trade Center or something else? This particular claim (the "controled demolition theory") has been so completely debunked by so many other reliable sources, that the claim that a reputable scientific journal is supporting it does strike me as being "extraordinary". It certainly would be a WP:FRINGE viewpoint. Whether that viewpoint should be discussed or ignored, however, depends on which article we are talking about. Blueboar (talk) 19:39, 14 April 2009 (UTC)

And I don't think any of this is enough to require a change in the policy. Blueboar (talk) 19:40, 14 April 2009 (UTC)

"Known adherent" instead of "self-identified" adherent

Maybe the sentence should read "known adherent" instead of "self-identified" adherent. "Known" would refer to knowledge established by WP:RS sources. That the person is an adherent may itself be an exceptional claim that would need an exceptionally reliable source (see the "my neighbor says Obama believes" example above). --Cs32en (talk) 18:39, 15 April 2009 (UTC)

I don't think "known" vs "self-identified" affects anything. Essentially, this could probably be discussed in the 9/11 conspiracy theories article (which already discusses the "contrloled demolition theory")... I don't think it belongs in the other articles (too much of a Fringe Theory, even if it isn't an "extraordinary claim"). In any case, I see no reason to edit the policy to address this issue... take it back to the talk page and argue about it there. Blueboar (talk) 19:06, 15 April 2009 (UTC)

Hi Blueboar, I've been discussing this on the talk page of the article World Trade Center controlled demolition conspiracy theories. My comments there have been deleted or collapsed (some are now restored), and people there (both sides) are citing excerpts from various guidelines more or less at random. It's just not possible to have a discussion focused on policy (instead of on content, or POV) there. (I didn't intend to include any info about people making extraordinary claims in any main article on 9/11, because of the notability issue you mentioned.) --Cs32en (talk) 19:33, 15 April 2009 (UTC)

Many self-published sources are press releases

(I renamed this section from its old, misleading title "Press releases are typically self-published"; see #Section title was misleading below.) Eubulides (talk) 01:02, 16 April 2009 (UTC)

WP:SPS lists several types of self-published sources, but it doesn't mention press releases. Let's add "press releases" to the list in that section, as press releases are a common and typical form of self-published sources. Eubulides (talk) 06:29, 15 April 2009 (UTC)

The Policy lists the most common types of self-published sources - as examples, not as an exhaustive and complete list. I would resist the temptation to add to this list... the larger the list, the more we support the mistaken idea that if something is not on the list it must not be a SPS. While I agree that press releases are self-published, I don't see a need to expand the list to include them. People need to think of our policies as principles, not as laws. We need to focus more on the intent of the Policy, and less on the exact language. Blueboar (talk) 13:43, 15 April 2009 (UTC)

In that case I propose to remove "personal websites" and add "press releases". This will make the examples more useful without increasing their length. It's unfortunately quite common for Wikipedia articles to cite press releases; it's less common for them to cite personal websites. Eubulides (talk) 19:47, 15 April 2009 (UTC)

I very much disagree with that... Perhaps this is a function of what topic areas we each edit, but I have rarely seen a citation to a press release, while I constantly come across attempts to cite personal websites. Blueboar (talk) 20:40, 15 April 2009 (UTC)

I disagree that press releases are self-published (as a rule). They are primary sources, but not self-published, because they're published (again, as a rule) by a company or institution, and written by an employee. Self-published means published by one person, or perhaps by a very small, unprofessional group, with no editorial oversight. Or it can refer to vanity publishing, where someone pays a company to publish their material. If you extend the definition to press releases and any material that a company has published themselves, then the New York Times becomes self-published by the people who own it. SlimVirgin ^{talk|contribs} 20:54, 15 April 2009 (UTC)

The problem with press releases is that they end up almost verbatim in newspapers. And I've certainly seen citations to newspaper articles that are clearly press releases. I can't convince myself that they aren't, at lest usually, self-published. Dougweller (talk) 21:17, 15 April 2009 (UTC)

OK, then let's replace "tweets" instead. Surely you are not constantly coming across attempts to cite tweets? There's a {{cite press release}}, but there's no {{cite tweet}} (or perhaps we should create one? :-). Eubulides (talk) 01:02, 16 April 2009 (UTC)

Most press releases are self-authored, probably, but most are published by secondary publishers such as PR Newswire or the American Academy of Sciences' EurekAlert. The degree of fact-checking that these publisheres do is probably minimal, but it is nevertheless untrue to say that such releases are self-published. Scan EurekAlert's breaking news and you will see that the vast majority of the authors of such releases scientific journals and universities. Companies and institutional press releases are often more reliable than newspapers because they are actually closer to the source. Many journalists will only read such sources. Companies are limited from making false statements in their press releases by financial regulations; institutions by their reputation (the same thing which maintains their editorial reliability). II | (t - c) 21:51, 15 April 2009 (UTC)
- I love it when angels dance on the head of pins here. Press releases, though technically not self-published, have most of the shortcomings of self-published sources. That is because there is a COI situation with respect to the claims made (i.e. they are always self-serving and frequently less than evenhanded), the goal of the press release is self-interest rather than truth (which goes to WP:V) or edification (which goes to WP:N), and there is little or any editorial oversight, fact checking, or accountability with respect to accuracy. As Imperfectly I. notes, there are things that make press releases (and, I would add, self published material generally) reliable in some contexts, so some common sense is in order. I would think it's uncontroversial to use press releases to verify the names of a company's officers, the location of its facilities, a M&A transaction, etc. It would not be correct to use a press release to source the claim that company X is the most popular or effective version of product Y, or even as a primary source to note that the company made that claim (because there is no hint from the press release why the claim is noteworthy). I'll add too that if you look closely at what it means to publish something, and what a self is, the distinctions will all break down just like the angels and the pins. There are do-it-yourself PR outlets that basically serve as web hosts. And there are publishers that function as legal entities and have a staff and more than one writer but basically serve as the mouthpiece of a single individual. Wikidemon (talk) 22:06, 15 April 2009 (UTC)

I'm not saying that press releases should be used for the claim that "company X is the most popular or effective version of product Y". :) I wouldn't even support that from NYTimes if they didn't have underlying data. The goal of the journalist is self-interest as well, and that may be why journalists (even at dedicated publishers like the New Scientist or Scientific American) are notorious for exaggerating scientific stories in order to make them more provocative. Compared to a journalist the person at the scientific institution is probably more likely to be close to the researchers and have a scientific background. The person writing the press release is unlikely to be on the scientific team, but there will generally be comments from the researchers on interpreting the results. Press releases are sometimes the only lay interpretations available for scientific stories. As far as finance, press releases from publicly-traded companies, like similarly self-authored SEC filings, should not make material misstatements. They're not great, but they're not always terrible either. The Form 8-Ks which often accompany them are a bit more cryptic. As far as PR outlets, most are equally unreliable. PR Newswire does press releases for even the miniscule microcap OTCBB stocks, as do its competitors Business Wire, Marketwire, and GlobeNewswire. Anyway, this a bit of WP:CREEP that's best handled on an a case-by-case basis. II | (t - c) 22:36, 15 April 2009 (UTC)

Section title was misleading

"I disagree that press releases are self-published" Sorry, I seem to have confused this discussion with an inaccurate title. I just now changed it from "Press releases are typically self-published" to "Many self-published sources are press releases", which is what my text and proposed wording say. The proposed wording does not say or imply that press releases are all self-published, any more than it says that books are all self-published (it mentions books in the same list as press releases).
"I have rarely seen a citation to a press release, while I constantly come across attempts to cite personal websites." No doubt it depends on the area. For what it's worth, {{cite press release}} is used by over 6,000 articles, and this is obviously much lower than the true number of Wikipedia citations to press releases. In contrast, I've never run across a Wikipedia citation to a tweet. (Maybe medical editors don't tweet? :-) Eubulides (talk) 01:02, 16 April 2009 (UTC)

I would say most of the press releases we use as sources aren't self-published, Eubulides. Can you give an example of the kind of press release you have in mind? SlimVirgin ^{talk|contribs} 01:13, 16 April 2009 (UTC)

Sure. Academy Awards, the first article listed in Special:WhatLinksHere/Template:Cite press release, contains the following press release, which was clearly self-published:

"ABC and Academy Extend Oscar Telecast Agreement" (Press release). Academy of Motion Picture Arts and Sciences. February 7, 2005. Retrieved 2007-04-13.

This citation supports the claim that ABC has the Oscar telecast rights through 2014. But the article could have cited the New York Times (a more-independent source, surely) to support that claim; see:

Cieply M (2009-02-01). "Oscars suspense: will people watch?". N Y Times. Retrieved 2009-04-16.

Another advantage of the New York Times URL is that it still works, whereas the Oscars press release's doesn't. Eubulides (talk) 01:46, 16 April 2009 (UTC)

That's a primary source, meaning a source very close to, or directly involved in, the events you're writing about. It's not a self-published source. You're right that, in this case, a secondary source (one not involved e.g. New York Times) might have been more appropriate, but sometimes primary sources are better. It depends on the context, and on which source is regarded as better informed. SlimVirgin ^{talk|contribs} 01:49, 16 April 2009 (UTC)

It's a primary source, yes, but it's also a self-published source, and its citation should therefore play by the rules for self-published sources. It could well be appropriate to make an exception for citing a self-published source to support a claim about the source's publisher, but that exception should apply to all self-published sources, not just to press releases. Eubulides (talk) 02:02, 16 April 2009 (UTC)

It's not what we call a self-published source. Are you saying that, if the New York Times issues a statement about some issue they were involved in, they become a self-published source? That wouldn't be right. They would be a primary source. There's a significant difference in terms of how they may be used. Press releases are fine to use btw. All you need is in-text attribution if you want to make clear that you're relying on the organization for information about itself. SlimVirgin ^{talk|contribs} 02:14, 16 April 2009 (UTC)

"Press releases are fine to use btw. All you need is in-text attribution" The topic here is self-published press releases, which are not at all fine. It would not be fine to add to Autism the text "Homeopathy has good results in treating autism, according to Dr. Harshad Raval" and cite Raval's self-published press release.
"if the New York Times issues a statement about some issue they were involved in, they become a self-published source" The definition of self-published source in WP:V would exclude the NYT, as the NYT has editors and a well-established reputation for accuracy and is not just a situation of "anyone can create a website". However, to respond to your example, WP:V #Reliable sources requires "third-party published sources", and so a Wikipedia article should in general not rely on NYT articles about the NYT. For example, the citations in Vicki Iseman to two NYT articles describing her lawsuit against the NYT should be replaced by citations to more-neutral third parties.
"It's not what we call a self-published source." Which "we" is that? Certainly WP:V makes no exceptions for press releases. A self-published source is a self-published source.

Eubulides (talk) 02:45, 16 April 2009 (UTC)

Certainly WP:V says nothing about self-published sources from government agencies, large institutions, and the like, because they aren't typically thought of under that umbrella. There's nothing reasonable in thinking that an institution can publish a journal and that information becomes "published" and therefore more reliable, but when it issues a statement, it is not. II | (t - c) 03:06, 16 April 2009 (UTC)

The Academy Awards press release you gave as an example above is in the same boat as the NYT, in that it has editors, publicists, lawyers etc checking its facts. It's definitely not what this policy means by a self-published source. In terms of whether it should be used as a primary source, who would know better than the Academy what deals the Academy has entered into, assuming there's nothing controversial about it? SlimVirgin ^{talk|contribs} 01:56, 17 April 2009 (UTC)

I think they are in fact self-published sources when used for information about themselves, just like any other organization. They can be used for factual information only. Otherwise, say, a University's statement "We are among the very best schools for this subject" unsupported by facts, could be used to document a statement that "They are among.... ". DGG (talk) 17:39, 19 April 2009 (UTC)

"It would not be fine to add to Autism the text "Homeopathy has good results in treating autism, according to Dr. Harshad Raval" and cite Raval's self-published press release." No, but it would be appropriate to use that source on an article about Raval. If Raval is well-respected in the homeopathy community, it may be appropriate to use it in an article on homeopathy, to substantiate a claim that some homeopathy practitioners claim good results in treating autism. Self-published sources are perfectly reasonable, as long as they are used in an appropriate way, that is: in an article about the author, or with caution to attribute them as opinions of the author in a topic where the author's opinion would be considered relevant. In this case, Academy Awards is using an article released by the awards' own parent organisation, which is analogous to the case of using Raval's press release on a page about Raval himself, and therefore the source is perfectly fine. I disagree that using the NYT article is better; for this kind of uncontentious information a source closer to the original source of the information is better. JulesH (talk) 18:09, 19 April 2009 (UTC)

DGG, following your logic, the White House is a self-published source when it releases material about the president. SlimVirgin ^{talk|contribs} 01:40, 20 April 2009 (UTC)

SV, the White House (or more specifically the office of the White House Press Secretary) is indeed a self-published source when it releases material about the President or his administration. It happens to be a very high quality SPS, (clearly passing the "expert" criteria) but it is an SPS never the less. Blueboar (talk) 13:06, 20 April 2009 (UTC)

It's a primary source, but not a self-published source. Imagine a situation in which the White House publishes a comment about a living person. What you are saying is that we would never be allowed to use that in that person's BLP, because self-published sources aren't allowed in BLPs. We would only ever be able to link to secondary sources that repeated what was said, but never to the primary source itself. To call White House statements self-published sources, when they are possibly checked and edited by more eyes, and in particular by more legal eyes, than any other statements or releases in America, if not the world, is just wrong. The prohibition on self-published sources is entirely to do with their not being checked, with there being no editorial or legal oversight, so that we can't risk relying on them. That does not, of course, apply to the White House, or to any other major institution. SlimVirgin ^{talk|contribs} 15:38, 20 April 2009 (UTC)

SV, the Verifiability policy does not define self published source, it just describes them, but does not state that any self-published source not fitting the description is not, for verifiability purposes, self-published. --Jc3s5h (talk) 14:41, 20 April 2009 (UTC)

It describes them in a way that makes clear what is meant: "Anyone can create a website or pay to have a book published, then claim to be an expert in a certain field. For that reason self-published media, whether books, newsletters, personal websites, open wikis, blogs, Internet forum postings, tweets etc., are largely not acceptable." There's no way the White House fits that description. I didn't follow your second point -- too many nots ("... does not state that any self-published source not fitting the description is not ..."). :-) SlimVirgin ^{talk|contribs} 15:38, 20 April 2009 (UTC)

Okay, sorry, got it. No, but it's a matter of common sense. "Anyone can create a website" etc does not imply the New York Times writing about itself; the White House releasing statements from the president; the High Court in London releasing one of its rulings. SlimVirgin ^{talk|contribs} 15:41, 20 April 2009 (UTC)

Primary source indicates that secondary source is wrong

Information in Wikipedia should be supported by secondary sources (WP:NOR). If a primary source indicates that the information presented in a secondary source is wrong, shall the information

still be presented, because the primary source is not admissible, or considered to be of lesser relevance, or
not be included in the article at all, or
be attributed to the secondary source, with a comment that the primary source contradicts the information given by the secondary source? --Cs32en (talk) 21:53, 20 April 2009 (UTC)

Without knowing the details this is impossible to answer... It really depends on the exact sources in question. I would say your last option is generally the best. The first assumes that Primary sources are not allowed, which is wrong (Primary sources may certainly be used, they simply have to be used with care... see WP:PSTS). I suppose the second works if the secondary source is contentious. Of course you leave out a fourth option... see what other secondary sources have to say on the matter and report that. Blueboar (talk) 22:06, 20 April 2009 (UTC)

The secondary source that the article currently refers to attributes the claim to a primary source ("A", for convenience). Other secondary sources either just contain the claim or contain the claim and a reference to the same primary source A. The primary source referred to above ("B") would be a WP:RS source, but does not explicitly state that the information in the secondary source is false. I have found no definite WP:RS secondary sources discussing the primary source B, or referring to it. (Of course, if there were such secondary sources, the whole issue would not have come up in the first place, probably.) --Cs32en (talk) 23:40, 20 April 2009 (UTC)

