Wikipedia talk:Verifiability/Archive 31

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

Archive 25

←

Archive 29

→

Translation from French source needed

I am dealing with a controvercial statement at the article on the Grand Orient de France (the largest body of Freemasonry in France). In talking about the Grand Orient's involvement in French politics, the statement reads: In addition, it expects its members who hold a public office to fight for its values; in order to forward its stated purpose of exerting an influence on ideas, it holds regular talks with elected representatives, including the Prime Minister.

The article does give a citation for the statement, but the source is in French, which I can not read. I need someone who understands French to check the source and let me know the source does indeed support the statement? Can anyone here do so? If not can anyone point me to someone who can? The source is: Interview of Jean-Michel Quillardet, Great Master of the Grand Orient de France, April 2007. Thanks, Blueboar (talk) 14:30, 14 October 2008 (UTC)

The transcript at the extlink does have M. Quillardet say that (in the response to the second question from "Roux"). Lupo 14:42, 14 October 2008 (UTC)

Thanks. Blueboar (talk) 15:16, 14 October 2008 (UTC)

BTW, I've generally found the national WikiProjects very helpful with these sorts of things. If it comes up again, you might ask at WP:WikiProject France for help. WhatamIdoing (talk) 02:14, 15 October 2008 (UTC)

Another option is automated translation by google (translate.google.com) which generally does an adequate if not impressive job of translating web pages between a few dozen supported languages. --Jtankers (talk) 03:15, 15 October 2008 (UTC)

Thanks again to all who replied. Or perhaps I should say, "Merci". Blueboar (talk) 14:40, 15 October 2008 (UTC)

WP:Update

See WP:Update for the September changes to all the Category:Wikipedia content policies pages (including this one) and also the most generally-used style guidelines (called, unsurprisingly, Category:General style guidelines). If anyone wants to take on the job of updating monthly content policy at WP:Update, please reply at WT:Update. Obviously, anyone can make any edit at any time, but regular updaters would be nice. - Dan Dank55 (send/receive) 18:06, 16 October 2008 (UTC)

What is an extremist source?

The policy states:

Questionable sources are those with a poor reputation for fact-checking. Questionable sources include websites and publications that rely heavily on rumors and personal opinions, are promotional in nature, or express views that are widely acknowledged as extremist or pseudoscience. Because of this, they can be treated similarly to the way self-published sources are treated. Questionable sources should only be used as sources about themselves as described below. Articles about such sources should not repeat any contentious claims the source has made about third parties, unless those claims have also been published by reliable sources.

There is a dispute on Talk:Press_TV as to whether Zionism is "widely acknowledged as extremist", so that the Jerusalem Post (being clearly a publication that supports Zionism) is a questionable source that should not be cited in Wikipedia. This seems incorrect to me - regardless of one's political view of Zionism, it clearly belongs in the mainstream of political discourse.

I'd be grateful for comments. LeContexte (talk) 15:11, 18 October 2008 (UTC)

Actually, there is no dispute at all: one editor, Causteau (talk · contribs), against all the others, including LeContexte. It's more a case of extreme POV-pushing. --RCS (talk) 16:11, 18 October 2008 (UTC)

Concrete example: Causteau always refers to United Nations General Assembly Resolution 3379 (equating zionism with racism), but never acknowledges United Nations General Assembly Resolution 46/86 (revoking it).--RCS (talk) 16:18, 18 October 2008 (UTC)

Why not truth?

Just coming here from slashdot (via a blog), but I've always thought this policy was dumb. If we know that "reliable sources" are actually wrong and someone's statement is more likely to be true, despite not being published in a "reliable source," why should wikipedia knowingly publish a falsehood? I realize that this policy protects wikipedia from people who can't tell fact from opinion, or who think their fringe theory is "the truth". But there are occasionally situations where wikipedians just can't say what they know is true, and have to go with what is false because the press got it wrong. The issue of Jimbo Wales' birthday comes to mind. He wanted wikipedia to go with the published birthdate, even though he had previously said his birthday was on another day, and never claimed that his previous statement was actually untrue. 140.247.249.238 (talk) 02:08, 22 October 2008 (UTC)

Although this is not the place for this kind of discussion, it would be impolite to ignore a request for clarification. To put it very simple, this policy allows us to focus on data compilation and not research. In other words, we reflect the investigations others have done (which is exactly what any encyclopedia would do).

Several times people how we decide what to cite and what not. For example, "The sky is blue because that is the color of oxygen." requires one ("because..." or two "is blue" and "because..."). You may find Wikipedia:You do need to cite that the sky is blue and Wikipedia:You don't need to cite that the sky is blue interesting (perhaps a bit amusing too). -- ReyBrujo (talk) 03:00, 22 October 2008 (UTC)

Tweak to reflect existing practice

There's a discussion on the BLP noticeboard about self published sources (see this historical link). It is apparent that some SPSs are OK in a BLP, as long as the SPS is generally recognized as a reliable source and the text is only critical of the person's views. This practice was not properly documented here - the wording (before I changed it) left very little wiggle room on SPS sourcing in BLPs.

So I've boldly modified the SPS section here to reflect existing practice, to avoid future confusion. I've opened the discussion here because it's a policy page and I wanted to explain my rationale. ATren (talk) 23:53, 22 October 2008 (UTC)

I've reverted your changes, because they do not reflect what is being said in the BLP/N discussion in question. The issue is part of a discussion on BLP/N because it's contentious, gray territory. Arguments are being made that are not necessarily consensus, and certainly not cause to alter the wording of this policy to indicate that a SPS is sometimes acceptable for information about a person, which it certainly is not. Mishlai (talk) 00:32, 23 October 2008 (UTC)

OK two points: (1) what I wrote on this policy page absolutely did reflect what was being said there, and (2) looking back, perhaps they didn't reflect consensus - it was just you and KDP who contributed to that argument (I honestly thought there were a few more who chimed in, but looking again, no). In any case, obviously we have a disconnect between your interpretation of the policy and mine, so please help me to understand why you think the SPS/BLP rule didn't apply there, and hopefully we can then modify this policy to avoid such confusion in the future. Right now, clearly, it is subject to differing interpretations, so it needs to be clarified. ATren (talk) 01:19, 23 October 2008 (UTC)

Please go slowly here... you are correct in saying that this policy should reflect what is said at BLP (and vise versa)... but don't change things based on discussions. Only make changes to this policy if the BLP policy itself is changed, and wait a bit to be sure that the change sticks and has consensus. Blueboar (talk) 02:12, 23 October 2008 (UTC)

OK, will do. My initial attempt was done in a fit of boldness, but I'll tread more carefully now. Thanks. ATren (talk) 05:04, 23 October 2008 (UTC)

Weakening

"Special cases may arise; and editors should be careful not to exclude a point of view merely because it lacks academic credentials."

