Wikipedia talk:Verifiability/Archive 35

Archive 30

←

Archive 33

Archive 34

Archive 35

Archive 36

Archive 37

→

Archive 40

Content notability

One debate I have seen in WP is that there is no content notability guidelines. Any time someone tries to bring up WP:NOTABILITY to show that suggested content is not worthy of inclusion in WP, someone brings up the fact that Notability only applies to articles, not material. But I know of no specific policy that mentions that fact. I feel we should add something that states that inclusion of content has a different standard than the creation of an article on a topic, and it seems that Verifiability is the closest content guideline for that aim. Angryapathy (talk) 14:38, 24 September 2009 (UTC)

The idea of judging content to decide what to include is exactly the point of WP:NPOV. — Carl (CBM · talk) 14:44, 24 September 2009 (UTC)

• The trouble with any proposal on "content notability" is that it is extremely undesirable to cut reliably-sourced material from Wikipedia. If it's reliably-sourced, it merits inclusion, and if it isn't, it doesn't. WP:NPOV is often but not always a matter of phrasing and emphasis rather than exactly what content to include.—S Marshall ^Talk/_Cont 14:51, 24 September 2009 (UTC)

No, "if it's reliably-sourced, it merits inclusion" is not true. The US census rolls for 1930 are a reliable source for who lived in Fort Worth, Texas in 1930, but it would be inappropriate to trancribe the name of every resident into the "Fort Worth, Texas" article. --Jc3s5h (talk) 16:32, 24 September 2009 (UTC)

AFAIK, the best we have is WP:WEIGHT which is a subsection of WP:NPOV. A Quest For Knowledge (talk) 12:26, 25 September 2009 (UTC)

Jc3s5h, I didn't say "every single detail from every reliable source should be included". I'm not denying that Wikipedia should be written in summary style. I'm saying that there should be a presumption against cutting reliably-sourced material from the encyclopaedia.

In your example, if some idiot added the name of every resident, you'd summarise that by saying "According to the US census rolls, (x number of) people lived in Fort Worth in 1930."—S Marshall ^Talk/_Cont 14:06, 25 September 2009 (UTC)

This thread is about content notability, which means whether content is important enough to include or not. Presumably if one is considering eliminating a claim because it isn't important, one has already decided that less subjective reasons for exclusion, such as lack of a reliable source, don't apply. If one is writing a policy and says, in a section about how important material must be to merit inclusion, "if it's reliably-sourced, it merits inclusion" then, according to Marshall's policy, the full list of 1930 Fort Worth inhabitants may be included. That's why including "if it's reliably-sourced, it merits inclusion" in a policy about content notability would be a disaster. --Jc3s5h (talk) 15:49, 25 September 2009 (UTC)

You aren't responding to what I said. You're addressing a straw man.

I said "it is extremely undesirable to cut reliably-sourced material from Wikipedia". That's what I meant.

You can move it, you can merge it, you can trim it, you can summarise it, you can cut it from huge swathes of content down to a paragraph, a sentence or even a couple of words. But only in exceptional circumstances (such as BLP violations) should you remove the link to the reliable source.—S Marshall ^Talk/_Cont 17:12, 25 September 2009 (UTC)

I have no patience for editors who expand and reword their ideas to prove they are always right. You are just wrong on this one, and I'm done discussing it. Any attempt to put this idea into policy will result on me calling upon others to take any necessary steps to remove it. --Jc3s5h (talk) 17:23, 25 September 2009 (UTC)

I think the main point Marshall is saying is that the "if it's reliably sourced, it merits inclusion," is pretty much the general unsaid rule of WP. The qualification to that is there are some rules that will cut down on which information is disallowed. Your example of the census data would not be allowed under WP:NOT, specifically WP:NOTLINK. But this argument here is a perfect example of why something should be included to address this issue somewhere. And I disagree with Carl that WP:NPOV addresses this issue, as NPOV refers to when two sets of information contradict or are in conflict with one another. Angryapathy (talk) 17:50, 25 September 2009 (UTC)

WP:NPOV says, "All Wikipedia articles and other encyclopedic content must be written from a neutral point of view, representing fairly, and as far as possible without bias, all significant views that have been published by reliable sources. " The key word there is significant. The key issue in "content notability" is undue weight: for some content, even though it is reliably sourced, the due weight is zero, and we should not include it in our articles. This includes not only tiny-fringe opinions, but also hyper-esoteric facts, detailed census data, etc. Of course editorial judgment is required to determine whether to include content in an article. The question, according to NPOV, should be framed in terms of how much weight to give each piece of information, with some content not even warranting a mention. — Carl (CBM · talk) 20:37, 25 September 2009 (UTC)

See, that quote would be applicable if it didn't have the word "view". A view is an opinion or one side of an argument. That is very different from information. The historical stock price for a company is not a view, but can be reliably sourced. So again, that policy still does not say "any significant content that is reliably sourced may be added to Wikipedia," only significant views. There is a still a hole is the guidelines where many arguments are fought about what makes content notable enough for WP. Angryapathy (talk) 20:54, 25 September 2009 (UTC)

Actually, contrary to your assertion, NPOV does say that. WP:WEIGHT notes: "Undue weight applies to more than just viewpoints. Just as giving undue weight to a viewpoint is not neutral, so is giving undue weight to other verifiable and sourced statements. An article should not give undue weight to any aspects of the subject, but should strive to treat each aspect with a weight appropriate to its significance to the subject." This text was there at least a full year ago,[1] so it is not a recent addition or new interpretation of NPOV. Vassyana (talk) 21:03, 25 September 2009 (UTC)

NPOV covers not only opinions, but all material in articles. It is possible that whoever wrote NPOV was using the widely-held convention that "all writing is argument". So I am saying that, if anything needs to be clarified, it is the section on "undue weight" in WP:NPOV. — Carl (CBM · talk) 21:00, 25 September 2009 (UTC)

Carl mentions one other key factor in determining content notability... editorial judgment. Whether a specific reliably sourced fact or viewpoint should be mentioned in a specific article is ultimately up to the editors of the article. It isn't a policy issue. Blueboar (talk) 21:44, 25 September 2009 (UTC)

Tertiary sources

I have come across numerous articles where contentious information is supported by tertiary sources. When information is supported by reliable secondary sources, it is possible to trace the sources and also to find academic criticism, neither of which is possible with tertiary sources. The problemn usually arises in political articles where editors use tertiary sources to define terms like "liberal", "left-wing", "social liberalism", etc. where definitions are disputed. It would be helpful to have a guideline that tertiary sources should not be used when they disagree with reliable secondary sources. The Four Deuces (talk) 18:27, 24 September 2009 (UTC)

See WP:PSTS. Angryapathy (talk) 18:31, 24 September 2009 (UTC)

Removing unsourced material

• ... how quickly this should happen depends on the material in question and the overall state of the article. Editors might object if you remove material without giving them enough time to provide references, especially in an underdeveloped article. It has always been good practice to make reasonable efforts to find sources oneself that support such material, and cite them.

1. Must every unsourced statement be removed on sight? If not, how long may a {{fact}} tag stay up?

   I ask this, because one editor has been following me around and methodically removing every word I write on a particular topic. Meanwhile, I've seen fact tags stay up for over a year in articles unrelated to this topic.

2. Should this other editor be making reasonable efforts himself to find sources that support such material? Or does he have the right to erase every statement I make until and unless I can find an acceptable secondary source?
3. Must all information about a church be provided by sources other than that church?

   For example, would a Catholic Church source be acceptable about the history, doctrine, organization or current activities of Catholics?

Please give me practical, in-depth answers to these questions. Thank you. --Uncle Ed (talk) 19:28, 27 September 2009 (UTC)

1. It depends on the assertion. If someone writes that the moon is made of cheese I'd expect a source immediately. But I usually wait at least a couple of weeks or a month if the assertion isn't extraordinary or controversial.

2. The burden is on the editor adding the material. While it's polite and shows good faith when an editor seeks out sources for an unsourced assertion, it isn't a requirement.

3. The Catholic Church has not always been a reliable source for its own activities. New religious movements are even more problematic as sources. I believe that official church sources should be normally treated as primary sources. They'd be fine sources for the exact wording of a prayer, for example. But articles should be based mainly on third-party sources.   Will Beback  talk  20:19, 27 September 2009 (UTC)

• Will, in my haste, I overlooked your reply. Thank you for it; it seems quite sensible, especially #1 and #2. I need further guidance on #3, as I am a member of a new religious movement and get most of my information about it, from it. I'm often astonished to find that other Wikipedians would question a Unification Church source about church theology or organization. For example, the church in America got a new leader (the founder's own daughter), yet when I quoted a church document as a source this source and the info itself were both immediately challenged.

• Can't members of the Unification Church of America be trusted to know who their own church president is? --Uncle Ed (talk) 14:12, 28 September 2009 (UTC)

Uncle Ed, I have never examined any of your edits, so understand this is hypothetical. If you were to repeatedly add the same, or similar, claims to various articles, had been repeatedly asked for sources, and never found a source, then you should just stop adding that claim or group of claims. The policy writers never really thought of the phenomenon of the same unsupported claim being added over and over to many different articles, but I think it is no different in spirit to adding the same unsupported claim to a single article over and over. That's called edit warring. --Jc3s5h (talk) 00:42, 28 September 2009 (UTC)

That's not what I'm talking about; obviously, as a Wikipedian of long standing, I know that it would be counterproductive to keep inserting the same challenged info - if I also failed to locate sources.

What I'm asking about, rather, is one particular user's practice of immediately removing information which (to me) seems not to require a source. Instead of following WP:BURDEN and adding a fact tag and giving me sufficient time to find sources, he just (1) removes the information and/or (2) lectures me on supposedly violating WP:BURDEN.

This user is holding me to a higher standard than anyone else at Wikipedia must follow, and the effect is to squelch WP:TEAMWORK.

So I ask again, is there a rule that requires each statement inserted in an article to be supported with a source? And is it within policy for a user to routinely delete all statements which lack a source? --Uncle Ed (talk) 14:05, 28 September 2009 (UTC)

This user Ed Poor (talk · contribs) seems to have a pattern of creating wholly unsourced articles [2], and also adding completely unsourced material to articles, including even to WP:BLP articles [3]. He may be seeking comments here to vindicate this inappropriate behavior. Cirt (talk) 14:07, 28 September 2009 (UTC)

Why not just add a source? If you can't find a source, don't add the information. WP:BURDEN puts the work on you to source what you've added. In cases of BLP-related articles and info, you absolutely should not be adding unsourced material. Any such additions should be immediately reverted. If the claim is contentious, whether BLP or not, it should be removed immediately. Lara 14:12, 28 September 2009 (UTC)

The problem is that I am writing about controversial topics - not biographies per se. Now, in the old days of Wikipedia, I never had any trouble with balancing articles by adding positive information when the article consisted primarily of negative information. This is in accordance with NPOV, which at that time urged writers to adhere to the highest standard, i.e., that no one reading the article (or even a particular edit) would be able to guess which side of the controversy the writer personally favored.

I am not seeking vindication (or I'd be in conflict resolution). I am seeking clarification of the new policies and guidelines. If things have changed (or if as is theoretically possible I simply am misremembering the old policies and practices in place before 2005), then I need guidance. Now it looks to me like one user is trying to remove all important and/or positive information about the Unification Movement and substituting his own personal (and unattributed viewpoints as if they were indisputable facts.

I on the other hand am trying to present a balanced view, conscious of the handicap of trying hard not to violate the conflict-of-interest rules. I could use some help and guidance. If no one wants to help me find sources for the valuable information I have, does this mean I have to do all the research myself or just keep out of it? I don't see this standard being applied to other articles, but if articles touching on Rev. Moon and his movement must follow this standard, I will follow it; but I expect all others to follow it as well.

Before I embark on a campaign of removing all poorly-sourced derogatory information and unsourced speculation and OR from all these articles, I'd like a clarification of policy.

I was an admin and bureacrat, and one of the most trusted members of this community. I have not changed, but if the community has changed I promise to change with it. Please tell me what the new rules are. (Or at least show me specific cases where I have messed up and explain what I should do differently.) The constant barrage of false accusations by one user - who has been repeatedly scolded by others for this - is not helping. I need guidance from experienced editors in good standing. --Uncle Ed (talk) 14:29, 28 September 2009 (UTC)

I think that trying to balance POV with unsourced statements is a bad way to go. If there are NPOV problems, unsourced statements should not be used as the solution. Angryapathy (talk) 16:59, 28 September 2009 (UTC)

Agree fully with Angryapathy (talk · contribs), but I would add in addition to unsourced statements to balance POV, statements from primary sources, and sources affiliated directly with the subject, should not be used - especially when an abundance of independent reliable secondary sources are available! Cirt (talk) 17:01, 28 September 2009 (UTC)

There is no need for an article to have a balance of positive and negative statements; what is required is that the article present a fair picture of what the reliable sources say. If all the reliable sources about X say he is an asshole with no redeeming qualities, then Wikipedia should say X is an asshole with no redeeming qualities. --Jc3s5h (talk) 17:44, 28 September 2009 (UTC)

First principles derivations do not need citation

I argued about that some time ago here or on the OR page. I was told to invoke WP:IAR if I were to get into trouble on the articles I've been working on, because it was unlikely that wiki policies would be changed. Now, User:OMCV is threatening to put the article Relations between heat capacities on AFD, simply becaue of a ciation issue, not because he has any argument about the content.

Now, this article is a sub article referred from a few articles on heat capacity. Interested readers can read about the technical derivations from first principles. So, the very existence of this article enhances verification. If you know the necessary math to understand the derivation but have not studied thermodynamics, you would not be able to easily verify the statements via the citations in the main article. You would have to get hold of the books on thermodynamics and then you would have to work through the book, studying thermodynamics in the process. So, it is not a straightforward verification at all.

Because of my background in theoretical physics and my teaching exprience, I am able to give self contained derivations where possible. The statements themselves can be cited, but the derivatins cannot. The theory behind the derivations can, of course, also be cited. The reason why the derivations cannot be cited, is because I've adaped them to make them suitable for incusion in here in wikipedia. Derivations in books will build on other statements made elsewhere. The writer of the book will assume that the student has a certain knowledge. Here on Wikipedia, I make different assumptions on that the reader already knows (I assume that the reader knows less) and I will have to do with wiki links to other articles. In some articles, others have added a citation for the derivation, but that's i.m.o. misleading.

Now, the state of the thermodynamics article before I started to edit them was abominable as I explained here last year. The wiki policies on verification, rather than helping to make the articles better, were actually helping to keep the articles in the flawed states. Flawed statements were seemingly supported by citations. People are less inclined to question a statement that looks a bit strange if that statement is supported by a citation. Citation are almost never checked out (as explained above, that would be a non-trivial exercise anyway, it could involve weeks of study). You'll find no trace on the talk pages of the affected artcles that discussed even quite trivialy flawed statements.

Recently, I wrote this proposed policies to address these sorts of problems, but I got little supoort for that. Count Iblis (talk) 15:51, 1 October 2009 (UTC)

The bigger issue I see with that page is not whether the derivations can be sourced, but that the page looks like a textbook. The issue of when to include derivations is not simple, and there are differing opinions, but in general the focus of the article should be on the theorems rather than the proofs, so to speak.

We do not require that every derivation can be cited word for word from a textbook. Actually, copying long derivations word for word would not fit into our rules about copyrighted text, and paraphrasing them too closely would be some sort of plagiarism even if it managed to avoid the copyright issue. So some element of originality will always go into writing derivations. The general standard (which is vague, not precise) is that the derivations we include here should be written in a way parallel to the way they would be written in the best textbooks in the area. — Carl (CBM · talk) 17:09, 1 October 2009 (UTC)

PS I added a first reference to a textbook with a derivation of one of the formulas. For standard textbook fare it is usually not too hard to find them. — Carl (CBM · talk) 17:18, 1 October 2009 (UTC)

I think this is an issue where it is probably best for WP policy not to be too specific. On the one hand, there are articles where it's valuable to have a reader-friendly derivation, and I wouldn't like to see sourcing rules get in the way of that. On the other, Talk:0.999.../Arguments is a pretty convincing demonstration of why it's not enough to say "sound derivations are okay" - too many people aren't competent to judge whether their derivations are sound. I don't think it's possible to handle this problem without making the occasional judgement call. --GenericBob (talk) 05:43, 3 October 2009 (UTC)

I completely support "First principle derivations do not need citations". One article that I liked was Stress (mechanics) because of it's derivations. If you care about "truth through verifiability" rather than "verifiability, not truth" then an inline derivation is much more verifiable than a citation. Jrincayc (talk) 12:27, 20 October 2009 (UTC)

Translations of primary source quotes

Lots of articles include quotes from ancient texts, including the Bible, Qu'ran, hadiths, and in the case of one article I'm working on, the Mahabharata. I know that these should only be there as a supplement to sourcing from secondary sources, but within these limits I believe consensus is that it's acceptable to include them. Say we use a standard English version of the text. Is it really useful, per this policy, to include either in the text or in a footnote the original Latin, Arabic or Sanskrit? If "chapter and verse" is given, then it should be easy for anyone who is able to read the original language to go and look it up. Itsmejudith (talk) 15:12, 3 October 2009 (UTC)

I'd say that this should be handled on a case by case basis. In my opinion, only if the original language version were somehow well-known in English (e.g., cogito ergo suum) or there were some verifiable dispute among translators would there be a need for the original language source; WP:NPOV would require acknowledging different translations if they were meaningfully different, I think. RJC ^Talk_Contribs 16:29, 3 October 2009 (UTC)

Unless there is a specific reason that including the original language quotation is necessary or particularly edifying, I don't see why it should be quoted. Additionally for Bible texts, what particular source texts are being used need to be known. Different Bible translations use different text sets. For example, one would need to know if the translators were using the Textus Receptus, Nestle Greek New Testament, or another source. Vassyana (talk) 13:29, 19 October 2009 (UTC)

Access to sources

Please see: Wikipedia:Reliable_sources/Noticeboard/Archive_45#pay_up_or_shut_up as an example of a recurrent frustration at WP:RS/N which resulted in Wikipedia:Reliable sources/Cost, both of which resulted in WP:RS/N editors adding "access to sources" to this. Fifelfoo (talk) 03:21, 7 October 2009 (UTC)

I have added a short section on how the accessibility of sources does not impact verifiability. This is due to getting repeated questions at RSN about "pay sites" and the like. We need to make it clear somewhere that verifiability does not neccessarily mean that "you... personally, must be able to access the source right now, for free, on line, without leaving your chair or doing any leg work" Blueboar (talk) 03:22, 7 October 2009 (UTC)

Agreed. Cirt (talk) 03:22, 7 October 2009 (UTC)

Agree, sources should be verifiable in principle, not verifiable sitting at home without expending effort or money. However, there may be extreme examples where it is necessary to rethink this, e.g. citation to an incunabulum written in archaic German accessible only by certified researchers kept in a museum in Budapest. LK (talk) 05:11, 7 October 2009 (UTC)

suspect Primary and Original Research cover that generally Fifelfoo (talk) 05:33, 7 October 2009 (UTC)

One saving feature is that if the only source to verify a statement is very hard to access, that source could not be relied on to establish notability nor anything contentious. Johnuniq (talk) 07:20, 7 October 2009 (UTC)

I think it is generally understood that sources must be accessible to whoever is willing to go to the right place and pay the required fee. It should not be necessary to go to work for a particular government organization nor to join a particular organization or religion. The case of having to be a bona-fide researcher isn't so clear. The researcher just has to demonstrate that he/she is serious; the researcher need not work for a particular employer or espouse any belief. --Jc3s5h (talk) 15:50, 7 October 2009 (UTC)

Many sections of the NY state archives (largest state archives in the US) are accessible only to certain members of the state govt or those researchers who are able to get special access. Many of those documents would be of great use for a article, even if they are primary sources (and yes primary sources are important, and of great use in a Wikipedia article). Should we say that anyone who does get clearance to see them can not then use them in an article because few others can? Especially if it is to dispute a claim put forth in a source that says "document A states..." and the document actually says the opposite when put in context. As for religion, yes we dont expect individuals to convert just to see a document or read a book; but there are Wikipedians of every religion out there who could verify FOR you, you dont have to be able to see it yourself, do we not trust each other? Same with foreign languages, find someone who can read the language and translate it for you. Verifiable by SOMEONE, not necessarily YOU.Camelbinky (talk) 16:28, 7 October 2009 (UTC)

The matter really needs to be addressed and it would be good if there were some definite provisions for referring to (extracting facts from) archive documents, because in many cases they are the *only* real way to get certain facts straight. Things like: what school did X go to? has X really held a manager position at company Y? when did they start teaching this or that subject at so-and-so school? Sometimes that kind of thing is vital to find a good and illuminating way to set something out - what person X says about it himself might not do, there could be a thousand reasons that he'd distort things or hush it up. Or in order to judge the value of what a third-party source is saying, perhaps on some other matter (if archive documents show that X was doing his military service at time T then it might prove that he can't have gone abroad and witnessed the event E at the same time, which he said later that he had). Telling people in interviews that he went to a more flashy school than he actually did and so on. Some in the WP community want to ban all use of archive documents, even public ones that are open for free to anyone, and also stuff with restricted access such as the NY archives, because they are - you guessed it - not easy to check for everyone and thus threaten to corrupt articles where they are used.

We need to develop some stance on this, at least some clear guidelines. It's not just archives: more and more news sources are erecting some kind of paywall again, or blocking access to some material unless you happen to be in the same country. For the last five or six years we've had relatively unlimited access to free news and features from every outlet that was online and with a very wide backlog, and sites like Wikipedia have drawn on this, but that era is coming to an end at least with many reliable newspapers and magazines. /Strausszek (talk) 11:39, 9 November 2009 (UTC)

We altrady have a stance on this... see: WP:Verifiability#Access to sources Blueboar (talk) 14:59, 9 November 2009 (UTC)

Uh, you slightly missed the point. I'm fine with that it doesn't mean "everyone must be able to check at a moment's notice" but some facts are actually hard to check, or verify in a fail safe way, for other than a limited swarm of people with special interest or special access to the relevant documents or books. And the range of sources that will demand special access (paid subscription etc) is growing, and very likely to keep on growing. Lots of WP articles use online newspapers and mags as their main verifications, but more and more papers are protecting a big part of their content behind paywalls. A few years ago the Wall Street Journal and the New York Times were the only US newspapers who did that, but there will be more of it and effectively it means people here could keep on challenging each others' claims if these are sourced from such a news source behind a paywall, as long as no one will check. And as long as the losing part isn't so suspicious he says "You made that up, you sided with him! I don't believe The Economist actually ran that argument!" /Economist now requires subscription to be able to read much of its content online, especially the backlog of articles/

---edit: okay, I can see we agree in principle on paysites, but suppose it's a site or a magazine that has a subscription fee that's much higher than normal. Some leading scientific journals have prohibitive subscriptions because they are effectively able to impose subscriptionship: if you're an elite university in that area you have to keep journals A,B and C to qualify. That's actually a major problem to university libraries these days. The publishers can command 2.000$ a year in subscription and no, they don't put things out online. Only on paper. Of course those mags will be accessible at the research libraries in question but it's still limited access. - end of added bit -

I've seen instances here on WP where people have been accused of "original research" and their matter deleted when they added something highly relevant to an article, even though their sources were all open and could be checked by anyone who took the trouble and they'd indicated (without any real synth input) exactly why this ripped apart certain claims made by the subject of the article (not first-person claims about X self, but central assertions that had been made in, for instance, a book he/she had written). Sometimes when you want to do an article that does its job, says what is relevant and central, you need to resort to sources with limited access to make sure you get it right. The current WP:Verifiability#Access to sources doesn't really deal with it in any thought-through way. It just presupposes everything that matters can be sourced and settled from places that are open and free./Strausszek (talk) 18:28, 9 November 2009 (UTC)

I am not sure how you can reach the conclusion that the section presupposed what you say it does... if anything it presupposes the opposite... that materials are often NOT free or are difficult to access... and they are just as acceptable as materials that are free and easy to locate... to quote:

• "... This does not, however, mean that any one can do so instantaneously, without any cost or effort. For example, some on-line sources may require payment to view; and some print sources may only be accessible in specific university libraries. The ease of access does not affect the verifiability of the information taken from such sources." (bolding added for emphisis)

This means exactly what it says... "The ease of access does not affect the verifiability of the information taken from such sources". Period, end of sentence. I don't see how we can get any clearer than that. Blueboar (talk) 19:16, 9 November 2009 (UTC)

(out)I'm not sure how you want things changed: I don't think you are saying "if its above x dollars a paragraph, then we can't use it?" Some portion of citable information will be difficult to access. Some (probably small) portion of that will be contentious. Some smaller portion of that will not have a secondary wikipedian readily available to check. In the small number of cases where it continues beyond that point, I don't think there is much we can pre-decree here. What SHOULD we say here about it? (John User:Jwy talk) 19:07, 9 November 2009 (UTC)

I guess one part of what I want nailed down is this: If we have a news story, a reported scientific discovery or the like and some journalists are making a factually overstated claim which sounds hefty, twists the story and sells copy, but which is not borne out if one takes a hard look at the primary (data results, a recent Ph.D. study that's cited as a brekthrough) or secondary (scientific survey reports, interviews in a specialized journal) sources these journalists would or should - have used, then those better sources must override what those journalists said, even though the papers and tv reporting are much easier to access and will get repeated by some wikipedians: "it says in this newspaper that so-and-so, and that other source you're waving about isn't really trustworthy, how can we know it says that - I can't access it?" It doesn't matter what newspaper or magazines printed those faulty articles, if they can be proven wrong those statements shouldn't be cited anymore. Very few news outlets are 100% accurate all the time today, or can claim they never print something that was engineered to create sensation or controversy, and so it might likely be overstated, abusive or wrong.