In this case one should be very careful that the primary source isn't being misinterpreted. I've seen cases on Wikipedia where such 'contradictions' were simply due to a lack of understanding of the primary source. For instance - source presents two related data items which look like they should be the same thing, but aren't because of subtleties in how they were collected. In cases like this, secondary sources are very important because (IMHO) we should not be relying on Wikipedia editors to choose the correct option from the primary source. --GenericBob (talk) 22:31, 20 April 2009 (UTC)

I have looked again at the matter. There would be a possibility to reconcile the information given by the sources, but it would involve a rather extraordinary assumption. Thank you for your comments! I was unsure how to respond to a certain line of argument in principle, and your comments have been helpful. Please understand that I'm not going into the details here, in order to keep the discussion on this page focused on policy, not content. --Cs32en (talk) 23:30, 20 April 2009 (UTC)

The problem is that policy and content can not always be divorsed. Blueboar (talk) 23:33, 20 April 2009 (UTC)

I agree with this observation. In this case, however, the argument I encountered was made explicitly with regard to policy, not content: "Accepting for the sake of argument that it says what you say it does, I don't think we can reasonably use Prosecution Trial Exhibit P200054, a primary source, to over-ride the secondary sources." (The exhibit was produced by the United States.) I think we can come to a conclusion with regard to the validity of this particular argument without looking at the content, while the content would, of course, be relevant for determining the appropriate way of presenting the information in the article. --Cs32en (talk) 00:58, 21 April 2009 (UTC)

"There would be a possibility to reconcile the information given by the sources, but it would involve a rather extraordinary assumption." – I think this is a key point. I believe something that happens quite often is that a writer of a secondary source (e.g. a newspaper reporter) gets confidential information which he has reason to believe is correct, and which puts the primary source into a new light. Let's say in a corruption affair in Bananistan several personal friends of the president were involved in the suppression of what looks like it must have been relevant information. The way a reporter writes about this may be influenced dramatically if he knows from a trustworthy but unquotable source that the suppressed information relates to shop-lifting by the president's daughter. For us it would look like an "extraordinary assumption". Of course we can't speculate in articles about such extraordinary assumptions. How to handle such a situation in WP depends very much on the circumstances. --Hans Adler (talk) 07:25, 21 April 2009 (UTC)

The first part of the argument I was referring to in my previous comment was "Accepting for the sake of argument that it says what you say it does,..." In the context of the discussion, this would translate into "Assuming that the primary source would prove that the secondary source is wrong,...". Therefore, while it is possible that the secondary source has some information that would invalidate the primary source, this is not the assumption of the editor who has expressed this argument. This editor, in essence, simply said: "Even if the primary source is correct, and the secondary source is wrong, we would have to use the secondary source."

For everyone who is interested in the complexity of the content, I'll try to explain it as concisely as I can: The question is whether a series of phone calls from a plane, that have been reported by the secondary source, have actually been made. The trial exhibit purports to list all phone calls from the plane, as well as from three other planes (all were flying on Sept. 11, 2001, just in case someone hasn't guessed this already). For all connected calls, the destination number is given, except for four calls, which are listed as unknown caller, unknown destination (but with time and duration). As the claim from the secondary source implies that the calls have been made to a well-known and easily identifiable call number in a public institution, these calls should have been listed in the trial exhibit with the destination given, if the calls actually took place. The fact that they were not listed indicates that the information given by the secondary source, which is supposedly supported by those phone calls, is actually unsupported, as those phone calls did not take place. --Cs32en (talk) 08:25, 21 April 2009 (UTC)

That's still an example of what I said. A reporter investigating these phone calls might get insider information that the calls in question a) appeared as unknown caller/destination, or b) were suppressed in the trial exhibit, or c) were erased from the records by accident or stupidity, etc. For an example of c), perhaps the software used to analyse call records for law enforcement purposes isn't particularly user friendly and can be used for manipulating information, as well as for reading it. Look at an entry, press the wrong button, and poof!, the record is gone. Normally at most 1 or 2 people look at such a record, but in this case it's likely that whoever was in a sufficient position of trust to be able to do so, did so out of curiosity. Including some who were not actually used to using the software. In case b), I suppose there would be a criminal act involved. The reporter might trust his source enough to ignore the apparent contradiction to the primary source, but not enough to mention the faked exhibit as such. Remember that the same kind of considerations that we discuss openly on talk pages, are likely to happen privately, often based on information that is never published. --Hans Adler (talk) 10:59, 21 April 2009 (UTC)

I agree with you that, in principle, any of these possibilities could be true. I wouldn't have brought this up here, if the argument would not have been "we use the secondary source, even if the primary source would prove the secondary source to be wrong".

Now, back to the content: The phone calls in question were the phone calls by Barbara Olson, journalist and commentator, to her husband, Ted Olson, Solicitor General of the United States, at the time. The secondary source is CNN, and has attributed the information to Ted Olson. The alledged phone call remains the only primary source to which the use of box-cutters by the 9/11 hijackers has been attributed, although the story was repeated multiple times by probably all mainstream media. In 2006, in the trial case United States vs. Zacarias Moussaoui, the U.S. presented the Exhibit P200054 (the primary source).

Obviously, multiple complementary methods have been used to compile the data. For example, callers have been attributed to a call whether they were calling from their row of seats or from somewhere else on the plane, and the recipients of the phone calls are also given. Anyone working on that exhibit would very probably have a tendency to include rather than to exclude the calls from Barbara Olson, and the exhibit was very likely checked at various levels during the preparation. For me, there is certainly sufficient doubt about the accurary of the secondary source to conclude that the information should not be presented in the article.

Any further investigation would amount to original research or could only be supported by sources which are not unquestionably WP:RS, and would therefore not be included in the article. In my view, the most likely case is that the prosecution, or the source that provided the prosecution with the data, had no indication about connected calls from Barbara Olson (the exhibit includes an "unconnected call", duration 0 seconds, to the DoJ). Then, as someone realized that the information would become public when presented in the trial, the "unknown" calls were added, giving the media enough room for speculation so that they did not need to retract the story. --Cs32en (talk) 12:43, 21 April 2009 (UTC)

At risk of jumping in cold, a conflict in sources would ideally mean that the question (whether the calls were actually made) would be unanswerable via source material and WP would remain agnostic, even if we personally don't remain so. Others would say that the exhibit echoes a fringe view, but that doesn't work because this is (presumably) an article about the fringe view. So, with citations, WP reports only that: CNN cites Olson, saying that call(s) to Olson occurred; exhibit says it lists all calls but lists no calls to Olson. And that: unknown calls are listed; Olson's primary phone is widely known. Anything more is WP:OR or WP:SYNTH; and you're bordering there to mention the phone number anyway, because apparently no reliable source has taken note of the discrepancy that you notice (and wouldn't it be so much better if one had done so? WP is not the place to blow gaping holes in prevailing 9/11 theory). I generally hold that such a "semicolon solution" can avoid charges of OR. But why would we be required to infer that Olson's main "easily identifiable call number" was necessarily the one used, rather than a private number, where "unknown" might be "legally accurate" language for "classified"? JJB 13:10, 21 April 2009 (UTC)

And see, this is why I don't get involved. The commission report linked from 9/11 attacks, chapter 1 page 8, distinguishes between "knives" (on all 4 flights) and "box cutters" (on 77 only), yet 9/11 attacks says the captain of Flight 11 and others were killed by "box-cutter knives", with both the Wikilink and the word "box-cutter" constituting claims unsupported by the cited sources and perhaps confusion bordering on WP:OR. Especially with the issue of WP remaining agnostic in the minority-view article, it should be very careful not to prejudice a claim without proper sourcing in the majority-view article. Reporting there. JJB 13:28, 21 April 2009 (UTC)

Well, if the number was classified, the exhibit could either state "call to Ted Olson, classified number", or "call to classified number", or at least "call by Barbara Olson, no further information can be provided". Actually, Ted Olson said that his office received the phone call, before putting the call through to him, so this was very probably a telephone number that could have been made public. Regardless of that, there is enough doubt about this information, so that it should simply be left out of the article. --Cs32en (talk) 14:03, 21 April 2009 (UTC)

I don't see why it has to be left out... just discuss the discrepancy neutrally... Olson has stated his wife called him. Phone records presented in the Zarakowi trial do not list such a call. Don't speculate on why the two sources disagree, or which source is right or wrong. Blueboar (talk) 14:13, 21 April 2009 (UTC)

Frankly I don't see this either. It's just that attempting to include anything that would not be a mainstream WP:RS source (i.e. a non-mainstream secondary source, or a primary source) would be very difficult, given the peculiar form of the editing process that is going on there. --Cs32en (talk) 14:58, 21 April 2009 (UTC)

Self-Published Sources - defined too restrictively?

It's increasingly the case that websites which are essentially no more reliable than self-published blogs publish articles by a number of authors. As the policy is currently drafted, those articles are not - strictly speaking - self-published. But it's surely not the intention of the policy that a blog with poor or no editorial standards can be cited just because the blog-owner's pal wrote the cited article and not the blog-owner him/herself.

This makes me think that it's not self-publication as such which is the problem. Does this need to be revised?KD Tries Again (talk) 16:18, 23 April 2009 (UTC)KD Tries Again

For a tertiary source encyclopedia self publication is a sweeping hint of a worry, but the pith of this has to do with the verifiable background and noted sway of the writer. Keep in mind, Wikipedia is not after truth (TM), but verifiability. Gwen Gale (talk) 16:38, 23 April 2009 (UTC)

List of gamelan ensembles in the United States

Editors are seeking help in regard to use of official sites as references in this RfC. The article includes over one hundered gamelans, only a few of which have their own articles. The question is: when is it appropriate to use the official sites for gamelans as sources? --Ronz (talk) 18:49, 1 May 2009 (UTC)

Which, as I understand it, is equivalent to using http://www.stuy.edu as a reference at Stuyvesant High School, or http://www.royalsociety.org references at Michael Faraday Prize, or the http://www.juniper.net references at List of acquisitions by Juniper Networks. They are all primary self-published sources, that are reliable about certain things. Hence, can be used with due care. -- Quiddity (talk) 20:23, 1 May 2009 (UTC)

From a quick glance at the RFC, it looks like this is more about notability than verifiability. I'd be happy to take a gamelan's official website as evidence for its existence and basic details, but that doesn't establish that the gamelan is noteworthy. With that many gamelans listed, there are also quality-control issues - the page is likely to turn into a large collection of dead links as pages vanish or move without Wikipedia being updated (at which point verifiability comes back into it).

One possible solution would be to move the list offsite - if one of the interested parties is willing to host such a list in their own webspace, the article could just link to that (along with some in-wiki coverage of the more notable ones). Now I check policy, this issue seems to have already been considered at WP:ELYES and WP:ELMAYBE. --GenericBob (talk) 00:02, 2 May 2009 (UTC)

non English sources

The guidance in WP:NONENG is specific to situations where both an English and non-English source exists. What about situations where only non-English sources can be found? I would think these should be treated as non-verifiable sources. Is there any additional guidance available on how non-English sources should be handled?--RadioFan (talk) 17:28, 2 May 2009 (UTC)

Saying that non-English sources are not verifiable (because some readers can't understand them) is like saying that professional journals on quantum physics are not verifiable (because some readers can't understand them). --Jc3s5h (talk) 18:01, 2 May 2009 (UTC)

Agree. Sources (in English) which are found in only a handful of university libraries (or even just one!) are also verifiable. Nobody said it had to be easy to verify them. I suspect that we, collectively, here on en: can cover ninety-some percent of the world's languages. - Hordaland (talk) 18:32, 2 May 2009 (UTC)

While this makes sense in the overall scheme of things, when faced with an article containing only non-english sources, what is the right thing to do? Is an article warranted in the English wiki because the sources are verifiable to someone? Or should that article be transwikied to the wiki of the langauge where those sources are from? --RadioFan (talk) 13:41, 5 May 2009 (UTC)

There is nothing inherantly wrong with an article using only non-English sources. Our policy is that English sources are preferred, not required. If there are any English language sources on the topic, see if you can improve the article by discussing what they say. If not, leave it. Blueboar (talk) 14:50, 5 May 2009 (UTC)

Published Photo of French Cruiser MONTCALM incorrectly ID'd

On your wiki page "French Armored Cruiser MONTCALM" there are 2 photographs allegedly of the ship. The photo in the text block is a correct photo of the ship described. The picture above the ship's specifications, however, is not of the same vessel. Points of difference: the dubious photo shows a smallish cruiser with plow ram -- extreme feature; two fat funnels grouped forward; 2 burly armored masts of the kind introduced c. 1890 in the French fleet. The actual MONTCALM of 1898 had a scaled down foremast as shown in the other photo and a pole main. The MONTCALM had two blocks of firerooms with the engine room in between, as common in the c. 1900-1912 generation of French armored cruisers. Changing the funnel distribution in a refit to the configuration shown in the dubious photo would have been impossible (no vent for smoke from the aft boiler rooms). The hull shape is altogether different. The MONTCALM as shown was a large ship of 9600 tons. The one in the questionable photo appears much smaller, perhaps 5000 tons (half the size of the other photo).

An inspection of JANE'S FIGHTING SHIPS of 1906-07 convinces me that the mystery cruiser shown is one of the AMIRAL CHARNER class of 1893. This 4,750-ton class followed the design of the DUPUY De LÔME on a scaled-down level. Layout of the armament, etc. in the photo appears to follow that shown in the Jane's schematic plan.

Since the ship is mis-identified in the printed source used (caption is shown in the scan), it is understandable that it could be included. Nevertheless, it is misleading. Whether or not the photo represents a CHARNER class ship or an earlier MONTCALM (none, however, is listed in JANES over the relevant years), it is clearly not the same ship as the article is about. The photo should therefore be removed.

I can supply a scan of the schematic and a photograph of MONTCALM from the port quarter if desired to replace the photo which should be re-labeled and removed from this article.

LT CDR Percy Poppycock, RN Principal Author, BigBadBattleships.com

Pictorial evidence to support my point: For my explication of the relevant classes of cruiser in the French navy of the period, visit: http://www.CityofArt.net/french_armor.html

The Dupuy de Lôme article is accessible from my French Navy main page: http://www.CityofArt.net/frameset6.html - on nav bar at left. The characteristics described above are clearly visible in the many photos shown. A photo of the AMIRAL CHARNER is linked through my page on the discovery of the DANTON, also accessible through the nav-bar. I have not found any photos of the ships in that class indicating the the two armored masts were removed or replaced. They fought in WWI in the Mediterranean, evidently with their archaic, 1890s look intact. This excess topweight may explain, in part, why they sank so swiftly when torpedoed (deficient watertight subdivision was surely another contributing factor, however.) —Preceding unsigned comment added by 74.60.12.6 (talk) 21:07, 4 May 2009 (UTC)

Is a FAQ a Policy?

There is a dispute on the status of a "Frequently Asked Questions" page for a Policy: should the FAQ of a Policy be a Policy itself, or a Guideline to the relevant Policy. I've set up an RFC to discuss the issue on the disputed FAQ. The Policy is WP:NPOV and its FAQ is Wikipedia:NPOV/FAQ. All input is definitely welcome! Dreadstar † 00:17, 6 May 2009 (UTC)

Ilan Pappe

I've posted a question on the RS noticeboard about whether the Israeli historian Ilan Pappe counts as a reliable source within the terms of the sourcing policy. Any input would be appreciated at Wikipedia:Reliable_sources/Noticeboard#Is_Ilan_Pappe_a_reliable_source.3F. Cheers, SlimVirgin ^{talk|contribs} 20:17, 6 May 2009 (UTC)

Neutrality enforcement: a proposal

I've started a proposal to enforce neutral editing on Israel-Palestine articles, which could be extended to other intractable disputes if it works. Input would be much appreciated. See Wikipedia:Neutrality enforcement. SlimVirgin ^{talk|contribs} 08:04, 8 May 2009 (UTC)

Clarification on SPS statement

Right now there is a debate on WT:VG about Final Fantasy VII (Famicom) which relies almost exclusively on primary sources (the manual, or a translation of it) and SPSs. The SPSs are considered experts, so I want to make it clear that for the most part they meet that criteria. However the non-SPSs are dubious if not outright unreliable. My question therefore concerns the following statement and what that means for a seperate article on a unoffiical remake of a game:

However, caution should be exercised when using such sources: if the information in question is really worth reporting, someone else is likely to have done so.