When did this awful sentence get included? Was there consensus discussion? Marskell (talk) 11:55, 24 October 2008 (UTC)

I see. It was Pma, no doubt up to his old tricks. Marskell (talk) 12:06, 24 October 2008 (UTC)

Per Sandy? That is over the line: this was my edit. SandyGeorgia (Talk) 14:16, 24 October 2008 (UTC)

Dealing with potentially outdated sources

I'd like to propose an addition to WP:V to address an issue that has been brought up several times on the reliable sources noticeboard - namely how and when to use old sources. This was discussed recently on the RSN in relation to the 1913 Catholic Encyclopedia. There are obvious problems with using such old sources, given that information in them may be out-of-date, superseded by newer information, or based on obsolescent assumptions or interpretations.

WP:V doesn't currently seem to address the problem of reliability depreciating over time. The reliable sources section of WP:V appears to assume that if a source is reliable at one time it will always be reliable. This clearly isn't the case - knowledge isn't static, it constantly evolves, and old knowledge becomes less useful as circumstances change and new knowledge is developed. To address this, I suggest adding a new section on "Old sources", as follows:

The reliability of old sources can be reduced by their age. Such sources may have been written to a high standard for their time and may present a substantial depth of coverage of a topic. However, sources that are decades or centuries old are unlikely to represent the modern understanding of a topic. They may present outdated information or be based on obsolescent assumptions or interpretations. This is particularly likely to be the case where the topic has been the subject of ongoing research and reinterpretation. Editors should treat such sources with care. An old source may usefully illustrate viewpoints that were current at the time of publication, but it should not automatically be considered a reliable source for modern scholarship or opinion. In judging whether an old source is still of current relevance, editors should consider whether the source is still referenced and discussed by modern sources. Care should also be taken in how old sources are presented, so that they are not misrepresented as indicative of current thinking.

Any thoughts? -- ChrisO (talk) 23:50, 22 October 2008 (UTC)

Chris, I think this is excellent. Your suggestion strikes exactly the right ballance. It allows for these types of sources to be used... as long as they are used appropriately. Well done. Blueboar (talk) 02:03, 23 October 2008 (UTC)

Seems good to me. You may want to look into updating Wikipedia:1911_Encyclopaedia_Britannica as well, which could benefit from your balanced view. Protonk (talk) 04:12, 23 October 2008 (UTC)

I have a hard time with the idea that older sources are unlikely to represent modern understanding. That may be, but then maybe not. The reliability of old sources can be reduced by their age...and may have been and may present outdated information....but then you go on to say that older information is unlikely to represent modern understanding...and particularly if there has been contemporary research and reinterpretation. Again, maybe, maybe not. In the hard sciences, we sometimes have an expansion of knowledge, but the kernel of information that something is built on is perfectly verifiable and valid. Mechanics, older versions not obsolete, if improved upon. Sometimes, time moves on and sciences like phrenology get thrown out. Perhaps it may take decades or centuries for a theory to be tested and proved or disproved. Modern understanding and reinterpretation, on the other hand, may be merely a fad or a phase which hasn't been so tested. Some subjects, for example in the fields of philosophy or art -- their interpretations might well be much the same today as when they were written. History can be "reinterpreted," but without being accompanied with newly discovered facts, these interpretations are unlikely to be accepted by all historians, ie they will be controversial. Holocaust Denial is one example of reinterpreting history, as are 9-11 Truth theories. Both are based on what they consider "ongoing research." (Re)Interpretations are theories. They can be and should be included in an encyclopedia, but they should be given no extra weight simply because they are new, just as older sources should not be denigrated simply because they are older. In fact, I would argue that more care should be given that new theories are not interpreted as representing modern understanding as newer theories have yet to be tested. Tundrabuggy (talk) 05:57, 23 October 2008 (UTC)

I didn't see the problems you mention; then again, I probably read what I wanted to see. Your concerns can be alleviated by ensuring it is clear that even "old sources," assuming they were published reliably, would have to be overruled by a preponderance of more recent, reliable sources. I've seen people try to rewrite a well referenced article based on a single, brand new RS, and this is the sort of thing I'm referring to that should be discouraged. Of course, I've also seen BS like your other concerns, but I think other policies (WP:NPOV, WP:FRINGE, WP:RS) already handle that. Someguy1221 (talk) 06:20, 23 October 2008 (UTC)

I'm happy with Chris O's suggestion -- I understand what Tundrabuggy is saying, but overall, I think 'unlikely' is correct although it may not apply to certain fields of knowledge. And I certainly don't agree that a century old source should be given more prominence than a modern one, or that his examples of revisionist/fringe views of historical events are relevant. Doug Weller (talk) 07:40, 23 October 2008 (UTC)

To Tundrabuggy: I don't agree one bit. It is a strawman to introduce holocaust denial and 9/11 truth as new "theories" when discussing the reliability of encyclopedias from 1900. Neither of those is a commonly accepted theory or explanation in any sense of the word. They are obviously false and they have only recency in common (though denial of the existence of repression of jews isn't really new)--also, since both 9/11 and the Holocaust postdate ~1900, how are we to use these as a basis for comparison? Let's be fair and consider how mainstream history and science was represented then and how it is represented now. My suspicion is that regardless of actual truth value, sources from the present day will more accurately reflect modern views of subjects than sources from a century ago. In some fields, modern sources will be a vast improvement over the older sources (how well do you think the 1911 britannica covers the Battle of Little Big Horn?). In some fields, we may see little change. Some subjects may be less well covered today then they were in 1911 or 1913 (as a result of changing tastes), but those would represent the exceptions where a 1911 source would present an appropriate look at a subject. Most cases we undertake a big hazard to use older sources without caution. Protonk (talk) 07:47, 23 October 2008 (UTC)

Scientific theories can be disproved by the accumulation of new evidence. That's why, for instance, books on planetary science that pre-date the spaceflight era aren't reliable sources for information on the planets - you would find, for instance, the hypothesis that Venus is covered by oceans (disproved by the space probes sent there in the 1960s). You could also consider Darwin's theory of evolution by natural selection. His Origin of Species is an old book, published 1859, and omits a lot of modern information - he knew nothing of genetics and was unable to explain exactly how natural selection worked. However, because the underlying theory has never been disproved and has been enormously useful to science, the book is still a highly relevant and very widely cited source despite its age.

In other fields such as history, interpretations may change because of improved understanding and different perspectives. To give you an example from my own experience, 19th century translations of the Cleopatra's Needle inscriptions are unreliable because of the incomplete understanding of Egyptian hieroglyphs at the time. Early translations are thus of less value - and may in fact be completely wrong, as I found - compared to modern ones.