It could be an old story as well; reporters and popular writers tend to parrot each other a lot, so once a certain story, a certain turn to a story, has taken root, it can survive for many decades even if it clearly misrepresents the facts. You get a thousand popular accounts that say or imply e.g. that the October Revolution was a popular rising undertaken by the ordinary people with lots of street fighting. No, it wasn't - there actually wasn't a huge amount of street fighting except at some key points and the ordinary people of Petrograd didn't participate. But thirty to fifty years ago the version which made it look like the French Revolution was current even among western historians who had zero sympathy for Lenin. Even in some school textbooks. They had bought the idea that the city was jampacked with people shooting with rifles and cannon and that everyone was aware that the old cabinet was going. Eight months before, at the February Revolution when they kicked out the tsar it was just like that, but that one had next to nothing to do with the Bolsheviks. There are lots of old legends like that, I'm citing it to point out how they survive.

The praxis we have today makes it very hard to stick to the authority of the sources who really know about a certain contested thing, plus many wikipedians hate authorities on principle. The catchphrase "verifiability, not truth" gets in the way, because to many people it means all third-party sources are equal. But that kind of "anything goes as long as it's quoted" attitude really jeopardizes the reliability of WP. /Strausszek (talk) 21:23, 9 November 2009 (UTC)

Ah... what you are talking about is "assessing the comparative reliability" of different sources... ie which sources should carry more weight ... not whether a specific source is verifiable or not. Wikipedia deals with this under WP:NPOV and WP:RS. Essentially, if two equally reliable sources say something different, NPOV tells us to present both view points. However, very often sources are not equally reliable... and WP:NPOV does not say that we must treat all sources as being equal. Obviously a scholarly source carries more Weight than a media source... and Wikipedia wants our articles to be based on the most reliable sources possible. However, judging between two sources can not be narrowly discussed and mandated by policy... because such determinations are always subject specific. They must be worked out on the article talk page. Wikipeida policies and guidelines are broad principles, not narrow laws. Blueboar (talk) 22:15, 9 November 2009 (UTC)

I plainly don't think you can say, en bloc, "the New York Times is always top of the list reliability-wise if it's in conflict with other sources" (the NYT can be replaced here with any high-end news source). Now, I regularly read NYT online and I would personally trust it >90% of the time, but that doesn't mean every word printed there is a gold standard and can't be challenged from facts (or from some other source's interpretation of something, and the facts cited there). And WP:RS just seems to imply that news sources that qualify as "highly reliable" should always be supposed to be right if they are cited. The trouble for me is that it permits no attempt to discuss the reliability of a specific article or statement. At least if we see WP:RS as a Law (not a rule to be tweaked by common sense) and if it doesn't matter what the primary source facts are or who wrote the news piece/op-ed piece in question. Unfortunately, some people here, even Admins, do see WP:RS and other rules as Laws to be clobbered into everybody's heads.

To lots of Wikipedians I've encountered, the established practice is that all statements from reliable sources are perfectly equal: there is no real source criticism or discussion of the relationship between different sources and accounts, unless you can borrow that discussion straight from a third-party source (and it still risks getting neglected). And in practice, when the high-end newspaper is set against other sources, these users tend to lump together Fox News and the rare scientific report that you have to go 300 miles to some college library to read, because Fox News isn't reliable (true) but the rare report isn't easily accessible. So they are both pushed out by, like, Daily Telegraph or CBS 60 Minutes or Larry King Live - which is a bit debilitating. A guy who is doing serious research wouldn't take such a sloppy view of his particular sources. WP isn't generally into research, but I think it should at least accept that we can avail ourselves of the means of research to check specific facts and get the picture right. /Strausszek (talk) 22:49, 9 November 2009 (UTC)

There SHOULD be discussions such as those you say and difficult disputes can appeal to Wikipedia:Reliable sources/Noticeboard. (John User:Jwy talk) 23:14, 9 November 2009 (UTC)

I think we are getting off the topic of "access" to sources and payment regarding reliability and getting into comparative weight. If that is true, I suggest we state a conclusion to the access part (status quo as per Blueboar's interpretation as codified into existing policy and the cost essay) and break this into a subthread regarding comparative weight.Camelbinky (talk) 00:29, 10 November 2009 (UTC)

It isn't that we got off the topic... I think we simply clarified what the topic actually is. Strausszek's concern isn't really an access issue... but one of comparative reliability. This is something that can not be dealt with through policy... the various sources need to be examined on a case by case basis by those who best know the specific subject and the sources ... ie on the article talk page. And if they need outside opinions... Jwy hits the nail on the head... WP:RSN was created, in part, to give editors a venue to get help with issues relating to reliability such as those Strausszek is concerned about. Blueboar (talk) 00:41, 10 November 2009 (UTC)

Citation for advanced topic of a profession

In short, is the article of such topic requires citation, although the contributor is strongly against it?

1. A topic(substance, theory, truth) of a particular professional field.
2. That topic isn't (likely to be) challenged today in that profession.
3. That topic isn't considered as a "common knowledge" by people outside of that profession.

So here's the story. Rencetly in ZH Wikiepdia Village Pump (Policy), there's a repetitious debate (假如一條條目沒有外連及指出參考文獻，理應掛{{unreferenced}}模板，可是……) stemmed from few users who aren't in the profession placed the {{unreferenced}} tag in 1,3-dipole. The contributor of that article insists that 1,3-dipole is an objective truth without dispute in chemistry and citing this EN policy page as his basis of argument:

All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation

While other users challenge that because they do not understand chemistry, particularly about that article being discussed, citing source is required. The dispute continues with each side repeating their same statement and no one seems to agree with the opposite. So I ask for suggestions here. Thx -- Sameboat - 同舟 (talk) 12:24, 19 October 2009 (UTC)

Why not just find a source... If something is "an objective truth without dispute in chemistry" it should be easy to find a source for it (for example in a basic chemistry text book). Blueboar (talk) 12:37, 19 October 2009 (UTC)

I've suggested the same. (As the contributor also stated such topic is available in chemistry textbook, so one {{cite book}} template would have simply settle the dispute.) But the point is, right now the policy doesn't mandatorily require citation for "an objective truth without dispute in that profession". The contributor seems to afraid that Wikipedia would become an encyclopedia full of unrefereneced tag or articles plagued by ref code (so making them edit-unfriendly.) Anyway, the opposite cares not the difficulty to have citation in article, but the policy which "if it isn't required, don't force it like it is disputed." -- Sameboat - 同舟 (talk) 12:55, 19 October 2009 (UTC)

You are correct in that we don't require every statement in Wikipedia to be referenced... however, we do require sourcing for any statement that "is challenged or likely to be challenged". We allow editors to write articles without having to spend time and effort locating sources for basic information ... information that is "not likely to be challenged". However, sometimes deciding whether something is "likely to be challenged" or not is a judgement call... and different people may disagree as to whether something is "likely to be challenged". What we can definitely say is that, when someone comes along and accutally does challenge the information (by requesting a citation), then there is now a burden to actually locate and add a source.

In other words, you don't need to worry about citing basic stuff unless someone asks you to do so... at which point you do have to cite it. Blueboar (talk) 13:32, 19 October 2009 (UTC)

That it correct. We don't want a citation for every mention of "sky is blue", but if an editor requests it, we should be willing and able to provide a source, at least on the talk page. Generally, scientific points that are "well known" to experts in that science can be sourced to a textbook, or a famous paper. If neither type of source exists, it's very likely not so "well known", and some source is clearly needed. Crum375 (talk) 13:43, 19 October 2009 (UTC)

As Crum says, you can also consider giving a citation on the talk page, for the sorts of things where putting a citation in the article would be silly.

Regarding the specific article 1,3-dipole, the issue there is that the article is a stub, with only two sentences. So it should be marked as a stub, which is stronger than just marking it as unreferenced. Adding at least one general reference to the article, just so that someone who wants to know where to look can find it, would be a good idea. — Carl (CBM · talk) 13:51, 19 October 2009 (UTC)

Actually I'm talking about the ZH version of zh:1,3-偶极体. Anyway I've found out that in the ZH WP:Verifiability there's an exclusively extra statement (特别是受合理質疑且确無可靠來源查證的內容) which can be loosely translated as "especially for content without citation of realiable source and received reasonable doubt". Which sets up the threshold for non-professional user/reader to question/challenge the verifiability of content without citation. I wonder if this statement is rational, it seems that the contributor defends by this reasonable doubt to avoid citation for a "truth" somehow not studied by the majority. -- Sameboat - 同舟 (talk) 04:22, 20 October 2009 (UTC)

Can I ask why are we discussing an article on the ZH version of Wikipedia on this page? Different versions of Wikipedia have their own policies and guidelines, and (when these are similar to those here on the EN version) have developed different interpretations of their policies and guidelines. What we say here does not relate to what goes on at the ZH version. Blueboar (talk) 13:43, 20 October 2009 (UTC)

There is a dispute in the Chinese Wikipedia and it is related to the interpretation of the policy. As many policies including WP:V in ZH Wikipedia are based on their counterparts in the EN version, the EN Wikiepdia is served as the supreme court of Wikipedias. --Quest for Truth (talk) 01:54, 21 October 2009 (UTC)

I don't mean that. Since the discussion there's almost stagnant, I ask for suggestions in EN. Alhtough I personally prefer unchallenged scientific article must provide at least 1 citation (either inline note or reference reading), I'm not going to use this disucssion to serve as a supreme judgment to my interest. Either way, WP:Scientific citation guidelines (not yet translated to ZH) does not mention any exception, I hope that's enough to support my basis. -- Sameboat - 同舟 (talk) 02:12, 21 October 2009 (UTC)

Quoting from a source that is not easily accessible.

My opinion is that sources that are not easily accessible or free should be quoted in detail. Perhaps as a foot note, so that readers are able to read the original. It would also prevent an editor from misquoting. Yogesh Khandke (talk) 13:55, 20 October 2009 (UTC)

How do you define "easily accessible"? Easy for whom? Keep in mind that many very high quality accademic sources are "easily accessible" by those living in major cities or university towns, ie places with major libraries, while these same sources would not be "easily accessible" by someone living in a rural area. Would you have us quote, in detail, every dead tree source that is not accessible over the Internet?

Also... extensive quoting can bump up against copyrite laws. Blueboar (talk) 14:23, 20 October 2009 (UTC)

If a x*^#@ editor would make up a citation, the x*^#@ editor would make up a quote too. However, having the quote would help detect cases where the source and quote actually exist, but don't quite support the claim presented in the article. --Jc3s5h (talk) 15:58, 20 October 2009 (UTC)

My opinion is all sources that cannot be counter checked right away, at a click, should be quoted in detail. I am not sure about the copyright angle. Just my opinion. Yogesh Khandke (talk) 03:23, 21 October 2009 (UTC)

This would also take care of the foreign language problem, as then one who wishes to check could do so using translation software, which are available without charge. Yogesh Khandke (talk) 03:25, 21 October 2009 (UTC)

This is simply not practical. To actually check a reference usually requires looking at large parts of a text, including parts much earlier than the actual part being quoted. This is because you have to check the context in which the quoted material is located. But even if we made editors type only the surrounding 3 paragraphs of the source, this would add an enormous burden to anyone who wants to add a source. There is no reason why all sources need to be verifiable online; sometimes people have to go to the library, and that's OK. — Carl (CBM · talk) 12:58, 21 October 2009 (UTC)

On the other hand, if you suspect that a specific source is being misapplied, taken out of context, or otherwise misused, there is nothing wrong with requesting a more detailed quote on the article's talk page. Blueboar (talk) 13:18, 21 October 2009 (UTC)

Of course. I'm just saying that it would kill my productivity if I had to retype a page out of the book every time that I make a citation. If a specific issue comes under discussion, then it starts to be more worthwhile for me to spend that time. Particularly because, when an issue is under discussion, it's usually necessary to look up several sources to make sure that the one being quoted is reflective of the overall literature. — Carl (CBM · talk) 13:47, 21 October 2009 (UTC)

I concur with Blueboar and Carl. Plus asking editors to type in large quotes from offline sources raises huge potential copyright infringement issues. --Coolcaesar (talk) 15:57, 21 October 2009 (UTC)

Concur, this is a non-starter in almost all cases. Better to work on improving the quality of citations so that other editors have a reasonable chance at finding the source in a major library.LeadSongDog come howl 20:08, 21 October 2009 (UTC)

The need to refer to books violates the utility of an online encyclopaedia. Well perhaps this view of mine is generated from my limitations. The nearest library for me is a world away. Yogesh Khandke (talk) 16:31, 22 October 2009 (UTC)

Just because you are unable to access the source does not invalidate it. The principle is that anyone must be able to do it, not everyone - that is an important distinction. All it takes to call into question a source is that only one person checks it and reports that it does not say what it is claimed to. Roger (talk) 19:05, 22 October 2009 (UTC)

I would go a step further and say, instead of "anyone" it actually means "someone" can verify it. Just as one person can claim the source doesn't say "it", it also takes just one person to verify that "yes, its OK". I feel bad for you, Yogesh, that the nearest library is a world away, however an encyclopedia's utility is not to exist as a tool to "find" sources, actual book encyclopedias rarely put their all their sources, or many of them, as the individual articles are mostly written by professionals on that topic (Einstein once wrote an article on Relativity and Physics for an encyclopedia). The sources used in a paper encyclopedia are even harder to access than those we use in Wikipedia. Your suggestion would easily open the door to invalidation of any book citations, newspaper articles, and other non-internet sources. Contrary to popular belief very little of total human knowledge can be located on the internet, even less is free.Camelbinky (talk) 19:55, 22 October 2009 (UTC)

Actually my views are from the perspective of an editor and not a casual reader, though in the case of Wikipedia they are often the same. I am not suggesting harsh steps taken in a hurry. I wonder how articles in the real world for example journals are reviewed? Does the writer of these articles provide the sources he quotes? I could check this out though. I have a few acquaintances who have written papers published in reputed international scientific journals. I would like to rephrase your ending remark, I would prefer it to be very little of total human knowledge is presently located on the Internet... Yogesh Khandke (talk) 21:36, 22 October 2009 (UTC)

As a grad student I can tell you that for most professors who publish it is his/her grad students who round up the various sources (I even worked for a professor who wrote out his entire paper, then told two of us "to go find sources to support what I wrote and throw in footnotes where needed"; so basically his opinions came first and then sources that supported his statements were thrown in; and yes it was published with no problem in a peer reviewed journal). Peer reviewed journals, at least in my discipline of Political Science, may often have completely opposing papers in the same journal often on purpose; "peer reviewed" does not mean, as some on Wikipedia seem to think, mean that the conclusions or data presented has met the approval of the writer's "peers" or that it is representative of the consensus in that community; on the contrary even professional academic journals do very little "peer reviewing" as I just defined. I agree with Yogesh's rewrite of my quote; hopefully by using non-internet sources and bringing more of them to Wikipedia we are increasing the amount of total human knowledge on the Internet every day.Camelbinky (talk) 23:09, 26 October 2009 (UTC)

As a publishing scientist (in computer science), I know the papers I reference. If I review something, I typically know the core papers that are being referenced. If not, I often get the abstracts to see if a claim is plausible, or, even the paper - either online, from the library, or from the scientist. But most computer science papers I read present substantially new material anyways - references mostly add context and lay the groundwork. Referencing is much less anal as on Wikipedia. --Stephan Schulz (talk) 23:02, 29 October 2009 (UTC)

I don't see that providing good context is much of a burden at all. Lets face it, copying even 3 paragraphs of text from a book you have in front of you is not that big of a deal. Considering how much time most of us spend, and how much text is written, on talk pages alone, those 3 paragraphs' value far outweigh the one-time cost. This is not about denying AGF, it is about honoring it, while allowing for pertinent information. I don't know that we should demand it for every offline source, I doubt it, but it might be worthwhile to at least introduce such behavior as a 'good idea'. Unomi (talk) 03:08, 27 October 2009 (UTC)

It might not be a big deal, but it's likely to constitute a violation of copyright policy, which requires that non-free text be used in accordance with WP:NFC. NFC forbids extensive quotations of copyrighted text. Three paragraphs is extensive even if it is the only text copied from a source, but if a source is used multiple times in one article or across multiple articles, then the infringement is compounded and we can wind up with a very substantial taking indeed. --Moonriddengirl ^(talk) 12:14, 29 October 2009 (UTC)

If the reference is about something contentious, where someone may doubt about what does the source exactly says, we can quote the text being referenced. But 3 paragrapahs, or even 1, are too much, and unneeded: a quote of the key sentences would be enough. And if the information is distributed in the way "important sentence - uninteresting sentences - another important sentence" we can use the quotation technique of "important sentence. [...] another important sentence." MBelgrano (talk) 15:55, 29 October 2009 (UTC)

{{Cite book}} is already provided with a field for the quote from the book. Having in mind the topics about copyright and not making a reference giant, it simply needs to be used when needed. MBelgrano (talk) 12:52, 28 October 2009 (UTC)

A demand that if the reference (I prefer to say "reference" rather than "source" because often the fact you want is not indisputably demonstrated in the reference text, it's just mentioned as an already verified thing, or even implied) is not easy to get hands on for any medium user then it must be quoted in full detail - that demand if enforced would make some articles almost impossible to write or to keep understandable. It would increase the scope for pointless edit wars too, because people could always say, if they wanted to, "this is not quoted in enough detail" or "I have another source which contradicts your source - I'll cut what you've written and put in my version".

Science isn't always unambiguous and scholars don't always express themselves in cut-and-dried statements of everything they've thought or concluded, so if we're not allowed to use some measure of judgment in handling references, appreciating what they want to say, where the facts come from and whether the reference author makes his case, there is really no point in trying to stitch together logical and sourced articles - much of the time it still won't lead to any stable result, just "back-writing" from the kind of sources you can wave in front of other wikipedians. Preferably sources that are online or easy to get at a public library, and preferably in English. That kind of demand would make this WP both more shut-in, more biased and more uncritical.

As Camelbinky put it, the source groundwork behind articles in paper encyclopaedias and scientific reviews is often even harder to access and check than the references we use here, and actually, the arguments in those professional texts aren't completely chained down to the data references. The reasoning and the "story" they make follow, to some extent, their own logic to win over the reader. I have become increasingly disillusioned with some editors here at WP who don't understand how complex it can be to "verify" beyond doubt an outwardly simple statement when you can't grab it in one piece off the bookshelf, and who will edit out everything that "doesn't have a fail safe source right now", no matter how dubious that "non-sourcedness" would be. Any statement that deals with the intentions of a person relating to something he did, or a social/historical/ literary movement, or which says that X likely did not know about Y or wouldn't have understood that Z, those and other statements can be challenged no matter how obviously true they are if you know the topic. They are really hard to verify before a non-peer audience which you can never be sure you may tell "bear with me and I'll explain": it takes several steps of reasoning to bring the grounds out and people who don't get it tend to say "you haven't shown us all the sources for every step of this - why can't you be like a physicist who points out exactly what he claims and proves it?". Er, if I did roll out the entire reasoning, they'd fall asleep!/Strausszek (talk) 04:00, 3 November 2009 (UTC)

That seems to have strayed a bit from the topic of this section, but "... arguments in those professional texts", etc. seek to do something which Wikipedia seeks to avoid doing. Wikipedia editors should not engage in synthesis of published material that advances a position. WP articles should not contain in original research. WP articles might report that Bob says ABC and Jane say DEF, citing sources supporting those assertions, and might report properly attributed conclusions drawn about what Bob and/or Jane said by the cited sources reporting what they said. WP articles, however, should not draw conclusions or seek to persuade. There is some leeway in that for editorial judgement; how far that leeway extends is established by editorial consensus on an article-by-article basis. Regarding cases where excessive detail is needed to properly support an article assertion, removing that detail to an explanatory footnote might minimize disruption to the flow of the article prose. Wtmitchell (talk) (earlier Boracay Bill) 00:26, 4 November 2009 (UTC)

If a source does some synth, we can and do use that synth that the source does and do not have to attribute that "source A says B+4=X" in the text. We simply say "X" and source it to A using a footnote, we dont have to say anything about how source or author A got to the conclusion or that it was a synth of sources B and 4. So I would disagree with most of Wtmitchell if I correctly understood him. However I think Wtmitchell also got off the point of this thread. The point is that WP:V's last section is quite clear as is the related essay linked to at the page (both of which were created at my request at the RS/N)- cost or ease of accessibility does not in any way affect reliability or verifiability. Exessive quoting is not only discouraged by current policies but is looks bad and is cumbersome to the reader and can get one in legal trouble as well. The quick answer to anything anyone may have regarding an opposing opinion is this- "verifiable does not mean by YOU, it means SOMEONE, we dont care if you personally cant verify it, tough luck". So basically I dont see any opposing opinions here that cant be answered back with that quote right there.Camelbinky (talk) 00:50, 4 November 2009 (UTC)

Wtmitchell, it's not about trying to brainwash people or push a POV or a fave theory on the subject of an article, but about producing intelligible and nuanced texts. That's a central part of what this site is about in my mind, and I wouldn't be here if I didn't think you could improve articles to some lasting results, a ground level of usefulness and accuracy where those articles (at least some of them) will remain for a decent length of time. But to get there, in very many cases the people participating have to be given the space to be specific and clear about what verifiably went on during, let's say, the Suez Crisis, the Battle for Britain or the evolution of MTV. Behind the scenes too (a part which is often vital and mostly not covered in easy-to-get news sources, which toss in anecdotes or legends instead). And they have to keep out the urban myths and falsehoods. It's just not possible to give a cogent account of the hows and whys, of what went on and how it's been retold and interpreted by historians (we're all our own historians to some degree) if the only allowed structure is to say "at date A, B happened -ref- - then at date C, D happened -ref- and RSource E says that F because G <quote or pseudo-quote from E here, indicating precisely F>G" etc. That way of telling something breaks down into a jumble as soon as those who are listening can't supply their own memories of the events, or a firm consensus doesn't already exist. It's neither scientific nor readable, and that kind of account leaves the door wide open for any ignorant editor to say "I don't believe it happned like that, I have this source that says it was P who invented it", "my source says Bush planned 9/11" and so on.