Does that mean that just using SPSs for the entire article is okay if nothing better can be found or are they okay if they have absolutely no place they can be merged (in theory this could be merged with the main FF7 article, though it would have to have its own seperate plot and gameplay sections as well as reception would also have to be divided for this game)?じんない 19:38, 7 May 2009 (UTC)

If all you have are primary sources and SPSs, I would say there is indeed a problem. A merger sounds like a possible solution. The real question that I have is this: has notability been established? I am not saying the game is not notable... I am asking if that notability has been (and can be) established in the article? Blueboar (talk) 23:17, 7 May 2009 (UTC)

Well the SPSs are all independent and talk about the game, so in that sense it has some notability although the SPSs aren't apparent to the average reader they are reliable. No mainstream reliable press has commented on the game, just individual experts on their blogs, but about 5 or so of them.じんない 02:28, 8 May 2009 (UTC)

Just jumping in here but tossing in that there's a discussion on Jinnai's talk page that elaborates further on the sources, with the initial discussion found between Jinnai and myself found here. To further clarify the sources for the article include Kotaku, Joystiq, Boing Boing, and Insertcredit, each of which *do* have editorial standards.

Also to add, the reliance on the manual (the only primary source) is isolated to the related gameplay sections and other sections of the article such as reception rely exclusively on third-party coverage of the game. The debate, as I'm understanding it from arguing with Jinnai, is less "is it notable" and more "if it wasn't covered in a magazine or journal should we cover it?", which seems rather extreme...--Kung Fu Man (talk) 05:28, 8 May 2009 (UTC)

I'm trying to clarrify if an article can be based solely on Primary and/or SPS sources because that phrase makes it sound like it really shouldn't be. In theory, we could ignore the rules with the amount of reliable SPSs if it wasn't the case, but I'd rather we have clearer wording and understanding of that phrase.じんない 02:41, 9 May 2009 (UTC)

there are two questions, notability and verifiability. As far as verifiability goes, yes, an article can be based entirely on primary sources if they are objective and reliable. for the straightforward description of the plot of a work of fiction a RS is the work itself. For interpretation and the like, a secondary source is needed. for the straightforward description of anything at all, an official source from the company or the person is acceptable also, if there's nothing controversial. The question you ask that is undecided relates rather to Notability. for components of a fiction, all I need say here is that views vary widely. but this is the article about the work as a whole, and it would be very highly unusual for the notability of an "unlicensed remake" like this to be accepted without secondary sources, though they might perhaps be informal, such as well-regarded blogs. Sometimes a licensed work in a well known series will be accepted as notable on the grounds of the previous works, but this would seem not to be the case here. AfD is the place to settle it; since there do seem to be some secondary sources. . DGG (talk) 02:55, 10 May 2009 (UTC)

I asked here because the way SPS is written its implying that one shouldn't use SPSs as the principle sources in an aritcle. If that's not the case I'd like it clarified.じんない 01:18, 13 May 2009 (UTC)

It's generally not good to rely upon a SPS, however every article is different, and context matters. Without addressing the question of notability, I would say if the editor consensus is that the SPS is the most accurate, and if readers would find that source credible on this topic, then go ahead and use it. For the details of a video game, I would find that arrangement unsurprising. There's also this statement

“

Self-published material may, in some circumstances, be acceptable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications.

”

and it sounds like, from what Kung Fu Man is saying [2] [3], that perhaps the sources in question meet that alternate standard. Cheers. Mishlai (talk) 07:39, 13 May 2009 (UTC)

Dealing with OR, V

[withdrawn and taken to WP:RSN, thanks]M 23:42, 11 May 2009 (UTC)

This page is for discussing changes to the Verifiability policy. You want maybe WP:RSN or WP:ORN. Please move this section there. cheers, Rd232 ^talk 23:10, 11 May 2009 (UTC)

Self published

Doesn't this tend to eliminate the citing of any sources from before the 20th century (i.e, almost all religious texts, almost all philosophy texts, almost all science texts, almost all history primary sources ?)

"Self-published or questionable sources may be used as sources of information about themselves, especially in articles about themselves" --InnocentsAbroad2 (talk) 00:22, 18 May 2009 (UTC)

Not at all... there were publishing houses around since at least the 1600s. However, since it is unlikely that we would cite such old sources for anything except a statement about themselves, the point is sort of moot, isn't it? Blueboar (talk) 00:28, 18 May 2009 (UTC)

Ahah. But those older texts are often (as with so much of our history) based upon manuscripts, inscriptions, and artifacts that have been since lost to war, theft, decay, etc.

So, in some cases, our doubts about an old text versus a new text are not very helpful.

--InnocentsAbroad2 (talk) 02:29, 18 May 2009 (UTC)

In historical articles, old sources can be very relevant. For instance, articles on ancient Roman topics cite the historians of the period heavily. In some cases where the accuracy is in question, those cites are presented as "X claims... but modern historians question this". But in a lot of other cases, those histories are presented as fact - e.g. a lot of Julius Caesar is referenced to Plutarch's biography.

Technically, most of these would probably count as self-published sources. But note that WP:SPS also says: "Self-published material may, in some circumstances, be acceptable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications..." The original publications may not have had editorial oversight, but the way they've been received by modern historians makes it reasonable to treat them as 'established experts' (except where those modern historians have identified them as unreliable, see e.g. Nero.)

By contrast, we generally wouldn't rely on Pliny's Natural History as a source for scientific claims, because he's not often cited by modern scientists. --GenericBob (talk) 02:50, 18 May 2009 (UTC)

To some extent sourcing is topic-sensitive. The point of wikipedia's standards is to prevent malarky from mascarading as good information in an encyclopedia entry. If relevant experts in a field consider a source to be valuable, then the article probably will too. The other point is that if the topic is notable, there is likely to be secondary analysis by modern experts, published in reliable sources such as journals related to their field. Primary source concerns, OR and Synth also come into play. I don't think that you could directly cite a civil war diary entry as a statement of fact (other than as the diary writer's opinion), for example, but peer reviewed expert historian opinions concerning such diary entries would certainly be relevant. That doesn't mean that an entry from a very notable figure's diary might not become a direct source in some way, but it would only be in the context that "Abraham Lincoln wrote X in his diary." To assert that Lincoln's claims were factual you would need to cite a wp:rs that came to that conclusion.

How these self-authored sources (changing language to make a distinction) get published today is also important. If you have a civil war diary from your great-great-grandpappy and post it online, that wouldn't be a very reliable source because we don't know that you didn't forge it. Diary entries that have been made available online by a reputable museum are assumed to have been verified to some extent as being authentic by the museum's own due dilligence. The sniff test is whether ~~the typical~~ a skeptical reader can be expected to take the source seriously as a verification of the statement it supports. Credibility is the essence of wp:v Mishlai (talk) 05:24, 18 May 2009 (UTC)

I agree that secondary analysis is vital to establish the credibility of a self-authored source - I'm not a historian by training and I'm not qualified to judge whether Plutarch/Tacitus/etc are reliable on a particular topic. However, I also think there comes a point at which it gets tedious to keep writing "Plutarch says X happened and modern historians A and B accept this account" - if they don't cast any doubts on his account and don't have anything to add to it, eventually it becomes preferable just to write "X happened (source = Plutarch)". If necessary, we can invoke A and B's work as evidence that Plutarch's 'work in the relevant field has previously been published by reliable third-party publications', but that's something that can happen on the talk page rather than in the article itself.

For me, the key issue is whether the source has been adequately exposed to professional scrutiny (and if such scrutiny has cast doubt on it, whether the article reflects that). WP:V/WP:RS puts a lot of emphasis on pre-publication scrutiny, and that's probably a good thing - it simplifies things by packaging the author's claims in scrutinised form from the beginning - but there are some cases where post-publication scrutiny should be acceptable. Assuming common sense, this is pretty much equivalent to the sniff test... unfortunately, policy mostly seems to be motivated by editors who lack the common sense to gauge what a skeptical reader would take seriously. --GenericBob (talk) 00:06, 19 May 2009 (UTC)

I agree that it would be tedious to write things that way. I think where secondary sources are available that the article would simply state "X happened" or whatever the secondary source interpretation is and then cite Historians A and B, while (optionally, if it adds value to the article) making mention of the primary source that they based their conclusions on. Mishlai (talk) 00:53, 19 May 2009 (UTC)

Emily Dickinson -- not even self published

To illustrate, if Emily Dickinson didn't publish her poems, then how do we know she wrote them ?

How do we know they weren't just a concoction of a random poet that lived after her ? --InnocentsAbroad2 (talk) 00:25, 18 May 2009 (UTC)

Wow! Your right... let's delete everything about this Emily Dickinson person!

Oh, wait... someone else published these poems, right? So that means that instead of being 'self-published' they fall under the category of just plain old "published"? Darn... I was hoping we could do without Dickinson ... oh well. Blueboar (still haunted by memories of high school poetry class). (talk) 00:33, 18 May 2009 (UTC)

So, if something is published it is true

No, the veracity of the proposition of Emily Dickinson as the author of the poems purported to be hers is based on a very distinct set of criteria.

Handwriting. Chain of possession of the manuscript. Witness of her contemporaries, in their own handwriting. And, the scale of doubt should also be based upon an analysis of motivation.

Who would be motivated to pass their own poetry off as that of a dead New England gentlewoman ? Little motivation.

Whereas, in matters of religion, politics, power, money, and unfortunately, academic credential, motivation is not only more powerful, but assertions therefore more subject to scrutiny. --InnocentsAbroad2 (talk) 02:26, 18 May 2009 (UTC)

For many reasons, but especially as a matter of economic interest, the data and publication of any article in a journal that derives its funding from a governmental or industry source seems dubious.

By the same token, citing new sources would seem to me to be shaky ground if the subject of the article concerns matters of economic or political interest.

Any analysis of a citation must include the analysis of motivation, unless one believes humans act in an objective or disinterested fashion.

Emily Dickinson did not publish her poetry because she had a very real motivation - a proper and accurate suspicion of publishers, academics, and the press. That on first blush, her work would neither be appreciated or understood. --InnocentsAbroad2 (talk) 02:37, 18 May 2009 (UTC)

Do you have a point to all this? Blueboar (talk) 13:45, 18 May 2009 (UTC)

This is... so...so... WRONG

Please, in your own context, listen - Sources over Truth. So...so...wrong. —Preceding unsigned comment added by 124.180.27.141 (talk) 11:52, 20 May 2009 (UTC)

Thank you for sharing your opinion. Blueboar (talk) 14:44, 20 May 2009 (UTC)

Meaning of "self-published"

I tried to add the following to the project page:

For purposes of deciding whether sources are reliable enough to verify information in the English Wikipedia, "self-published" means published by a natural person or an entity under the control of a natural person. Information published by governments, standards organizations, and other organizations with editorial controls are not considered self-published.

However, User:Cs32en reverted with the edit summary "Please rework that. Sources cannot verify, and what do we do with statements from North Korea or PR material from Microsoft?)".

First, the problem: whenever someone comes to this talk page saying that information from ANSI, the U.S. Supreme Court, or the like, is self-published, they are shouted down with the claim that only individuals self-publish, not big entities. But this understanding is not expressed on the project page. Indeed, Cs32en's remark illustrates this difference in understanding: North Korea and the Microsoft PR department have good editorial control, it's just that some people think they're biased.

As it stands now, everything published by the same legal person (which could be a government or a corporation) as wrote it is self-published. I don't care if the talk page fills every storage device on the planet, if it isn't in the project page, it doesn't count. --Jc3s5h (talk) 16:00, 20 May 2009 (UTC)

People involved in this debate could use another look at WP:COMMON. Adhere to the spirit, not the letter, of the policy and these problems go away. --causa sui ^talk 19:14, 20 May 2009 (UTC)

Indeed, focus on the intent of this policy (and all of our policies and guidelines) more than on the exact wording and you will never go wrong. There is a wonderful essay on this: see: WP:The rules are principles.

The term self-published with it's vague meaning may have a chilling effect on editors who wish to use publications from sources that wrote what they publish (such as Microsoft documentation). This chilling effect is not visible because the chilled editors never wrote anything. --Jc3s5h (talk) 23:31, 20 May 2009 (UTC)

I can appreciate why you want to clarify it, but putting that clarification in the form of a rule might be counter-productive - if people aren't willing to think about the purpose of the rules, adding more rules provides more fodder for rules-lawyering. (eg: If a husband and wife jointly own a website and co-produce content, it's not under the control of one 'natural person', but the usual SPS concerns would still be relevant. If Rupert Murdoch owns 51% of a publishing house, it's "under his control", and there might be concerns of bias, but it would be absurd to treat it as a SPS.)

Instead of "for the purposes of this policy, self-published is X", I'd lean towards something like this: "To determine whether these concerns apply, ask yourself - who, aside from the author/s, is responsible for quality control?" --GenericBob (talk) 01:16, 21 May 2009 (UTC)

Suppose I were to replace the paragraph at the head of this thread with this:

The kinds of self-published sources we are concerned about are those where a few people control both writing and publishing; when dealing with corporate authors or governments with a system of editorial control, it isn't useful to think of them as self-publishers.

Of course, corporate authors and governments can be biased, but there are other parts of the policy to deal with that. --Jc3s5h (talk) 03:09, 21 May 2009 (UTC)

Works for me. --GenericBob (talk) 04:05, 21 May 2009 (UTC)

Wikipedia:Perennial_proposals#Define_reliable_sources. --causa sui ^talk 10:01, 21 May 2009 (UTC)

Hi all — I have removed the recent addition to the policy page. I'm not opposed to any addition or clarification there, and I probably agree, at least partially, with the intent of the addition. At the same time, it seems to me that the wording of the addition would have rather added to already existing confusion about the meaning of policies and guidelines. I'll try to explain this:

A decision of the Supreme Court is not a source, but a document. It changes the social reality, while a source (i.e. a scholarly paper) just describes such a reality. For example the Gold confiscation order made it illegal to possess gold, and you would have gone to jail if police would have found gold in your possession. If a scientist writes that gold is poisonous and that possessing gold therefore leads to early death, that would not change reality (whether true or false, and no matter how reliable the source is that reported on this). The website of the Supreme Court, presenting its decisions, would be a primary source for the document. So the relevant policy is WP:PRIMARY, not WP:SELFPUB.
Whether we want to source things to North Korea or Microsoft depends very much on the context. We probably wouldn't write "Satellites are used to transmit media broadcasts,^{[Source A]} weather information,^{[Source B]} and revolutionary music.^[KCNA][4]. However, in an article about North Korea, we would want to let the reader know that the country's government claims to have put a satellite into orbit that transmits revolutionary songs.
Likewise, an editorial process may avoid minor mistakes, but this does not make the source reliable, if major misstatements are left in the text or added to it during this process. So it's good that media check on the statements published by parties that have a stake in the content of the information, and these parties often include governments. (See, e.g. "EPA misled public on 9/11 pollution. White House ordered false assurances on air quality, report says" [5], which shows how an elaborate editorial process can work in some instances.)
The important point about self-published sources is not that they are often not published by professional people. The essential characteristic of self-published sources is that there is no independent publisher that puts his reputation behind the accuracy and relevance of the published material. An independent publisher is a person or institution whose reputation does not depend on the outcome of a particular dispute. Most common examples are of course, news outlets, who specialize in distributing information, but do not engage in disputes. (Some newspapers do, of course, and if they do it in a responsible way, they keep the news coverage and the editorial board separate. That's also the reason for which opinion pieces and editorials are not the best sources for Wikipedia.)