Underlying assumptions also change over time. Many 19th century historians and archaeologists approached their work from a religious or classical perspective, seeking to find archaeological evidence of Biblical history or to verify ancient Greek and Roman accounts. Heinrich Schliemann at Troy was a case in point. Most modern historians take a different approach, seeking to understand ancient societies in terms of the circumstances of the time, rather than through the filter of later perceptions. For instance, it used to be traditional to interpret ancient Persia in the light of (often hostile) Greek sources. Modern historians take a different and perhaps more sympathetic approach - this was the point of a major exhibition called "Forgotten Empire" that the British Museum hosted a few years ago.

9/11 truthism and holocaust denial aren't in the same category - those are fringe viewpoints, always have been and probably always will be. The issue there is fringe theorism and unreliable (non-academic) sourcing, not revisionism. (Anyway, revisionism isn't automatically a bad thing, as shown by the case of the Persian exhibition I mentioned.) -- ChrisO (talk) 08:17, 23 October 2008 (UTC)

Excellent idea. Since the objection is based on fringe theories I'm not sure I see a problem. We deal with fringe theories as fringe theories, whether they are 200 years old or from the past decade. I do have a question about the proposal. Are there limits to how old a source must be in order to start losing its reliability, and/or could this differ greatly between contexts? For instance the proposal reads:

This is particularly likely to be the case where the topic has been the subject of ongoing research and reinterpretation.

Some topics are subject to much more vigorous "ongoing research and reinterpretation" within much shorter spans of time, meaning that old theories become outdated rather quickly. This is by no means a perfect example, but take Satanic ritual abuse for instance. The late 1980s through 90s contain a massive amount of sources that represent a now thoroughly debunked and academically fringe position. Proponents of the fringe position use these relatively recent but contextually dated sources to claim that there is still an ongoing academic "debate". Where does something like this fit into the equation here? Are there better examples of a quick turn around or establishment of a scholarly consensus? Thoughts?PelleSmith (talk) 11:33, 23 October 2008 (UTC)

Ultimately this is about representing POVs equally, older POVs and newer POVs, there's no reason you shouldn't represent older POVs in an article simply because there are more modern POVs, this would be to suggest that modern POV are 'correct' which is fundamentaly against NPOV policy. Older POV are certainly of encyclopedic interest. If a new source contradicts the older source, you can add that cited info to the article too. In the end it isn't editors' job to decide that modern views are more reliable, as it isn't always the case. To use the Egyptian hieroglyphs translation example above, you simply state both translations atrributed to a source. We cannot say the modern one is correct and the old one is wrong because we do not know this and it would not be neutral to say so, however we can say that modern egyptologists believe it to be correct at the current time, providing it can be sourced so obviously. --neon white talk 12:16, 23 October 2008 (UTC)

I think that represents the most extreme interpretation of "verifiability not truth" out there, and I'm not sure its implemented to such extremes usually or that it should be so implemented. The notion that editors would be the ones deciding that modern views are more reliable is a misrepresentation. Scholars are deciding this for us by preferring modern interpretations to antiquated ones and sometimes even by more clearly refuting them. I think the suggestion here is that in this context the relative age of a source is a good indicator of its perceived reliability within current scholarly consensus. I don't think that in any way contradicts any current policies or practices. Of course if a much older source does not contradict current thinking in the first place then why not use the more recent ones? Is there any way to suggest, within reason, to simply stay away from much older sources, when possible?PelleSmith (talk) 13:10, 23 October 2008 (UTC)

Just to give everyone an idea of how extensivly we cite outdated encyclopedias, take a look at Wikipedia:Catholic Encyclopedia topics. I am, by no means, saying that all of the articles listed there use the source inappropriately (In fact, I would say quite the opposite. After a quick glance through the list, my guess is that most of them use the CE appropriately)... but it does show that we have a lot of articles that are primarily based on a source that is almost 100 years old. Blueboar (talk) 13:46, 23 October 2008 (UTC)

Yeah, by my guess, we have something like 13,000 articles that cite the 1911 britannica extensively (and have the template transcluded). That's not a lot in terms of 2.5 million articles total, but it is a more than one person could hope to work through. Protonk (talk) 14:57, 23 October 2008 (UTC)

I think that something much shorter and simpler might be better. Perhaps something like:

An older source may usefully illustrate viewpoints that were current at the time of publication, but it should not be assumed to agree with modern scholarship or opinion, particularly in areas of active research, such as medicine. Care should also be taken in how old sources are presented, so that they are not inappropriately presented as indicative of current thinking.

I also add that this is really only an issue for certain uses of a source (presenting something as a currently agreed-upon fact) and only in certain fields. What Octavian decreed in 7 CE is still what he decreed, no matter whether the source was printed in this century or the previous one. WhatamIdoing (talk) 16:56, 23 October 2008 (UTC)

the interpretation may well not be. there have been major developments in Roman historiography since 1911-- I would certainly not accept a 1911 discussion of the nature of the Patricians or the Senate, or whether the Kings were historical or mythical. This also applies to the basic facts, though I believe mainly the earlier periods. But even for the classical period monuments are still being found. I would not assume that in any field whatsoever there have been no progress unless one actually checks for this, and the only way to do so is with modern sources. DGG (talk) 04:11, 24 October 2008 (UTC)

Yes, I think that's right. Ultimately it comes down to a question of what we're trying to achieve here on Wikipedia. Our goal should be to present up-to-date information to our readers, presenting a contemporary understanding of topics (meaning both facts and interpretations). That can't be achieved if we either depend too heavily on outdated sources or misleadingly present such sources as equivalent to, or actually as being, modern interpretations. -- ChrisO (talk) 08:35, 24 October 2008 (UTC)

I've added the proposed text to the policy. I've made a few amendments to mention the need to consider changes in circumstances - for instance, we would not want editors to present a source like the 1911 Britannica as an up-to-date reference for issues such as politics, demography or economic matters. -- ChrisO (talk) 12:36, 25 October 2008 (UTC)

In a general form I agree with the addition, though in my opinion an older source is probably just fine as an initial source, so I don't think it should just be removed automatically. It should, however, quickly be set aside if newer sources are brought in. For example, if there's a stub that's relying on the 1911 EB, and no one is challenging it with newer sources, then it should be left alone. If there's a dispute between using the 1911 EB or a 2001 academic journal though, the 2001 source is preferable, and the 1911 source should either be removed, or information from it should be qualified in the article, for example, "According to the 1911 EB, <whatever>, though more modern sources say <stuff>." But, if there are no other sources contradicting what the 1911 EB is saying, it's probably fine to use. Lastly, if someone sees an article that's sourced only to the 1911 EB, and they think it's got bad info but they don't have the time at the moment to find newer sources, they can tag it with:

{{out of date}}
{{time-context}}
{{update}}
{{old}} (for inline citations)

--El on ka 17:10, 25 October 2008 (UTC)

Even this can't be a hard & fast rule. Important factors will be

How much older?
What different POVs are these sources coming from?