To get to a reasonable account, those "isolated fact points" have to provide some structure and the structure mostly means some amount of synthesis in many places - statements that are not just brute A+B=C. Ordinary encyclopaedias and handbooks are full of it. One of the reasons they don't fall into the kind of insane POV wars we see here is that people who write for the Britannica, for newspapers or whatever highly reliable sources, have to think of their reputation. If they publish gravely biased or unsourced drivel, they risk ruining their own good names. But here at WP, broadly speaking, no one is conditioned by the need to protect their rep and no one can point to his own authority in that way - so to some extent we have to find other ways of handling it. The "reference everything" position is one way of doing it, but a bad way I think, because it doesn't really match how reliable and intelligible texts are produced. /Strausszek (talk) 03:21, 4 November 2009 (UTC)

Peer review

Camelbinky I too did not suffer from the misconception that peer review meant that the peers agree to the conclusion but that, one purpose of peer review is to check the references, for example if I write that rats are a staple source of proteins for Indians in the summer and quote a source, the peers check whether the quotation is there, the interpretation is correct and the source is of a good standard. Which is one job of editors here too. I hope I am clear. Yogesh Khandke (talk) 05:47, 28 October 2009 (UTC)

Editorial oversight

"Questionable sources are those with a poor reputation for checking the facts, or with no editorial oversight."

What if you have a newspaper that has a good reputation for fact-checking and editorial oversight and the journalist being sourced in WP is quoted for an international story, but the journalist specializes in local and human interest pieces? Would the newspaper be likely to provide editorial oversight or fact checking in such a situation? This recently came up where the writer interviewed an individual with a POV on internationals matters in relation to an upcoming local event. The journalist makes a blanket statement in relation to what "the Arab world" thinks. I believe the journalist just picked up on something the person he was interviewing said but did not quote him, merely expressed it as a fact. Some say it is acceptable to use this person's quote as RS for the thinking in the Arab world, since he is under the auspices of a RS. However the journalist clearly does not write about international issues, just local and human interest (handbags and shoes and murders and such), so I contend he would not be subject to editorial review in terms of what he says about international events.

Can we discriminate this issue? Am I making a reasonable argument? What would WP say? I'd appreciate input. Stellarkid (talk) 06:25, 1 November 2009 (UTC)

I agree that the byline and the "tone" of an article (a source) need be taken into consideration. An example: Newsweek is generally a RS, I think, but a recent article referring to "our muslims" (immigrants to the US are well-integrated as opposed to those in Europe) had me seeing red. Sounds to me just like "our gypsies" and "our niggers". Never mind that the handfull of Muslims I know are way different from one another (Uyghur, Somali, Iraq (Arab), Iraq (Kurdish), Bosnia, Chad, Egypt -- in no particular order).

I'd rather we avoid any sign of patronizing or racism, or just in general sloppiness. - Hordaland (talk) 13:23, 1 November 2009 (UTC)

• Would I be right in thinking you're talking about Gaza War and the article in the South African Sunday Times? My answer would be that the South African Sunday Times is a reliable source and it is permissible to use it as a source in that article. However, the relevant phrasing should not give undue weight to Lauren Cohen's opinion, and must conform to a neutral point of view.

  In this case I would think the phrasing you removed was not inappropriate and it might have been preferable to leave it in. I'm sorry if that wasn't the answer you were looking for.—S Marshall ^Talk/_Cont 13:24, 1 November 2009 (UTC)

You are correct about the Gaza War article, although I think the question is a broader one or I would not have put it up here. There are other issues involved in its removal. Does it really qualify as an "alternative name"? As a citation in the lede and contentious, we could expect more RS etc. and the issue of POV. So while I would hope that the answer supports me in this case, I really think the broader question is important for background. If a newspaper is reliable does that editorial oversight policy include every writer for every topic? Would the cartoonists be covered under the policy for example? (not trying to be funny here) Stellarkid (talk) 18:17, 1 November 2009 (UTC)

This sounds like another situation where in text attribution would clarify things. Suggest that you re-phrase the sentence to read something like: "According to a news story printed by <name of reliable source X>, the Arab world thinks 'blah blah blah'.<cite to reliable source>" In text attribution alerts the reader to the fact that the information comes from one single source and might not accepted by other reliable sources. Blueboar (talk) 14:19, 1 November 2009 (UTC)

• (In response to Stellarkid) I don't think it's possible to generalise here. For commonsense reasons a cartoon or humour piece isn't to be treated the same as a serious article, but I think in these matters of the detailed content of an article or the prominence to be given to a particular phrase, a local judgment needs to be made and it would not be possible to give a rule that wasn't stating the blindingly obvious.—S Marshall ^Talk/_Cont 22:03, 2 November 2009 (UTC)

Citations to media sources (books, TV, Movies, etc)

My synthesis of the WP:NOR and WP:V/Access to Sources rules suggests to me that, when writing about a piece of publicly published media--a novel, movie, song, or TV series epsiode--a formal citation to the source is not necessary for factual assertions made about that primary source in the context of that article about that source... but we don't actually say that, and I've seen people tag such assertions with FACT, which I geneerally take back off. Anyone have thoughts, pro or con?
--Baylink (talk) 07:47, 8 November 2009 (UTC)

You are essentially correct; But a lot depends on exactly what is being said in the article. For a basic statement of plot discription, such as "The Headmaster of Hogwarts is Professor Dumbledore" we obviously could cite to the Harry Potter books... and since that citation is so obvious, we do not require that you actually do so.... the citation is assumed.

However, we do have to be careful... the statement: "Professor Dumbledore is the homosexual Headmaster of Hogwarts" is not something that could be cited to the books (or left with an "assumed" citation)... as the sexual orientation of that character is not actually mentioned in the books. For that fact, we need to cite another source, such as the author's statement where she disclosed Dumbledore's orientation.

Also, we can not use the books for a citation (or an assumed citation) for an analytical statement about the books, no matter how obvious we think the analysis might be ... As per WP:OR, we would need to cite a source that directly includes the analysis. Blueboar (talk) 14:32, 8 November 2009 (UTC)

Yeah, I'm clear on the distinction. I was talking about people citing for direct content, as when someone alleges a character on a TV episode speaks a certain line of dialogue; that appears to be something which a reader can themselves independently verify given the context, and which therefore some *other* editor should not FACT tag, which was the thrust of my inquiry. Should we work up some language on this point for WP:V?
--65.34.94.81 (talk) 15:04, 8 November 2009 (UTC)

We had something to this effect a while ago... and it caused problems (mostly over people misunderstanding the differences I layed out above) and so it was removed. I think it best to leave it as an "unwritten rule"... as an interpretation of the policy (specifically, the concept that what we require is that things be verifiable, not that they be verified.) The fact is, sometimes those pesky FACT tags are valid, and correctly draw attention to something that does need to be cited. So, it is better to let people add the tags freely, check them, and simply remove those that are mis-applied. If they add the tag back, ask about it on the talk page. Blueboar (talk) 19:16, 8 November 2009 (UTC)

I agree with Blueboar's conclusion, but not his expression of it. There are many ways to cite a work, for example, footnote, as a general reference, or described in the text. The title of the article Harry Potter and the Philosopher's Stone constitutes a general reference to that book, and everything in the article that can be confirmed directly by reading the book is properly cited. I prefer not to use the word "verified" in this discussion, because it leaves open the question of who has verified it, the editor who supplied a citation, or the reader who might or might not have taken advantage of the citation. --Jc3s5h (talk) 19:36, 8 November 2009 (UTC)

Fair enough. Blueboar (talk) 20:09, 8 November 2009 (UTC)

A somewhat related discussion is going on at the Village Pump (proposals) regarding spoiler warnings and it has evolved to a discussion also about whether or not detailed "plot" descriptions based on the book or movie itself and not on secondary sources. I thought some of you may be interested in it.Camelbinky (talk) 21:11, 8 November 2009 (UTC)

Ok. Then I will continue to Be Bold in removing such FACT tags when I think they're unwarranted for this reason, and not be too concerned that we're not explicit about it in the policy. Thanks, folks. --Baylink@en.wp —Preceding unsigned comment added by 65.34.94.81 (talk) 05:18, 9 November 2009 (UTC)

Organization of articles pertaining to both a virus and a disease

Have started a discussion pertaining to the organization of articles to address a conflict between WP:VIRUS and WP:MED Wikipedia_talk:Manual_of_Style_(medicine-related_articles)#Organization_of_articles_pertaining_to_both_a_virus_and_a_disease Doc James (talk · contribs · email) 13:41, 8 November 2009 (UTC)

Question concerning self-published and third party statements

Is there agreement that these are correct statements of Wikipedia guidelines?

• The limitation on self-published statements that are unduly self-serving is primarily a limitation on using such a statement to establish truth. It is not a limitation on using it to state and define what a person’s beliefs are. For example, a scientist may have gained notoriety because he claimed to have produced cold fusion. He may be quoted as saying “I have produced cold fusion” as long as this is presented as his statement and belief and not as the generally accepted belief.

• In an article about Scientist A, who claims to have produced cold fusion, if a statement is made by Scientist B denying the claim and asserting that Scientist A is a “crank,” then Scientist B is not considered a “third party” for purposes of restricting rebuttal by Scientist A. --Swood100 (talk) 15:05, 10 November 2009 (UTC)

Note: this is being answered at WP:RSN... lets keep the discussion in one place please. Blueboar (talk) 15:26, 10 November 2009 (UTC)

Right. Let's discuss it there.--Swood100 (talk) 16:32, 10 November 2009 (UTC)

Self-published sources

This was added: "Self-published sources should only be used as sources about what they claim rather than what they be or have done."[4] I think this directly contradicts the statement above that "Self-published or questionable sources may be used as sources of information about themselves, especially in articles about themselves, without the requirement that they be published experts in the field." I think this needs discussing before it can be considered policy. Fences&Windows 23:30, 13 November 2009 (UTC)

You're right, it's a contradiction, and doesn't even make much sense. SlimVirgin ^{talk|contribs} 00:25, 14 November 2009 (UTC)

• Needs an example, and taking it to extremes is the best way to illustrate the need.

  I could easily produce a self-published source that says "S Marshall is the most important person in the universe." If someone were dumb enough to start an article about me, they might then write "S Marshall claims to be the most important person in the universe" and cite my self-published source, and that would be okay. But they could not write "S Marshall is the most important person in the universe" and cite it to the self-published source.

  Make sense now?—S Marshall ^Talk/_Cont 00:31, 14 November 2009 (UTC)

I think that's already stated or implied in the policy. We would use in-text attribution, and we probably wouldn't use that anyway under the principle of "not unduly self-serving." SlimVirgin ^{talk|contribs} 00:44, 14 November 2009 (UTC)

• I did say it was an extreme example. Real-world examples wouldn't be so clearly self-serving.

  I'm all in favour of keeping policies as succinct as possible (in the hope that eventually they'll get so short that some of the people who need to read them, actually will), but I also think this a matter that it's a good idea to pin down precisely. So if we can find a succinct way of encapsulating that idea then I'd rather like to see it in.—S Marshall ^Talk/_Cont 00:49, 14 November 2009 (UTC)

We could simply add, "using in-text attribution as appropriate," but it's a hard thing to legislate for, as so much depends on context. I saw a recent example of someone who didn't like a BLP adding, "A says on his website that he has a B.A. from X, and also says that he studied journalism as a postgraduate. According to A, he also obtained a degree from ... etc," so that the attribution reached the point of almost being a BLP violation. :) SlimVirgin ^{talk|contribs} 01:02, 14 November 2009 (UTC)

By all means change or tighten up wording, but policy needs to be self-consistent. Your point is that we shouldn't believe everything someone says about themselves, but lots of things that people say about themselves get reported in the media, like EssJay's credentials.[5] Fences&Windows 01:09, 14 November 2009 (UTC)

We also shouldn't believe everything people say about others, not only about themselves. When to use in-text attribution is a matter of editorial judgment, and it's not restricted to self-published sources talking about themselves. I'd be worried about adding a clause that we can't predict the consequences of, given that editors have a tendency to cling to the precise wording of policies. SlimVirgin ^{talk|contribs} 04:21, 14 November 2009 (UTC)

God, I wish they did...—S Marshall ^Talk/_Cont 11:10, 14 November 2009 (UTC)

More accurately, they do when it suits them. :) SlimVirgin ^{talk|contribs} 08:41, 16 November 2009 (UTC)

Burden when removing reliably sourced material

I added (and SlimVirgin swiftly removed) the following underlined clause to the article:

The burden of evidence lies with the editor who adds or restores material; however, the burden of evidence also lies with the editor removing published or otherwise reliably sourced information.

Slim says this is a major policy change. I don't see that. Wikipedia is a verifiable encyclopedia which anyone can edit; if you add something with a reasonably reliable source, the advantage should be on you and not someone who just doesn't like the material. The remover should have to show that the information doesn't work because (1) it is inaccurate, (2) it is likely to be inaccurate (the source isn't really reliable), or (3) it violates policy through being undue weight or something similar. This is how Wikipedia does in fact operate, and this why I can't simply troll through articles deleting sentences which contain information I personally don't like. II | (t - c) 08:15, 16 November 2009 (UTC)

I think we'd have to be very careful how we worded anything like that, and I'm not sure it's worth it, because the issue is a matter of common sense. If a source really is good, appropriate, and is being cited correctly, it will tend not to be removed. I know that a consensus of editors sometimes forms on talk pages and agrees to remove material that strictly speaking is fine, but our adding something to the policy isn't going to stop that. SlimVirgin ^{talk|contribs} 08:43, 16 November 2009 (UTC)

• Not only is it common sense, it's also already enshrined in policy. The relevant shortcut is WP:PRESERVE. I see little benefit in repeating that here.—S Marshall ^Talk/_Cont 10:49, 16 November 2009 (UTC)

I have always had a slight problem with WP:PRESERVE... I understand and approve of the general intent... but it does not take into account edits that are intended to preserve the information, but improve the wording of a passage; and it does not take into account edits that replace one source with a better, more reliable source. We should make it clear that it is the information that is to be preserved... not the exact wording and not the source. Blueboar (talk) 15:38, 16 November 2009 (UTC)

I think it's important to preserve all three, each for different reasons, but that's a matter for a different talk page! :)—S Marshall ^Talk/_Cont 17:17, 16 November 2009 (UTC)

Not entirely... it does relate to this discussion. The proposed addition says: "... the burden of evidence also lies with the editor removing published or otherwise reliably sourced information." My point is that, unlike adding information (which needs to be verifiable), removing information does not always require "evidence". Information can be removed from an ariticle for many reasons... it might be out of date (superceeded by more modern sources), the statement as written may be overly POV, or may constitute Original research. I can agree that, in these situations, there is some degree of "burden" on the editor removing the information to explain why he/she is removing it... but an explanation is not the same as presenting "evidence". Indeed, sometimes there is no "evidence" to present... an editor might simply think the information is trivial or irrelevant to the topic of the article. Sometimes the determination of what should be mentioned in an article is simply a matter of editorial preference. Yes, the reason for removal should be stated (either on the talk page or in an edit summary), but that is not the same as saying that there is a "burden of evidence". That term implies presenting some degree of "proof", and presenting "proof" is not always applicable. Blueboar (talk) 18:03, 16 November 2009 (UTC)

Another valid reason to remove information is the information is too detailed for the intended audience of the article. It would be inappropriate to give, in the September 11 attacks article, a detailed description of the Boeing 767 mechanical systems, unless they related directly to the hijacking. --Jc3s5h (talk) 19:56, 16 November 2009 (UTC)

Which is covered in WP:UNDUE. Fences&Windows 20:54, 16 November 2009 (UTC)

In other words... This policy talks about there being a "burden of evidence" when adding material, because providing sources is directly connected to the concept of verifiability. However this policy does not talk about there being an equivalent "burden of evidence" for the removal of material because there are a lot of reasons for removing material that have nothing to do with the concept of verifiability. The various reasons for removing material are dealt with elsewhere. Blueboar (talk) 21:04, 16 November 2009 (UTC)

I have long railed against the "burden of evidence" being put on the person who wants to put back material that was unsourced due to the fact that if an editor can take time to remove information then they should take time to do a google search first and either cite it themselves or put in their edit summary that they couldnt find a source for it. It should be their burden to do their due diligence before removing it and causing others to clean up their mess.

• With that said, I have also supported there being extra measures in place to put the burden on those who remove sourced material. I see no harm in having an extra sentence here saying that the editor who removes sourced material has to have extra load of burden on their shoulders to prove WHY they removed it and not to throw the burden on others; since this is the policy most often quoted by those who go around removing information it stands that this policy does need to cover it as well even though it may be more under the pervue of another policy, perhaps link to the relevant policy.Camelbinky (talk) 22:59, 16 November 2009 (UTC)

Language is important.. I feel strongly that we do not want to use the word "PROVE" in connection to the removal of material ... we need to "EXPLAIN" why we removed it. Sometimes this will be a simple edit summary... sometimes it will require discussion on the talk page. Blueboar (talk) 23:21, 16 November 2009 (UTC)

I agree with Blueboar about both not wanting to use the word "prove", and with that sometimes it can be a simple edit summary. However the proposed new sentence II wants to add doesnt say "prove" and if we can all agree that it is already existing policy and therefore his sentence doesnt contradict policy nor hurt Wikipedia; my question is- why cant we agree that it is ok to have here? If it isnt hurtful then I see no reason in not having it in. It will be a good reminder to those who like to quote this policy when recklessly removing information they disagree with, sourced or otherwise and who probably ignore WP:PRESERVE anyways. As I stated I cant stand the existing wording anyways, would love to see a discussion on changing it but I know how that would go as too many editors already think the status quo must always be defended as if it was written by a race of super-Wikipedians who handed it down to us verbatim and we're too stupid to rewrite it because they knew better than we do.Camelbinky (talk) 23:32, 16 November 2009 (UTC)

Ok... then let me rephrase... I strongly oppose using the words "burden of evidence" or similar phrasing in conjunction with removal... for all the reasons before said. I conceptually support the idea that "any time editors remove information they should explain why they are doing so (either in an edit summary or on the talk page as appropriate.)"... but (unlike the burden to provide sources) I don't think this statement has anything to do with the concept of Verifiability and, therefore, I don't think such a statement belongs in this policy. Blueboar (talk) 00:16, 17 November 2009 (UTC)

My attitude is that people who edit an article should watch the article for at least a few days to see how others react. Since the editor who added material knows where to find the source, and is (or should be) watching the page, it is acceptable to remove any unsourced addition within a few days of the addition, just because the editor who added it didn't bother to provide source information. --Jc3s5h (talk) 00:11, 17 November 2009 (UTC)

I do not like the proposed new wording. The point of "burden of evidence" is to stop fringe or POV claims being dropped into articles, possibly with an edit summary or note on the talk page that the information is well known, or treated in common text books, or available in a search, or whatever. Another editor can remove the claim with edit summary "unsourced". It is not up to the removing editor to prove that no source is available: that is what "burden of evidence" means. The question of what to do when an editor removes a sourced statement needs to be dealt with somewhere else, but isn't it just WP:BRD? Johnuniq (talk) 00:46, 17 November 2009 (UTC)

• • I don't really have a strong interest in playing out a long debate on this. I do think this is related to WP:V in that it expresses the fundamental idea that Wikipedia is made up of verifiable information and that such information should not be removed without good reason. It's true that this is expressed in other policies such as WP:NPOV (all notable positions should be expressed) and perhaps WP:PRESERVE, although PRESERVE feels more technical and does not feel like it communicates the idea - that reliably-sourced information pertinent to the topic can't just removed based on someone's opinion.
  • Blueboar: maybe "evidence" doesn't seem like the correct word, but citing a policy is certainly a form of evidence. Explanation doesn't feel like it conveys the message, as it makes it sound like any old excuse can do, when the point is that the editor removing should show that the information doesn't work (that something has been shown is subject to interpretation, obviously). It's a philosophical thing which I don't imagine would change much, but often individual editors don't seem to realize that when they block the inclusion of a small bit of well-sourced information based on their opinions, they are not doing right by Wikipedia.
  • Jc3s5h and Johnuniq, this is targeted specifically at well-sourced information, which I presume to be largely works which reference their sources and are published by decently well-known institutions (although the particular details of that can be debated at another forum). There's nothing in this wording which could promote fringe or POV claims; to the contrary, this puts in words the spirit of Wikipedia and reduces the ability of individual people to block content they don't like - a form of POV pushing. I recall that this sort of behavior was quite evident at ADHD not long ago, when a pro-drug person really didn't like the negative research sourced to reliable, recent, PubMed-indexed reviews and would repeatedly try to cut it out citing "undue" or somesuch. However, the problem runs across the Wikipedia. Typically both groups will call the other fringe and try to keep out content, even content sourced to recent, academic, secondary sources. Admittedly, it's not always academic sources (and academic sources often cannot describe all views, given that they have a particular bias) - in politics newspaper articles might be the best available, and in environmental issues public interest groups might be the only ones covering the information from the particular perspective. In finance, asset managers and corporations will often have a relevant perspective. II | (t - c) 02:43, 17 November 2009 (UTC)

The standards for adding and removing information are much different. To add information, there must be a source or it must be unlikely to be challenged. To remove information, reasons such as "it makes the article too long" or "it isn't interesting" are adequate reasons. Of course, deliberately suppressing reliably sourced information because it does not agree with an editor's personal opinions is not a good reason to remove information, but it is wrong to write a policy as if every deletion is a deliberate attempt to suppress the truth. --Jc3s5h (talk) 02:54, 17 November 2009 (UTC)

"Not interesting" is a reason to remove information? Since when, and according to whom is the information "not interesting"? That is a subjective call that should first go to the article talk page. Same with "it makes the article too long", since if the article is too long the information should be preserved but split into another article per WP:PRESERVE (does anyone read that policy anymore?!), this can be done by veteran editors who know the article and not a passerby removing unsourced or sourced material based on their own decisions.

• If information is factually correct, and sourced then you really need to have a damn good reason to remove it, and if someone returns it what "burden of evidence" are you expecting of the person returning it?, it is already sourced! So by default the "burden of evidence" must fall on the remover, they need to in some manner show why the information is harmful (too long or not interesting is not harmful and is not a valid reason btw, imo). If it is not sourced, why not look up in two minutes a source for it?, alot of editors get really pissed by that as, evidenced by the support I got at the VPP when I brought this very question up. In fact another user created a warning template to use on editors who removed sourced material without reason. This very policy warns you that it is common courtesy to first find a source yourself prior to removal, for the very same reason I give here, it pisses people off.
• Other policies spell out why this crap about "unsourced information added by IPs and newbies needs to be taken out" is crap- ignorance of our way of doing things, procedures, and our policies do not negate a good faith edit. If an IP or newbie adds new things, they might not watch a page, good information is lost when all it could have taken was two minutes at Google for a veteran editor. Are these types of editors who remove stuff that busy on Wikipedia they cant be bothered? If so why are they trolling to other articles just to remove unsourced material? Be constructive, not destructive.Camelbinky (talk) 03:46, 17 November 2009 (UTC)

By your logic, if I were to add the statement "Barak Obama is the current President of the United States<ref>http://www.whitehouse.gov/administration/president-obama</ref>" into the article on My Favorite Martian, it should not be removed... dispite the this factually accurate and sourced fact to be completely irrelevant to the topic.