I hope that these thoughts can help to clarify some of the problems with the comment that I have removed, and possible also help to develop the relevant policies and guidelines. Cs32en 11:59, 21 May 2009 (UTC)

I don't think we can clearly define Self-publication. Lets review some different situations:

Printed Works: It is not always easy to know whether a printed book has been self-published or not. You have to know something about the author and the publishing house.

Film and Video: Where do we draw the line with film and video? I hope we would agree that a film produced by a major studio is not self-published, and that a home movie is... but what about Independantly produced films?

Web Pages: I think we would all agree that personal webpages are self-published, but we have a harder time when it comes to organizational webpages? It really depends on the organization... anyone can create a webpage and claim to be an "organization" after all.

In other words, there are all sorts of shades of grey when it comes to self-publication. So... I think we are better off NOT defining what we mean by "self-published". Leave it with an unwritten "we know it when we see it." Yes, this will often lead to disagreement... No one said editing Wikipedia is easy. Blueboar (talk) 14:55, 21 May 2009 (UTC)

Something else to consider... I think a lot of the "confusion" as to SPS stems from people mis-using it in content disputes. A lot of people seem to think that the policy is saying don't use self-published sources... but that is not the case. We definitely do not ban their use. Blueboar (talk) 16:16, 21 May 2009 (UTC)

Not to rant, and certainly to be concise, we might wish to refer to the recent Craigslist controversy. If I advertise a used lawn mower, there is a fact -- an advertisement occurred. Is it a fact I advertised the lawn mover ? That would be to determined by a secondary inquiry. Is there really a lawn mover ? Same thing. --InnocentsAbroad2 (talk) 21:48, 21 May 2009 (UTC)

To know what "self-published" means, look to the consequences of being self-published in our policies and guidelines. You can only use self-published sources for information about themselves, or if they are acknowledged experts who have published in third-party independent publications. The consequences are described in terms of a natural person, not a corporate person. If we are going to consider a corporate author to be a self-publisher, all are policies need to change.

Furthermore, corporate author-publishers are held to account in the same way that regular publishers are: if they produce crap, they either go out of business (unless the crap involves lots of scantily-clad people). Likewise, democratic governments are held to account by the voters. Is the accounting perfect? No, it isn't, neither in the case of large author-publishers, nor in the case of regular publishers. But they all have something to loose, unlike the anonymous author of a personal website. --Jc3s5h (talk) 16:46, 21 May 2009 (UTC)

In my view, the best way to determine the policy intent behind the WP:SELFPUB guideline is to try to define the principle of independent sourcing: An independent publisher has built up a reputation for accuracy that is independent of the outcome of any particular dispute about facts or opinions presented in a single publication. (This is also the underlying reason for the exception for self-published recognized expert opinions, and if there is no dispute about a fact in the society as a whole, the details of sourcing policy are not that important after all.) This should apply to corporate and government self-published sources as well. We shouldn't cite tobacco companies on questions of human health, except in cases in which their views are relevant to the article's subject and are attributed to them. Most standards or other specialized government information that is not reported by the general media is primary material and should be used in connection with appropriate reliable secondary sources, per WP:PRIMARY. Cs32en 22:09, 21 May 2009 (UTC)

Using Wikipedia as a source

MyWikiBiz references an apology from Jimbo. This reference is only to a post by Jimbo in a Wikipedia talk page. It's also clearly appropriate as a self-published source.

Therefore, I suggest changing

Wikipedia itself is self-published. Therefore articles and posts on Wikipedia, or on websites that mirror its content, may not be used as sources.

to

Wikipedia itself is self-published. Therefore articles and posts on Wikipedia, or on websites that mirror its content, may only be used as sources in situations where self-published sources are allowed.

Ken Arromdee (talk) 16:13, 14 May 2009 (UTC)

WP:SELFREF already addresses the subject.LeadSongDog come howl 17:12, 14 May 2009 (UTC)

That's about something different--it's about not writing an article in a way that only makes sense if the article is on Wikipedia. It doesn't really address using Wikipedia as a source.

The problem is that that section of WP:V contradicts other policies. It's written so that Wikipedia can't be used as a source at all because it's self-published. But elsewhere we don't bar the use of self-published sources at all. In fact we allow them in limited circumstances. If the rationale for restricting the use of Wikipedia as a source is that it's self-published, then it should also be allowed in those limited circumstances. Ken Arromdee (talk) 20:56, 15 May 2009 (UTC)

Wherever, the point is that WP should not be relying on itself for verification - it's a recipe for self-supporting fallacies. A vandal or even an erroneous good faith edit introducing a fallacy to one article is a limited problem. But if other articles rely on that article for verification, the problem becomes much larger in scope. So we don't do that.LeadSongDog come howl 21:07, 15 May 2009 (UTC)

This is a very cumbersome way to put it. You are effectively saying "Don't do X, except when X is allowed". It may be best to leave it the way it is and the let people use their common sense to determine when to make exceptions. The policy isn't a firm rule after all. --causa sui ^talk 21:28, 15 May 2009 (UTC)

I do understand what Arromdee is trying to say. Perhaps a better way to express it is:

Wikipedia itself is self-published. Therefore articles and posts on Wikipedia should not be used as sources (An exception may be made when sourcing articles about Wikipedia). Also, to avoid inadvertant self referencing, websites that mirror Wikipedia's content should never be used as sources.

Would that clear the confusion up? Blueboar (talk) 00:26, 16 May 2009 (UTC)

I'm still worried about enumerating the exception, since it creates the impression that other (unforseeable) exceptions couldn't be made without also editing them into the policy. That opens the door for instruction creep by creating the impression that all possible exceptions are listed in the policy and everything else is forbidden. Try this:

Wikipedia itself is self-published. Therefore, articles and posts on Wikipedia should generally not be used as sources. Also, to avoid inadvertent self referencing, websites that mirror Wikipedia's content should never be used as sources.

--causa sui ^talk 00:54, 16 May 2009 (UTC)

That works for me... but then I tend to focus on intent rather than obscessing about exact wording (ie wikilawyering). Blueboar (talk) 01:13, 16 May 2009 (UTC)

You're 1) justfying the limit on using Wikipedia as a source by saying it's self-published, yet 2) putting stricter limits on it than on other self-published sources. Those don't go together. And in my original post I gave an example of an article which uses Wikipedia as a source right now in an obviously useful way.

As for the possibility of a vandal introducing an error and then referencing it somewhere else, it would still be subject to other limits concerning self-published sources. A self-published source by the vandal is only allowed if it's about the vandal and there's no doubt the vandal published it. Your scenario, then, would be that the vandal mentions something about himself on a talk page, signs it in such a way that nobody doubts it's from himself, then uses that talk page as a reference for what his own words are. I somehow don't think this is a big problem, and if a vandal wanted to do this, he doesn't need Wikipedia--he could just use his own blog instead (which would be a self-published source too). Ken Arromdee (talk) 18:26, 17 May 2009 (UTC)

We may not express the why' all that well... but there is clear consensus that we should not use one Wikipedia article as a source within another. WP:SPS is but ONE reason. Another is that Wikipedia is subject to rapid change, so something we are citing to in one of our articles may not be in the article the next time someone tries to verify it. To me, the most important reason is that doing so would set up a self-reference. The only time we should allow a citation to Wikipedia is in the context of an article about Wikipedia. Blueboa r (talk) 18:56, 17 May 2009 (UTC)

Wikipedia being subject to rapid change is a non-issue. You link to a specific version, which doesn't change. And to clarify, are you saying that the reference in MyWikiBiz ought to be removed? Ken Arromdee (talk) 00:53, 18 May 2009 (UTC)

In the case of MyWikiBiz, the article clearly involves Wikipedia, and so could be cited for certain kinds of information. However the statement "when [name deleted] was banned from Wikipedia" could not be supported just with a citation to Wikipedia, because the true identity of Wikipedia editors is seldom known. --Jc3s5h (talk) 01:39, 18 May 2009 (UTC)

I understand and agree with the objections to the existing phrasing. Self-published sources are not completely unusable, so saying that wikipedia is a SPS is not an adequate explanation for "never" (with few exceptions) using it. Self-reference is the real issue. It's also fairly pointless except as described above. Why cite a wikipedia article when it would be more helpful to the reader to cite that article's sources directly? Mishlai (talk) 00:45, 19 May 2009 (UTC)

Are we all in agreement that Wikipedia should not be used as a source (except in articles about Wikipedia)?... if so, then it is just a matter of properly explaining why this is so.

To toss in another reason... Wikipedia is not considered a reliable source (being an open Wiki) and we should not cite unreliable sources. Blueboar (talk) 00:55, 19 May 2009 (UTC)

But that's still no different from other self-published sources. Other self-published sources aren't considered reliable either. And if they're web pages, the author can change them at any time. In fact, that's worse than Wikipedia, since on Wikipedia you can link to a specific revision and then the fact that the Wiki is open is no longer relevant (since nobody can change the specific revision).

I just don't see any justification for treating Wikipedia as worse than other self-published sources. And the current wording seems to do so without good justification; it basically says "Wikipedia is as bad as other self-published sources, so it must be treated worse". Ken Arromdee (talk) 22:24, 21 May 2009 (UTC)

Ken, are you arguing that we should be able to cite one wikipedia article in another wikipedia article? If so, I can assure you that will not fly. We may be having difficulty expressing why we don't allow it... but it is fairly clear that consensus is that we shouldn't. Blueboar (talk) 00:34, 22 May 2009 (UTC)

All information in a Wikipedia article is supposed to be sourced to somewhere else. A self-published source is only useful if it gives information about the author that can't be found somewhere else. So even if we allow Wikipedia as a self-published reference, it would never be possible to use a Wikipedia article as a self-published source in another Wikipedia article (unless you're talking about their activities in Wikipedia).

However, this reasoning isn't true of talk pages or other parts of Wikipedia. Certainly if someone apologizes to another person in a talk page that's just as good a source as if they put it on their blog instead. (That's the scenario that led me to write this in the first place: someone apologized to someone else on a talk page. We'd have accepted it if it was in a blg instead.) Ken Arromdee (talk) 19:55, 22 May 2009 (UTC)

There are several problems with citing Wikipedia: (1) Wikipedia does not publish, but is (for the most part) a collection of single edits that in most cases cannot be regarded as published text. (2) It's hard to determine whether an edit expresses the true opinion of an author, or even his or her real assessment of a certain fact, or whether it is an attempt to find a compromise. (3) Some Wikipedia content gets deleted, of course, including some page histories. (4) Talk page edits are often not intended for the general public, but are part of a specific conversation between involved editors. Almost all of Wikipedia thus cannot be regarded as a self-published source. Cs32en 01:23, 22 May 2009 (UTC)

1) Such a single edit was created by one person and was put up with the full knowledge that everyone on the Internet can see it. It's beyond me how you can say that that isn't self-published.

2) That is basically saying "Wikipedia is worse than other self-published sources because people on Wikipedia are more likely to express opinions that are not truly their own." I find this unbelievable. (Also, compromising isn't the same as lying.)

3) This is no different than using any web page as a self-published source. All web pages can be edited or deleted; we still accept web pages as self-published sources.

4) See 1. All Wikipedia edits are, in some sense, intended for the general public; someone who edits Wikipedia knows very well that the general public can see it, and does it anyway.

Ken Arromdee (talk) 19:55, 22 May 2009 (UTC)

How about WP links that double as informal non-WP cites? Some examples, from the Ninth Amendment article:

However, Justice William O. Douglas rejected that view; Douglas wrote that, "The Ninth Amendment obviously does not create federally enforceable rights." See Doe v. Bolton (1973)... Justice Antonin Scalia has expressed the same view, in Troxel v. Granville (2000)... Justice Arthur Goldberg (joined by Chief Justice Earl Warren and Justice William Brennan) expressed this view in a concurring opinion in the case of Griswold v. Connecticut (1965)...

IMHO, naming the case should satisfy citation requirements - it's not a conventional 'footnote' cite, but it gives the reader all the information they need to go check out 'Doe v. Bolton' for themselves. However, the USC opinions are heavy and painful reading, so it's user-friendly to point the reader to a WP article that discusses them in plain English. Combining both these things in a single link seems like an efficient and readable way to do it. --GenericBob (talk) 01:29, 22 May 2009 (UTC)

Linking is not the same as citation. In a link we are not using the article as a source for information, we are mearly saying "see here for an article on this topic". So far we have been talking about using Wikipedia as a source. Blueboar (talk) 01:48, 22 May 2009 (UTC)

A citation need not be in a footnote. If the information needed to find the source can be conveniently put in the running text, that's fine. Probably the most common examples of this come up when writing about books and films; just mentioning the title of these is an adequate reference in some cases. If Wikipedia happens to have an article about it, it can be linked to the article too, but it is the named work that is serving as a source, not the Wikipedia article. --Jc3s5h (talk) 01:52, 22 May 2009 (UTC)

New inline tag proposal

In editing an article which seem to have to have some poor use of SPSs i wondered if an inline tag might be useful to allow editor to find better sources. User:Neon_white/self-published --neon white talk 10:44, 19 May 2009 (UTC)

In general if I found an SPS supported statement where such a source was not acceptable I would just remove it, reword it, or resource it to conform with policy. You could also just strip the source and fact tag it.—Preceding unsigned comment added by Mishlai (talk • contribs)

True but a specific tag, allows an editor to see what exactly is considered wrong with the current source expecially if there is a dispute. --neon white talk 10:15, 22 May 2009 (UTC)

What about subjects where the only sources are inherently unreliable

A particular one comes to mind - the article on Gurdjieff largely relies on recollections and autobiographical tales as told by himself. Is it really acceptable? Especially since most of it is simply not true. —Preceding unsigned comment added by Kotika98 (talk • contribs) 14:52, 2 June 2009 (UTC)

How do you know it's not true? That's not a rhetorical question; books which debunk Gurdjieff are likely to be useful sources in heading towards neutrality. Septentrionalis PMAnderson 14:46, 4 June 2009 (UTC)

Rules like this...

...are the reason that no one trusts Wikipedia as a reliable source. Verifiability over truth? Jesus. The concept speaks of idiocy. Basically what this article is saying is that, even if something is widely known as true, it is not wiki-worthy unless some source specifically states it.

This article also seems to indicate that even if a source has printed faulty information, faulty information that comes from a cited source is better than true information that comes from common sense, common knowledge, or individual knowledge. There is no rhyme, reason, or logic to this concept.

This rule should be changed in order to make Wikipedia look like less of a joke to the average person. 76.217.89.51 (talk) 14:16, 24 May 2009 (UTC)

So, how would you propose to tell the difference between true information that comes from common sense, and false information that comes from what somebody believes is common sense? --GenericBob (talk) 14:50, 24 May 2009 (UTC)

How about Verifiability in search of truth?-- Legeres (talk) 17:12, 24 May 2009 (UTC)

Buried deep in this pages history, you could find that same consternation posted by me. I think in large part, the opening statement is what often leads to this topic of discussion.

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

I think that many people see that bold "verifiability, not truth", and take it out of context of the entire paragraph. (personally, I'd lose the bold immediately). I'd also change it to verifiability, of truth. Just a thought as I wander through the 'pedia. — Ched : ? 17:58, 24 May 2009 (UTC)

Verifiability of truth sound much better.-- Legeres (talk) 12:00, 25 May 2009 (UTC)

Verifiability and truth aren't alternatives here. Truth (on WP) is established through verifiability. That could probably be expressed clearer in the policy. The phrase "verifiability, not truth" is certainly pretty unhelpful. Rd232 ^talk 12:41, 25 May 2009 (UTC)

I think it can be helpful, but it's easy to misunderstand. Maybe I'm in the minority, but when I read it, it immediately made sense to me: what it told me was that I should only put information in articles that I knew to be verifiable, and not information that I knew to be true. The former standard is much higher than the latter because it includes the concept of truth in it. Not only must it be true, but it must be verifiably true. I think that's key. The concepts aren't mutually exclusive because one is contained in the other. The policy wants to make clear that the higher standard is our standard, because we want everyone reading the articles to be able to check the facts for themselves. Maybe the policy could be reformed to communicate that idea more explicitly. --causa sui ^talk 13:30, 25 May 2009 (UTC)

I meant it was unhelpful because once you understand it, it's meaningful, but to people trying to understand the policy, it's easily misunderstood. I'd be quite happy to do away with the phrase! Rd232 ^talk 14:05, 25 May 2009 (UTC)

I'd be in favor of replacing it with something better. How about "The standard for inclusion is not only truth, but verifiable truth." Or something to that effect? --causa sui ^talk 14:15, 25 May 2009 (UTC)

(outdent) Currently it says

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

Suggest

The standard for inclusion of information in Wikipedia is verifiability: information should not merely be true, but verifiably true, with reference to reliable external sources. Both readers and other editors need to be able to verify information in Wikipedia, and this can only be done when relevant sources are appropriately cited. Editors should provide a reliable source for quotations and for any material that is challenged or likely to be challenged, or the material may be removed.