We do not want people assuming that an opposing side's argument can be removed simply because the supporting side comes from a somewhat more recent source. It certainly possible that new developments would fundamentally invalidate an old assertion, however. The best approach I think is to encourage editors to give due weight to how current or dated a source is in the context of the subject. Age of the source is only a single factor. Mishlai (talk) 17:49, 25 October 2008 (UTC)

I agree in general terms. I've deliberately avoided setting any cutoff points; there are certainly examples of old sources (Darwin's Origin of Species is a case in point) which are still highly relevant in the context of the subject. You can't simply say (for instance) "all sources published before 1900 are unreliable". Editors will need to consider sources on a case-by-case basis - that's the point of the statement that assumptions should not be made automatically. Age is likely to mean that an old source is less reliable, given the development of new knowledge and interpretations and changes in circumstances on the ground. But this does not automatically mean "unreliable"; it just means "less likely to be reliable". -- ChrisO (talk) 18:35, 25 October 2008 (UTC)

I also oppose establishing a general cutoff date, because it implies acceptance of anything that is even slightly more recent. For example, I'm in an ongoing dispute with an agenda editor who is cherry-picking medical texts from the early- and mid-20th century to promote a particular POV. There are (a few) sources from the mid-20th century that are still relevant for medical topics, but he's deliberately picked a 1951 cardiology text because all of the more recent editions destroy his pet theory. The right "cutoff date" varies dramatically by field. WhatamIdoing (talk) 21:16, 25 October 2008 (UTC)

Exactly - this is a good example of what I'm referring to in my comment to Jayvdb below, where a historical source is misrepresented as being indicative of modern knowledge or opinion. -- ChrisO (talk) 00:48, 26 October 2008 (UTC)

This sounds like a problem best tackled using WP:UNDUE. If the 1951 source isnt a currently respected viewpoint, it should be either omitted if it wasnt ever a significant viewpoint, or moved into the "History" section if it was a significant viewpoint for a while. John Vandenberg ^(chat) 01:35, 26 October 2008 (UTC)

I like the idea of essays and guidelines covering the difficult task of discerning the utility of older sources, but this page is a policy page about verifiability, not truth, and it isnt helpful to introduce that passage into this policy.

Also, this could lead to the wrong approach. We should be citing the most appropriate work, rather than encouraging people to overlook old sources out of fear they might contain information that is "not acceptable by policy". Whenever there has been significant differences/improvements in the understanding of a topic over the last 100 years, there is also room in this paperless encyclopedia for a "History of topic" topic. Following on from the example of On the Origin of Species, readers and editors should know where that text fits into Evolutionary biology by reading our article History of evolutionary thought. Likewise for Catholic topics that are causing concern at Wikipedia:Reliable_sources/Noticeboard/Archive_20#Citing_old_encyclopedias; there should be either a "History" section or a "History" subtopic where these older sources can be appropriately used. In this way we can give historical sources ample room to be covered, enhancing our readers understanding of history of human thought. John Vandenberg ^(chat) 00:00, 26 October 2008 (UTC)

I would like to second what WhatamIdoing said, since I have had the exact same experience: Dealing with a POV-pusher who was cherry-picking older sources (some going as far back as the 1300s), and arguing that per WP:NPOV, their theories needed to be given equal placement in the article along with more modern research. So I would very much like to see something spelled out in a policy such as WP:V which clarifies that our job is to provide a summary of prevalent views and scholarship, and not a detailed account of every outdated view that's ever been held for the last several hundred years. --El on ka 00:26, 26 October 2008 (UTC)

I disagree with your reversion; I think you've misunderstood the point of the passage. It isn't about truth, it's about reliability, which is very much within the scope of this policy. As I said at the start of this discussion, Wikipedia:Verifiability#Reliable sources doesn't take into account the depreciation of reliability over time. My addition is specifically intended to fill this gap in the policy. It doesn't in any way prevent a "History of <topic>" article. I took that into account: the wording says, "An old source may usefully illustrate viewpoints and circumstances that were current at the time of publication" As a historian myself, I'm very conscious of the value of older sources in showing how views or circumstances have changed over time. That's an entirely appropriate usage of older sources. But what I'm trying to address here is the possible misuse of such sources, whether accidentally or deliberately, to misrepresent historical ideas and circumstances as being current. -- ChrisO (talk) 00:32, 26 October 2008 (UTC)

This pillar is about verifiability. Tacking on other concepts does not help. KISS. I understand what you are driving at, and it is a good idea; but, I do not believe it belongs in this policy. How these reliable sources are utilised in a different concept. WP:RS and WP:NPOV (WP:UNDUE especially) seem the more suitable resting ground for your proposed addition. John Vandenberg ^(chat) 01:27, 26 October 2008 (UTC)

OK, if I moved it across to WP:RS, I presume you wouldn't object to me adding a line to Wikipedia:Verifiability#Reliable sources to refer to it? -- ChrisO (talk) 01:32, 26 October 2008 (UTC)

The section already says "For a guideline discussing the reliability of particular types of sources, see Wikipedia:Reliable sources (WP:RS)." and "Tiny-minority views and fringe theories need not be included, except in articles devoted to them." I would object to labouring on the age of a source as one measure of reliability - there are many, and the age of a source is not more important than the bias of the editors, the target audience of the work, etc. John Vandenberg ^(chat) 01:41, 26 October 2008 (UTC)

I agree that the discussion should be at WP:RS and not here. I do not agree that this is 'labouring' about anything, there is an issue about old sources that should be addressed. Doug Weller (talk) 05:48, 26 October 2008 (UTC)

I had meant that WP:V need not specifically talk about the age issue (labouring). If it is comfortable to work in the age aspect without labouring the point, I'll not complain, as the trailing sentences that point to WP:RS and WP:NPOV are not perfect.

On RS and NPOV, the issues to do with various sources, including old ones, can be fleshed out fully. The rule of thumb that old sources are bad for discussing current worldviews is ... self-evidence, but as some people have difficulties grappling with that .. I've no issue with RS and NPOV spelling it out. My only concern is that "V" is a simple pillar that says that assertions must be attributed to verifiable source. John Vandenberg ^(chat) 08:19, 26 October 2008 (UTC)

Articles about vs. articles citing supporting sources

When it popped up on my watch list, I thought that this edit made good sense. However, I see that it has it has been challenged here. Yes, the change alters the meaning of the sentence—to the betterment of the article, IMHO. I support the change. -- Boracay Bill (talk) 22:41, 26 October 2008 (UTC)

Strengthening a bit?

I haven't looked at WP:V in a while, but after re-reading it now, it seems that the general language used there is a little weak. The phrase "any material that is challenged or likely to be challenged" seems a bit too weak, too amorphous and subject to too much interpretation. I have always thought that any material of factual nature, related to people and events (e.g. when someone was born/died/get married, etc, when and where a particular event took place, etc) need to be referenced to a reliable source (although not necessarily by an in-line citation). The reasons for this go beyond simple accuracy and verifiability, but also to point the reader to specific sources where more can be learned about the subject and where more detailed information can be looked up. I think that the text of the policy should say something along these lines, right after the sentence about "any material that is challenged or likely to be challenged". E.g. something like this:

"To the extent reasonably possible, a Wikipedia article should provide, either by in-line citations or by a general list at the end of the article, reliable sources for all the basic factual information given in the article, even if this information is not likely to be challenged".