Or perhaps relevancy qualifies as a "damn good reason" in your book? Blueboar (talk) 04:24, 17 November 2009 (UTC)

Any decent writing guide will emphasize the need to make any piece of writing concise. Camelbinky does not seem to accept this principle. I will always feel free to rip out crap that does not contribute to the theme of an article, and give that as my reason for doing so. --Jc3s5h (talk) 04:17, 17 November 2009 (UTC)

I don't accept the premise that "If information is factually correct, and sourced then you really need to have a damn good reason to remove it," and I doubt that most Wikipedia editors do. I'm currently reading a biography of Alice Sheldon that's almost 500 pages long. If I were to insert every accurate, sourced fact from that book into the Wikipedia bio it would become the longest article on Wikipedia. The article on Meryl Streep, for good reason, doesn't cite every review mentioning her for every film she's ever appeared in. Outside of their immediate families, the odds are that nobody cares who Barack Obama's fourth-grade classmates were, and turning up a class yearbook or list wouldn't justify inserting the list of names into an article about him. Wikipedia would be greatly improved, on balance, if all the sourced reports of incidents embarassing celebrities were removed; te signal-to-noise ratio there is minuscule. There are many adequate reasons for removing sourced content from articles that have nothing to do with verifiability, and everything to do with writing a useful encyclopedia article that satisfies the needs of its users. Hullaballoo Wolfowitz (talk) 16:53, 17 November 2009 (UTC)

I can see what you mean. I don't think that adding a statement that one has to make a good argument for removing well-sourced material by itself contradicts summary style, staying on topic, undue weight, or other rules, but I can see how it could be thought of that. Perhaps such a statement is more trouble than it's worth. II | (t - c) 21:04, 17 November 2009 (UTC)

Support Inclusion While we may agree that not 'every' factually correct and sourced bit of information need be preserved, this is solved simply by achieving consensus for whether that bit of information falls under say WP:WEIGHT. We do need this line in there, and it is not a major change, it is in fact captured by WP:TEND and precedence has been established in this arbcom principle. The arguments put forward by Wolfowitz and others above are sound, but they still need to be made on a case by case basis on a relevant talk page. Unomi (talk) 20:22, 17 November 2009 (UTC)

I have boldly reinstated the language that holds that the expected behavior is to make a good faith effort to find sources, rather than merely good practice. Unomi (talk) 20:46, 17 November 2009 (UTC)

I object to this change... I could barely accept saying this was "good practice"... I strongly disagree with saying that it is "expected behavior". We have options...

If an editor thinks that some bit of unsourced info is accurate... he can either a) leave it unsourced, b) find a source himself, or c) tag it so that someone else will provide a source.

If an editor thinks that some bit of unsourced info is inaccurate, he is not (nor should not) be "expected" to waste time searching for sources that he has every reason to think probably don't exist... he should either a) tag it or b) remove it (depending on what the statement is). Yes, he could be wrong... a source might exist... but it is up to those who want to keep the challenged info to provide it and demonstrate that the removing editor is wrong.

As for Sourced information... I think there is a general consensus that "Sourced info should not be removed without cause"... and I don't have a problem with this concept. What I do question is the appropriateness of talking about it in this policy. The reason why this policy talks about the "burden of evidence" for unsourced information is because that is an issue directly related to the concpet of Verifiability. However, the reasons for removing sourced information usually center on issues that have nothing to do with to Verifiability (and the vast majority of removals relate to some other issue). Thus, while the concept probably should be stated in some policy, I think it is inappropriate to state it this policy... in WP:Varifiability. Blueboar (talk) 21:01, 17 November 2009 (UTC)

I'd like to understand why I get "ripped on" for stating that "if you remove sourced material you better have a damn good reason", but then Blueboar states "general consensus is that 'Sourced info should not be removed without cause". How is that any different than what I said?! Blueboar's example about Barack Obama is an example of having a damn good reason, exactly what I wanted; everyone is reading way too much into this. Why cant it be "remove sourced information only if you have a good reason, and you need to state it"; exactly what I said, but somewhow another editor says "Camelbinky doesnt understand", I dont understand what exactly?!

• As for Blueboar's three "options" an editor can do with unsourced info, why is three even an option? If you have time to tag, find a source, if you're too lazy then dont be going around tagging (you as in the editor, not referring to Blueboar, whom Ive always respected and am sad to be not be agreeing with him right now). The problem I see with removal is- it can get buried under subsequent edits and therefore not be at an editor's watchlist for him/her to see offhand, per wp:preserve if it isnt hurtful or illegal then what harm does it do to keep it? If you cant find a source for it or its obviously not true then go ahead and remove it, but dont be lazy and not look for a source, I have seen LOTS of edit summaries by other editors that say "added source, took 2 mins" and sometimes it gets ruder and almost a personal attack on the editor who removed the information. I myself- if someone removes information and I have to put it back and it took less than 2 minutes to find the source...I go to their talk page and I let them know that and ask that in the future they take the 2 mins themselves instead of robbing me of that 2 minutes I couldve spent doing other good things.Camelbinky (talk) 03:10, 18 November 2009 (UTC)

removing unsourced information and "best practice"

Here's why it is an option... any reasonably intellegent person can read an article on just about any topic and think... "hmmm, I think this statement needs a source". I do not need to be familiar with the topic to say this. However, I do need to be at least somewhat familiar with the topic to know whether a particular source is reliable or not.

I could try to source the statement myself, but I know my own limitations ... I know I run the risk that, in my ignorance, I will inadvertantly cite a source that is in fact completely unreliable. Alternatively, I can identify the problem statement with a citation tag ... so that other editors (editors who are familiar with the topic and who do know what the the reliable sources are likely to be) can fix the problem instead.

In other words... I prefer to tag, not becuase I am too lazy to look for a source... but because I do not feel qualified to provide one... and I hope someone else will be. Blueboar (talk) 23:01, 19 November 2009 (UTC)

Tagging, fine; but removing should be a last resort used only if the editor has looked for a source and either a- couldnt find one in a reasonable time (a few mins is all I'm asking) or b- found a source that completely contradicts it. I am of course only talking about normal writting, not vandalism, issues unique to BLP (that dont apply to all articles), etc. I dont see why we cant remove the whole "burden" and "good practice is to take time to look for a source" and replace it wholesale with the following-

• If information is uncited either find a source to cite it or tag it with the appropriate template. More specific policy applies to biographic articles, see WP:BLP for those.Camelbinky (talk) 03:05, 20 November 2009 (UTC)

I more or less agree with you when it comes to articles on topics that I am not familiar with (and why I don't remove unsourced information in most articles)... but, when it comes to articles on topics that I know well (admittedly, a small number... but there are some), I could not disagree more. In those few topics, I am very familiar with the reliable sources, and I am also familiar with what they say on the subject. If I come across a unsourced statement in an article on one of those topics, and I think it is unlikely that any reliable source will support it, I have no hesitation in removing it. I am not going to waste even a few seconds searching for a source that I am sure does not exist... nor should I be required to. It is up to those who wish to keep the questionable statement in the article to demonstrate that my well informed removal was, in fact, incorrect.... by providing a source when they return the information. That is the essence behind the Burden of Evidence section of this policy, and I support it full square.Blueboar (talk) 04:34, 20 November 2009 (UTC)

And I agree with you on the fact that if you (or me or anyone) removes information on articles they know about then it cant be misused; and I similarly have a small core topic that I work on and know about, and I too do not remove information on articles I dont know about, you and I are very similar on that. However, it isnt editors like you and me who misuse the burden clause and then quote it as justification. Editors who troll around to articles they dont know about are the problem. Their actions and mindset of course wont be changed by changing the wording here, but perhaps you can work with me on a wording compromise/change on here to make it so they would be less able to justify removing information they know nothing about? I believe you once wrote you didnt like the "best practice" wording that suggested one should find a source oneself first, do you think there is a better way we can address my concern or are the trollers a necessary evil I have to live with?Camelbinky (talk) 04:48, 20 November 2009 (UTC)

I really don't think this is as much of a problem as you make it out to be. I don't think many editors "troll around to articles" removing unsourced information just for the hell of it ... oh, there may be a few... but the vast majority of editors only remove information from articles on topics they think they know something about; and they only remove information they think is incorrect in some way. Yes, they may be wrong on both counts... they may not know the topic as well as they think they do, and the information may actually be correct, but it is up to those who want to keep the information to demonstrate that they know better and that the information is, in fact, correct... by returning the information with a source... a source which should have been there from the start.

That is my beef with the best practice statement... the best practice is for editors to research their topic and to have their sources lined up before they add material to an article... and to cite those sources when they add it. This is what we should be telling editors to do. If more editors followed this best practice, we would have far fewer unsourced statements for anyone to challenge and remove. Best Practice is: "Never add unsourced information to Wikipedia, especially if that information is likely to be challenged". Blueboar (talk) 13:59, 20 November 2009 (UTC)

Self-republished sources

I've added the following text to the section on self-published sources, which I believe is noncontroversial and presumably reflects the intent of the existing policy:

"On occasion, a book previously published by a commercial house may be reissued in a self-published edition after the rights have reverted to its author. Such a later reissue does not mean that the book is no longer a reliable source, although if the text of the later edition is revised, claims unique to the later edition may require independent sourcing."

As commercial publishing (at least in the US) continues to shrink, it's becoming increasingly common for authors to reclaim their rights to out-of-print books and reissue them through print-on-demand publishers or other publishers providing self-publishing services. This is probably most common among genre fiction writers, but also includes various types of nonfiction, and will almost certainly become more common as e-book publishing expands. Since the reliability of the source can't be undermined by its republication, there's no policy justification for treating a self-published reissue as less reliable than the trade/commercially published original. I would have thought this went without saying, but I've seen contrary editing on a few occasions. Hullaballoo Wolfowitz (talk) 21:05, 16 November 2009 (UTC)

That's fine if the print-on-demand publisher is reliable, or if the author is reliable enough to be taken at his/her word that no changes were made, or that the changes were confined to particular areas. However, if the author is not reliable, then a reliable source would have to state that the republished version is the same (with certain described exceptions) as the original. --Jc3s5h (talk) 21:19, 16 November 2009 (UTC)

I agree with the concept here... but I am not sure why it is necessary. If a book is reissued under self-published but was previously published by a commercial publishing house, just cite the commercially published edition.

The only time I could see having to cite the self-published re-issue would be cases where some bit of information was not included in the commercially published edition... in which case, that specific information is self-published (with all the caviats and restrictions that apply). Blueboar (talk) 21:22, 16 November 2009 (UTC)

Blueboar wrote "If a book is reissued under self-published but was previously published by a commercial publishing house, just cite the commercially published edition." I can't do that if I don't have the original publication. If I don't have some means, outside of the republication itself, to establish that the republication is reliable, I can't use it. Jc3s5h (talk) 21:28, 16 November 2009 (UTC)

My point was that if you can locate the commercially published version, and avoid using the self published re-issue, then you don't need to worry about the issue. You can often find older editions through google books. Blueboar (talk) 22:28, 16 November 2009 (UTC)

burden vs BLP?

I was involved in a now-settled dispute which was somewhat complicated by the fact that WP:BURDEN and WP:BLP don't seem to exactly agree. WP:BURDEN says "Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living persons or organizations, and do not move it to the talk page." WP:BLP doesn't have an equivalent statement that I can find. The closest I can see is: "Contentious material about living persons that is unsourced or poorly sourced—whether the material is negative, positive, or just questionable—should be removed immediately and without waiting for discussion."

So WP:BLP says that contentious unsourced material should be immediately removed (this word is used multiple times), and WP:BURDEN says damaging unsourced material should be immediately removed. But these aren't the same thing at all; something can be damaging without being contentious, or vice versa. Indeed, the language above from WP:BLP explicitly does not take the positive or negative nature of the information into account, and the rest of the article in general doesn't imply that positive and negative material should be treated differently (except when dealing with people who are relatively unknown, in WP:NPF).

I'd think at least one of these pages should be changed for consistency (and possibly WP:RS too which uses WP:BLP's language). It seems to me that the sentence in WP:BURDEN is intended to summarize WP:BLP, so WP:BURDEN is what should change. Gruntler (talk) 06:58, 17 November 2009 (UTC)

The concern about BLPs has always been that incorrect negative or damaging information is inserted destroying people's lives or careers, and leading to lawsuits against Wikipedia. There is no more urgency to remove untrue positive information about a person than there is to remove untrue positive information about a country or a company. As such, I think that the BLP rule should be made consistent by stating that unsourced questionable material should be removed from BLPs if it is 'negative or damaging'. --LK (talk) 08:48, 17 November 2009 (UTC)

That's not quite right. The Wikimedia Foundation's resolution regarding BLPs [6] also stresses the inappropriateness of inaccurate/unverifiable positive claims in BLPs, and calls for "special attention to the principles of neutrality and verifiability in those articles" -- not limited to content of any type. That concern underlies the "whether the material is negative, positive, or just questionable" language in WP:BLP. Hullaballoo Wolfowitz (talk) 15:47, 17 November 2009 (UTC)

Untrue positive information included in the biography of one person could harm a person who competes with the biography subject. This might happen in elections, or if several actors are competing for the same role. --Jc3s5h (talk) 19:13, 17 November 2009 (UTC)

In this case, BLP should take precidence over BURDEN. If one of them needs to be conformed to the other, I think it is BURDEN that needs to be modified to match BLP. Blueboar (talk) 19:17, 17 November 2009 (UTC)

Only in the case of biographies though, right? I for one would not want to have BURDEN match BLP in order to take BLP's stricter requirements and apply them to all types of articles. I've seen editors (who obviously dont know what BLP stands for) quote BLP in an argument about an article that wasnt a Biography of a Living Person (or a biography at all for that matter, it was a building).Camelbinky (talk) 03:16, 18 November 2009 (UTC)

Correct... although I suspect that my understanding of the normal Burden requirements are far stricter than you would like, I agree that there is a difference between them and BLP. Blueboar (talk) 20:54, 21 November 2009 (UTC)

Reliable source - overall guidance

The current Reliable Sources section contains overall guidance starting in its second paragraph. This section is possibly the most important part of the article in getting practical instruction across to the editor trying to get appropriate verifiablity into an article. Unfortunately, some of the language is hard to follow, such as run-on sentences, going from specific to general instead of the other way around, mixing topics across paragraphs, etc. I'd like to propose the following change. Currently reads:

In general, the most reliable sources are peer-reviewed journals and books published in university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. Electronic media may also be used. As a rule of thumb, the greater the degree of scrutiny involved in checking facts, analyzing legal issues, and scrutinizing the evidence and arguments of a particular work, the more reliable the source is.

Academic and peer-reviewed publications are highly valued and usually the most reliable sources in areas where they are available, such as history, medicine and science. Material from reliable non-academic sources may also be used in these areas, particularly if they are respected mainstream publications. The appropriateness of any source always depends on the context. Where there is disagreement between sources, their views should be clearly attributed in the text.

New proposed text:

Here's a good rule of thumb for determining the level of reliability of your source material: It is best to choose material that has undergone a great degree of scrutiny in checking facts, analyzing legal issues, and scrutinizing its evidence. Anything that has a greater degree of scrutiny in these areas is typically more reliable than material that had less scrutiny in these areas.

Usually, the most reliable sources are peer-reviewed journals and books published in university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. Electronic media can be valid sources, subject to the same reliability criteria. Academic and peer-reviewed publications are highly valued and usually the most reliable sources in many areas, such as history, medicine and science. Material from reliable non-academic sources may also be used in these areas, particularly if they are respected mainstream publications.

The appropriateness of any source always depends on the context. Where there is disagreement between sources, their views should be clearly attributed in the text.

Opinions? Dovid (talk) 10:56, 17 November 2009 (UTC)

I don't think it works as presented, Dovid. I didn't quite get what you were saying about the current version being hard to follow. Is it just that you want the "rule of thumb" part to go first, or is it more than that? SlimVirgin ^{talk|contribs} 13:19, 17 November 2009 (UTC)

More than that. Yes, rule of thumb should probably go first, as it sets the table for what follows and is generally more applicable. But here are other things I've attempted to fix:

• I turned two paragraphs into three. The older two paragraphs really have three topics (general guidance, specific best types, and context/conflicts), so it makes sense to have three paragraphs. Previously, the middle topic on best sources was split artificially between the two paragraphs, and the rule of thumb was in between the two parts of best specific examples. This breaks the train of thought and is somewhat disorganized, and I fixed that.
• Electronic media sentence is misleading, as it leaves the impression that it continues from the list of most reliable sources. That really still depends on the publisher/author, so I added the "subject to..."
• The rule of thumb was an awkward sentence. It was a long conditional with a short conclusion (the greater the BLAH BLAH BLAH BLAH the more reliable the source is). Makes it hard to follow and connect the two "halves." Plus, ending in "the source is" raises form issues, similar to the prepositional ending. I'm not a stickler for those, but the first three times I read the sentence, my brain was looking for what comes after the "is."

So, when you say it doesn't work as presented, I hope you were refering to my explanation, now made clearer, and not my text. Dovid (talk) 14:05, 17 November 2009 (UTC)

How about this?

The appropriateness of any source depends on the context. In general, the best sources have a professional structure in place for checking or analyzing facts, legal issues, evidence, and arguments. As a rule of thumb, the greater the degree of scrutiny given to these issues, the more reliable the source. Where there is disagreement between sources, their views should be clearly attributed in the text.

The most reliable sources are usually peer-reviewed journals and books published by university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. Electronic media may also be used subject to the same criteria. Academic and peer-reviewed publications are highly valued and usually the most reliable sources in areas where they are available, such as history, medicine, and science. Material from reliable non-academic sources may also be used in these areas, particularly if it appears in respected mainstream publications.

SlimVirgin ^{talk|contribs} 14:44, 17 November 2009 (UTC)

I like this last version SV. Blueboar (talk) 16:15, 17 November 2009 (UTC)

Thanks. I think it flows a little better. Dovid is right that the current version is a bit choppy. SlimVirgin ^{talk|contribs} 16:18, 17 November 2009 (UTC)

Proposed third paragraph to supplement Slim's two:

If a source comprises user-submitted content, such as messageboard posts, blogs, IMDB, or Wikipedia itself, then it is not a reliable source.

Might as well spell it out.  :)—S Marshall ^Talk/_Cont 16:40, 17 November 2009 (UTC)

That's incorrect. We do allow user submitted content in some situations, e.g. from established experts who have been published elsewhere. Crum375 (talk) 17:20, 17 November 2009 (UTC)

Change "not" to "not usually" in the appropriate place, then.—S Marshall ^Talk/_Cont 22:00, 17 November 2009 (UTC)

I would suggest this change:

The most reliable sources are usually peer-reviewed journals; and books published by university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers.

Otherwise readers might think peer-reviewed journals are normally published by university presses, and that other peer-reviewed journals are not among the most reliable sources. Statement of potential bias: I am a member of the Institute of Electrical and Electronics Engineers, a publisher of many peer-reviewed journals. --Jc3s5h (talk) 19:22, 17 November 2009 (UTC)

Done, good point, though it opens up the problem mentioned in the section below, of some peer-reviewed journals being a bit dodgy -- making sure they were published by university presses at least protected us against that. I don't think it's an issue we can get into here, though, because as you say, there are good peer-reviewed journals that aren't published by universities. SlimVirgin ^{talk|contribs} 07:26, 18 November 2009 (UTC)

It's always been implied that the order that the sources are given in is important. i.e. from most reliable (peer-reviewed) first. Can we spell that out as: "The most reliable sources are usually (in order of preference): peer-reviewed journals; books published by university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers." LK (talk) 03:20, 18 November 2009 (UTC)

The order isn't important, and no one should see a hidden implication that it is. All the sources we mention are regarded as reliable in principle, and the way in which they ought to be used depends on context. SlimVirgin ^{talk|contribs} 04:43, 18 November 2009 (UTC)

Starting with Slim's, ah, version. The "context" sentence shouldn't be the lead, it is more of a final modifier, i.e., the basic idea is show how to rate sources to begin with, and only then say "now that you have figured out how reliable your source is, use context to decide whether that is reliable enough for the subject." I also would prefer not to use the phrase "professional structure," it is presumtuous. The previous language was more neutral about how the fact checking was done, just that more checking is better. And "disagreement about sources" is still breaking the flow between the rule of thumb and the specific application (journals, etc.). So, here's my edit to SV's:

In general, the best material comes from publishers that are careful and consistent about checking and analyzing for evidence, facts, legal issues, and arguments. As a rule of thumb, the greater the degree of scrutiny given to these issues, the more reliable the source. The most reliable sources are typically peer-reviewed journals and books published by university presses; university-level textbooks; magazines, journals, and books published by respected publishing houses; and mainstream newspapers. Electronic media may also be used, subject to the same criteria.

Where there is disagreement between sources, the opposing views should be clearly shown and attributed in the text.

The appropriateness of any source depends on the context. Certain topics will demand higher value sources, while other subjects accept a somewhat lower grade of source. For example, subjects such as history, medicine, and science are rigorous. They also have top notch sources available, such as those described above. For these subjects, lesser sources, such as corporate web sites, blogs, or brochures, would be inappropriate. However, for less rigorous subject areas, such as company information, travel, and geography, a wider variety of sources might be considered reasonable, such as corporate web sites and government travel brochures. So, there are many degrees of subject, and many degrees of source reliability, and editors must find the right match and balance.

Opinions, again? No need to nitpick, the tweaks could go on forever, but I do want some quality and consensus before posting this in Dovid (talk) 06:22, 18 November 2009 (UTC)

Is the phrase "peer-reviewed journals and books published by university presses" intended to mean that peer-reviewed journals are published by university presses? --Jc3s5h (talk) 06:27, 18 November 2009 (UTC)

Dovid, I disagree with your version. I'm happy for it to be copy edited, so long as the new text really is better written, but not a rewrite and an expansion, and the "top notch" sources thing takes us into very subjective territory. I'm going to add the basic copy edit I suggested above, because it doesn't change the content but flows better than the current text. If you want to argue for a change of content, that should be done separately. SlimVirgin ^{talk|contribs} 07:14, 18 November 2009 (UTC)

Problems with pseudo-NPOV/Suggest greater qualifications

As we all know (I hope...) even the best of sources can contain errors of various kind, including typos, poor fact-checks, out-dated information, and a number of other problems. This is an unavoidable problem and something that we have live with.

However, this is complicated by two other issues, namely:

1. Some of the sources often considered reliable (at least by the proponents of a cited statement) are not. Major newspapers, e.g., tend to by abysmal at fact checking, yet are often taken as infalliable with statements like "X is the leading newspaper in Y! This is a reliable source! I am right and you are wrong!" (Note that this is not a policy issue, per se, but rather a lack of judgement in the individual editors.)
2. The current style of writing typically amounts to a claimed fact followed by a reference, claimed fact followed by a reference, etc.

As long as we write on topics that are sufficiently uncontroversial and have a near-consensus, this works well. However, when we move on to controversial subjects (typically something were strong ideological, religious, nationalistic, or similar feelings and POVs come into play), this becomes very dangerous, and articles tend to be written in a pseudo-NPOV manner, where claims are stated as unequivocal facts followed by a reference (as in 2.) and any attemp at altering or critiqueing these statements are met by claims that they are referenced claims. (The hitch is often that radically different, often opposite, claims with equally many references could be provided---which brings the issue down to what faction has more time and editors.)

In order to combat this problem, I suggest that the recommendations
are changed to unequivocally require that disputed statements be written
in forms like "According to x ..." rather than as in 2. Doing so would
go a long way towards reducing both the POV problems that we do have and
the many, many fights over what articles are POV, NPOV, who is right, who
is wrong, ...

(Exactly how to define "disputed" goes beyond this post, but in first sketch: a) A sole editor disagreeing with a referenced source is likely too little. b) A number of editors, or a sole editor with own, reputable references should be enough. c) Cases were a mere majority of all scientists agree, as opposed to a near-consensus of scientists, would typically be disputed.)