Rd232 ^talk 14:32, 25 May 2009 (UTC)

Sounds great to me. --causa sui ^talk 15:18, 25 May 2009 (UTC)

lol .. I'll go along with that one Ryan! — Ched : ? 18:43, 25 May 2009 (UTC)

hey, I took a shot, revert at will ;) — Ched : ? 18:46, 25 May 2009 (UTC)

re: revert - np ... glad to see some work on this. — Ched : ? 19:08, 25 May 2009 (UTC)

I like what Rd offers .. "make it so" if you please. ;) — Ched : ? 19:26, 25 May 2009 (UTC)

Think before you act, please... which of the following verifiable statements is "true":

Jesus was a wise man
Jesus was the Masiah
Jesus is the Son of God
Jesus never actually existed
Jesus was a Prophet

Perhaps now you can understand why we include the language we do?Blueboar (talk) 20:18, 25 May 2009 (UTC)

It has only now dawned on me that the "verifiability" requirement is profoundly anti-scientific. The scientific method seeks assertions (hypotheses) which are falsifiable, but have never been falsified. If you can't test the hypothesis with an experiment, you can't falsify it. It is then not science, but faith. Nothing is ever verified by scientific experiment, only falsified. The weight of many experiments that do not falsify a proposition collectively support it, but one experiment is all it takes to falsify it. LeadSongDog come howl 20:27, 25 May 2009 (UTC)

I think LeadSongDog is mixing up scientific research with reporting of science, which is what WP does. In reasearch a hypothesis can only be falsified, as LeadSongDog said. However a WP science article reports "A wrote that X" or "[The current scientific consensus is] X", which are verifiable from published documents. Even if WP simply says "X", "The current scientific consensus is ..." is implied by WPNOR and WP:NPOV. --Philcha (talk)

I really have no idea what this Jesus example is supposed to demonstrate. None of these claims is verifiable (and probably they are all also false, by the way). Anyway, LeadSongDog (talk · contribs) is definitely confused about what we are doing here. This is not scientific method because Wikipedia is not a project of discovery. It is a tool for reporting information. In that way we rely on verifiable truth such as "XYZ study found a correlation between A and B and researcher Dr. Q concluded that there is a causal relationship..." and so on. --causa sui ^talk 20:48, 25 May 2009 (UTC)

That rather clarifies my point. If we only reported explicit consensus positions we would miss most content. Far more often, scientific consensus is established by the simple lack of contradiction. When a published result stands unchallenged for several publication cycles it begins to acquire weight. It is rare for someone to publish a verifiable paper stating that "Dr. Q's conclusion has not been challenged after five years", and WP:NOR would prevent WP editors from making that statement. So the consensus around Dr. Q's conculusion would, to a WP reader, appear simply as one researcher's conclusion.LeadSongDog come howl 06:44, 27 May 2009 (UTC)

Blueboar, I am really sorry , but I'm going to have to fall squarely in the Ryan camp here. Word one, Statement one .. Jesus... While I do firmly believe in my heart certain things as a Christian (WP:OR, from an encyclopedic point of view: there are no WP:RS which can verify word one. Hence, all proceeding comments fall into the null and void category. As far as the "scientific vs. other" points of view, those would be best taken up at the respective talk pages of the articles in question. Just IMHO — Ched : ? 21:14, 25 May 2009 (UTC)

My point was this: for each of the above statements, there are people who believe that the statement is "True" (and, conversely, that the others are "false"). "Truth" is far to often a subjective concept. People can disagree over whether something is "true". We can not verify a statement such as: "Jesus is the Son of God"... even though millions of Christians around the world will tell you that the statement is "True". The language "Verifiability not Truth" makes it clear that we don't care about true or false... what we care about is verification. Blueboar (talk) 21:33, 25 May 2009 (UTC)

I think you're confusing verifiability with NPOV. Verifiability applies to things which can reasonably be verified. Transcendental and religious truths do not fall into that category - so we don't try to verify their truth, we only try to verify who thinks what about these things, and keep a neutral point of view about whether it's true. Anyway, I don't see really what you're saying in respect of my proposal - is my proposal less clear on this to you? Rd232 ^talk 22:04, 25 May 2009 (UTC)

My point exaclty... "Truth" is a matter for WP:NPOV... this policy should only deal with Verifiability... as in "The threshold for inclusion is Verifiability, not Truth." Blueboar (talk) 22:18, 25 May 2009 (UTC)

At the risk of repeating myself... "I don't see really what you're saying in respect of my proposal - is my proposal less clear on this to you? " As far as I can see "information should not merely be true, but verifiably true" covers the same ground, just more clearly. (for one thing, "verifiability, not truth" risks implying we can verify things which aren't true - the policy clarifies this, but why obscure it in the first place with this problematic phrase?) Rd232 ^talk 23:01, 25 May 2009 (UTC)

Would it help any to substitute "accurate" for "true"? Rd232 ^talk 23:05, 25 May 2009 (UTC)

Yes, accurate is a better word... we do strive for factual accuracy, and accuracy is not subjective the way "Truth" is. My main concern not to open even a hint that we base inclusion on whether something is "true" or not. Do that and we will quickly get people arguing over truth and falicy... when none of that matters to wikipedia. We can report on things we think are "false" (for example, we have an article on inteligent design, even though the consensus of just about every scientist is that it is a false, pseudo-sciencific theory). We don't care whether something is true or false... we care about Verifiability. Blueboar (talk) 00:09, 26 May 2009 (UTC)

It takes a pretty foolish reader to think that an encyclopedia endorses the subjects it describes. And your example about "having an article" is an issue of notability, not verifiability. Rd232 ^talk 00:17, 26 May 2009 (UTC)

I also think you're forgetting we have a whole policy page to explain exactly what's meant. The issue is only whether the intro can be made clearer. I think "verifiably true" is not problematic in the way you describe, because what exactly is "verifiability" verifying, if not truth? It's just unpacking the noun for clarification. Rd232 ^talk 00:17, 26 May 2009 (UTC)

At the risk of being overly philosophical about this: Something can be verifiable (according to our definition) and still be false. Our sources may not know it yet. For a historical example: consider Newton about relativistic behavior. The 1800 Wikipedia article on physics would be verifiable but "false." We are to report what "qualified" sources say is true, not what IS true. Hopefully there is high correlation between the two, but they are intersecting sets of "information," neither a subset of the other. Arguing for inclusion should be based on verifiability. (John User:Jwy talk) 00:27, 26 May 2009 (UTC)

At the risk of pointing out the obvious, that's a problem with the policy, not the lead section under discussion. And since WP is a work in progress, when new evidence comes to light, previously presumed verification can be challenged. Rd232 ^talk 10:06, 26 May 2009 (UTC)

Rather than getting bogged down in this semantic debate, can we get an up or down vote on rd232's proposal above? I think most of us are supporting the change. --causa sui ^talk 00:29, 26 May 2009 (UTC)

support — Ched : ? 01:00, 26 May 2009 (UTC)

Strongest possible oppose -- who is going to test for truth? In how many subjects is this virtually impossible? What is the 'truth' in history, archaeology, even science? No scientific theory is 'true' for instance. Many historical claims can never be shown to be true. I've been looking at the battle of Mons Graupius, generally considered to have taken place as described but there is some current thinking that it was either trivial or didn't take place. What's 'true' here? This would be a fundamental change in policy and in the direction of Wikipedia. Last year I ran into a problem with an editor when I used a very good source that stated that the positions of the finger bones of corpses suggested violence -- I was told that the statement had to be verified to show it was true. As John says about, "arguing for inclusion should be based on verifiability". And wouldn't it mean that a lot of our articles could no longer show any disagreement between experts? Dougweller (talk) 07:32, 26 May 2009 (UTC)

oppose - and I say this as someone involved in articles where I often encounter information which is "verifiable" by WP standards, yet I know to be false. While I appreciate the reasoning behind the recommendation, such a change may make it MORE likely that verifiable but false information is disemminated as true - because contrary to the OP, people DO trust wikipedia. Heck, in another forum I've been getting text from Wikipedia quoted at me as evidence of something, and the irony is that (unbeknowns to them) I wrote that text, albeit poorly. This change says "if it's in Wikipedia, it must be true!". And we all know that isn't always the case. --Insider201283 (talk) 09:26, 26 May 2009 (UTC)

Section break

Look, my clarifying rephrasing wasn't intended to be a change of meaning, as some suggest. Bottom line, we need to unpack slightly more what "verifiability" and "verifiable" means to provide understanding more easily to the passing reader / new editor.

OED on verifiability: (a) "The fact of being verifiable." (b) "Philos. The fact of being capable of verification." The latter meaning is clarified by a quote: "1967 Encycl. Philos. VIII. 241/1 We shall understand the verifiability principle as claiming that the cognitive meaning..of a sentence is to be determined by reference to the verifiability..of the statement expressed by the sentence."
OED on "verifiable": "That can be verified or proved to be true, authentic, accurate, or real; capable, admitting, or susceptible of verification."
From current WP practice, I draw the conclusion that we are attempting to verify truth, in two senses. (a) For simple facts, we seek to support them with a reliable source in a way that implies we think it's true; and that's generally enough unless it's disputed, in which case some truth debate takes place, albeit displaced onto arguing about the reliability of competing sources. (b) For opinions or claims that can't reasonably be verified or where reliable sources disagree, we seek to support them by linking them to reliable sources in a way that makes it clear we're not judging whether it's true (WP:NPOV), eg by direct quotation and explicit attribution in the body text. The truth we seek to establish here is not the fact in the statements, but the fact of the statement (that it was actually made). Example a: X was born on such a date (footnote). Example b: X said that battle Y took place (footnote); Z said it didn't (footnote).

In conclusion, Dougweller asks, quite reasonably, "who is going to test for truth?" The answer is WP editors, collaboratively when and to the extent that they can. This is not in any way, shape or form a change to current practice, just a clarification of it. I don't see anything wrong with that practice, but those who do should really start by supporting clarification of it, as that will it make it easier to challenge. (However, conceptually and practically, and I don't see any realistic alternative.) Rd232 ^talk 10:01, 26 May 2009 (UTC)

Actually, no... WP editors should never be in the position of "testing for truth". We look to reliable sources to do that... WP editors should mearly report what the sources say. Blueboar (talk) 13:52, 26 May 2009 (UTC)

If no source ever contradicted another on any topic, that would be a tenable position. If all sources were equally reliable, that would be a tenable position. If it wouldnt' make articles unreadable to explicitly attribute in the body text every single piece of information, that would be a tenable position. Rd232 ^talk 13:57, 26 May 2009 (UTC)

Seriously, people think the sky is falling here and I'm not sure why. Nobody is altering the meaning of the policy or proposing that we change actual practice. I would like to know how it's possible for something to be verifiable but not true. Think about it carefully and you'll discover that it's impossible. --causa sui ^talk 12:59, 26 May 2009 (UTC)

"Verifiable" in WP nomenclature only means the info comes from a reliable source and that someone else can go to the source and verify that's what it says. It does not mean the source is correct! Take mass media for example - newspapers and magazines are generally considered WP:V, but I doubt anyone suggests they're always true! The problem with the proposed rewording is that it implies that if it's Wikipedia, it must be true. In reality, if it's in Wikipedia the best that can be concluded (and only in a well edited article!) is that it's based on reliable, verifiable sources. --Insider201283 (talk) 13:24, 26 May 2009 (UTC)

Ryan, the problem that I see is that the proposed changes do alter the meaning of the policy. By saying that something must be "verifiably true" to be included, you place this policy in direct conflict with WP:NPOV. It is not the job of Wikipedians to verify whether something is true or false... Our job is simpley to make sure that our readers can verify that someone ELSE (ie a reliable source) said something. That's it. In fact, according to WP:NPOV we have to include viewpoints that Wikipedians might think are not true. Very often, what we are verifying are opinions and not facts.

As for verifying something that is not true... this is very simple... Take Newton's law of gravity as an obvious example. We can verify it, even though we now know that it has flaws (ie it is not true). Blueboar (talk) 13:47, 26 May 2009 (UTC)

You can't verify it because it's not true. What you can verify is that Newton said x, y, and z. That's what we do at wikipedia. We report what other people are saying. --causa sui ^talk 15:17, 26 May 2009 (UTC)

I seriously think people are taking issue here with the flaws inherent in the policy (Wikipedia:Verifiability). The lack of clarity of the present intro hides those flaws a bit, but (see my examples above) this doesn't affect current practice. Verifiability, in practice, means (a) X is true, according to these sources (referenced info) and (b) these contradictory sources say certain things (referenced, explicitly attributed info). Address this practice please, because that's what we've currently got! Rd232 ^talk 13:55, 26 May 2009 (UTC)

No... it can be verifiable that: (c) X is false, according to these sources (citation) or (d) We don't know whether X is true or not (citation)? Verifiability simply means that we must point the reader to a reliable source that supports whatever we have written. Blueboar (talk) 14:54, 26 May 2009 (UTC)

(c) is a variation of my point (a): we are stating that it is true that "X is false", based on certain sources. (This isn't necessary that often, but there are cases - outdated scientific theories spring to mind.) (d) is a variation of my point (b). What exactly is your disagreement with me? Rd232 ^talk 15:10, 26 May 2009 (UTC)

"Verifiability simply means that we must point the reader to a reliable source that supports whatever we have written. " - yes, and what we have written necessarily contains some kind of truth statement, even if it's only truth claims about the sources, remaining neutral on content. Refer back to the OED definitions above. Rd232 ^talk 15:12, 26 May 2009 (UTC)

WP doesn't focus on truth, although clearly when we write, "A said X," we hope that A really did say X. Nevertheless, what we focus on is attribution, not the truth of X. It's unfortunate that the policy was called "verifiability" many years ago, which implies that we go out hunting for the truth. I've tried a few times to have the name changed to avoid that connotation e.g. with Wikipedia:Attribution, but with no success. Regardless of the name, all we do is look for reliable attribution. SlimVirgin ^{talk|contribs} 15:27, 26 May 2009 (UTC)

rd232 answered my question about who will test for truth by saying " The answer is WP editors, collaboratively when and to the extent that they can.". A number of our articles - good ones at least -- will say that there are these reliable sources that say X is the case, another set that says Y is the case, and another set that says we simply don't know. They hopefully do not suggest that any of the alternatives are the truth. In fact, when we see an editor with 'truth' in their name, alarm bells often go off. Now if all Rd232 means is 'truth claims about the sources', the word for that is probably 'accuracy', or something akin, and not 'truth'. Dougweller (talk) 15:40, 26 May 2009 (UTC)

The problem is, as I said above with my (a) (b) examples, that there are simple, uncontroversial things which we don't attribute in the text. For example type (a), we say X was born dd/mm/yyyy (footnote). To the average reader, that means we think it's true (the reason we think it's true being given in the footnote). For example type (b), yes, we can, do and should attribute sources explicitly. Overall, "accuracy" is probably closer to what's going on than "truth", but we're kidding ourselves if we deny how closely the two are related; it still leaves the same "hard" problem, which is that, at the end of the day, statements in the authorial voice of the encyclopedia are implicitly statements that are claimed to be true. When those statements explicitly attribute claims to sources, that's fine - the truth claims are about the sourcing. When the statements are backed by footnotes, it's a problem - the truth claim appears to be about the content. Rd232 ^talk 15:50, 26 May 2009 (UTC)

Section Break 2 - Truth

Ah... I think I see where Rd232 is comming from (correct me if I am wrong)... Yes, Verification does mean demonstrating that "it is true that reliable source X says Y". However, what the "Verifiability not Truth" language is focused on is that when it comes to adding (or removing) information, Wikipedia does not really care whether Y is "True" or not. More importantly, the "Verifiablility not Truth" language is designed to make it clear to editors that they should not add something mearly because they think it is "True" ... they have to show that someone ELSE (a reliable source) thinks it is true. Blueboar (talk) 16:11, 26 May 2009 (UTC)

I've agreed with Blueboar throughout this conversation. Mishlai (talk) 10:25, 27 May 2009 (UTC)

That would be more helpful if I felt that Blueboar had addressed my arguments. We seem to be slightly talking past each other. We agree, obviously, that information shouldn't be added "just because an editor thinks it's true, even though there's no reliable source for it". The question basically is about what happens with statements supported by reliable sources (eg in footnotes) but stated as fact in the text (X is true), not as reported opinion (A says X is true). In the former case, which inevitably and probably inescapably is common, WP is saying, in effect, that X is verifiably true. (It's saying "it's true - we verified it from this source, and you can too.") In the latter, it's saying that "A says X is true" is verifiably true. (The fact of A saying X is true has also been verified from the source given.)