I think this both reflect the current best practices and the general expectation of what a good article should be. Nsk92 (talk) 16:19, 28 October 2008 (UTC)

I think that your proposal, while well enough in its way, doesn't actually address what I understand to be the point behind the existing section. I understand the existing section to mean "If you add a direct quotation, or anyone complains about a specific claim, then you need to cough up a very specific inline citation to support that quotation or fact." For these circumstances, "Here's a list of twelve long, poorly indexed books, and I'm sure the quotation/fact/claim is in one of them" is wholly inadequate. WhatamIdoing (talk) 18:47, 28 October 2008 (UTC)

I agree but I wanted to make an additional point, beyond the point of the existing section: namely, that in addition to what the text currently says regarding material "that is challenged or likely to be challenged", all basic factual info in an article needs to be sourced in some way. The current text of WP:V does not seem to say this anywhere, while I think that it should. I recently had a discussion with a new WP user who created an article about her father (who was a sufficiently prominent painter to pass WP:BIO despite WP:COI concerns). She put a number of facts about his life (of not particularly controversial nature: dates of birth and death, details regarding education, moving from one country to another, etc) that were based purely on her own personal knowledge. When I said that this info needs to be sourced per WP:V, she read WP:V (which I myself had not done in a while) and told me that, based on her understanding of what WP:V actually says, only potentially controversial material needs to be sourced and the sort of info we were talking about should be accepted on faith. It was hard to convincingly counter this argument, based on the current text of WP:V and especially in view of WP:AGF. That's why I'd like the text of the policy to be more clear and explicit on this issue. Nsk92 (talk) 19:33, 28 October 2008 (UTC)

Perhaps your additional point should be made separately. WhatamIdoing (talk) 19:59, 29 October 2008 (UTC)

OK, I am giving it a try and have added a separate sentence after the paragraph about quotations, material that is likely to be challenged, etc. Nsk92 (talk) 15:11, 30 October 2008 (UTC)

I have deleted Nsk's addition, not because I disagree per say, but because I think this is a topic that is going to need a lot more discussion and concensus. More importantly I forsee a conflict developing with what is being discussed at our other other policy pages. For example, while Nsk's language here said that even small uncontrovercial facts need citation, there is proposed language being discussed over at WT:NOR#Verifiable versus Verified saying exactly the opposite... that simple statements of uncontrovercial fact don't need citation unless they actually are challenged.

I don't really care which direction we go (there are good arguments for both), but I do care about having different policies say different things about the same concept. That simply leads to confusion and edit wars. Blueboar (talk) 18:00, 30 October 2008 (UTC)

I don't think this is necessarily a conflict. Nsk92's language says that everything "should" be cited, as much as possible, for all basic factual points. That does not go so far as to say that every fact must be cited from the get-go, even if non-controversial, or else it will be immediately removed--which would be a wrong-headed policy. As an aspirational goal, citation is a good thing, but that doesn't mean citation is mandatory and failure to cite is a license to delete verifiable content. COGDEN 19:10, 30 October 2008 (UTC)

Correct: I certainly did not mean it as grounds for deletion or removal from the article (unless the material is specifically challenged) but rather as a standard to be aspired to, to the extent possible. Moreover, I did not mean the language to insist on in-line citations for every fact. But I do think it is desirable that for every basic fact there should be a source somewhere in the article (which may be cited as an in-line citation for something else or just given as a reference without any in-line citations referring to it) that supports that fact. Nsk92 (talk) 20:07, 30 October 2008 (UTC)

Hmmm... OK, I read it a bit differently, but now that you comment on it I can see that this might be the intent. If this is indeed Nsk's intent, perhaps the best solution is have a joint discussion with the folks at NOR, and work out common language that can be used in both policies. Blueboar (talk) 19:58, 30 October 2008 (UTC)

Updated for primary sources

30-Oct-2008: For years, this policy page has omitted the use of primary sources, which are well-explained under WP:NOR. To avoid the typical instant condemnation of primary sources, I am copying the WP:NOR policy text (about proper use of primary sources) into WP:VERIFY, as follows:

"Primary sources should be used to support statements by definition, such as descriptive statements issued by the subject of the article. As stated in WP:NOR, primary sources that have been published by a reliable source may be used in Wikipedia, but only with care, because it is easy to misuse them. For that reason, anyone, without specialist knowledge, who reads the primary source should be able to verify that the Wikipedia passage agrees with the primary source. Any interpretation of primary source material requires a reliable secondary source for that interpretation. To the extent that part of an article relies on a primary source, it should:

only make descriptive claims about the information found in the primary source, the accuracy and applicability of which is easily verifiable by any reasonable, educated person without specialist knowledge; and
make no analytic, synthetic, interpretive, explanatory, or evaluative claims about the information found in the primary source.

For example, hurricane wind speeds and storm surge levels could be quoted directly from the U.S. National Hurricane Center (NHC) as a primary source; however, stating, "The storm has become more dangerous" must be traced to that source or another reliable source, rather than stated as a conclusion of the Wikipedia writer. Opinions must be traced to reliable sources.

Articles should also rely on reliable, third-party published sources with a reputation for fact-checking and accuracy, for statements beyond mere definition about the subject."

(Text ~~added~~ not added yet to WP:VERIFY 30Oct08)

On numerous occasions, I have had to untag or debate the legitimate use of primary sources, due to a lack of the logical notion of "proof by definition" from the primary sources. IMHO, this has been a key problem in logical thinking, imagining that every statement must be defended with extensive debate, rather than using the simple "proof by definition" by citing the definitions or descriptions in the primary source. If a corporate website says, "Our CEO is John Q. Public-Doe", then there is no immediate need for another source for that text (unless other reports contradict that). I hope the added explanation about primary sources will help reduce future debate about the article sources. -Wikid77 (talk) 14:25, 30 October 2008 (UTC)

I fully agree with your goal here, and your quoting of NOR. I too, have seen too many situations where a valid source is improperly challenged becuase it is a primary source. NOR makes it clear that Primary sources can be used, and in some circumstances they are the best source possible and thus should be used. They simply need to be used with great care.