Note that a somewhat similar issue applies with regard to NPOV: Editors often argue "NPOV means that the article should reflect the majority view in its main formulations [or views according to followers, or similar]; therefore, this statement should be made as an absolute."---which is not (necessarily) true: If three out for dentists prefer X and the fourth Y, then it is not a good practice to claim that X is the truth, but instead statements like "three out of four ..." should be used. (Note that this is, again, not as much a policy issue as poor judgement in editors.) 88.77.135.55 (talk) 23:12, 17 November 2009 (UTC)

This is not a verifiability issue — it relates more to WP:NPOV. When making a non-contentious and properly sourced statement, it can simply be stated as fact: "X was born and raised in Jackson, Mississippi.[1]" If there is some controversy, the statement can be presented with an in-text attribution, for example: "According to the Jackson Register, X was born and raised in Jackson, Mississippi.[1]" If there are conflicting sources, the ones representing the majority and larger minority views should be presented: "According to the Jackson Register, X was born and raised in Jackson, Mississippi, although a biography published in 1972 states he was born in New Orlean, LA and his family moved to Jackson when he was four.[1][2]" This is all part of presenting sources neutrally, in a proper balance, and the proper format should be decided by the editors on the page. Crum375 (talk) 00:08, 18 November 2009 (UTC)

Also, more complex source conflicts and/or apparent errors can be discussed in footnotes, so as not to detract from legibility. Crum375 (talk) 00:13, 18 November 2009 (UTC)

New paragraph

I don't know when this was added, but I just noticed it, and it's problematic:

Care should be taken in evaluating the quality of journals, as journals created to promote a particular viewpoint may claim peer review, but have no meaningful peer review outside of adherents of the viewpoint they promote. Such journals generally represent the consensus view among such adherents, but may otherwise be considered unreliable; for instance, the prominence and notability of their views should be ascertained by using other sources. Examples include, The Creation Science Quarterly, Homeopathy, and Journal of Frontier Science (the last of which claims blog comments as its peer review process).

This seems too general for the policy, and somewhat POV, given the particular journals that are mentioned, which may very well be reliable sources in articles about the issues they cover. In addition, all journals arguably exist to promote a certain view of the world, and all peer reviews will tend to include only adherents of those views; anyone strongly disagreeing will likely not be part of the group regarded as "peers." As this opens up a can of worms, I think we're better off not saying anything about it and leaving it to editorial judgement. SlimVirgin ^{talk|contribs} 07:22, 18 November 2009 (UTC)

I've no problem with the statement but I think it belongs in the reliable sources guideline rather than a policy. Rather than being too general I think it is too specific for a policy. Dmcq (talk) 09:00, 18 November 2009 (UTC)

I can understand why it was added. The policy explicitly mentions that "The most reliable sources are usually peer-reviewed journals..." and this leads some editors to the (incorrect) assumption that anything with the word "Journal" in it's title must be considered a reliable source. The fact is, not all "Journals" are equal, and some definitely do not have a reputation for fact checking or accuracy.

That said... I agree that this probably belongs in WP:RS rather than here. As far as this policy is concerened, I think we can get across the idea that there are exceptions to the statement that "The most reliable sources are usually peer-reviewed journals..." if we simply highlight (ie italicize) the word "usually". Blueboar (talk) 13:56, 18 November 2009 (UTC)

I think the paragraph has been there for at least a few months, and hence many people have viewed it and found it acceptable. I support it's inclusion, at least as a footnote, as there really are a lot of 'journals' that are written and reviewed only by a particular fringe group, and viewing them as reliable is a serious problem. LK (talk) 14:10, 18 November 2009 (UTC)

I think there may be some confusion between reliability (for WP purposes) and neutrality, or representation of the mainstream. We can have a fringe group which believes the Earth is flat, and publishes its peer viewed journal, Flat Earth Monthly. The articles in that journal, if we assume they are vetted by representatives of the group, are reliable for WP purposes as representing of the views of that group. OTOH, they are so far removed from the mainstream, and such a tiny minority, that their views should not be included at all in the Earth article, because of WP:UNDUE. But if we were writing about that group in the Flat Earth article, that particular peer reviewed journal would be preferable to some individual member's blog. So as bottom line, for reliability or verifiability purposes, a peer reviewed journal, even if all peers are adherents, is generally a better source (for the views of a group) than somebody's website. But WP:NPOV must also be considered as part of the equation. Crum375 (talk) 14:41, 18 November 2009 (UTC)

Um, I see your point. I wonder how to put that. After all journals besides Flat Earth might disagree with the earth being flat and would not be considered reliable by flat earthers, and yet Flat Earth would be reliable source or sorts about them. Dmcq (talk) 17:21, 18 November 2009 (UTC)

The policy presently states "The most reliable sources are usually peer-reviewed journals...." Almost anything, even a scrap of paper pulled out of a waste bin at a metro station, can be used as good evidence of something. Journals published by fringe groups should not be considered "among the most reliable sources". The most reliable sources are those that can be used with confidence for any relevant purpose, and will be accepted by everyone except fringe groups. (Exception: in the area of religion, few adherents accept anything outside his/her belief system, including those who believe the supernatural does not exist.) --Jc3s5h (talk) 17:38, 18 November 2009 (UTC)

I think there is still confusion between verifiability/reliability of a source, and the neutrality of its presentation in relation to a given topic. The Flat Earth journal could be an excellent verifiable and reliable peer-reviewed source to describe Flat Earth adherents views about the Earth, and why they believe so, in the dedicated Flat Earth article. It would not belong in Earth per WP:UNDUE. To use the word "reliability" on WP as a synonym to "closest to the truth" is plain wrong. That is not what we mean by it here; our goal is to present what sources say about a subject, not to strive for "truth". For us, reliability only means a well-vetted published source relating to a topic. We do need to gauge the relative prevalence and acceptance of sources by the mainstream, so we can decide which sources to use and how to present them per WP:NPOV and WP:UNDUE. Crum375 (talk) 19:50, 18 November 2009 (UTC)

There are at least two aspects to reliability: quality and breadth of application. While a journal published by a fringe group might be of high quality in the sense that it accurately portrays the views of the group, it has a very narrow breadth of application. Nature, on the other hand, not only has well-vetted articles that are as accurate as can be expected, but the articles can be used in any relevant Wikipedia article without having to think about whether the claim made in the article falls within some very narrow application area. We can say that Nature is more reliable than any fringe-group journal because it at least matches them for accuracy, and beats them for breadth of application. --Jc3s5h (talk) 20:19, 18 November 2009 (UTC)

If the topic is Flat Earth, then Flat Earth journal may be a better source than Nature. So while I agree with you that if I wanted to get a mainstream view on a scientific topic I would use Nature, I would not necessarily see it as the best source for cults or fringe groups. So the point is that it's relative: the most appropriate source depends on the subject matter, independent of any "intrinsic" reliability. Crum375 (talk) 20:34, 18 November 2009 (UTC)

You are essentially arguing that because a source is reliable for the claims of the group publishing it, it therefore is a reliable source. This is incorrect. For example, self-published material has long been accepted as a reliable source for the claims for the individual or group publishing it, but this does not mean that self-published material is a 'reliable source' (per our policy). A 'reliable source' is one that is reliable for generic statements of fact, (e.g. "the Earth is an oblate spheroid"), not for qualified statements of the beliefs of a particular group (e.g. "according to the Flat Earth society, the Earth is disk-shaped"). LK (talk) 06:07, 20 November 2009 (UTC)

LK, I think you have a basic misunderstanding of the concept of RS on Wikipedia. A WP source is not RS because it "is reliable for generic statements of fact". On WP, a source is reliable if it has been published and we have reason to believe that it has undergone some vetting by third parties. In general, the more and better the vetting, the more reliable the source for WP purposes. It could still be extremely "POV", and claim that the Earth is made of cheese, and people made of marshmallows, which would make it a tiny minority view, in which case it would be unacceptable due to NPOV/UNDUE, unless we were writing an article about a particular group that holds that view. To be able to assert a statement without in-text attribution, e.g. "the Earth is round", we'd also require a source which we consider mainstream and part of an overwhelming majority, with at most a small fringe dissenting. I think your confusion stems from conflating neutrality and reliability, which are two distinct concepts on WP, though both influence what we may include and how we present it. Crum375 (talk) 16:34, 25 November 2009 (UTC)

That is not my, nor I feel anyone else's, reading of 'reliable source'. Ask anyone in academia what a peer-reviewed journal is, and they'll have very definite criteria – criteria that publications like Creation Science Quarterly do not meet. An in-house magazine put together by a small group, is not a peer reviewed journal regardless of the claims of the group. You are essentially arguing that there is no significant difference between a peer-reviewed journal like the Quarterly Journal of Economics and Creation Science Quarterly, this is something that i cannot believe the majority of Wikipeidians believe. If what you say is correct, what is to stop me and my cousin from founding an association, and starting the Bi-monthly Journal of Economics, start printing it ourselves, claim that it is a peer reviewed journal, and then start citing from it on articles here, as a 'most reliable source'? — Preceding unsigned comment added by Lawrencekhoo (talk • contribs) 12:34, 25 November 2009

LK, WP policies are self-contained, and define their own terminology, which may differ a bit from usage elsewhere. In our case, RS is not a source we accept as being "more correct", or even "more mainstream". The only criterion for WP RS is that it is published and properly vetted. It may represent the view of a tiny minority and be completely contrary to mainstream views, and still be RS for WP purposes. Your confusion, as I noted above, is that you seem to conflate NPOV/UNDUE with V/RS, which are distinct concepts on WP. Crum375 (talk) 17:42, 25 November 2009 (UTC)

I disagree. A RS is a source with a reputation for fact checking and correctness. We usually accept them as authoritative at least within their domain. The hypothetical Flat Earth Journal is not making claims about Flat Earthism, it makes claims about geology. It is not a RS for anything except the position of the journal. Nature, PNAS, or, for more general news, the NYT or The Times, are reliable in the sense that we trust them to usually get it right. --Stephan Schulz (talk) 17:52, 25 November 2009 (UTC)

"An RS is a source with a reputation for fact checking" - agreed. That's another way of saying we know that it is well vetted. As far as "correctness" — what is that? How do we gauge "correctness" of a source? And which side does the gauging, if it's controversial? I believe "correctness" is part of NPOV, i.e. representing a certain faction or view: mainstream, majority, minority, etc. Since WP takes no sides in controversies, per NPOV, we must focus on verifiability, which means that if we state X said Y, any reader can verify our statement, and we ensure that X is a source with a reputation for fact checking, i.e. performs good vetting of its published output. Crum375 (talk) 18:01, 25 November 2009 (UTC)

And let me add that we address the "mainstreamness" part of the equation by following the NPOV/UNDUE policy requirement, which tells us to exclude tiny minority views, and present and weight information according to its prevalence among the reliable sources. Crum375 (talk) 18:06, 25 November 2009 (UTC)

No. That is an unrealistic and practically useless definition of NPOV. We do make and can make absolute statements if supported by sufficiently reliable sources. Figuring out what "sufficiently reliable" is, is part of our job as an editor. From our current FA: "Count Nikita Moiseevich Zotov (1644 – December 1717) was a childhood tutor and life-long friend of Russian Tsar Peter the Great." No "according to X here, and it would not be helpful. A simple item of news can be sourced to a reliable newspaper, and an uncontroversial statement of scientific fact can be sourced to PNAS. But no absolute statement can ever be sourced to the Flat Earth Journal. NPOV is good, but if we fall into the postmodernist trap (all reality is a subjective construct of society) we lose all relevance. --Stephan Schulz (talk) 18:29, 25 November 2009 (UTC)

We can make "flat" statements, or assertions, if the editors decide that a certain view is non-controversial, after considering the available reliable sources. So we can say "life on Earth began 3 billion years ago", without in-text attribution (but with an inline footnote) because the editors of that article decided that the minority's opinion that life began 5,000 years ago is so tiny and non-mainstream that it can be relegated to a footnote, or its own article about biblical views or creationism. This is exactly why we have both RS and NPOV: the former to tell us when a source can be accepted in principle, and the latter to tell us how to present it, if at all. If NPOV tells us it's mainstream with no significant opposition, we can present it as an assertion. If there is controversy with a non-trivial opposing view, we can use in-text attribution, etc. If it's tiny minority, it disappears from view except perhaps in its own article. The important point is to realize that NPOV and RS are separate and distinct concepts and policies, which should be both used to tell us how to present information. Crum375 (talk) 18:44, 25 November 2009 (UTC)

The consensus here seems to be that the paragraph is need, and I'm taking silence here as consent. Does anyone have a problem with adding the paragraph back in as a foot note? If not, I'll stick it in tomorrow or so. LK (talk) 05:17, 22 November 2009 (UTC)

I fear that the above 18 November comment by User:Dmcq (edit | talk | history | links | watch | logs) has been overlooked not been fully addressed. This para is about RS, not V. It belongs in WP:RS not WP:V.LeadSongDog come howl 16:02, 25 November 2009 (UTC)

This policy spends a considerable amount of time discussing reliable sources, so the section belongs here as much as at WP:RS. Also, this policy makes a very big claim that peer-review journals are usually the most reliable, so it makes sense to qualify that statement immediately instead of being redirected to another policy. Angryapathy (talk) 17:45, 25 November 2009 (UTC)

My point, evidently put too tersely, is that this page already has far too much discussion of reliable sources to be good policy. Good policy should be simple and readily understood. It certainly should not need to use weasel words such as "usually". That is the path to becoming just a guideline. We should leave the interpretive guidance to WP:RS.LeadSongDog come howl 18:04, 25 November 2009 (UTC)

This policy specifically refers to WP:RS for more details about evaluating reliable sources. Saying that it makes a claim is a misunderstanding of how these things work, policies and guidelines describe the consensus of editors. Whether something is true or false is not the point and statements in policies do not need justification according to WP:V WP:N or anything else. Dmcq (talk) 18:07, 25 November 2009 (UTC)

I think there is some confusion about what peer-review means. It does not mean reviewed by a group of people who hold the same opinions but by people who are acknowledged experts in the field. The Flat Earth Journal could not be peer-reviewed because it advances the theory that the earth is flat and does not acknowledge that even were the flat earth theory generally accepted by geographers, that the consensus could change. An article advancing the flat earth theory would never be accepted for publication in a peer-reviewed journal like The Geographical Journal and even if it was the consensus of other contributors would discredit the theory. The Four Deuces (talk) 18:15, 25 November 2009 (UTC)

I think another way to look at it is that "peers" are all those who have carefully studied the field in question, not just those who have studied the field and come to a particular conclusion. A peer-reviewed journal is one which submits papers to review by a group of peers who reasonably represent the views of all peers as a group, in the sense that a jury represents the views of the community from which they are drawn. A jury drawn from those who are members of the Klu Klux Klan or who live in fear of the same would be no jury, and a trial at which they served would be no trial. --Jc3s5h (talk) 18:47, 25 November 2009 (UTC)

Agree. The fact is, there is a world of difference between reliable peer-reviewed journals and something printed by a fringe group that claims peer review. As the place in policy to go to for what is or is not a reliable source, this basic fact should be noted in WP:SOURCES. LK (talk) 09:48, 26 November 2009 (UTC)

I feel this policy already says more than enough about the whole subject. WP:SOURCES already says stuff about this in more than one way. Any more and it'll be reiterating things even more. Policies should not reiterate and they should not duplicate. It is definitely not the way to make clear policy. The place to look for more details about reliable sources is WP:Reliable sources. Dmcq (talk) 11:23, 26 November 2009 (UTC)

Second opinion

I am in dispute with Someidiot at List of red-light districts about sourcing. Everything on the list is currently sourced and I think a source should be added immediately when adding an entry. He disagrees and thinks (among other things) that a lot of entries are common knowledge therefore don't require a source or at least not immediately. To me that's degrading the article. See this discussion. Could an uninvolved editor give some input? Garion96 (talk) 10:32, 20 November 2009 (UTC)

Subtle Vandalism

I am concerned about minor vandalism that might have profound effects in the future. Recently, I ran across and editor who was adding unsourced middle names to a variety of BLP articles. It appeared he was just making them up as he went along, as in one case, in two different edits of the same article he had inserted two different middle names. The only reason I caught this was it happened on a page I was watching; I reverted, look at the editor's history, and found this going back a couple of months. With so many sites mirroring Wikipedia content, many of changes had already propagated, which could lead to a circular reference. This type of vandalism is hard to detect, particularly not caught immediately. I am not sure what the answer is; I just wanted to bring it to everyone's attention. --Seaphoto^Talk 18:01, 21 November 2009 (UTC)

Adding middle names without a source is not vandalism... but purposely adding incorrect middle names (ie knowingly adding false information) is. I would report this to WP:ANI... they will be able to look into the situation and assuming it is vandalism, may be able to perform some form of blanket revert of all this user's edits. Blueboar (talk) 20:51, 21 November 2009 (UTC)

Source does not support assertion

When a reference is given but does not actually support the article text citing it, what is the template to tag it with instead of ? --JWB (talk) 07:42, 25 November 2009 (UTC)

When I see those situations, I copy the ref to talk and state why I've removed it from the article. If the bit of text really needs a ref, add cn tag. - Hordaland (talk) 11:35, 25 November 2009 (UTC)

Template:Failed verification. --GenericBob (talk) 12:24, 25 November 2009 (UTC)

However, quite a lot of spam links are added to certain kinds of articles simply to promote websites (see WP:REFSPAM). If you notice one of those, do not say "failed verification", just delete the reference with edit summary, for example, "remove [[WP:REFSPAM]]". Where appropriate, you might also insert {{cn}}. Johnuniq (talk) 23:07, 26 November 2009 (UTC)

BYU Studies

All: Please do not respond further to this section. Go to Wikipedia:Reliable_sources/Noticeboard#BYU_Studies and reply there.

I am engaged in a discussion at Talk:Kirtland Safety Society about the use of BYU Studies as a source. My claim is that it should be noted inline when a source makes a claim that supports Joseph Smith and the source is published by the church. I am asking for help here regarding whether:

• That's just silly, the source can stand on its own
• Use of the source should include an inline note that it was published by a LDS-funded journal
• The source should be used sparingly, as it is a WP:SPS

My claim is that BYU Studies is a WP:SPS for the following reasons:

• The explicitly stated mission of the journal is "We want to share these publications to help promote faith, continued learning, and further interest in our LDS history with those in the world who have a positive interest in this work." [7] (emphasis mine)
• The submission guidelines are explicitly religious, not just scholarly. [8] They cite LDS scripture by chapter and verse. They also say: "BYU Studies strives to publish articles that openly reflect a Latter-day Saint point of view and are obviously relevant to subjects of general interest to Latter-day Saints, while conforming to high scholarly standards." (emphasis mine)
• The staff of the journal and its Church history editors are employed by BYU [9]. BYU staff are subject to dismissal for contradicting Church doctrine, see Academic freedom at Brigham Young University
  • Richard E. Bennett
  • Brian Q. Cannon
  • Kathryn M. Daynes
  • Steven C. Harper
  • Steven Sorensen appears to be dead
  • Frederick G. Williams

A thought experiment I propose is to imagine if we would be willing to use a source that made the same statements regarding its partiality, but with the opposite polarity. If a university journal had the explicit purpose of attacking or debunking Mormonism, while claiming to also be scholarly, would we not be very cautious about using it as a source?

The counterpoints to this offered on Talk:Kirtland Safety Society include that the journal is a university journal that is peer reviewed and that at least one of the authors who wrote the one of articles in question is a credible history scholar. (The others may be as well, they have not been discussed)

I claim that the existence of pockets deep enough to support universities and their attendant journals should not exempt any organization from WP:SPS, especially when the journals admit being POV on certain issues.

Thoughts? WhyDoIKeepForgetting (talk) 14:39, 27 November 2009 (UTC)

I tend to agree. This looks like a very insular publication. Unless the journal is widely (and not just as an example for weird claims) cited in the wider academic literature, I would reject this as a RS. Looking at Google Scholar, the most frequently cited article from BYU Studies has 11 citations, and many have none at all - that means the journal has essentially no impact. BTW, I suggest you move the question (and my reply) over to WP:RSN, where it is more apropos. --Stephan Schulz (talk) 15:06, 27 November 2009 (UTC)

At your suggestion, copied there. All: Please do not respond further to this section. Go to Wikipedia:Reliable_sources/Noticeboard#BYU_Studies WhyDoIKeepForgetting (talk) 15:10, 27 November 2009 (UTC)

RfC on RS

Wikipedia_talk:Reliable_sources#RfC_on_page_move. Should Wikipedia:Reliable sources be moved to Wikipedia:Verifiability/reliable sources to become a subpage of the sourcing policy, WP:V? There would be no change in either page's status: the policy would remain policy, and RS would retain its status as a guideline. 23:40, 27 November 2009 (UTC)

Wikinews

I've read the archive discussion on Wikinews? Was there any further developments? We have a Wikinews template, very intrusive like Wikicommons, that directs the reader to external sources, not at the end, like Wikicommons, but within the article itself. This seems intrusive and probably wrong.

But why should a wikinews source be given more visibility than say, some other link (at the end) which might actually be of more value even though less "up-to-the-minute"? Student7 (talk) 15:10, 28 November 2009 (UTC)

We're generally not linking to Wikinews as a reference, we're linking to it as a sister project. Using it as a source would need to be discussed on the article talk page. Fences&Windows 22:22, 28 November 2009 (UTC)

Clarification on SELFPUB?

Here's an example situation that seems to happen quite frequently:

1. A new editor (let's call him Alex) arrives and, in good faith, makes a pseudoscientific statement in an article (let's say "The Earth is a cube" in Earth)
2. A more experienced editor (let's call her Bea) sees the article and removes that statement, with a concise edit summary ("unreferenced")
3. Alex readds the statement, now with a citation to Joe Bloggs, one of the numerous "Cubic Earth" blogs out there: "The Earth is a cube<ref>Joe Bloggs</ref>"
4. Bea removes the statement again, this time with reference to WP:SELFPUB, with an edit summary along the lines of "Joe Bloggs is unreliable, as it is a self-published source"
5. After reading WP:SELFPUB Alex reinserts the material, this time with an attribution: "According to research by Joe Bloggs, the Earth is a cube".
6. Following this a long discussion ensues. Bea says that the inclusion of Joe Bloggs fails WP:UNDUE. Alex typically makes the argument: "But according to SELFPUB, it's OK to use Joe Bloggs as a source on what Joe Bloggs thinks about the Earth"

Does WP:SELFPUB need clarification? Or is the example above something that's supposed to happen? The generic "They should not be interpreted in isolation from one another" in the lead does not appear (to me) to be sufficiently clear. Gabbe (talk) 09:32, 30 November 2009 (UTC)

The wording of the opening sentence of SELFPUB has been changed a few times due to being either too strict or too lax compared to the actual accepted practice. It used to say "only in articles about themselves", but that is far from an absolute in practice. There also used to be a restriction that the material cited should be "relevant to the notability" of the subject, but there wasn't even agreement on what that meant. Certainly WP:UNDUE applies, but how to work that into the wording of SELFPUB I don't know. PSWG1920 (talk) 10:05, 30 November 2009 (UTC)

Would it be wrong to add a sixth proviso, "inclusion of the material does not give undue weight to the self-published source"? Gabbe (talk) 10:22, 30 November 2009 (UTC)

Do you have a recent example where this has come up? Unomi (talk) 11:10, 30 November 2009 (UTC)

Yes. But is it necessary to "name names" here, so to speak? Gabbe (talk) 12:11, 30 November 2009 (UTC)

I am a bit wary with essentially duplicating WP:UNDUE here, and even more so with broadening its 'reach' as it were. A frequent refrain within the sphere of articles mentioned on fringe theories noticeboard is that relatively weak sources (such as blogs and other SPS) are allowed as 'real scientists' rarely write to scientific journals debunking (insert pet peeve here) as such, SPS are allowed for color commentary on what is expected to be scientific consensus. Wording such as you propose may disallow some of those sources as it could be argued that the 'source' is given undue weight rather than focusing on the 'material' relative to our appreciation of what scientific consensus may be. In the example you mention the inclusion of joe bloggs on Earth falls afoul of WP:UNDUE due to what we expect scientific consensus to be. I expect that the cases you have met are less black and white, WP:NPOVN and WP:FTN might prove useful for the future. Unomi (talk) 12:22, 30 November 2009 (UTC)

I agree my proviso isn't very good, as all it says is "Also, don't violate other policies". I'll see if I can hammer out a better suggestion. I've tried the noticeboards, but posts at WP:FTN often go unreplied. Gabbe (talk) 12:49, 30 November 2009 (UTC)

Actually, in your example, Bea should have pointed Alex to WP:SPS... the section immediately before WP:SELFPUB, as the problem is that the source being used is an unreliable blog... (unreliable in that Joe Blogs is not likely to be a published expert). Besides, we are not really dealing with a self published statement where the author is talking about himself (Joe Blogs), but a self published statement where he is talking about something else (Earth), so SELFPUB does not apply.