Either way, when we say we are concerned with "verifiability", that cannot meaning anything other than that we are concerned with is information which is "verifiably true". All of the concerns above, for me, boil down to the accepted and fairly obvious WP practice that if in any particular instance we can't agree on saying "X is true" (as fact), then saying "A says X is true" (as reported claim) is intended to be neutral, and not a hidden claim that X is true (and it certainly doesn't require us to claim that X is true in order to say that "A says X is true"). The question is how to best reflect that practice in the lead, to people not already familiar with it, and "verifiability, not truth" doesn't fit the bill. "Truth through verification" is clumsy but would be closer. "verifiable information, not unverifiable truth" would be inelegant, but closer. Rd232 ^talk 11:03, 27 May 2009 (UTC)

I don't think we are saying that "X is verifiably true"... we are simply saying that X is verifiable (even if we have not actually included a citation to a source that verifys it). I don't have to know the truth of information to add it. I have no idea whether it is "true" that someone was born on a particular date... all I know is that reliable sources tell me the person was born on that date. It is possible that every single reliable source gives the wrong date of birth... perhaps someone in the hospital where the person was born made a clerical error, and all the sources have repeated that error. The point is, we have no way of knowing the "Truth" of the matter. However, we do know what date is given in the sources, and because that date is verifiable that is the date we use on Wikipedia. Blueboar (talk) 13:20, 27 May 2009 (UTC)

That's remarkably close to sophistry. If you add someone's birthdate, based on a "reliable source", you are not doing so (and should not be doing so) if you think the information isn't accurate. Equally, if you add that Source X gives the birthdate as such-and-such, you are not doing so with no opinion as to whether Source X actually gives the birthdate as that. You are making truth claims in both cases, whether you like it or not, or realise it or not. You are saying "I think this is true, because these sources say so." Furthermore, to return to the OED definition above, to say that "X is verifiable" is to say that X is true. (The reliability of reliable sources themselves is a red herring in this discussion - we're relying upon them, so we hope they're accurate, but if we think they're not, then we don't think they're reliable sources, and that's a different topic.) Rd232 ^talk 15:07, 27 May 2009 (UTC)

"We hope they're accurate"... exactly. We don't know they are. "Truth" implies definitiveness. Verifiability does not. Blueboar (talk) 15:23, 27 May 2009 (UTC)

I actually agree that a lot of what I am saying is sophistry... but I use sophistry to counter your sophistry as to things being "verifiably true". This is the problem with using the word "Truth" at all... sophists have been arguing about the nature of "Truth" since at least Classical times.

So, to cut the sophistry... What we mean when we say "The threshold for inclusion is Verifiability, not Truth" is simple... you can not add something simply because you believe it is true... you need to be able to cite it. You can sometimes get away with not actually including a citation (ie the information is not controvercial), but you must be able to cite it if needed. A correlary to the verifiability not truth statement is that you can not remove cited information because you believe that it is not the "Truth"... you can challenge it (for example, you can argue that the source is not reliable, or that the information is irrelevant to the topic, etc.), but you can not summarily remove it based purley on your belief as to what the "Truth" is. Wikipedia avoids arguments as to "Truth"... it focuses on citation... ie Verifiability. Blueboar (talk) 16:17, 27 May 2009 (UTC)

You're basically restating the policy, which we know we agree on. The question is why should the lede of the policy falsely and misleadingly oppose verifiability and truth? Question: what does "verifiable" mean? Rd232 ^talk 16:53, 27 May 2009 (UTC)

That's just it... I don't think we do agree on the policy. You seem to want Truth to be part of the threshold for inclusion in Wikipedia, and I wish to keep it at Verifiability. "Verifiable" means that we are able to cite the information. No more, no less. It does not mean that the information we are citing is nescessarily "Truth". Also, I don't think the policy opposes Verifiability with Truth. They are related but distinct concepts. Blueboar (talk) 18:33, 27 May 2009 (UTC)

Now we're definitely talking past each other. Again, I do not "want Truth to be part of the threshold for inclusion". Truth is infuckinescapably part of what Wikipedia is trying to communicate: it is trying to create a collection of information which its editors believe to be true, that belief being justified by certain cited sources. We do not create random collections of words like monkeys writing (possibly) Shakespeare; we do not create collections of deliberate untruths! Now please answer my question above: what does "verifiable" mean? Check back to the OED. Rd232 ^talk 21:15, 27 May 2009 (UTC)

OED on "verifiable": "That can be verified or proved to be true, authentic, accurate, or real". Therefore, the statement "verifiability, not truth" makes no sense. Please, let's try and come up with something that does. Rd232 ^talk 21:33, 27 May 2009 (UTC)

I agree that Rd232 is supporting existing policy. He/she is just trying to reword it so that the very jarring (to an outsider) "we don't care about truth, only verifiability" can be replaced with a phrase that both encapsulates current editing practice and also communicates to newbies and outsiders that we are actually trying to be accurate. I'm not ready to propose a change through, I'm way too tired to do it well. Semantics are tripping this conversation up, because "Truth" has become a bad word to wikipedia insiders. If the wording is changed it needs to be absolutely clear that we're not talking about a policy change. Mishlai (talk) 21:27, 27 May 2009 (UTC)

Yes, I think "truth" has become a loaded word on WP in this area, implying that information should be included merely because an editor thinks (or "knows") it's true, and that information can't be included unless editors agree it's true; that sort of thing. Of course in practice now we instead rely on "reliable sources". The issue is therefore not "verifiability vs truth" but "reliably sourced vs not". We ought to be capable of making that clear... (perhaps we should merge WP:V and WP:RS...) but I'm tired as well. Anyone else want to chip in here? Rd232 ^talk 21:41, 27 May 2009 (UTC)

I'm afraid I have no idea what Rd232 is saying. All I can say is that truth = accuracy = facts = everything that is the case. "Verifiability, not truth" means that we do not comment on the truth/accuracy/factual nature of the content. All we do is guarantee that, if we say A said X, A did say X and we provide a citation so that anyone can check it. This is what we mean by "verifiability." In addition, we try to make sure that A is not a kook or an entirely non-notable source, so that what A says is deemed to matter. That is what "verifiability, not truth" means. SlimVirgin ^{talk|contribs} 22:35, 27 May 2009 (UTC)

Well I'm starting to feel like a kook myself :) ... because we seem on the one hand to agree on the obvious (current practice, situation B: we rely on reliable sources and when there are competing sources we present them neutrally - WP:NPOV) and on the other hand to disagree on the obvious (current practice, situation A: we rely on reliable sources and when there is no disagreement, or only one source but no likely disagreement, we simply state the fact). More generally, we're not creating a random collection of words, we're bringing together information which we believe is accurate, whether that information is "X is true" (situation A) or "A says X" (situation B). If we really thought our better articles had no more truth value than a random collection of words, why would we be bothering? The answer, of course, is that we don't think that. We think good articles have value because they have valuable information; and that information is only valuable when it is accurate/true (even it is accurate in reporting, say, a lie, which is to say a claim shown to be untrue). It is, of course, accurate/true to the best of our knowledge, as demonstrated by the verifiable reliable sources given. Rd232 ^talk 23:11, 27 May 2009 (UTC)

Rd, if you want to revisit the idea of merging RS and V, I would strongly support it, because they both say the same thing i.e. it wouldn't be a merge so much as a redirect. RS was created as a fork of V, and it really should have been deleted at birth, but unfortunately lived to cause confusion. Part of what we did with WP:ATT was try to merge two, and also try to replace the idea of "verifiability" with "attribution."

WP:ATT did become policy (ATT was a combination of V, NOR, and RS), and things were going really well -- I was even getting emails from people to say they understood the content policies for the first time. Then one day Jimbo saw it and strongly objected. We had to hold a WP-wide poll. We achieved a majority, but not the 75 percent needed for consensus, so we had to go back to having V, NOR, and RS as separate pages. I doubt we'll get consensus to merge V and NOR, but I do think RS should be redirected to V, given that they're identical. Or someone needs to write up RS as a page where different kinds of sourcing are discussed. SlimVirgin ^{talk|contribs} 22:41, 27 May 2009 (UTC)

Well I'm generally in favour of discussing possible merger of policy pages because in general policy spreads ad hoc like a weed, and it takes a special effort, occasionally, to cut it back to something more manageable (and understandable for newbies). Feel free to resurrect a proposal or create a new one or whatever. Rd232 ^talk 23:11, 27 May 2009 (UTC)

cf my efforts to merge WP:3RR into WP:EDITWAR. (WT:3RR#RFC) Rd232 ^talk 23:19, 27 May 2009 (UTC)

(undent) I like the "attributable" notion. I'll need to read that more carefully later. Mishlai (talk) 23:36, 27 May 2009 (UTC)

mm. it suddenly seems to me that replacing "verifiable" with "attributable" might just solve all these issues. Rd232 ^talk 23:43, 27 May 2009 (UTC)

The ATT page got messed around a bit. This might be a version you could take a look at. I may try to work on it on a user subpage and see if we can relaunch it. Donning my NBC suit as we speak. :-) SlimVirgin ^{talk|contribs} 00:05, 28 May 2009 (UTC)

The policies are pretty entrenched as they are and people fear change, so I'm pessimistic about the chances for a second try. But looking at this, I can't believe it wasn't adopted. I'd like to make some minor adjustments ("not merely whether it is true" etc) but even failing that I'd sign the petition for making this revision into policy exactly as it is. It's always good to reduce fragmentation of the policy and reduce the ruleset into a smaller, more parsimonious set of general principles. Please do keep me updated. --causa sui ^talk 01:58, 28 May 2009 (UTC)

I will. I'm thinking I'll try to get the writing as tight as possible and then I can show you. It was policy for quite a few weeks. Then one day Jimbo was arguing with someone about NOR, so he went to the NOR page and found the redirect, and had a fit (even though he'd been told about it). He thought that NOR was being deprecated, though in fact all that had happened was it had become a section of ATT -- because the idea of needing a source for your edits covers NOR. So he said we needed to have a poll, which we did, and we got a majority, not counting neutrals (424/354/102), [6] which I think was a lot of support considering Jimbo had come out against it. But, of course, we needed consensus, so the majority wasn't enough. Jimbo suggested setting up a working party after the poll to decide whether to adopt it, but by that point we were fed up with it. It left quite a lot of bitterness because people had put an enormous amount of work into it over several months, and some of the opposes clearly didn't realize what the vote was about (and that probably goes for some of the supports too) -- you just can't make policy at that wide a level. But that's Wikipedia. :-) SlimVirgin ^{talk|contribs} 02:18, 28 May 2009 (UTC)

It might help to give it a long breathing period as merely a summary of the relevant policies, and making the existing policies effectively daughter articles of it. Let people get used to that for a few months, hopefully gradually switching to using its shortcuts because it's more helpful; and then talk about merging (with redirects now possible to the appropriate section to keep previous shortcuts, it should be more palatable). Rd232 ^talk 11:03, 28 May 2009 (UTC)

Before you attempt to create "Summaries" of policy... you should know that we have tried the idea before and found it controvercial. When WP:ATT failed, it was left (at Jumbo's suggestion) as a "summery" for quite a while... and constantly attacked because "summary" was a concept that editors could not understand. Did it have weight in disputes? Was it policy or mearly a form of essay? No one seemed to know. Personally, I did not see any problem with having policy summaries, but others did and it was a constant bone of contention. Blueboar (talk) 15:20, 2 June 2009 (UTC)

WP:ATT contains a lot of good ideas, and many ideas which were too controversial to be consensus. (We may disagree on which is which.) Nevertheless, having them discussed for inclusion here is a good thing. Septentrionalis PMAnderson 14:52, 4 June 2009 (UTC)

Questionable sources, their uses, and associated ambiguity

Hi all.

I've just been reviewing this page as I haven't been around in a while and I observed the following about questionable sources:

Questionable sources are those with a poor reputation for fact-checking. Such sources include websites and publications expressing views that are widely acknowledged as extremist, or promotional in nature, or which rely heavily on rumors and personal opinions.

I'm presuming this means any source that isn't reliable, including my personal botch-job of a website which I might have uploaded just yesterday (hypothetical scenario of course!). We read onwards:

Questionable sources should only be used as sources of material on themselves, especially in articles about themselves.

Questionable sources can in principle be used in articles about themselves, yes. But this phrasing allows for questionable source usage in articles not about themselves. Which means my lovely well-meaning advocacy website just uploaded yesterday could be used as a source in another article, so long as the cited material:

1. is not unduly self-serving; 2. it does not involve claims about third parties; 3. it does not involve claims about events not directly related to the subject; 4. there is no reasonable doubt as to its authenticity;
5. the article is not based primarily on such sources.

Therefore, a sentence or two like:

Some hardline animal rights protestors have opposed wind farms on moral grounds.<ref>Link to my brilliant advocacy webpage about animal rights</ref>

... would technically be fine in an article on Wind farms. I don't see anything clear in the above extracts which negates this sort of use of sources. To me, verification of the above example requires a reliable source which has made that observation, if it even is indeed noteworthy. The above wording ultimately allows for use of patently unreliably sourced material in articles so long at is presented in a way that does not contravene the mentioned caveats. I think this is against the spirit of Wikipedia policies on reliable citation and ensuring material is reliably verifiable.

Perhaps it's just the ambiguity of the wording which needs to be resolved? Or maybe there's a more substantive issue at hand here? ITAQALLAH 18:23, 26 May 2009 (UTC)

First, your personal botch-job website would not be a "questionable source"... it would be outright unreliable. This refers to sources where there is a question as to its reliability. And it reflects the fact that questionable sources are not limited purely to articles about themselves... at least not as far as verifiability is concerned. If an advocacy group has issued a statement as to its stance on Wind Farms, and it is judged important to mention this stance in the Wind farms article, we obviously would cite the groups website as a source for that stance. Now... it isn't always appropriate to mention what an advocacy group says... but that isn't a verifiability issue. The are lots of other policies (such as the UNDUE section of NPOV) that might impact whether we do mention what the advocacy group says. However, if we do mention it, then we do have to verify it.

To take another example... It might be appropriate to quote something that Hitler wrote in Mein Kamph in a section of an article about Ethnic cleansing. We have to be able to cite to Mein Kamph if we discuss it. so, while Mein Kamph is clearly a questionable source, we are not limited to only citing it in the Mein Kamph article. Blueboar (talk) 19:06, 26 May 2009 (UTC)

Thanks for your response, Blueboar.