That said, I am confused by your wording in the first sentence of your proposal... what do you mean by "...used to support statements by definition"? Also, I think we do need to make it clear that while primary sources may be used where appropriate, articles should primarily be supported by secondary sources. In other words, the bulk of the material in any article should be supported by secondary sources. Blueboar (talk) 16:04, 30 October 2008 (UTC)

In formal verification steps, or formal proofs (as in mathematics and traditional logic), the term "proved by definition" is an age-old idiom that indicates direct reference to the primary source. Search Google for "proved by definition" (which Google can actually match, verbatim), and note that the idiom is used often, but Google only indexed about 129 webpages with that phrase. Because of the tie to "formal proofs of correctness" (also Google that phrase), I used the phrase "by definition" in the above text for WP:VERIFY. As you see by the red-links, articles about "correctness" should be expanded in this vein as well. Wikipedia still omits vast areas of age-old knowledge (as people bicker over sources!). -Wikid77 (talk) 17:47, 30 October 2008 (UTC)

OK... I think I get what you are saying... the problem is that Wikipedia Policy needs to be written in plain English that any idiot (such as me) can understand without having to look the term up. See if you can dumb your first sentence down for the rest of us. :>) Blueboar (talk) 18:28, 30 October 2008 (UTC)

I don't like having text copied from one policy to another simply for the sake of redundancy. The result is a synchronization nightmare (and if you can't imagine how, then I invite you to spend a month working on WP:MOSCO's problems). Why can't we just say "The use of primary sources may be acceptable in certain circumstances, so go read WP:NOR"? WhatamIdoing (talk) 20:32, 30 October 2008 (UTC)

31Oct08: I think a good balance about primary sources would state more than one short sentence, but give details beyond the WP:NOR wording. I suggest the following text be added instead:

"Primary sources can be used to support statements of simple description, such as definition of terms or measurements, but avoid subjective, broad claims, such as about actual effectiveness, marketplace rank or social impact. Any interpretation of primary source material requires a reliable secondary source for that interpretation.

For example, hurricane wind speeds and storm surge levels could be quoted directly from the U.S. National Hurricane Center (NHC) as a primary source using airplane or buoy readings; however, stating, "The storm has become more dangerous" must be traced to that source or another reliable source, rather than stated as a conclusion of the Wikipedia writer. Opinions must be traced to reliable sources.

Articles should also rely on reliable, third-party published sources with a reputation for fact-checking and accuracy, for statements beyond mere definition about the subject."

(Text not added yet to WP:VERIFY 31Oct08)

In the balance, the combined goal is to state the general use of primary sources, as well as give some details (such as avoid "marketplace rank") and include long-term examples, such as hurricane data using NHC as a primary source rather than wait until network news reports. It has been a confusing issue, to get people to understand use of primary sources; otherwise, there would not be so many debates about rejecting primary sources. Also, I agree, there is no need to tie the wording to WP:NOR, since the policies will be used together, regardless of the wording of either. -Wikid77 (talk) 11:27, 31 October 2008 (UTC)

This is the policy on verifiability, not truth. It is here to ensure that people complement their editing with citations of where they are obtaining the information from. This policy does not try to guide editors to do the right thing with various sources. Please do not complicate this policy with the more complex matter of when it is appropriate to use which sorts of sources. That is best covered in more depth in other policies and guidelines. John Vandenberg ^(chat) 15:37, 31 October 2008 (UTC)

Refs in Burden of evidence section

The first sentence in the Burden of evidence section contains two references that are actually excellent explanatory comments. I suggest they not be "hidden" in the references section, but be made a part of the content.

Current version:

The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged should be attributed to a reliable, published source using an inline citation.^[1] The source cited must clearly support the information as it is presented in the article.^[2] The source should be cited clearly and precisely to enable readers to find the text that supports the article content in question. Editors should cite sources fully, providing as much publication information as possible, including page numbers when citing books.

Revised version:

The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged should be attributed to a reliable, published source using an inline citation. When content in Wikipedia requires direct substantiation, the established convention is to provide an inline citation to the supporting references. The rationale is that this provides the most direct means to verify whether the content is consistent with the references. Alternative conventions exist, and are acceptable if they provide clear and precise attribution for the article's assertions, but inline citations are considered 'best practice' under this rationale. For more details, please consult Wikipedia:Citing_sources#How to cite sources. The source cited must clearly support the information as it is presented in the article. When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested should be provided as a courtesy to substantiate the reference. The source should be cited clearly and precisely to enable readers to find the text that supports the article content in question. Editors should cite sources fully, providing as much publication information as possible, including page numbers when citing books.

What think ye? -- Fyslee / talk 16:01, 1 November 2008 (UTC)

The text is too long when formatted as a single paragraph. It would need to be broken into pieces if the footnotes were removed. But there's nothing wrong with having explanatory comments in footnotes. — Carl (CBM · talk) 16:14, 1 November 2008 (UTC)

I have no problem with the footnotes. They are still part of the policy. Blueboar (talk) 16:33, 1 November 2008 (UTC)

Request for outside opinions

A user has been stonewalling progress in removing unsourced fancruft from Dr. Strangelove. Here is my preferred version, here is the version it has been reverted to. It would be great to get some fresh eyes on the situation. Thanks in advance. --John (talk) 19:06, 2 November 2008 (UTC)

And it would also be great if the fresh eyes would come unprejudiced by John's rather injudicious use of pejoratives such as "stonewalling" and "fancruft"!! Thanks. Ed Fitzgerald _{t / c} 04:26, 3 November 2008 (UTC)

Marking primary sources differently

Everybody in his right mind knows that citing sources in Wikipedia is very important.

Everybody in his right mind knows that citing secondary sources in Wikipedia is very important and that secondary sources are better than primary ones.

Sometimes, though, primary sources are not terrible. Sometimes they are acceptable. But even when some people think that they are acceptable in a certain case, other people may - rightly - doubt it.

Now that is a very common argument between Wikipedia editors; but what about the readers? Readers should know which sources they should trust more. It may be surprising for Wikipedia editors that are used to discussing reliable sources all the time, but not all readers actually know the difference between primary and secondary, or peer-reviewed and self-published.

So, here's the idea - mark primary sources in different color or something. Or some other visual way. So readers don't mistake a footnote with a primary source for a reliable source.

This idea really comes from WP:PAPER; it's one place where we can be different and better from a paper encyclopedia.