That said... I think it would be a good idea to mention WP:UNDUE somewhere in either SPS or SELFPUB... we don't need to repeat the entire thing... just include a one sentence note with a link should be enough. Blueboar (talk) 13:22, 30 November 2009 (UTC)

Is there any particular reason why we say "may be used as sources of information about themselves" rather than "may be used as sources of information about their author(s)"? It seems to me that the latter would be clearer, and would avoid the inherent problem outlined here — that any claim can be transformed into a statement about the source, hence making the policy ineffective. I'm trying to think of a reason why "about their author(s)" would be insufficient in a legitimate situation... Jakew (talk) 13:45, 30 November 2009 (UTC)

Exactly. The argument from Alex at point 5 is that the statement is a piece of information about Joe (namely, what Joe's opinion is), rather than about the Earth. Gabbe (talk) 15:45, 30 November 2009 (UTC)

And it is at this point that we cross the line from discussing a reliability/verifiability issue into discussing NPOV/Undue weight issue... even if we accept that Joe's blog is a reliable source for verifying a statement as to Joe's opinion, we have to ask the question of whether Joe's opinion is noteworthy enough to mention in the context of the Earth article. That is going to depend on who Joe is. Blueboar (talk) 16:06, 30 November 2009 (UTC)

I would hope that WP:UNDUE covers the issue, although I acknowledge the real problem posed by Gabbe. I would argue to Alex that the article is about the Earth and not about Joe, so Joe's opinion on the shape of the Earth is only relevant if WP:RS is satisfied. Also, I have found that mentioning WP:REDFLAG can be helpful: an extraordinary claim needs extraordinary proof, not just someone's opinion. Johnuniq (talk) 22:43, 30 November 2009 (UTC)

WP:SOURCES#Questionable sources

Can this part be reworded? Because of this sentence, "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", someone said that reviews aren't excepted at all. Or can you give me a good way to explain it? Joe Chill (talk) 23:53, 30 November 2009 (UTC)

It isn't saying that all personal opinions make a source "questionable" ... those of recognized experts, for example, are allowed. Also, personal opinions that appears in a reliable source such as the New York Times Op Ed page... or a book review in a similar publication... would be OK as well (if phrased as opinion and not as fact). However, a publication that was nothing but the personal rantings of some Nut Job with an axe to grind would definitely be a "questionable source". Blueboar (talk) 02:41, 1 December 2009 (UTC)

what is unsourced?

Everything is different from its original in some degree if it is not copied, to be more accurate, including its context also! So anything can be named "unsourced" or misleading if it does not copy exactly something else.

Does that contribute to an original article writing community? — Preceding unsigned comment added by 62.38.141.211 (talk • contribs)

Pardon? Statements of fact need to be backed up by a source. What's your point? Fences&Windows 20:56, 7 December 2009 (UTC)

In an extreme interpretation, 62.38.141.211 is correct. Anything not exact an exact copy will involve some interpretation and some form of "original research" to extract information. You could even say an exact copy is in some ways an "interpreted" alteration of the original. But this is obviously not the intent and a less strict interpretation is agreed by most the community: one checking the reference should easily be able to recognize the presence in the source of the referenced-from information in wikipedia. (John User:Jwy talk) 22:36, 7 December 2009 (UTC)

Hermeneutics is that way. Fifelfoo (talk) 22:41, 7 December 2009 (UTC)

from "62.38.141.211" (excuse my formatting difficulties - not much accustomed yet)

"Fences and windows" - Statements of fact originally come from experiment and perception/witnessing and not by just copying a source. If we would only copy somehow, there would not be even one bit of info written anywhere. I am astonished to the extent the community is used in copying. Ok, if we are writing about what Albert Michelson wrote, sure this can and should be verifiable. But this could not apply to Albert Michelson, otherwise he (and then we) would not have anything to write.

"Jwy" & "Fifelfoo" - Hermeneutics is surely a more flexible way of targeting the issue. I agree.

But the issue is not complete with the Hermeneutics extension. It is using what the scientific community uses for centuries now : personal accounting of the info provided (which contradicts our ip and nicknaming in some way). Until there is some, we may only wish for the best sourcing available but not insist on absolute sourcing or original info will be lost. Just like everything else around. Have you seen any book sourcing each paragraph? No, because we trust the writer. How come here everything not sourced is a deletion candidate (according to some above) ? Good faith is what carried wikipedia to here.—Preceding unsigned comment added by 62.38.141.211 (talk) 10:24, 8 December 2009 (UTC)

Is this some roundabout and complicated way of saying there is something about the editing of the article about Albert Michelson that you disagree with? Or is there some specific place in this policy where you think it should be changed? Dmcq (talk) 11:24, 8 December 2009 (UTC)

from "62.38.141.211" - nothing to do with the Michelson article of course, only an example in clarifying my view on the issue of verifiability, hoping it is useful. To make it even more clear, he (Michelson) made an experiment, the outcome of which was never before written anywhere. He was witness to some original info. This is my point. There is original, not referenced anywhere, info. Should it be excluded? The policy has worked wonderfully till now. The only reason i am writing this is because i have seen a different trend, to a more strict environment.

If his experiment has not been published, then it should be excluded. If he has published it, then we can discuss what he says about his experiment in that publication. See: WP:NOR. Blueboar (talk) 13:48, 8 December 2009 (UTC)

Thank for the clear answer. I will stay by this written policy I didn't see earler. Therefore i will not again write about my village, my company, our commons, until someone publishes something about them. —Preceding unsigned comment added by 62.38.141.211 (talk) 14:14, 8 December 2009 (UTC)

Glad to help... and when someone does write about your village, company etc, stick to summarizing what the source says on the subject... don't add your own observations, analysis, or interpretation. Blueboar (talk) 14:22, 8 December 2009 (UTC)

If a source is reliable, does that mean it is also independent?

Following on from discussions at WT:Notability#RFC: Self promotion and indiscriminate publicity, there seems to me some doubt that are reliable source is also an independent source.

The general notability guideline goes to some length to make it clear that independent sourcing is a pre-requisite for article inclusion, yet nowhere in WP:V is independence made an explicit requirement. Similarly, WP:RS is also silent about independence, and the nearest thing to a guideline on this issue is contained in the essay Wikipedia:Independent sources.

Should this policy be amended to make it clear that a reliable source means that it is also independent? Appologies if this question has been asked before. --Gavin Collins (talk|contribs) 11:02, 10 December 2009 (UTC)

No. E.g. WP:RS makes it very clear that a self-published source is a reliable source for statements about itself. Technically in official wiki speak there is no such thing as an absolutely reliable source. Reliability or otherwise always depends on the claim that it is used for. WP:RS must not be changed just because it doesn't exactly fit the way people try to re-use it for notability questions. That's not its purpose. Hans Adler 11:18, 10 December 2009 (UTC)

No, reliability and independence are separate (though often related) issues. You are a indisputably reliable source about how to spell your own name - your lack of independence in the matter is irrelevant. Roger (talk) 11:54, 10 December 2009 (UTC)

(Asked to comment here) A source can easily be both not-independent of something, and yet reliable about it (verifiability). What is not so easy is to be not-independent and yet be reliable evidence how much notice the world at large takes (notability).

Example: An football club's official publication and annual report can be a verifiable source that company X sponsors football, but however much it highlights X's around its pitch, its players' clothing, and its publications, its mentions of X is unlikely to be good evidence that the world at large takes note of company X (because their motive is likely to be affected by their sponsorship deal). FT2 ^{(Talk | email)} 12:13, 10 December 2009 (UTC)

Some of these answers are a bit obtuse for me, but what about media in countries that do not have a free press? (ie are not independent) Or is this another question altogether? How about media in countries that do not have a free press and are involved in conflict or hostilities with another country. Can their words be taken as reliable with respect to the conflict? Stellarkid (talk) 16:57, 10 December 2009 (UTC)

No simple rule can be written to determine whether a source is reliable. Issues must be decided on a case-by-case basis on the article's talk page. --Jc3s5h (talk) 17:13, 10 December 2009 (UTC)

Agreed, no simple rule can written to determine whether a source is independent either. But is reliablity mean that it is also independent? And if so, where does it say this?
A good example may be found in the article Kender, where creators of the fictional characters are a source of significant coverage about their creation in the form of annotated commentary, which has been provided to be used by teachers. The commentary itself is effectively a reliable secondary source, but is clearly not independent. Like the expert commentary that accompanies a classic text, it contains commentary and analysis but is clearly linked with the primary text.
In this case, is the commentary reliable, even though it is not independent from the creators and publishers of the primary work? --Gavin Collins (talk|contribs) 16:14, 11 December 2009 (UTC)

Plenty of RSes are not independent. It's generally reasonable to take people's word on the details of their own life and work, with certain exceptions (e.g. claims that are self-aggrandising, claims that have been contradicted by other credible sources, and cases where the person has a record of dishonesty). --GenericBob (talk) 07:55, 12 December 2009 (UTC)

Going back to Stellarkid's question, editorial judgement would at least require cautionary disclosure in such cases. In-text attribution to the source might read along the lines "State controlled Soviet news agency Tass reported on 30 February that Stalin had won re-election by a landslide 99.7% of the popular vote." Balance then would require an additional reference to independent sources: "In the West, both Reuters and the Associated Press noted the Tass report, observing further that there had been only one choice on the ballot." (Insert suitable citations for my fictional example). LeadSongDog come howl 16:00, 12 December 2009 (UTC)

To try and flush out what I am getting at, I would have to ask the question, should we amend this policy so that there a reliable source is also an independence source? and if so, what would the best form of wording? --Gavin Collins (talk|contribs) 10:53, 14 December 2009 (UTC)

Editorial opinions or points of views are not a problem regarding to Verifiability. Keeping a neutral point of view is not a requirement for a source to be reliable (as long as keeping a point of view does not make it do other things that would turn it unreliable, such as lying, difamate, use hoaxes, etc.). In those cases, what we have to do is to identify the point of view of the source and point as such, pointing as well the different approach by media with conflicting viewpoints.

As for the level of free press in the country, that's a too generic reasoning. The media must be checked on it's own, as it's possible that some media may decide to remain independent despite of government pressure. And it must be checked as well which topic is being referenced: a media that's completely "loyal" to the government may not be reliable for making a neutral description of an scandal the government is involved into (such as the lack of free press), but may be perfectly reliable for other non-politic topics MBelgrano (talk) 13:44, 14 December 2009 (UTC)

I think what everyone is getting at is this: A source does not need to be neutral (or independent) to be verifiable. The key is that we need to be neutral in how we use the source. There are times when a non-neutral, non-independent source will be appropriate, and times when it won't be. Blueboar (talk) 14:41, 14 December 2009 (UTC)

So when this policy says "no reliable, third-party sources can be found for an article topic, Wikipedia should not have an article on it", does that mean that the third party source does not need to be independent of the subject, e.g. excludes works produced by those affiliated with the article topic including (but not limited to): self-publicity, advertising, self-published material by the subject, autobiographies, press releases?
I understand that a source does not have to be neutral, but does the requirement for "third-party sources imply independence? --Gavin Collins (talk|contribs) 15:41, 14 December 2009 (UTC)

AFAIK, yes. A Quest For Knowledge (talk) 17:37, 14 December 2009 (UTC)

"Third-party source" and "independent of the subject" are both difficult to interpret, and are bad wording to allow to remain in the policy, or to introduce to the policy, respectively. "Third-party" has little meaning outside of contract law; the fact that Gavin.collins has to ask questions about it shows how bad it is. "Independent of the subject" only applies when the subject of the article is a person, corporation, government, or similar organization. It makes no sense to talk about "independent of the subject" if the subject is the solar system, algebra, or ancient poetry. "Disinterested" is a better word to use. Jc3s5h (talk) 17:59, 14 December 2009 (UTC)

When this policy says "no reliable, third-party sources can be found for an article topic, Wikipedia should not have an article on it", it is repeating what is stated at WP:NOTE, which requires that notability be established through sources that are not directly connected to the topic. But the bulk of the policy is dealing with verifying individual statements and paragraphs. Once notability is established, we certainly may use sources that are directly connected to the topic.

To give an example... say we are writing an article on a religious movement. First, we need to determine if Wikipedia should have an article on it at all... we need to establish that it is notable by our standards, that someone other than an adherant of the movement has noticed the existance of the movement... we do that by discussing and citing what reliable secondary sources have to say about the movement. However, having done that, it would also be appropriate to mention and cite what the movement has to say about itself. For that we can use non-independant sources. Blueboar (talk) 19:43, 14 December 2009 (UTC)

Can I get a clarification? What exactly does "independent of the topic" mean? Can I establish the notability of the band Led Zepplin using cite from Rolling Stone Magazine? Is Led Zepplin the topic or is music the topic? A Quest For Knowledge (talk) 20:09, 14 December 2009 (UTC)

In the case of the article entitled "Led Zepplin" the band is the topic of the article. In this case material written by members of the band, or by those speaking on their behalf (such as a press release from their record company) would not be "independent of the topic", while an article in Rolling Stone on the other hand would be "independent of the topic" (since Rolling Stone has no direct connection to the band). Blueboar (talk) 22:20, 14 December 2009 (UTC)

In broader terms, the answer is that "independent of the topic" does not mean "non-specialized media". Even if Led Zeppelin had appeared just in magazines or media specialized in music (and not in newspapers or vanity press, for example), as long as those media are not directly related with them we can consider the band notable. MBelgrano (talk) 22:28, 14 December 2009 (UTC)

(edit conflict)

• Independent means that the content of the published article cited is not controlled (to any significant extent) by the person or organisation who is the subject of that cited article or of the WP article in which the citation is to be used.
• The topic is what the first sentence of the article says the topic is. So far as possible this coincides with the title of the article.
• Yes, citing a substantive article on any rock band in Rolling Stone Magazine is more than sufficient to establish its notability. Perhaps not so much for an 18th century symphony orchestra (although that could make for an interesting discussion).LeadSongDog come howl 22:37, 14 December 2009 (UTC)

(unindent)As LeadSongDog says, "independent of the topic" only makes sense if the topic is a person or organization. If the topic is global warming, and we required the reliable source to be independent of the topic, we wouldn't be able to use anything published by a publisher located on earth. --Jc3s5h (talk) 22:54, 14 December 2009 (UTC)

It can be extended to other things as well... If the topic of the article is something like a theory... then "independent of the topic" would mean that the theory needs to have been discussed by someone other than its creator or a small group of adherants. For a well known theory like "global warming", notability has been established by reference to a multitude of independant sources (it has been discussed by the press, by other scientists, by politicians, etc.) But for some less well known theory, we might not have enough to establish notability (see: WP:FRINGE for more on that). This does not mean that the theory is right or wrong... just that it isn't notable enough for us to discuss at this point in time. The situation may change in the future (and if it does, then it would be appropriate to write an article on it)... but right now, no.

All of this is better explained at WP:NOTE and at the various topic area guidelines. Blueboar (talk) 00:04, 15 December 2009 (UTC)

I think questions about whether a source is independent or not are answered at Wikipedia:Reliable sources/Noticeboard as a matter of routine, yet strangely there is no mention of independence in this policy, nor WP:RS. I disagree that the origin for the independence requirement originates from WP:NOTE, because my understanding is that the general notability guideline (WP:GNG) is more or less based on, or expands upon WP:BURDEN. The question still remains: should a defintion/requirement indpendence be spelt out in this policy? --Gavin Collins (talk|contribs) 10:58, 15 December 2009 (UTC)

There's a statement above by Bluebore that needs to be addressed: 'When this policy says "no reliable, third-party sources can be found for an article topic, Wikipedia should not have an article on it", it is repeating what is stated at WP:NOTE, which requires that notability be established through sources that are not directly connected to the topic. This is incorrect; NOTE is more stricter than WP:V's statement.

Let's put it this way: given a topic and a source, there are (at least) 4 types of axis that we use to judge that source:

• First or third-party: If someone directly part of the topic (such as the person themselves in a biography article) wrote the article, that's first-party. We prefer third-party to avoid COI-type issues. (that's a make-or-break WP:V issue - articles based only first-party works will not pass it)
• Reliability: Can we trust that source? That's a core part of WP:V, same as above
• Independence from the topic: How much relation did the third-party have with the topic in creating the work? For example, a biography of a person which is extensively based on interviews with that person is not an independent source. (Note that there is no such thing (that I can think of) of a first-party, independent source.) Now, here, it's arguable that this doesn't pass WP:V's muster. If a reliable third-party source has created a dependent work from the topic (say, the person of a Barbara Walters interview), then that seems to be fine for WP:V, though it may be questionable by WP:N's sake. This probably demends a lot on the type of subject and the nature of the third-party to the topic (eg all those town/village stubs are based on the presumption that government listings are reliable, third-party, but dependent sources).
• Primary, secondary, tertiary - In what manner is the topic covered: simple factual details, analysis and commentary on the details, or a overarching summary of the topic? This is not WP:V's perview but instead the added criteria from WP:N in that we would like secondary sources on a topic.

It's important that there are several ways at looking as a source and how it is to be treated for a given topic. But WP:V is strictly demanding that articles at least have third-party, reliable sources; WP:N seeks independence and secondary information from them. --MASEM (t) 14:30, 15 December 2009 (UTC)

Masem, I disagree on one phrase and approach another word from a different point of view. "Third-party" is only fit for use in contract law. It has no place in this policy.

I approach "independent" from the reward point of view. Some rewards to the author and/or publisher (I'll just say "author" for brevity) of a source are unavoidable: using a source in Wikipedia provides some attention to that source. Also, everyone, including the author, benefit from the existence and improvement of Wikipedia. So the question becomes, will the author receive an undue reward if the source or subject receives more public attention? This could be increased sales of a commercial product, influencing public opinion toward a belief that has no merit, or just getting one's "fifteen minutes of fame" for no good reason. If Barbara Walters interviews someone, she isn't likely to become much richer or much more famous than she already is, no matter how much attention the interview receives, so she is independent. --Jc3s5h (talk) 15:28, 15 December 2009 (UTC)

Enough of the wiki-lawyering here... "Independent Third-party source" is an established "term of art" here on Wikipedia, one that has nothing to do with any legal usage... it has a simple meaning: a source written by someone who is not the subject of the article, or directly connected with the subject of the article. What we are saying is simple... if the only sources that discuss a given subject (be it a person or an idea) are those who are directly connected to him/her/it then Wikipedia should not have an article on it. This is a restatement of concepts laid out in more detail at our notability guidelines and at WP:Fringe. Blueboar (talk) 18:00, 15 December 2009 (UTC)

I reject the whole concept of terms-of-art in policies. I reject "third-party" forever. --Jc3s5h (talk) 19:11, 15 December 2009 (UTC)

Well, in that case you are probably in for disappointment, since it is terminology that seems to have consensus. Sorry. Blueboar (talk) 20:52, 15 December 2009 (UTC)

For clarity, is the disagreement Jc3s5h expresses solely with the wording, or is there an underlying disagreement with the concept that people or organizations should be unable to establish their own WP:N? LeadSongDog come howl 21:54, 15 December 2009 (UTC)

What to do when a reliable source is wrong?

What to do when a reliable source is wrong?

See for example this UK Daily Telegraph article Wikipedia hits 3 million articles where tey say

It was formally launched on January 15 in 2001 by Ward Cunningham and Richard Stallman and was initially intended to feed into Nupedia, an earlier now defunct project to produce a free online encyclopedia. Wikipedia soon overtook Nupedia and became the largest free global encyclopedia.

This is easy to fix, there's lots of other reliable sources. But what about BLPs, or obscure topics? Do we just stick with a printed source that's disputed? ருந்தவனிட (talk) 12:28, 18 August 2009 (UTC)

In the specific instance that you mention, it's a matter of not utilizing aberrant reports or tiny minority views. (As you mention, there's a lot of other reliable sources.) I cannot provide an opinion on your questions, because I'm not exactly sure what you are asking. what do you mean by disputed? Do you mean that there are some people on the net that dispute the report? Do you mean that the subject of the article, or a person close to the subject, disputes the source? Do you mean that there is a conflict between sources? Or do you mean something else entirely? --Vassyana (talk) 12:49, 18 August 2009 (UTC)

For empirical science topics, the problem is more serious. In the end, the "verifiability not truth" policy is questionable, because unlike in the humanities, which are mostly a game of arguing PoV, the "hard" sciences may not have absolute truth but do have absolute untruth. And ever so often, this throws a spanner in Wikipedia's works if one argues strictly according to policy.

Three examples:

Kinnareemimus - "The name "Kinnareemimus" was first mentioned in a 1999 paper by Sasithorn Kamsupha, and (as "Ginnareemimus") in a publication by Ryuichi Kaneko in 2000". Actually, the sources as presently cited only support the latter half of the sentence. I kinda happened across the Kamsupha factoid. But as long as we don't have anyone with sufficient Thai skills to the rescue, we can hardly cite an article titled ไดโนเสาร์แหล่งใหม่ที่จังหวัดชัยภูม and published in a (non-peer-reviewed, I presume, but a scholarly RS nonetheless) journal named การประชุมเชิงปฏิบัติการและสัมมนาวิชาการกองธรณีวิทยา...

Archaeopteryx. The problem revolves around the actual specimen to which the name Archaeopteryx is attached. Each and every source that deals with the problem - except one source, that is - asserts that the holotype is one of the skeletons.

But it isn't. It is the (indeterminable) feather, and thus Archaeopteryx is technically the most widely-used nomen dubium (a name that cannot be reliably referred to a specific genus that is known from specific and unequivocally identifiable skeletal remains, in this case).

However, the only source that really digs into this matter is in German and not available online as fulltext, so the barrier to it getting read by editors are almost as high as in the Kinnareemimus example. It has been ignored by other scholarly authors on the topic. As far as I could trace the matter, someone fucked up royally decades ago and didn't read the original source (in this case, ICZN Opinion 607), but nobody noticed. The low-down can be summed up in the following comment:

The fact is that Archaeopteryx the taxon (not ichnotaxon) prevails ovewhelmingly in the literature, and while you are right about all this, it's original research until published. In other words, nothing needs to be changed in the article itself here ;)

So policy seems to force us to spread old an old canard, simply because the source is obscure - and even though in this case it is a peer-reviewed scholarly RS (a source of the highest quality) and overwhelmed by a mass of RS that in this particular point are anything but R.

(Note the commentator himself fell into this trap. The OR claim is strictly speaking nonsense, as anyone can check out now that the secondary sources are online. Still, the number of "RS" that get it wrong is more than tenfold the number of RS that are accurate, so if you want to play a game of "I have more RS (even though they may be unreliable)", it is indeed OR)

Platycryptus (spider). I'll save you the details; they are not especially pretty. Suffice to say I spent some 5 bucks on getting shitty photocopies of the obscure French and German sources needed to resolve this issue. None of the sources, however, explicitly state that the spider genus has an invalid name, because spider and ichneumon wasp researchers have more important things to do than scour each other's publications. But the ICZN Code - the "law" for all of zoology - leaves no wriggling space.