Re: your first point, there needs to be a distinction then between what is 'unreliable' and what is, if I understand your clarification correctly, 'less unreliable' (or questionable). This is however a distinction I am unfamiliar with, and I haven't found it in the respective guidelines. The pages on reliability distinguish between what is reliable and unreliable. Some sources will be more reliable than others, but the cut off point is mentioned in WP:RS. Where do all of the crankish websites, polemical tracts or conspiracy media fit between 'questionable' and flat out 'unreliable'? The truth is that they are all unreliable - even if of varying severity - and by very definition of the word should not be relied upon.

There is a degree of arbitrainess when you say '... it is judged important to mention ...', '... isn't always appropriate to mention ...' - the notion of 'importance' and how such is judged is often subjective if sources are not used to determine this. If for example we find the grievances or views of the hardline protesters in the example above mentioned by a local paper, then we can say yes this is of significance and report what is written by the paper. If not, then its significance must rightly be called into question. To rephrase: if an appropriately reliable source hasn't mentioned it, on what basis does Wikipedia mention it?

As for whether it's a verifiability issue, the notion of reliability underpins all of the core content policies and their prescriptions. I only raise it here because questionable source usage is mentioned here. Yes, UNDUE does discuss weighting and balance, but this is purely based upon opinion distribution and degree of mention across reliable sources. Unreliable sources are not taken into consideration when determining this.

I think there some OR issues to contend with in your example, but presuming it is applicable for the sake of argument, then if it were to be cited it would need to occur along side a reliable source asserting its significance/relevance/contribution to the topic at hand. ITAQALLAH 20:16, 26 May 2009 (UTC)

I think it might help to unpack what we mean by 'reliable source', because there are two different concepts involved here. In general, WP:RS discusses reliability as if it were an attribute of sources - e.g. CNN is a reliable source, an anonymous blog is not. Even the policy name, 'reliable sources', ties in with this. We might also think of these as 'reputable sources', and it so happens that coverage by a reputable source doubles as a reasonable test for notability - if CNN's given it substantial coverage, it's probably important enough for a WP article.

But when it comes down to detail, what WP:RS is really concerned with is reliability of a source in the specific context where it's used. While we'd accept a CNN cite as reliable on most topics, there might be specific scenarios where we'd reject it (e.g. it is contradicted by a later story, or the journalist who filed it is caught faking his own sources). Conversely, while we would not rely on Gene Ray to tell us the time of day, he's an excellent source when we're making statements about Gene Ray's own claims.

IMHO, the scenarios you're discussing above are cases where the source is 'reliable' in a specific context but not in general - hence, while it would satisfy WP:RS as a citation for that specific claim, that isn't a good enough argument for notability. --GenericBob (talk) 01:59, 27 May 2009 (UTC)

Bob is correct in saying that context is important. No source is 100% unreliable in every context... even the nuttiest of crank websites will be a reliable source for a statement as to what that crank website says. Granted, it is highly unlikely that there will ever be a situation where it would be appropriate to discuss what such a crank website says. But, if the editors at a particular article deem it appropriate, then they have to be able to cite the website to support the statement. Blueboar (talk) 04:17, 27 May 2009 (UTC)

I agree with GenericBob too. This issue isn't that your own crappy website about your opinion on a matter can't be used to state that "an activist" holds a view, it's that the view of a group so small that it can only be found at such a place is not generally notable enough for inclusion, and violates wp:undue. If the viewpoint is so small that it hasn't been commented on elsewhere, then why mention it in an encyclopedia? Such a practice would result in Wikipedia being used as a place for tiny, fringe views to find voice, and that is not Wikipedia's purpose.

On the other hand, the blog of a movie producer might be used to cite his opinion on a movie that he made in an article about that movie. Even if the producer himself were not notable generally, he could be considered notable in the context of the movie and his blog - otherwise unreliable - would be a usable source for citing his opinion. Does that help?

Also, there's some wp:weasel going on in "Some ... activists" - when you go through the effort to answer the weasel-squashing questions - Who? How many? and so on, then suddenly "Bob Smith said on his website that he believes..." becomes very obviously not worthy of inclusion, because Bob is not notable. This might change if, for example, Bob has been arrested in a high-profile animal rights case that is mentioned in the article - now Bob's self-sourced view of the situation is possibly inclusion-worthy. Citing many individual non-notable websites and creating a chorus would be wp:or in most cases, and again if the viewpoint as a whole is notable then it will have been mentioned in a better source. Some discretion is required, but this is the general feel of it. Mishlai (talk) 04:29, 27 May 2009 (UTC)

Help in defense of WP:V

I could use a little help convincing some of the regular motorcycle-enthusiast editors at lane splitting that the burden for finding sources that substantiate challenged material in an article falls fully on those adding and supporting that material, not on those challenging it. My exuberance in making this point has turned into an issue in and of itself, including two ANIs and now an RfC. I feel I've lost all credibility with them, no matter how persuasive my arguments may be. They're not persuasive to them, because they are mine. Or maybe they're not persuasive to anyone... you tell me...

The real issue ultimately centers on this edit which, among other things, added the following statement to the article intro: "[lane splitting] is illegal in the US except in California". The sources indicate, directly and indirectly ("Lane Splitting" - Not referenced in Administrative Code or Statutes), that the legality of lane splitting is a gray area, at least in those states (like CA and MT) where lane splitting is not explicitly banned by statute.

But my argument is simply that the "is illegal in the US except in California" statement is not supported by any citation, much less one that is verifiable, reliable and authoritative, that I challenge its veracity, and, even if I'm wrong, the burden on them is to prove me wrong (by finding a source). The regulars there who disagree with me, primarily Dbratland 1, Tedder 2 and Biker Biker have argued, or at least implied through their edits, that it is the one challenging the material that has the burden, or at least a share of that burden, to find sources supporting their view.

I understand why someone might be reluctant to step into this quagmire, but I'm hoping that those of you who watch this page would have an inherent interest in conveying the importance of editors adhering to this policy page, and, in particular, the WP:BURDEN section:

The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation

For those of you who look into it, I think it can only improve Wikipedia, and how much editors take WP:V as seriously as it needs to be taken. Thank you. --Born2cycle (talk) 06:07, 4 June 2009 (UTC)

I left a comment. [7] SlimVirgin ^{talk|contribs} 07:12, 4 June 2009 (UTC)

Excellent opportunity

There is an excellent opportunity currently available for those of you believe it's important to emphasize the importance of this policy to editors who seem to not understand or fully appreciate it. There has been an Rfc filed against me because of my stubborn exuberance in trying to convince a handful of editors at Talk:lane splitting that the WP:burden is on them to provide sources for the claims that add and support in articles, not on the challenger.

If you would weigh in on this matter, here, and in particular whether I was out of line for making this issue so important, I would appreciate it. Thanks. --Born2cycle (talk) 20:11, 4 June 2009 (UTC)

Inline citations

I'd like to remove the footnote after the sentence: "The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation."

The footnote says: "When content in Wikipedia requires direct substantiation, the established convention is to provide an inline citation to the supporting references. The rationale is that this provides the most direct means to verify whether the content is consistent with the references. Alternative conventions exist, and are acceptable if they provide clear and precise attribution for the article's assertions, but inline citations are considered 'best practice' under this rationale. For more details, please consult Wikipedia:Citing sources#How to cite sources (my bold)."

There really aren't any alternatives to inline citations that are acceptable now on Wikipedia, and there's no reason not to use an inline citation (which can be footnote style or a Harvard reference, usually the former).

Does anyone object to its removal? SlimVirgin ^{talk|contribs} 07:18, 4 June 2009 (UTC)

I concur, it does seem the case that only inline citations are acceptable and the text should reflect that. Unomi (talk) 09:45, 4 June 2009 (UTC)

I agree, requiring inline citations has consensus. It used to be OK to just include a bibliography at the bottom of the article... no more. Blueboar (talk) 13:33, 4 June 2009 (UTC)

Not really an objection, but the only potential caveat I see is that if some challenged idea that is substantiated by sources is repeated, perhaps worded slightly differently, or implied multiple times in an article for some reason, there should not be a requirement for an inline citation at each instance. SlimVirgin, thanks, by the way, for your support above. For example, in that Lane splitting situation, the challenged statement in question is made in the intro of the article. But if it was substantiated and cited there, would it be a problem with it repeated later in the article without citation? --Born2cycle (talk) 14:02, 4 June 2009 (UTC)

It's best to repeat the citation if it's something likely to be challenged. It's very easy to do that; you just name the reference the first time you write it out e.g. <refname=Smith>Smith, John. "Blah Blah," ''The New York Times'', June 4, 2009.</ref> and thereafter you just type <refname=Smith/> and that reproduces the citation. SlimVirgin ^{talk|contribs} 14:08, 4 June 2009 (UTC)

In an ideal world where all editors of an article is fully cognizant of all sources involved it should not be necessary, in reality even putting in an html comment referring editors to where the source is first cited to support a related passage is not always respected. This unfortunately engenders a situation where at one hand we wish to avoid a 'sea of blue' and 'over citation' but at the same time each statement should be properly attributed each time it is used in the article. I know of no easy solution to this but would welcome if we were able to address it. Unomi (talk) 14:16, 4 June 2009 (UTC)

Inline citation is not clearly defined. Some people are inclined to read everything they see narrowly, and will use this to object to anything except an endnote created with <ref></ref> tags. I would suggest modifying the footnote to say "inline citations include footnotes, Parenthetical referencing, or full description of the source in the running text". Jc3s5h at 14:22 UT June 4, 2009 [Signature generation does not seem to be working.]

I concur with the original text; Jc3s5 describes two alternatives, but there are others: chiefly, whole articles, or sections of articles, may be drawn from a single page or couple of pages, in a single source. Most of these will be stubs, or articles of dubious notability, but not all; there was a long period in which almost all of Homotopy groups of spheres was drawn from a single source, except for a couple of sentences and a table, all duly noted. The source had a caveat in the bibliography This entire article derived from pp. 530-1 (or whatever it was) except as noted.

Policy pages should prescribe what end we need to reach, not how to get there. So yes, I object to the removal. Septentrionalis PMAnderson [15:02?], 2 June 2009 (UTC)

I don't quite follow your point, PM. It would be a trivial matter to add an inline citation to an article that was based entirely on one source; for example, at the end of each paragraph. Also, your second point—that policy pages shouldn't tell us how to reach our goals—that is exactly what policy pages do e.g. use good sources, don't engage in OR; don't be rude etc. SlimVirgin ^{talk|contribs} 16:55, 4 June 2009 (UTC)

There are two points here.

It would be relatively easy to add footnotes to a one-source article; but there is absolutely no reason to do so. It doesn't serve the reader (provided the source is indicated), and it doesn't affect the verifiability of the article. It would therefore be instruction creep to require it.
Don't be rude is what we want to achieve, and belongs in policy; it's a goal, on which we agree. A list of forbidden words doesn't belong in WP:CIVIL; although we would probably come closer to agreeing on one than in most questions of how to do something. So here; we agree that the source of assertions that are challenged, or are likely to be, should be clearly and precisely indicated. How this is done is not the proper concern of a policy page. Septentrionalis PMAnderson 18:08, 4 June 2009 (UTC)

Regarding your first point, there's no reason not to add a reference in the form of an inline citation when there's only one source. If you have the source and intend to add it anyway, it so easy to do it as an inline citation that I don't think there could be a good argument in favour of not doing it. Your anti-instruction-creep instinct is commendable, but it's rising to the level of ideology here, I think. :) SlimVirgin ^{talk|contribs} 18:14, 4 June 2009 (UTC)

There is a reason not to add an inline citation when there is only one source: I don't feel like it. If I feel like explaining how the source supports the claims of the article in some other fashion, that's my prerogative. Not only is it instruction creep, but it is putting an instruction in a place where no one who understands the nature of Wikipedia policies would expect to find it. (I think this is the 3rd time I've agreed with Septentrionalis this week; how bizarre!) --Jc3s5h (talk) 18:22, 4 June 2009 (UTC)

And, as Jc3s5h already said, there are other perfectly reasonable practices which a tight reading of the proposed language would prohibit. There are alternatives; they should not be ruled out by policy. (It may be that I am overestimating the chances of obstructive idiocy; but then I come here from an ArbCom case....) Septentrionalis PMAnderson 18:25, 4 June 2009 (UTC)

Let's talk about a specific article where this actually came up when it was featured and then again in November in its FA review: Great Fire of London. Most of the details about this event that we know now come from two notable diaries: that of Samuel Pepys and that of John Evelyn. Like the article, these sources are ordered chronologically, so it is no problem to look something up. Because of the great number of editions it makes little sense to quote page numbers. The article's solution is described in footnotes 14 and 34. The FA review was clearly based on a misunderstanding.

I agree that it would make some sense to ignore this example and create a prescriptive rule that forces us to repeat the same footnote ever again in this article. This would avoid this kind of misunderstanding. But it seems that this would not be descriptive policy, and I don't think it would be a good thing. If there is a loophole to be closed it's probably not in this article, so perhaps we should point to this article as an example of acceptable use of non-inline citations.

The main problem that I see with non-inline citations is that it starts with one global reference, then somebody adds a second, overlapping one. Then people add more material that may or may not be covered in one or two of these references. And then suddenly it's very hard to convert to inline citations. This article's citation system doesn't have this problem. --Hans Adler (talk) 18:57, 4 June 2009 (UTC)

It does have a lot of inline citations, though—it wouldn't have been awarded FA status otherwise—so I wonder whether I'm missing your point. SlimVirgin ^{talk|contribs} 19:05, 4 June 2009 (UTC)

There are entire paragraphs that don't have footnotes because they only talk about Samuel Pepys' or John Evelyn's observations on specified days. --Hans Adler (talk) 19:38, 4 June 2009 (UTC)

Integrating WP:V into WP:EP

There is an interesting discussion on how best to integrate (and even whether to integrate) WP:V at WT:Editing policy#What is the problem?. Please drop by and opine. Blueboar (talk) 17:29, 4 June 2009 (UTC)

Question about burden and "anti-material"

I don't see this issue addressed on the policy page...

WP:BURDEN states, rather unambiguously: "The burden of evidence lies with the editor who adds or restores material". The implication is, "no matter what the material is".

What if the "material" is "anti-material"? To put this point in the most general terms, consider an assertion X which can only be true or false. Clearly we need a source to say either "X is true" or "X is false", but if there are no sources that substantiate either assertion, do we need a reliable authoritative source that essentially says "the truth of X is unknown" in order for the article to say it? Since the burden is clearly on anyone who inserts material that says X is true or X is false, how can the burden also be on anyone who inserts material that says it is unknown whether X is true or false?

Now, a specific example. One of the main points of contention at the Lane splitting article is whether lane splitting is definitely illegal in all 50 states besides California (there is no dispute about the fact that safe and prudent lane splitting is at least sometimes, probably usually, and perhaps even always, legal in California). In some states there is no law that clearly prohibits lane splitting, but there is also no state authority that has stated (like the CHP in California did) that "safe and prudent" lane splitting is legal. It seem to me that anyone adding the statement that lane splitting clearly is legal, or clearly is always not legal, in those states, has the burden to provide a source. But is there any burden to provide a source for a statement like, "It is unclear whether lane splitting is ever legal in these states". I mean, can't we say, without citation, "the editors of this article cannot find reliable sources that clearly establish the legality of lane splitting in these states." But wouldn't the previous statement really convey the same thing, without being unnecessarily cumbersome?

In short, is it acceptable to add material to an article that conveys the absence of knowledge about a given relevant matter, without having to provide a source that states that knowledge is absent? I mean, isn't the appropriate way to challenge a claim of absence of knowledge by citing a source for that knowledge? If none of the sources tell us how many games Willie Mays played in some year, can't we say the number of games he played in that year is unknown, without citing a source that state that statistic is unknown?