It's just a blueprint idea; feel free to discuss it and propose something better. --Amir E. Aharoni (talk) 17:42, 31 October 2008 (UTC)

Oppose. It's a good thought, but I don't think it's a workable idea. Not only does the average Wikipedia editor not know the distinction between a primary and secondary source, most good Wikipedia editors don't know either. And even if they do, two reasonable editors can legitimately disagree on whether a source is cited primarily or secondarily. In fact, any two historians or experienced journalists might similarly disagree. Plus, whether a given source is primary or secondary often depends on how it is cited. Such color-coding would be one more excuse for NPOV violations: an editor who doesn't like a source might call it primary, while an editor who likes the source might call it secondary. Plus, depending on what the source is being used for, many primary sources are far more reliable than many secondary sources, and such a coloring scheme might give the wrong impression. COGDEN 18:52, 31 October 2008 (UTC)

It's a nice idea, but what counts as primary and secondary is not always a clear-cut case, and I can see it leading to arguments. Also, primary sources do often count as reliable sources, depending on how they're used. SlimVirgin ^talk|edits 19:52, 31 October 2008 (UTC)

Yeah, nice idea, but too complex and prone to abuse. ≈ jossi ≈ (talk) 21:50, 31 October 2008 (UTC)

Nice idea.... could be used in only clear-cut cases? Primary sources are often the best- Peer-reviewed. You mean, I guess, that we should mark self-published sources? [1] ——Martin^phi ☎ Ψ Φ—— 04:49, 1 November 2008 (UTC)
Self-published and not peer-reviewed, yes. --Amir E. Aharoni (talk) 20:05, 3 November 2008 (UTC)

Proposed DYK rules change

There have been incidents of questionable sources used to support DYK hooks. A change has been proposed to the DYK rules to address this. Wikipedia_talk:Did_you_know#Proposed_rule_change. Thank you. Kablammo (talk) 16:03, 3 November 2008 (UTC)

SELFPUB guideline backwards?

The first clause in WP:SELFPUB seems exactly wrong to me. It says we can use material from a dubious, or self-published source provided that the material used is relevant to the notability of the subject being discussed.

Assume the following two statements are published on the hypothetical website foobase.org:-

The Foobase Group believe foo to be quite a Good Thing, but only in moderation.
The Foobase Group offer expert advice on how best to use foo for fun and profit.

According to the above guideline, the article on The Foobase Group can contain the second statement, but not the first. This seems to be at odds with what I've always understood to be the principle behind SELFPUB, which says that you can use a group's own website to cite that group's views, but not much else.

I'm not sure what the intent of this line is. It could be any of the following:-

the material is not being used to establish the notability of the subject being discussed
the material used is notable, relative to the subject being discussed
the material used is notable and relevant to the subject being discussed

Does anyone have comments about any of these options? Of course, if I have completely misunderstood the intention of this, please enlighten me! Regards, SHEFFIELDSTEEL^TALK 20:31, 28 October 2008 (UTC)

I think this is merely an instance of (slightly) obfuscated writing. Since the list is introduced with "may only be used as sources about themselves, and then only if", it is clear that "the subject" must be the source itself. So you can use foobase.org to establish that "the Foobase Group believe that foo, in moderation, is good", or that "the Foobase Group offer advice on the use of foo, in a discussion of the Foobase Group. In your example the notability condition is automatically satisfied if the Foobase Group is notable for promoting the use of foo. It is not satisfied if the Foobase Group is notable for racist remarks made by their president. In other words: "relevant to the notability of the subject" means "relevant to what the subject is notable for", not "relevant as to whether we consider the subject notable". --Hans Adler (talk) 21:37, 28 October 2008 (UTC)

I agree with SheffieldSteel that this section is highly problematic, if not incomprehensible, and I don't recall anyone ever using it. The "relevant to what the subject is notable for" interpretation is not and should not be obeyed either. Frequently one wants to use self-published sources for things which have nothing to do with notability, but are obviously relevant, encyclopedic information, like basic biographical data about a person, e.g. date of birth, which might only be available from a self-published source, like a personal website, or the website of followers of a religious figure. And even if the foobase group is notable only for racist remarks, it is still common sense to tell people about its foo-related views just to tell us what on earth the foobase group is. If it is just saying "relevant to the subject of the article" like SheffieldSteel 's second and third interpretations above, it is OK, but just restating the obvious and WP:NOR. SheffieldSteel's first interpretation is covered by the notability guideline's insistence on independent reliable sources. So I think this section should be removed or clarified.John Z (talk) 07:33, 6 November 2008 (UTC)

Three kinds of facts: Historical, Experimental, Mathematical

There are basically three kinds of facts: Historical, Experimental and Mathematical. These are not perfectly separated, but the burden of proof is quite different for them.

Historical facts are statements of what happened. These can only be verified by establishing a chain of proof all the way back to the original event. The authenticity of a historical fact is the authenticity of the worst link in the chain. For these kind of facts you must have reliable sources, or multiple truly independent sources for the fact (truly independent mean that whatever parts are less reliable are repeated by different people or different methods). The Verifiability page as it currently stands (Nov 2, 2008) is primarily aimed at Verifiability of Historical facts, which do require reliable sources.
Experimental facts are statements of the results of some kind of repeatable experiment. For example, gravitational acceleration is about 9.8 m/s**2 on Earth is an experimental result. To verify this fact, you can either verify an experiment as a historical fact, or the experiment can be replicated (As I and many others have done in a physics lab class). Unlike a historical fact, you need not prove that anything happened, just what happens when you have the same conditions. For example, Galileo may or may not have dropped different weights of spheres off the leaning tower of Piza, but since the experiment (and similar ones) has been replicated, Galileo's theory of motion is not called into question. Many of the types of facts that will appear on wikipedia will be this type. For example, Windows 95 works better on a 16 MB of memory computer than Ubuntu 8.10 (which will not even boot). I'm not going provide any evidence for this claim other than that I have tried both, and you can too if you have the appropriate equipment. For some types of experiments, replication requires access to such complicated or expensive equipment that the easiest way to verify the fact is just treat the experiment as a historical fact.
Mathematical facts are basically proofs that some set of postulates can be used to prove some other postulate. For example 134978+80245=215223 is a mathematical fact (in base 10 at least). You could find that written in chalk on a sidewalk in New York city and you could verify it as true with just a piece of paper and a pencil. You could try searching for the rest of your life and you might never find a citation in a reliable source for it. For verifiability, all you need is sufficient details of the proof and to check each step. Of course, some proofs are complicated enough that few people are capable of checking them. However, currently even easily checkable mathematical types of facts are considered unverifiable based on this policy.

Basically, I think the verifiability policy is correct for historical facts, overly conservative for experimental facts and ridiculous for mathematical facts. I propose that alternative methods of verifiability should be allowed for experimental facts and mathematical facts. Jrincayc (talk) 16:42, 2 November 2008 (UTC)

I think you are on solid ground as to what you call "historical" and "mathematical" facts... but you run a-foul of several Wikipedia policies when it comes to what you are calling "experimental" facts. The problem is that while we can verify that other people have reached certain conclusions from their experiments (by citing to reliable sources that discuss these experiments and conclusions), we can not report on conclusions we reached in the process of conducting our own experiments (to do so would violate WP:No original research). Blueboar (talk) 19:52, 2 November 2008 (UTC)