We have raised the matter among some spider experts, but it'll be time - at least one more year - til we gonna get a RS (the reason is that it is being suspected that the "Platycryptus" are not a distinct genus, in which case the issue resolves itself). Til then, the fact remains that Wikipedia is the only source where enough spider and ichneumon wasp material is brought together to actually notice that the spider genus name was never valid: the wasp genus was accidentially left out of Nomenclator Zoologicus, and so every researcher was bound to think that Platycryptus was still an available name. (BTW nobody even think about doing a renamer. A lot of spider experts would be pissed off, and thus their Statement of Intent according to ICZN Code Appendix A(2) is herewith duly repeated.)

The matter can be expanded; there are even more extreme cases where two (or more) RS cannot be reconciled. Typical situations involve high-impact papers that present arguments based on statistical inference, when there is material evidence that flat-out contradicts it available but in a RS which receives little or no publicity (due to being extremely specialist, or non-English, usually)... but that evidence, being from another field of science entirely (say geosciences versus molecular biology) is overlooked by the team of scientists making the high-publicity claim. This is especially problematic because publications like Nature and Science get a lot of media presence but curtail authors' page allowance severely these days, preventing comprehensive coverage of all angles of some new discovery (see here for some professional scientists' comment on this problem).

A hypothetical but easier to understand example:

Suppose one gives somebody's death date according to the available RS. Suppose one then finds another RS that does not concern itself with the person specifically, but merely contains a picture of said individual alive and kicking and holding in their hands a newspaper with a headline regarding some event that only happened after the presumed death date.

Since a person cannot be alive and dead at the same time (at least not when being visible to a camera), clearly, the bulk of RS would not be as "reliable" as it seems. But in a strict interpretation of OR, claiming that all usually-cited sources are wrong and presenting the photo as counter-argument can be already considered OR, since the "dissident" source does not actually make the explicit claim that all others are wrong. And if this person is not especially notable (a long-dead minor celebrity of their time perhaps) the odds of a RS getting published that sets the matter straight for Wikipedia purposes are pretty slim.

Basically, the reasons for the problem are twofold:

• the fact that uncritical use of tertiary sources runs rampant even in RS of the highest quality, and
• the fact that Wikipedians are collectively able to check out many, many more sources than any single expert ever could, and are people with fields of interest so different as to never meet and exchange information in the context of professional research.

Current policy forces us ever so often to assert things as if they were "verified fact", when they are actually empirically, demonstrably and unequivocally wrong.

Still however, the rule allows us to avoid conflicts of opinion that otherwise would be prevalent in all cases where there is no "binary" truth, where there is a grey area of individual opinion that is larger by far than any black or white - i.e. basically all of the humanities. It is one of the reasons why Wikipedia can work at all. It works well insofar as it was intended to work.

Yet it assumes a world more perfect than reality. And it seems to have been authored by people who have little to no experience in epistemology in empirical fields where falsifiability is not just Popper's pipe-dream, but rather the gold standard that can in many instances be met.

We should perhaps ponder changing the rule to verifiability before truth.

That way, we can cope with the fact that Wikipedians as a collective are better (non-original) researchers than any individual scientist could ever be, and will ever so often stumble upon source conflicts that no RS has yet discovered. And we would still maintain the original intent, i.e.: a claim on Wikipedia that has no (R)S to back it up is deletion material if anything, or material at best, but certainly not something to tout or FA-worthy.

As policy is at present, however, it enforces a lower standard of quality than what could be achieved at no detriment to function.

(Also, it looks shitty. We are the only encyclopedia ever to pride itself on telling its reader "not truth", or so it looks - really really bad PR it is. Especially as we do manage quite well, by now, to tell the reader things that are interesting, verifiable and true!) Dysmorodrepanis (talk) 10:27, 25 August 2009 (UTC)

I told MY Truth about threshold some time ago. I feel like we are concerned by the same issu. Take a chair and sit with me ! I don't like the idea but it's sort of sacred ground here; Nothing will realy change until someone realy apply this rule exactly as it's written and get in serious trouble. --Iluvalar (talk) 14:18, 26 August 2009 (UTC)

In the cases supplied above, was there a serious dispute among Wikipedia editors that the best source was the obscure one? If not, then I don't see the problem: just cite the best source, even if it's obscure. And if so, what were the disputes? Eubulides (talk) 14:57, 26 August 2009 (UTC)

Today, I saw a quote that "Verifiable, in Wikipedia, means that a user can find the source. Not -- read the policy -- repeat not that the information is true."...04:52, 17 December 2009 (UTC)

I believe that the verifiability, not truth, has been misused. I presume the author meant "even if you think the edit is true, it can't be used unless the fact is verified". However, the modified converse has been put into policy ("verify, even if not true"). Because of this, I propose the following word change:

Before (beginning of policy):

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.

Proposal for slight modification of wording:</ref>

The threshold for inclusion in Wikipedia is verifiability, not unverified 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 with the caveat that false information is never allowed to be intentionally included.

Suomi Finland 2009 (talk) 23:15, 18 December 2009 (UTC)

Independence: proposed amendment

I propose making the following change to the wording of WP:SOURCES:

Articles should be based upon reliable, third-party published sources with a reputation for fact-checking, accuracy and indpedence.

Independence means that a source is free from pressures associated with a strong connection to the subject matter (such as, but not limited to family relationships, close political affiliation, business dealings or other benefical interest) that may compromise, or can reasonably be expected to compromise, the sources reputation for reliablity.

As I understand it, the issue of independence is closely related to WP:NPOV, as an indpendent source is less likely to be biased than one that is free from preasures associated with a strong connection to the subject matter. Independence is alos an issue for self-published sources, as WP:SPS identfies some promotional sources that are not independent as being more or less self-published. But because independence is not defined explicitly in either policy, perhaps it is worth making the requirement for indepence explicit in this policy? Comments, criticsm and suggestions most welcome.

I should warn you of one issue: that although the requirement for indpendence seems entirely reasonable, this additional reqirement may been seen as instruction creep. It may be that WP:N may be the appropriate venue for the issue of independence to be dealt with, but whether it is or is not, I do not have the answer, but it does occur to me that a source cannot reasonably be expected to be reliable if it is not independent. --Gavin Collins (talk|contribs) 10:31, 17 December 2009 (UTC)

I don't agree with this change because I guess it creates more problems than it solves. I appears to me that with your wording it would be impossible to use a published autobiography in an article on its subject. In my opinion it should always be permissible to base the article about Anne Frank mostly on her diary. In her case there is enough secondary literature to make it unnecessary, but for others it may not be the case. Another example of the top of my head would be Samuel Pepys. Hans Adler 11:13, 17 December 2009 (UTC)

I think the proposal goes a bit too far... I could live with it if we made it clear that this refers to the article as a whole and not to citing specific statements... ie that the bulk of the citations should be to reliable independent sources, but that non-independent sources are allowed where appropriate for specific statements.

To give an example... In the article Adolf Hitler, the article as a whole is, and should be, based upon what respected historians tell us about Hitler and his life... (ie the bulk of the sources cited are to to respected historians)... but if a section were to discuss (in a neutral tone) Hitler's views towards Jews, it is, and should be, appropriate to cite a non-independent source such as Mein Kamph to verify those views.

The exact context of what is being discussed in the article is important when determining what the best source for something is. Most of the time, an independent source is going to be the best... but not always. Blueboar (talk) 14:42, 17 December 2009 (UTC)

"... but if a section were to discuss (in a neutral tone) Hitler's views towards Jews, it is, and should be, appropriate to cite a non-independent source such as Mein Kamph to verify those views." No way. I've never accepted the idea that just because a writer espouses or publishes a view that we should then take for granted that he believes it. People lie, and particularly him. Hitler and his gang raised disinformation and propaganda to a science. We can use Mein Kamph to established what his published views were, but we should not take that to mean that he believed them. LeadSongDog come howl 20:08, 17 December 2009 (UTC)

My point was simply that sometimes it is appropriate to cite a non-independent source, not to get into an argument over when to use Mein Kamph. Blueboar (talk) 20:13, 17 December 2009 (UTC)

So could we have a less problematic example, please?LeadSongDog come howl 20:18, 17 December 2009 (UTC)

Hi LSD, I think the proposal contradicts itself, so it would be difficult to give an example of what it means. The proposal is that we say, "Articles should be based upon reliable, third-party published sources with a reputation for fact-checking, accuracy and independence." That a source is described as "third party" already signals a degree of independence, in the sense of having been published independently, or probably being a secondary source. To introduce a separate idea of "independence," so far undefined, would make no sense. SlimVirgin 20:30, 17 December 2009 (UTC)

Not necessary. As long as we understand them to be reliable for the topic at hand, a dependent source is perfectly acceptable as a verifiable source. There are likely cases of dependent sources that by virtue of the relationship can never be reliable (eg a PR firm in a press release for one of their customers).

Now, if you're talking about an article only sourced by dependent sources, notability is that way where this is spelled out. --MASEM (t) 14:57, 17 December 2009 (UTC)

I would oppose that change. Whether the most appropriate source is primary or secondary, self-published or published by a third party, independent of the topic in hand or in some way connected to it—that kind of editorial judgment depends entirely on context. SlimVirgin 18:37, 17 December 2009 (UTC)

The burden of identification on the challenger to identify what claims are alleged to be OR and whyso (Logicus becomes a Wiki-policywonk ?)

User Logicus proposes the introduction of some new rules of the following ilk to Verifiability ‘Burden of evidence’ policy to prevent the unacceptable but policywise currently possible abuse by some editor or clique of editors in excluding possibly valid and valuable material they dislike just because it conflicts with their biased POV simply by deploying the tactic of condemning that material as OR without ever having to identify where, why and how it is such. For if the onus of proof is on its editor to persuade the challenger(s) otherwise, this leaves its editor unable to locate where, why and how it might possibly be in breach of NOR and thus unable to determine where and how it might possibly be revised and improved to remedy any possible breach, or indeed even to assess whether the challenge is valid.

Such a possibility is clearly injust and unacceptable in Wikipedia and likely to bring its openness policy and the encyclopedia into even further public disrepute than it is already. For on the one hand it permits situations analogous to the Kafkaesque scenario of being charged with a crime without being told exactly what the crime was and where, when and how it was committed, and thus of being unable to defend oneself. And it also permits an injust regime of 'guilty until proven innocent' if the onus is on the defendant to persuade the challenger(s) that it is not OR in such non-specific OR allegations, but they are systemically unable to prove their innocence by virtue of not knowing where, why and how they are said to be guilty.

And on the other hand it enables any bunch of cranks who dominate an article simply by virtue of their hyperactivity and Wiki-nerdiness to protect their crankish POV in the article from correction simply by making nonspecific OR allegations against those who try to correct it, by refusing to accept any demonstration that it is not OR or that any revision of the material is not OR, and then by making further accusations of DE and TE and even invoking meritless RfC's, all to harass and subdue their adversary. Thus NOR policy originally introduced to exclude cranks also becomes a policy for protecting and entrenching cranks.

So Logicus proposes the introduction of the following three rules to prevent the injust and unacceptable interpretation of ‘burden of evidence’ policy by users Durova and David J Wilson in the recent RfC against Logicus in Section 3 of Talk:Celestial spheres that effectively makes OR of whatever material anybody says is OR and who refuses to be persuaded otherwise, without ever having to identify where, why and how it is such.

Whilst the burden of providing verifying sources for specifically identified claims that are validly challenged is not gainsayed, this must not be extended to placing a burden on the challenged editor to guess where and why and how the material is alleged to be OR.

Three new rules for the 'Burden of evidence' policy section

RULE 1: 'The challenger must at least identify the specific claim(s) made in any material that they allege is OR, and by some appropriate form of tagging of the claim(s), such as the or tags for example.'

The purpose of this first rule is to prevent whole blocks of material, or even just one sentence, containing many claims being challenged as OR without also identifying any specific claim(s) that is/are alleged to be OR. For without any such identification the editor whose material is challenged cannot possibly know where the challenged material needs to be sourced and/or revised to remedy any agreed breach of NOR according to the challenger, or indeed cannot even determine whether the challenge is valid.

This is a 'non-specific OR allegations' ban. Hence, for example, by virtue of such a rule it would be impermissible to condemn a whole article or block of material making multiple claims as OR and to delete it without identifying any specific claim made in it that is alleged to be OR. Thus it would ban non-specific OR allegations such as those made in a recent RfC against Logicus's material that had previously been accepted in articles for over a year.

RULE 2: 'And in the particular case where one or more sources have already been provided for the claim(s) challenged, the challenger must show why the quotation(s) in it/them that is/are alleged to support the claim made is in fact a failed verification that fails to support the claim made by identifying in what respect the claim made in the challenged material goes beyond what that/those source(s) claim(s).'

RULE 3: 'The challenger must also identify what rule of NOR they allege is breached by the specific claim(s) made. For example, they must specify whether it/they make(s) an illegitimate use of a primary source, or else commits Original Synthesis, or whatever rule(s) of NOR policy is alleged to be breached.'

Logicus would be most grateful for other users' views and comments on the possible merits or defects of this proposal. --Logicus (talk) 17:11, 18 December 2009 (UTC)

Too many Rules... Instruction Creep at its worst. Also, I think this discussion belongs at WT:NOR and not here... WP:BURDEN is talking about unsourced information. Blueboar (talk) 17:50, 18 December 2009 (UTC)

I've seen instances of editors who delete factually accurate material for apparently no other reason than it lacked a source and the editor didn't appear to make a good faith effort to find a source themselves. But I'm not sure the rules idea is a good solution to this problem. Seems to me that it could lead to a lot of WP:Wikilawyering. A Quest For Knowledge (talk) 18:04, 18 December 2009 (UTC)

(unindent) For the system to work there must be genuine communication between those who want to add material and those who want to keep it out. A proponent of exclusion could make OR claims that are too vague to address, just as a proponent of inclusion could insist that no matter what the other "side" writes, the complaint isn't specific enough. Adding more rules will not force those who don't want to communicate to do so. --Jc3s5h (talk) 18:16, 18 December 2009 (UTC)

I agree that this is the wrong place to discuss OR. It's also instruction creep at its worst, as Blueboar also says. I think it would also be to the detriment of the project as it would make it unnecessarily difficult to deal with OR. Logicus is unhappy with the outcome of an RfC and seems to have come here because of that. Dougweller (talk) 19:31, 18 December 2009 (UTC)

Ok, then lets take this to OR. Because I for one have stated many time before (and gotten shouted down) about how its ridiculous that people can tag and declare OR on anything and require a citation to the most ridiculous common sensical things just because "they want a citation". Policy is clear that a citation is needed for "anything likely to be challenged", since I dont know what is going to be challenged I often dont cite every sentence (though some say that must be done for FA, but they lie and that is not true). I have seen someone say "Im sure its true, in fact I know it is, but I want a citation for it"; well you cant go and admit you know its true and then remove the information from the article and declare that it must have a citation to be in the article; that's not challenging the information, that's being a prick. At least the parts I care about in this proposal deserve to be talked about in WP:V; but of course we have several times before and not enough people agree with me, so...Camelbinky (talk) 23:01, 18 December 2009 (UTC)

Not quite right. Camelbinky... the policy is clear that a citation is needed for anything "challenged or likely to be challenged." It is a two part requirement and you are only mentioning the second part. How does it work in practice?... 1) if you think something is likely to be challenged, cite it pre-emptively when you add the information... and 2) if something you didn't think needed a citation ends up being challenged, then cite it after the fact.

Perhaps I am wrong... but I think what you are really talking about is distinguishing a "legitimate challenge" vs. an "illigitimate challenge". I don't think this is distinction that we can make. Any challenge is legitimate. Blueboar (talk) 23:32, 18 December 2009 (UTC)

I dont know if I'd go so far as to say every challenge is legitimate, certainly not every challenge is done with common sense or is needed to be done. But I guess you cant actually say "everyone knows that" even though everyone does seriously know that Rome has historical buildings for example. Where would you take a challenge you felt was spurious and the editor wouldnt drop it? Do you just give in and actually present a citation for something like "Cincinnati is a river city" or "Antarctica is cold"? My personal opinion is that overcitation is distracting and those that promote the idea that every sentence must have a citation are actually hurting the encyclopedia as much as those that undercite TRUE sentences.Camelbinky (talk) 01:22, 19 December 2009 (UTC)

Everything which is challenged or likely to be challenged needs a reliable source. Beyond that, it's just editorial common sense. Footnotes are never a distraction for readers if they are combined into one per sentence, or one per paragraph. Crum375 (talk) 01:43, 19 December 2009 (UTC)

Yes, but what should we do about editors who remove factually accurate material on grounds no more sophisticated than "I disagree"? A Quest For Knowledge (talk) 02:31, 19 December 2009 (UTC)

If you mean they are removing material which they are challenging as unsourced, then unless they are clearly irrational (in which case it would be a behavioral issue, not an editorial one), the easiest and quickest solution would be to add a source. In most cases, it is much faster to find a source for something most people consider "obvious" than to argue about it. Crum375 (talk) 02:36, 19 December 2009 (UTC)

But in the case that A Quest mentions, which I've seen many many times, where an editor removes something for an irrational reason, Crum your solution seems to be- dont worry, just add a source and they'll go away. Well, even considering it should be easy, it still takes up an editors time that they had to spend on this, and an average editor's first gut reaction is to say something politely along the lines of "its commonsense, do I really need a source? Are you serious?" and a discussion may/may not start over it. Your basically encouraging bad behavior by saying "oh, just source it", because now that editor thinks "I'm doing good, Im on the right side if I go to any article and start questioning and removing anything unsourced. Its my right". I say slap bad editing behavior on the nose and say "bad editor bad! no you cant!". I'm laizze faire about talk pages and being allowed to swear and humiliate other editors; but I'm strict and a fascist about idiots showing up at articles they know nothing about and messing with them creating more work for others to clean up, even if it is only 2 mins; those minutes add up and can be frustrating.Camelbinky (talk) 03:25, 19 December 2009 (UTC)

Wikipedia is open for everyone to edit, which means we meet here all behavioral types on the planet. Some are productive and collaborative, and some are not. We can't eliminate the bad types by writing more rules; all that will cause is more confusion for everyone. It's better to have simple, clear and logical rules, as we (mostly) have now, and use our collective common sense to fill in the gaps. When we encounter behavioral issues, which is inevitable, we need to deal with them, which is part of the challenge of being an editor here. Getting to the specific issue you raise, I don't think that by saying "let's supply a source to everything which is challenged, unless it's clearly irrational" we are encouraging bad behavior. As I noted above, in virtually all cases, adding a source to an "obvious" statement is easier than arguing about it. Crum375 (talk) 03:39, 19 December 2009 (UTC)

I have to agree with Crum here. We can not legislate what does and does not constitute a legitimate challenge to unsourced information. We have to remember that something that is "obvious" to one person may not be so "obvious" to another. If someone thinks a bit of information needs a source, then we should provide one. Blueboar (talk) 15:33, 19 December 2009 (UTC)

In the case I'm referring to, the editor removed material that was not only factually accurate, could have been sourced had the editor done a good faith Google search to find sources. But they had literally gutted much of the article which forced me to spend hours and hours to find sources. (I wasn't the author of any of the material in question. I was just trying to save material that other editors had written.) When I asked the editor what their objection was, the response was always "I don't think that's true". A Quest For Knowledge (talk) 04:38, 20 December 2009 (UTC)

You can either revert to the previous version and ask that editor to please identify specific issues, or start with a stub and add sources as you build it up, if the original was very poorly sourced. Some editors are more helpful than others, and you have to learn to deal with all types. If you keep your articles well sourced, you'll have minimal problems. Crum375 (talk) 04:51, 20 December 2009 (UTC)

Logicus wrote:

"So Logicus proposes the introduction of the following three rules to prevent the injust and unacceptable interpretation of ‘burden of evidence’ policy by users Durova and David J Wilson in the recent RfC against Logicus in Section 3 of Talk:Celestial spheres that effectively makes OR of whatever material anybody says is OR and who refuses to be persuaded otherwise, without ever having to identify where, why and how it is such."

This is a blatant misrepresentation of what transpired in the cited RfC. The subject of that RfC was this huge, poorly sourced edit of Logicus's. Far from failing "to identify where, why and how" this edit was original research, I did in fact identify precisely just some of the material which I knew to be in conflict with at least some authoritative sources. Even so, I was only prepared to identify a single item as most likely original research. The rest I simply categorised as either a non-neutral point of view or original research, depending on whether or not reliable sources could be found for it.

After a response from Logicus, I provided a a further detailed list of significant differences between the treatment contained in the edit, and the treatment given in a source he had cited as supposedly showing that the disputed edit was not original research. After "kindly" asking me to consider whether the points I had made were "anything more than rather nitpicking objections which may be easily avoided by a few revisions" Logicus then described some such proposed revisions. However, even on a brief perusal of the proposed revisions I could see that the text was still riddled with unsourced passages that represented at least a non-neutral point of view. I then subsequently declined to engage in any further discussion on the matter.

It seems to me that the rules here proposed by Logicus would provide POV pushers with a powerful weapon for wasting other editors' time—namely, by continually demanding "proofs" that a never-ending succession of poorly sourced revisions of dubious material is original research. I therefore strongly oppose the adoption of such rules as Wikipedia policy.

—David Wilson (talk · cont) 17:44, 19 December 2009 (UTC)

Non English sources

Where this section does say that non English sources are OK but English are preferable, is this section in support of articles with only citations in a foreign language, personally I would say it isn't and that the section in relation to an article supports one or two foreign citations but not all or the majority, any comments. Off2riorob (talk) 09:50, 19 December 2009 (UTC)

No, there is nothing inherantly "wrong" with an article that cites only non-english sources... In some cases, it may simply be that the topic is not covered by reliable English ones. We still want to have an article on the topic, and we want the information in the article to be verifiable. And as far as Verifiability goes, the language of the source does not matter... what matters is that there is a source.

That said... if there are english language sources, please add them to the article and discuss what they have to say on the topic (especially if they present a point of view that is different than the one presented by the non-english ones). Blueboar (talk) 15:18, 19 December 2009 (UTC)

Blueboar puts it well. The appropriate distribution of English to non-English sources depends on the subject. To point out a couple of examples from featured articles, S. A. Andrée's Arctic balloon expedition of 1897 uses mostly Swedish sources its subject is covered best in that language. Pericles uses Greek and Latin primary sources and mostly English secondary sources with three German and French secondary sources, although obviously English speaking scholars don't dominate the field to such a degree. English is preferable when there's an acceptable source in English, but other languages are acceptable when needed. Durova³⁸⁶ 15:51, 19 December 2009 (UTC)

While I generally agree, caution is necessary in cases where the article discusses conflict between groups using different languages, particularly to ensure that the viewpoints associated with the opposing sides different languages are reasonably balanced for WP:NPOV. Otherwise we risk presenting one POV as verified fact. Conversely, use of English sources alone can introduce systemic bias in favour of the side which most uses English. LeadSongDog come howl 17:13, 19 December 2009 (UTC)

Well these comments here move me to want to change the wording in the policy to add what has quite strongly been commented here, that any citations in any language are good as long as they are reliable, the English is preferable bit has actually received little support in these comments, if it did get any support it was as a side issue. The question was wide ranging but at this time related to this BLP Odette_Krempin and its second deletion discussion, please feel free to come there and pass comment in the AFD, thanks for all comments here. Off2riorob (talk) 17:31, 19 December 2009 (UTC)

Agree with LeadSongDog: for instance, US-Mexican immigration issues would probably be best represented with sources in both English and Spanish, due to the different POVs likely to be represented in the respective languages. If a particular point is raised equally well by sources in both languages, though, the default preference would be to cite the English source. Spanish language sources would be reserved for facts and interpretations that have no equivalent value alternatives in English (either material not present in English, or not covered in as much depth, or not argued as well). Durova³⁸⁶ 22:21, 20 December 2009 (UTC)

• i have made a bold alteration to the policy that imo reflects the discussion here. Off2riorob (talk) 17:43, 19 December 2009 (UTC)

I've restored the long-standing English-language preference. It would need a lot more consensus than this to have it changed. This is the English WP, and our readers should be able to understand most of the sourcing we use. When there are no English equivalents, then of course other sources are welcome, but if there are English-language equivalents, it's hard to see why we wouldn't prefer them. SlimVirgin ^{TALK contribs} 18:11, 19 December 2009 (UTC)

Just to clarify, the section currently reads:

English-language sources are preferable to sources in other languages so that readers can easily verify the content of the article. However, sources in other languages are acceptable where an English equivalent is not available. Where editors translate a direct quotation, they should quote the relevant portion of the original text in a footnote or in the article. Translations published by reliable sources are preferred over translations made by Wikipedia editors.