I think this issue should be addressed under WP:BURDEN, by either clarifying that even for lack of knowledge material the burden is on the one inserting the material, or by establishing that lack of knowledge material is an exception, and explaining why. Thoughts/comments? --Born2cycle (talk) 06:59, 5 June 2009 (UTC)

It is a substantial research project (which should be published somewhere other than Wikipedia) to figure out which, if any, states forbid a particular act. The act may be forbidden or allowed under many different sections of the law, or perhaps be governed by an administrative rule or court precedent. If it seems very likely that the act is illegal everywhere, and there is no large financial incentive to find a state where it is legal, then most likely no one will be inclined to undertake the project. But just because no one has proven the act is illegal in every state does not mean the legality of the act is unclear; it may be quite clear to everyone that the act is illegal. That is exactly why the research was never undertaken, because the project is unlikely to yield an interesting result.

The best course of action when something is most likely true, but it is too burdensome to check with each entity with authority over the matter, is to find a good source that makes an appropriate generalization and quote it.

Another areas where this sort of thing comes up is dates. Does the US really use the Gregorian calendar. Who's in charge, the feds or the states? Has anyone ever checked the laws in every state to see if they have formally adopted it? Probably not. But would it be fair to say "it is unclear whether the Gregorian calendar has been adopted throughout the United States." Of course not. Just find a good quote. --Jc3s5h (talk) 07:26, 5 June 2009 (UTC)

"...is it acceptable to add material to an article that conveys the absence of knowledge about a given relevant matter, without having to provide a source that states that knowledge is absent?"... Yes, it is acceptable to add such a statement. However if that statement is subsequently challenged, then a source has to be found (or at least a consensus on whether a source is needed must be reached on the talk page). This is true for all information on Wikipedia, and statements of "it is unknown" are no different. 14:24, 5 June 2009 (UTC)

That makes sense. I'll wait a few days to make sure there is consensus on this point on this page, but if it appears there is, I'll go ahead and add something to this effect (incorporating points made by both of you) under WP:BURDEN. Thanks. --Born2cycle (talk) 14:51, 5 June 2009 (UTC)

Done [8]. --Born2cycle (talk) 03:08, 11 June 2009 (UTC)

Notice of related RFC

Please see current RFC related to this policy at Template talk:Unreferenced#RFC: Should the template employ "unverifiable" or "unverified"?.--Fuhghettaboutit (talk) 12:16, 6 June 2009 (UTC)

A fascinating discussion. --Born2cycle (talk) 19:41, 6 June 2009 (UTC)

Wikipedia:Articles for deletion/Sam Blacketer controversy

There have been suggestions at Wikipedia:Articles for deletion/Sam Blacketer controversy that WP:V is overly restrictive when it comes to matters close to Wikipedia - where applying proper sourcing can actually resulting in providing incorrect content. Could we find a way to circumvent that by altering WP:V in a way that would allow our internal sources (ie ARBCOM/signpost) to take precedence over lets say the press when it comes to internal matters. Agathoclea (talk) 17:36, 9 June 2009 (UTC)

Or we could simply delete the article, which would remove the need to amend WP:V to deal with it. — Carl (CBM · talk) 02:40, 10 June 2009 (UTC)

V and N vs OSE.

Which is more important: wp:v and Wikipedia:Notability's requirement for sources, or the precedent set by other stuff existing without sources?: Please comment at Wikipedia:Deletion review/Log/2009 June 6#List of MXC episodes. -- Jeandré (talk), 2009-06-10t11:39z

WP:V is a core policy... Other stuff exists is not a valid argument. Blueboar (talk) 14:11, 10 June 2009 (UTC)

Truth about threshold

I'm also concerned about that VNT interpretation. This realy mean to me : "As soon as a reliable source make a mistake, we at Wikipedia make every thing that is possible to present it as a fact (Even when all of us know it is false, it doesn't matter, realy). So we make sure we are less reliable then the less reliable of our sources". I started reading this pretty long archive to understand why this still in the policy. I didn't finished yet but, I retain two distincts points :

This is pretty useful against people who want THE truth to be told to anyone. (pov pusher).
The fact that truth is not enought : it as to be relevant.

But I'm more concerned about sounds peoples that might not get those subtilities, and get to that interpretation : include as much false statements that matches those criterion : "..." . Which, I hope (realy hope) it's not the true goal. So I come up for this draft, especialy written for sounds persons. my goal being : "Telling truth about threshold".

The threshold for inclusion in Wikipedia is verifiability and relevancy. Moreover, not only those statement can be included, but users should strives to include all of them according to their relevancies (see neutrality of point of vue). Sourcing a statement is barely always the best way to demonstrate the verifiability and relevancy of a statement (as much as we can), and so sources are relevants themselves. The more a common reader (note:those readers should know the basic of the subjet[so he shoudn't doubt about merely everything] but, still, he should learn something new) might doubt about a statement, the more sources become relevants and so, sources quickly get more relevant then the statements they support and should then be added first. Such statements without sources should be deleted reasonnably quikly (more or less fast according to the size of that doubt). This is specially true for living persons and others actuality subjects (where unverifiable or unrelevant statement can causes serious wrong).

I know, talking about relevancy of a source is quite new, but hell, sources are SO relevants :D . hope my spelling is not too bad. Again, this is just a draft ;) . Iluvalar (talk) 14:40, 9 June 2009 (UTC)

There is nothing in WP:V that requires any information to be included in an article, even if it is sourced. In other words, the ability to give a reference is necessary, but not sufficient, to include something in an article. We routinely remove claims that are incorrect even when they are sourced. For example, even if some seemingly reliable source said the atomic number of hydrogen was 15, we would not include this in the article on hydrogen. A fundamental part of our mission to be an encyclopedia is that our articles must present facts in an accurate way.

The wording of WP:V is a consequence of articles in areas where it is more difficult to ascertain the "truth" of the matter, for example articles about tenets of religious faiths. Unfortunately, some policies are written in a way to make them useful on the few controversial articles at the expense of being inaccurate about best practices for the vast majority of articles where there is no real disagreement about the facts at hand.

I agree that it would be nice if the policy were more explicit that verifiability is not a sufficient condition for inclusion, but I don't think it is likely to happen, for various reasons that wuold lead us off topic here. — Carl (CBM · talk) 14:54, 9 June 2009 (UTC)

I mostly agree with you. But...

"A fundamental part of our mission to be an encyclopedia is that our articles must present facts in an accurate way. " yeah and that "fundamental" is usualy called "Verifiability" and if the wording of this page was just close as clear as your statement I would not be here right now.
"There is nothing in WP:V that requires any information to be included in an article" indeed but WP:NPoV does. And both of them clearly state they should be intepreted jointly. So the way I read this page, I feel like it want to force me to include any single mistake I could find without even caring.

Please old users, forgive what you know about "the type and quality of material that is acceptable in Wikipedia articles" (I know you know) and try to interpret jointly WP:NPOV, WP:V and WP:NOR from scratch. You'll probably understand what I want to mean. You'll see what ought to be said that actualy is not. Iluvalar (talk) 00:27, 10 June 2009 (UTC)

Yes, you're right, but it's too hard to fix the wording at the moment. — Carl (CBM · talk) 02:32, 10 June 2009 (UTC)

So there's one last thing to do : Just sit and wait the next barnraising. Iluvalar (talk) 15:07, 11 June 2009 (UTC)

Reliability of Sources - Newspapers are often not reliable sources

I have noticed that speculative claims are often backed up by a reference to a single newspaper article, despite there being no real evidence for the claim.

What you tell me three times is true: I propose that all references where a newspaper article is cited should require three consistent sources.

At the moment, citations fall into (at least) three groups:

Those where a news-piece is referenced for the hard-to-find statistics it contains: either these are adequately referenced in the news-piece, in which case the citations should be of the original material, or they are not, in which case the information is not reliable without further confirmation. Newspapers are not subject to a review process, so a bald assertion of numbers by a journalist carries little weight.

Those where a news-piece has been referenced for something which is common knowledge - e.g. a football score, or an election result: if the matter is common knowledge, it will have been reported by more than one source, and usually a reliable primary one; if not, is it suitable for Wikipedia anyway?

Those where a news-piece has been referenced for the opinion it contains: this is more problematical, but surely no opinion that is not the consensus should be used in a Wikipedia article? If it is consensus, congruent sources can usually be found, or if it is proper expert opinion, it has probably been properly published as a book or in a peer-reviewed journal.

Any thoughts? 92.234.8.173 (talk) 00:34, 10 June 2009 (UTC) Josh

Nobody cares? Does that mean I should go ahead and make the change without further input? Surely not... 92.234.8.173 (talk) 19:37, 10 June 2009 (UTC) Josh

I find this not necessary. Newspapers from what I understand have to be reviewed by editors. What about articles which are from adapted from books of a peer reviewed journal. Does that mean that those should not be included? Plus, newspapers make it easier for events or other things to be verified. Also, I think this question needs to be in a different place for actual attention. Try the talk page on WP:Reliable Sources as it seems more appropriate there. Deavenger (talk) 20:09, 10 June 2009 (UTC)

The whole concept of this thread should be squashed. By only accepting sources higher in quality than newspapers (peer-reviewed journals, perhaps?) Wikipedia editors will have to be affiliated with major universities and the topics will have to be limited to the kind of subjects studied at major universities. Also, note that "Deep Throat" couldn't be mentioned in Wikipedia until someone published an academic paper about it, since he gave his interviews to a newspaper. --Jc3s5h (talk) 21:36, 10 June 2009 (UTC)

I think you misunderstood the suggestion, Jc3s5h - I'm not saying that newspapers shouldn't be used, but that they shouldn't be used alone as evidence of something that is not common knowledge. Deep Throat was picked up and reported by all the US papers, and papers around the world. It would be easy to find three sources saying the same thing. In the specific area which has been annoying me lately - footballers' wages and transfer fees - the figures printed in papers are just the guesses of journalists, in a lot of cases. It is very hard to find three papers printing the same figure, usually.

Deavenger, thanks for your input, but in the case of an article adapted from a peer-reviewed journal, wouldn't the better reference be the original? As for moving this - should I just put a link there to this page, or delete this and repost? Thanks. 92.234.8.173 (talk) 23:04, 10 June 2009 (UTC) Josh

Or we could just use the newspaper where it can be viewed by everybody. Plus, some newspapers are known for being reliable sources. For example, footballer wages or whatever, there a good chance that football player and the organization will tell a newspaper how much they're getting paid. We already have a thing of how certain articles such as sports, or other articles should have academic sources like books, peer-reviewed journals, etc. However, newspapers can be used, but it is preferred that other sources like peer-reviewed journals be used also, depending on the subject. Deavenger (talk) 00:22, 11 June 2009 (UTC)

Hmmn. I understand what you're saying, with regard to deficiencies in my suggestion, but I think the problem is still there and needs dealing with in some way - possibly by limiting the application of the suggestion to specific circumstances. As we're already talking about footballers' wages, let's stick with that; the problem is that although some wages are released, some are not. Sources that would otherwise be 100% reliable (on a political story, for instance) are known for printing every football rumour in existence, regardless of veracity.

I can see an argument that this kind of speculation isn't for Wikipedia anyway, but it seems to be an ever-present in articles on the subject. Considering they're all biographies of living persons, I assume Wikipedia wants to be especially strict on accuracy? 92.234.8.173 (talk) 00:58, 12 June 2009 (UTC) Josh

I don't think the number of newspapers reporting a particular claim is a good barometer of the story's reliability. Once one vaguely-reputable paper runs a juicy story, the others are quite likely to assume that it's been fact-checked and rerun it without much further checking - true or not, it helps them sell papers, and if it turns out to be wrong they can always blame whoever published it first. OTOH, an accurate story published by one paper might not get picked up anywhere else simply because it's of limited interest in that medium. Plus, there's the complication of determining what counts as 'different newspapers' - often several papers produced by the same publisher will share content.

It's hard, and IMHO inadvisable, to come up with blanket rules that cover all scenarios. But IMHO, if two sources contradict one another, that's a pretty good indication that at least one of them is unreliable (or, alternately, that one of them has been misunderstood). In this particular case, I'd be inclined to report the range, with both ends cited - "John Smith accepted a contract reported at between $2m (cite source 1) and $5m (cite source 2)". This tells people what claims have been made, and the range should make it obvious that these figures are fuzzy. --GenericBob (talk) 05:44, 12 June 2009 (UTC)

Scholarly Sources/Self-Published Sources/Established Experts

Given that a self-published article - posted on a blog for example - by an established, cited expert in the relevant field qualifies as a reliable source, surely an article published in a journal by an established, cited expert in the relevant field would qualify as reliable, even though the journal itself was of questionable reliability. The only argument against this I can come up with is that an established expert ought to be able to publish his/her work in a reliable journal. Any thoughts?KD Tries Again (talk) 18:44, 10 June 2009 (UTC)KD Tries Again

I agree with most of what you say. Self-published expert opinion is no more reliable than material published in a non-reputable journal. I'm not sure I agree with the conclusion, though. I think you've demonstrated why the first category is unreliable, rather than that the second category is reliable. 92.234.8.173 (talk) 19:36, 10 June 2009 (UTC) Josh

I think it has to be examined on a case-by-case basis. Why couldn't the expert get published in a reputable journal? Maybe because it was a minor correction or clarification of previous work, and wasn't important enough for a journal article. Why publish in a non-reputable journal? That's a harder question, and would have to be examined for any particular case. --Jc3s5h (talk) 21:39, 10 June 2009 (UTC)

What do you mean by self published? [9], [10] If I understand this correct are self published (If I'm wrong, please correct me), but both of these books are written by some of the top people in their field, and have been used as references in other academic work or referred to other academics. Deavenger (talk) 00:25, 11 June 2009 (UTC)

Deavender, no, Friedman's book is published by Doubleday. I didn't check the other one. --Jc3s5h (talk) 02:01, 11 June 2009 (UTC)

The other one is done by Norton and Co, so I guess those don't count as self published. Deavenger (talk) 02:09, 11 June 2009 (UTC)

I just meant self-published as defined in this article. The example I have in mind is an article on a topic which few reputable journals care about, but on which the author is unquestionably an expert.KD Tries Again (talk) 14:27, 11 June 2009 (UTC)KD Tries Again

Wikipedia:IAR - Does that answer your question? 92.234.8.173 (talk) 00:49, 12 June 2009 (UTC) Josh

Questions about WP:V in biography David Sheffield Bell

Couple of questions.

Is the following material properly sourced? One source is his website and the other a professional organization he belongs to:
He is a graduate of Harvard University, class of 1967, with an AB degree in English Literature. He graduated in 1971 from Boston University with an MD degree, and in 1976 completed post-doctoral training in pediatrics including Pediatric Behavior and Developmental Disorders. [11][12]

Also, is the source for the following material considered a primary source or secondary source, and does it meet WP:RS for his bio? (the source does state the material)
Bell also served as Clinical Associate Professor of Pediatrics at the State University of New York at Buffalo, in Buffalo, New York. [13]
Thanks for any help. Ward20 (talk) 04:53, 11 June 2009 (UTC)

Point number 5 at WP:SELFPUB says, those sources are oK when "the article is not based primarily on such sources." David Sheffield Bell is based on primary and selfpub sources, it had a notability tag since January and Source tag since February. When no reliable third-party sources about a person, like a newspaper article about them or a book biography about them it is hard to do an article on them w/o original research, Google Scholar has 629,000 articles by "DS Bell" it is a common name. RetroS1mone talk 18:33, 11 June 2009 (UTC)

RetroS1mone is an involved editor that has removed primary and secondary sources from the article.[14] I would appreciate an uninvolved editor's opinion on my original questions. Thanks. Ward20 (talk) 19:25, 11 June 2009 (UTC)

On the first para, I think the sourcing given meets WP:V, and indeed the first link alone (his website) would be adequate - unless there's some reason to doubt his credentials, it's not claiming anything very remarkable. So as long as there are also suitable non-self-published sources in the article that establish notability, I'd think it would be OK to use this SPS to fill in the basic biographical details. --GenericBob (talk) 05:56, 12 June 2009 (UTC)