SlimVirgin ^{TALK contribs} 18:15, 19 December 2009 (UTC)

Off2riorob, nothing anyone has said so far contradicts the the idea that English sources are prefered... what the policy is saying is that non-english sources are allowed. There is a difference. The current language makes the difference clear. Blueboar (talk) 18:28, 19 December 2009 (UTC)

Well that was not what I felt from the comments here, someone said that all English sources gave an all English pov and I have an article where it is cited to all German citations, which there seems to be strong support for, imo there actually isn't support for what is in the section..or at least I have not found much or any support for the position, my position is that the majority of citations are required to be in English and a few or some can happily be in foreign languages but I found no support for this position and the text as stands is unclear and vague, as it stands I dispute that the text supports articles with all foreign citations if I am wrong then I feel the text requires altering to support the actual reality of support. Please feel free to read the discussion at the AFD I linked to Odette_Krempin as it is related to this issue and comment there. It is my thinking that a person is not really notable in relation to this wikipedia if there is not at the least some citations in English to support that claim. Off2riorob (talk) 18:50, 19 December 2009 (UTC)

• Is there any support here for articles that are cited using citations that are all in foreign languages, as in not English. If there is imo this fact, that it is ok to do that should be added to the policy as it is unclear. Off2riorob (talk) 19:05, 19 December 2009 (UTC)

The policy already says this is OK if there are no english language sources available... but it also says that if there are english language sources for the same information, these are preferred. The best is often a blend of english and non-english sources. There is no majic ratio, however... it all depends on the specific topic, and the nature of the sources available. Blueboar (talk) 19:33, 19 December 2009 (UTC)

There may be issues, stories, concepts that no English-language source has written about, and which may be interesting and informative. I don't think it would be helpful to the project to say that such articles aren't allowed, and I don't think such a proposal would gain consensus. It would also be a notability issue, more than a verifiability one, though the two are connected, of course. But my guess is that, even if consensus were gained here to add such a clause, the rest of the project would ignore it.

Is there a particular article that has triggered your interest in this, Off2? SlimVirgin ^{TALK contribs} 19:17, 19 December 2009 (UTC)

Yes SlimV, this one and its AFD, Odette_Krempin . Off2riorob (talk) 19:43, 19 December 2009 (UTC)

Its very important that foreign language sources are allowed; otherwise we cannot possibly cover subjects concerning cultures that are non-English; we have never pretended to be the encyclopedia of the English speaking world, our goal is much wider. Are you also suggesting we dont use scientific language that cannot be understood by almost everyone; that would impede our coverage of science subjects. We dont need a dumbed down encyclopedia and a change like this can be lightly done without extensive input. Thanks, SqueakBox talk contribs 19:24, 19 December 2009 (UTC)

no, I am only suggesting that if what you also seem to agree with, that foreign languages sources are allowed and totally fine, that this position actually be reflected in the policy, as I reads it doesn't as I understood it, that point seems unclear, it says English citations are preferred but it also seems that there is no opposition to an article being totally cited in foreign language citations, this if it is true requires a comment to support that position as policy, as I see it it is a bit soft and unclear to say..oh English citations are preferred, what value is that actually when an article is considered to be totally fine with all foreign citations, it seems to be a false comment to me. The policy should then clearly reflect the reality. 19:36, 19 December 2009 Off2riorob (talk) 20:12, 19 December 2009 (UTC)

Are non-english sources allowed? YES. Are english sources preferred? YES. Is that clear enough for you? Blueboar (talk) 19:44, 19 December 2009 (UTC)

No it's not Blueboar and shouting at me won't help discussion, Squeak box has gone there and voted keep and actually referred the the references in his support, Keep as ref'd I would like the policy to reflect this reality, tell me how is a reference preferred..it is either good or not good, the text should reflect this and as I have seen discussed recently that the articles are mostly written and the only expansion is to start writing foreign articles then the reality that any citation is as good as another whatever language it is in needs to be reflected in policy, I as an independent neutral read the policy as written that one or two citations if they were irreplaceable in English are OK in another language but this is not the reality, any citations that are wikipedia reliable are fine, I want to add a small comment to reflect that, something like...although citations in English may be easier for users to read, any citation in any language is acceptable as long as it is wikipedia reliable, an article can be cited completely in foreign languages. Off2riorob (talk) 20:11, 19 December 2009 (UTC)

I went to the afd from here and nopt the other way round. Thanks, SqueakBox talk contribs 19:59, 20 December 2009 (UTC)

I think if it is entirely sourced with foreign citations that editors adding such citations should be willing if asked to provide translated quotes. Maybe we should also say that we would expect some translated quotes in the article. Dougweller (talk) 20:08, 19 December 2009 (UTC)

This wouldn't help where the issue is notability, which seems to be the issue that the initiator of this discussion is concerned with, rather than verifiability. Notability requires substantial coverage, but the posting of a translation of such substantial coverage would, by definition, go beyond fair use, and so, unless the source is public domain or under a free copyright, be a copyright violation. Phil Bridger (talk) 21:48, 19 December 2009 (UTC)

Not so; substantial coverage in terms of notability means coverage by an important source or coverage in a number of sources of moderate importance. It would not be necessary to translate a large enough passage from any one source to be a copyright violation. --Jc3s5h (talk) 22:02, 19 December 2009 (UTC)

• New offer for inclusion to clarify the reality position..Although it is preferable to have citations in English it is also acceptable to use citations in foreign languages. Off2riorob (talk) 20:31, 19 December 2009 (UTC)

I think the current wording is very clear and I don't see anything which needs to be changed. Crum375 (talk) 20:54, 19 December 2009 (UTC)

I do not think that the comment there is clear at all, on my reading it does not tell me that it is OK to create articles with totally foreign citations that or that it is ok to completely source an article in any language you want, it does not say that to me, and yet that seems to be a reality. What is the problem with clarifying the position? Off2riorob (talk) 21:21, 19 December 2009 (UTC)

We already say that "English-language sources are preferable to sources in other languages so that readers can easily verify the content of the article. However, sources in other languages are acceptable where an English equivalent is not available." How is that unclear, or preventing anyone from relying on foreign sources, if there are no English sources available? Crum375 (talk) 21:36, 19 December 2009 (UTC)

• It seems strange that someone without the basic English comprehension to understand the wording of this policy should be so insistent that we should have sources in English. Phil Bridger (talk) 21:29, 19 December 2009 (UTC)

"Squeak box has gone there and voted keep and actually referred the the references in his support..." Ah, so this is inspired by Wikipedia:Articles for deletion/Odette Krempin (2nd nomination)? Blueboar (talk) 21:34, 19 December 2009 (UTC)

Blueboar, I think I'm having deja vu, all over again. Seems you had to deal with this question not that long ago. Is the policy really that unclear that people have to keep asking the same question? Perhaps bolding the words and putting in the policy or top of the talk page that "Non-English sources are ok, if you have a question about them being acceptable, the answer is they are acceptable". Perennial questions get perennial answers, lets be proactive.Camelbinky (talk) 22:09, 19 December 2009 (UTC)

Yeah... I suppose so... OK, have added a sentence about it in the lede. Feel free to remove if you don't think it necessary. Blueboar (talk) 22:29, 19 December 2009 (UTC)

The AfD nomination is interesting and made me think about BLP articles. We've got a pretty strict policy about BLP violations, and if I look at an article and see what I think to be possible BLP violations and the cited source is not in English, I would probably delete the offending text until a source was provided that I or another editor could read to make sure there was no BLP violation. Dougweller (talk) 22:24, 19 December 2009 (UTC)

It strikes me that here, as at the AfD, Off2 is in a distinct minority (of one?) in reading into this policy language that simply doesn't exist, and in failing to read the policy as written. I don't think that there is any need to change the policy, but appreciate the comments here as supporting the comments by many (other than Off2) at the AfD in question. I join Off2 in inviting others to enter their views there, if they see fit. As to the BLP issue, we have had a number of German readers indicate that at the AfD that the sources support the article's statements, and nobody who reads German has disputed that.--Epeefleche (talk) 22:27, 19 December 2009 (UTC)

I wasn't thinking of the AfD article at all, just the general potential problem. And whatever we may think about it, we do treat BLP articles in a different way from ordinary articles. Dougweller (talk) 07:09, 20 December 2009 (UTC)

Doug, I would disagree with you on the BLP issue because we have made it clear several times that an editor's inability to see or read a source is not an issue at all. It should be tagged and someone who speaks the language should be contacted. Removing it just sends the message that it isnt equal to an English language version. A BLP article is often treated as being more important than anything else and if something might be a violation of BLP policies then that trumps any other rights or actions that are supposed to be taken according to other policies that would cover the same issue if it wasnt a BLP article. All articles are created equal in my opinion. But I dont work on BLP articles and therefore dont see the big fuss anyways. Living people annoy me.Camelbinky (talk) 01:33, 20 December 2009 (UTC)

It's true that dead people are often less annoying. :) SlimVirgin ^{TALK contribs} 01:38, 20 December 2009 (UTC)

It's rare that I actually physically laugh at something written on Wikipedia; but that definitely made me laugh. Thank you SlimVirgin for that. It made my day and brightened up my Wikipedia-universe.Camelbinky (talk) 03:19, 20 December 2009 (UTC)

Hungry person: "I'll have a cheeseburger." Waiter: "what kind of cheese?" Hungry person: "I prefer american to swiss; what kind of cheese do you have?" Waiter: "We only have swiss so I'll get you a hamburger." Chalk me up as another who finds the present language crystal clear on the framed issue and am having trouble understanding what is possibly ambiguous. There's nothing about saying "we prefer.. but allow" that implies that we don't allow exclusively, and to get there you have to go through some logical flights of fancy.--Fuhghettaboutit (talk) 04:58, 20 December 2009 (UTC)

Dougweller, I suggest you go round deleting any scientific refs that you don't understand as well, just for good measure. Indeed there is an argument that anything you dont personally understand already should be deleted. I, on the other hand, could report as a vandal anyone who went around deleting good refs merely because they don't have the education or the inclination to understand them. Thanks, SqueakBox talk contribs 13:56, 20 December 2009 (UTC)

• Thanks for adding the comment in the introduction, I think that does help clarify the position, I also do agree with Dougweller that in BLP articles and any articles if there is controversial content that it could be more of an issue, anyway I learned a few things and it was a good discussion and thanks to all who joined. Off2riorob (talk) 15:21, 20 December 2009 (UTC)

I'd like to see something that indicates a good non-English source is better than a poor English source. Any objections? Any suggestions for wording? (John User:Jwy talk) 16:29, 20 December 2009 (UTC)

I would object. If you have a reliable foreign source and a reliable English source, include them both. I see no reason to suppress either one, and arguing over which is "better" is counterproductive and a waste of time for everyone. Crum375 (talk) 16:49, 20 December 2009 (UTC)

I agree with John and disagree with Crum375. For one thing, we have to make judgments about situations which sources are better than others all the time; a review article in the Journal of the American Medical Association about a medical matter is so much better than People that People isn't worth mentioning. Furthermore, this page is not just about "how to win edit wars", it is also about informing editors of how to make choices about sources in the privacy of their own computer room. A bilingual editor should feel free to cite a non-English source and ignore an English source if the non-English one is clearly superior. Here I'm talking about choosing a source that has a more qualified author, many fewer obvious factual errors, etc.; I'm not talking about suppressing a point of view held by a significant number of qualified people. --Jc3s5h (talk) 17:26, 20 December 2009 (UTC)

A lot depends on the specific statement that you are supporting with the citation. My feeling is that if two sources equally support a statement, one in English and one in another language... then use the English one. However, if the two sources both support the statement but do so with slightly different "spin", it is better to cite them both. And if those "spins" are significantly different, it is probably necessary to discuss what each source says in more detail, so the choice is moot.

We can go into hundreds of "what ifs" on this... but we should not write rules based on "what if". I think what we currently say is clear enough. Blueboar (talk) 18:11, 20 December 2009 (UTC)

For delicate blp issues a foreign language source may well be more reliable, as indeed it may for any article concerning other cultures than those that speak English; often English language sources have been generated from foreign language sources in which case both should be used; we should certainly never get rid of a good foreign language ref merely because there is an inferior English language ref in place; besides, with so many wikipedians who speak foreign languages (all the major ones are covered) and good computer-generated translation tools around the verifiability of almost all foreign language refs is simply not an issue; if there really was an issue a wikipedia-group/project dedicated to resolving the issue would be called for; a change in policy would not. Thanks, SqueakBox talk contribs 19:40, 20 December 2009 (UTC)

I think offtoriobob might be basically arguing that the notability for purposes of the English Wikipedia should require sources in English because the enWP should cover only what is notable in English-speaking countries. That's a position which has been thoroughly rejected, and there will be no mention at all to be found to that in any policy or guideline. It has been considered in all discussions that the standard of notability is universal, and anything or anyone that can be shown to be notable anywhere is the world is appropriate for coverage here. True, we do better with ones in countries where the principal sources are in languages everyone here can read, but the conclusion from that is that we simply need to involve people who understand the other languages and can access the sources written in them, in order to do those topics justice. DGG ( talk ) 22:28, 20 December 2009 (UTC)

[ec] Rather than try to pre-empt every 'what if', it might be worth tweaking the language slightly, so it's clear that, all things being equal, English sources are preferable because they are more accessible, not because they're better. That is, that an English source is not considered more reliable than a German source, say, and writing "my source is in English" (or, equally, "my source is in Hebrew") cannot be used to win edit wars. --hippo43 (talk) 22:38, 20 December 2009 (UTC)

Each source must be judged by its content alone - language is irrelevant to judging its quality. A source in English is not necessarily superior to one in any other language only because it is in English. To do so would put many topics at risk of systemic bias (E.g. The vast majority of contemporary news sources covering events in the Soviet Union in English were written by people who regarded the Soviet Union as an enemy.) We cannot set a higher standard for acceptability of a source just because it is not in English. Its not Wikipedia's fault if a particular reader can't read Tagalog or Zulu - for most languages there are enough editors that can read them and verify the sources. Roger (talk) 09:38, 21 December 2009 (UTC)

(unindent) I would suggest the following change (the italics added):

English-language sources are preferable to sources in other languages so that readers can easily verify the content of the article. However, sources in other languages are acceptable where an English equivalent of the same or better quality is not available. (John User:Jwy talk) 03:14, 22 December 2009 (UTC)

I think that would be wrong. First, it would cause needless arguments on whether source X is equal or slightly worse than source Y, because this wording would mean that the latter is unacceptable. Second, although we prefer English sources, we accept all reliable sources, regardless of language. Every source (unless a carbon copy or direct translation) adds something, perhaps a new angle or perspective, perhaps some more details. There is no reason not to include all relevant reliable sources which add information, regardless of their language, while generally preferring the English ones. Crum375 (talk) 03:29, 22 December 2009 (UTC)

Amen brother Crum.Camelbinky (talk) 03:41, 22 December 2009 (UTC)

Okay, fair enough. If that's the general feeling, then the change is unnecessary. Proposal withdrawn. (John User:Jwy talk) 04:00, 22 December 2009 (UTC)

Who verifies the non-English sources?

Dodger67's post of 09:38, 21 December 2009 (UTC) raises a side issue. While we can fairly presume that most editors on English WP can read English references, the same is not true of other language sources. Particularly in the case of lower-visibility articles and uncommon languages unconnected to the article subject the risk that an inappropriately cited source might go unverified can be high. While a Zulu language source in an article in Category:South African politicians is likely to be verified, a Tagalog language source in the same article might never be checked without some control measure. How can we actively ensure that someone has in fact verified the cited references?LeadSongDog come howl 16:26, 21 December 2009 (UTC)

The same can be said for scientific articles. Articles like Sylow theorems are a foreign language to me and there are thousands of such obscure articles. Thanks, SqueakBox talk contribs 18:17, 21 December 2009 (UTC)

First, we don't verify sources, we use (cite) sources to verify information. That said, I do understand what you mean. Remember that we don't actually require that information be verified... we require that information be verifiable... there is a slight but important distinction there. With this in mind, I will take the liberty of slightly rephrasing your question: "Who is able to verify the non-English sources?" ... any editor who understands the language the source is written in. Blueboar (talk) 18:22, 21 December 2009 (UTC)

Not quite what I meant. WP:FACR calls for verifiable citations to reliable sources. Until we know that someone has actually read and understood the source, how do we know that it is reliable for the purpose of supporting the assertions for which it is cited? We don't presently require reviewers to be polyglots. Neither do we track that each assertion's citations were specifically checked by a credible editor. Yet somehow we still see articles promoted to FA. If we are to accept other languages, as I believe we must, then we need a system that addresses the assessment gap.LeadSongDog come howl 18:51, 21 December 2009 (UTC)

We don't require every source to be easily available to every editor. Should I, in England, refuse to accept English-language sources that are only available in libraries in the United States as valid because I can't verify them? Of course not, and the same argument applies to sources that only a subset of our readers and editors have access to because of the language in which they written. The consequence of following this logic would be that the only acceptable sources are those freely available online and written in non-technical English, which would lead to a completely dumbed-down encyclopedia. Part of the point of an encyclopedia is that its writers take information from sources that may not be accessible to the general reader and present it in a comprehensible way. Phil Bridger (talk) 19:14, 21 December 2009 (UTC)

(edit conflict)Of course we don't. But there is a reasonable presumption that some editor will look at each cited source before the article is promoted to FA. The more obscure the source, the less likely it becomes that this will actually happen. We should at minimum have a way of letting the reader know that it hasn't. We could, for instance, flag such citations with {{dubious}}. An editor removing the tag would then in effect be saying "I've checked this source and it supports the statements against which it is cited." Other possible mechanisms could of course be devised to similar ends. The alternative is to accept that any IP can make up obscure RS citations out of thin air to back false assertions. There are limits beyond which the assumption of good faith should not be taken. LeadSongDog come howl 19:32, 21 December 2009 (UTC)

LeadSongDog, you need to take up FARC assessment issues at FARC. Blueboar (talk) 19:23, 21 December 2009 (UTC)

This is a V issue pure and simple, not a FARC issue. LeadSongDog come howl 19:32, 21 December 2009 (UTC)

As someone mentioned, I also don't see how this differs from a scientific or math source, which most readers can't fathom. We provide the source, and since it's verifiable, if there is a problem, we assume someone will say so. No need (or justification) to separate out basic language issues from techno-speak. Crum375 (talk) 19:49, 21 December 2009 (UTC)

I think it is reasonable to apply more scrutiny to particularly rare* sources, including sources written in languages with a very small number of speakers. (*"Particularly rare" doesn't mean it requires a library trip or is difficult to acquire in a particular country. It means very rare. Primary source material only contained in a one library in the world, such as a famous author's journal, is an extreme example.) The rationale for such scrutiny is not only the very high difficulty of verifying such information, but also a matter questionable weighting. On the other side of the spectrum, widely-available sources in widely-spoken languages should not be a problem. For example, a French source available in a large number of Francophone libraries should be perfectly fine as such. A related caveat is that machine translation should never be used for verifying citations. It's just terrible. (See: User:Vassyana/badmachine.) Find a fluent speaker. WikiProjects are helpful for this. You can also find assistance at Wikipedia:Translators available and Wikipedia:Translation. Vassyana (talk) 19:56, 21 December 2009 (UTC)

LeadSongDog wrote:

"We should at minimum have a way of letting the reader know that it hasn't. We could, for instance, flag such citations with {{dubious}}. An editor removing the tag would then in effect be saying "I've checked this source and it supports the statements against which it is cited." Other possible mechanisms could of course be devised to similar ends."

Something along these lines seems like a good idea to me. I don't favour using a tag that would be visible on the rendered page, but something like a hidden comment could be used instead. Also, I think it would be better for editors who independently check a citation to leave a definite indication that they have done so, rather than simply removing a pre-existing indication that the citation had not previously been independently checked. With the latter mechanism it would be very cumbersome to tell whether an untagged citation was untagged because it had been independently checked, or because it had simply never been tagged in the first place. While this problem could be somewhat mitigated by having a bot do the tagging, I would still prefer to have some positive record that a citation had been independently checked and identifying the editor who checked it.

—David Wilson (talk · cont) 23:38, 21 December 2009 (UTC)

All of this sounds like instruction creep to me. Blueboar (talk) 00:12, 22 December 2009 (UTC)

Dubious tags should be only on information that in addition to needing to be citing are in fact "dubious" by definition; as in- highly unlikely. Uncited information can be very likely true, but simply lacking a citation. There can also be cited information that is dubious, because of the source; that would be a RS issue and should be taken to that noticeboard if not resolved on the talk page adequately. Dubious tags should not be used on a foreign source just because you cant read it unless you truly absolutely dont think the information could be factual; ie- you think it is dubious. If you want to on your own start using a hidden comment for other editors letting them know "hey, check this out if you can read Xhosa and let me know if it read does say what the article says"; that's your own business but it should not be a guideline or added to a policy where certain editors will take it as mandatory that they MUST start flagging every single foreign source and then take it upon themselves to remove any source that has been tagged (even invisibly) for a certain amount of time. Too many people read our policies as mandates to start tagging and removing, which is bad. The less we encourage removing factual information the better. Blueboar hits it on the head when he says this is instruction creep.Camelbinky (talk) 00:44, 22 December 2009 (UTC)

I would certainly oppose any policy or guideline which could reasonably be misunderstood as an "instruction" to perform such tagging, or as a licence to remove sourced material on the grounds that no-one except the provider of the source had checked it. Nevertheless, it seems to me that some indication that a citation had been independently checked by another editor or editors would be a useful service, especially in the case of sources which might not be easily accessible to a large number of other editors. I know from experience that in the articles of interest to me, sources cited for obscure facts fail verification sufficiently often to make it worth taking some pains to check them for accuracy. If I had some indication that another editor whose opinion I respect had already verified a particular citation I could then avoid wasting my time trying to do the same.

If such a tagging system could be agreed upon, I don't see that any changes would necessarily need to be made to official policies or guidelines at all. At most I would envisage a simple notice that such a system was available for the use of editors willing to adopt it—somewhat along the lines of the notice about citation templates in the style guideline on citing sources.

—David Wilson (talk · cont) 05:00, 22 December 2009 (UTC)

No other comment than that I agree {{dubious}} is an inappropriate tag, but {{verify source}} is perfectly appropriate for any request/need for citation verification. Vassyana (talk) 10:09, 22 December 2009 (UTC)

Excellent! The very clue we needed. It seems almost any of the V issues with obscure non-English source verification could be covered by using either {{verify source}} (sometimes adding |type=reprint) or {{request quote}}. Thank you, Vassyana. LeadSongDog come howl 15:52, 22 December 2009 (UTC)

Absolutely! And resolves the issue without trying to rewrite the policy. Blueboar (talk) 15:55, 22 December 2009 (UTC)