I am not sure about how wikipedia should handle verifiability for experimental facts. I am not proposing that Wikipedia allow original research. I personally think that wiki's are superior to traditional peer-review (I often wish I had an edit this page or dissusion link for peer-reviewed papers), but Wikipedia is aiming to be an encyclopedia. For Wikipedia I am mainly thinking of things that can be verified by performing some simple real world experiment. For example here is a statement: "Firefox 3.0 does not have a direct link to the license in the about box and the help menu" Verification experiment: "get Firefox 3.0, Go to the help menu, notice the lack of link to a license. Go to the about dialog, and the only buttons are Ok and Credits." Basically, if the average wikipedian can verify the fact experimentally much faster than they can find a reliable source, it might make sense to allow the fact in Wikipedia. How exactly this would be done is not clear to me, but I think maybe the citation could include a reference to a non-reliable source along with list of wikipedian editors who have experimentally checked it. Jrincayc (talk) 02:56, 3 November 2008 (UTC)

Sometimes I feel that the No Original Research rule is taken to ridiculous extremes. Anyone with an IQ above room temperature can see for themself that the sky is blue. Roger (talk) 20:49, 3 November 2008 (UTC)

It sounds like you are touching on the subject of falsifiability, IOW such a statement could stand until someone replaced it with a reference that disproves it, but that might not fly. OTOH it might, since self-obvious facts don't need references. -- Fyslee / talk 03:12, 3 November 2008 (UTC)

Your Firefox example is similar to verifying the plot of a movie or a book... the statement is verified by the thing itself. Normally such self-verifying statements do not need citation, but if someone actually does challenge the statement that "Firefox 3.0 does not have a direct link to the licence in the about box and help menu", and insists on a citation, I would simply add the following: "<ref>Firefox 3.0 help menu and about box</ref>". Blueboar (talk) 20:27, 3 November 2008 (UTC)

This is interesting, honest. But I do not think verifiability is about burdern of proof. In fact, I think our V policy is meant to ensure we avoid these discussions, which are best left to historians, mathematicians, scientists, and of course philosophers. Our motto is "verifiability, not truth" so the means for establishing truth or veracity or facticity can't be part of verifiability. I think the idea is that what is important is that we can establish to everyone's agreement that someone (and someone who passes some test of notability) believes in something, regardless fo who they think bears the burden of proof or how it is met. Slrubenstein | Talk 22:07, 3 November 2008 (UTC)

Hmmm... SLR, you just wrote a word that I like a lot... veracity. Another way of saying our mantra is "verifiability, not veracity". Verifiability means demonstrating that someone said something. Veracity, on the other hand, means demonstrating that what they said is "true". I know that the "verifiability, not truth" statement has often caused confusion and controversy... perhaps changing it to "verifiability, not veracity" would make what we mean to say clearer and be a less controvercial wording. Just a thought for people to consider. Blueboar (talk) 14:05, 4 November 2008 (UTC)

Well, far be it for me to dissuade anyone from quoting me (though this has usually been done for very different purposes!) especially when it may help improve the clarity of our policies. Goodness knows, I would love it if we wouldn't have to explain that we really are not arbiters of what is true, right or factual, once a month. But I leave it to other editors to decide whether your proposed formulation would accomplish that. I certainly don't object! Slrubenstein | Talk 14:56, 4 November 2008 (UTC)

I agree that verifiability is very important. If what is written in Wikipedia can not be verified by subsequent editors and readers who need the correct result, then Wikipedia can only be used for things that don't really matter (like Pokemon and Harry Potter). I propose that the matra is "veracity through verifiability". I think that sometimes the best veracity that we can have is to verify that someone notable said something. But any reasonably simple method of providing the means of verifying a fact should be acceptable. I once asked [2] if using the birth date and the death date of a person could be used to establish the persons age and I was told that I should find a source for the age and not calculate it myself. I propose that the ultimate test for verifiability is that serious readers of the article should be able to verify every fact in the article by some method that is as simple as reasonable. It could be by checking the sources and it can also be by doing the experiment or by following the careful reasoning in the article or footnotes. I believe that for mathematics articles providing the proof of a proposition is more verifiable than providing a citation for the proof. The proof can be checked by anyone who knows the mathematics, but the citation can only be checked by someone who can get access to the place where the proof is written. So I firmly believe that if we care about Wikipedia becoming a serious resource, our mantra must not be "verifiability, not veracity" but "veracity through verifiability". Jrincayc (talk) 13:36, 5 November 2008 (UTC)

No, providing a mathematical proof without a citation constitutes Original Research, which we do not allow (see WP:NOR) ... one of the fundamental aspects of Wikipedia is that we base our articles on what other people (reliable sources) have written or said. Blueboar (talk) 15:08, 5 November 2008 (UTC)

The problem with writing your own mathematical proof is that you might get it wrong. You really don't want to prioritize "the self-evident (to me) proof I turned in as geometry homework this morning" over a proof published by a well-respected mathematician. There's nothing about WP:V that prohibits you from including the mathematician's proof; it only requires that you also provide a reliable source for the proof. Think "both/and", not "either/or". WhatamIdoing (talk) 18:45, 5 November 2008 (UTC)

Commonsense is required here. Wikipedia is the encyclopedia that anyone can edit. We therefore must take a minimalist, lowest-common-denominator approach to our readership and editors. Simple arithmetic seems reasonable to assume for anyone capable of reading Wikipedia but higher mathematics is not. Mathematical proofs are generally inappropriate material for our articles per WP:NOTTEXTBOOK and WP:NOT PAPERS. Colonel Warden (talk) 23:51, 6 November 2008 (UTC)
- That's an excessively sweeping opinion. In many cases where mathematical theorems themselves are discussed, brief proofs or proof sketches are perfectly encyclopedic. For example, Diagonal lemma. We've never developed a guideline for when to include proofs and when to omit them, so there are certainly some proofs in articles that should be omitted and some proofs that are omitted that should be included or sketched. The use of proofs in Wikipedia articles, and the difference between an encyclopedia and a textbook, is too complex a topic to fit into a thread about the verifiability policy. — Carl (CBM · talk) 00:03, 7 November 2008 (UTC)

Like many mathematical articles, Diagonal lemma does not conform to our policy: A Wikipedia article should not be presented on the assumption that the reader is well versed in the topic's field. Introductory language in the lead and initial sections of the article should be written in plain terms and concepts that can be understood by any literate reader of Wikipedia without any knowledge in the given field before advancing to more detailed explanations of the topic. While wikilinks should be provided for advanced terms and concepts in that field, articles should be written on the assumption that the reader will not follow these links, instead attempting to infer their meaning from the text. Verifiability is not possible if any literate reader is unable to grasp what is being said. Colonel Warden (talk) 00:32, 7 November 2008 (UTC)
- It's hard to see how the lede of Diagonal lemma could be less technical. Lower down, there is footnote number 2 that gives several different references where the lemma is explicitly discussed (and where the proof can be found), so that verifiability is not in question. That is my real point here - that verifiability, per se, is not the issue. Whether WP:NOT is out of sync with actual practice is a different issue. — Carl (CBM · talk) 01:34, 7 November 2008 (UTC)

"challenged or likely to be challenged"

The first sentence in the Burden of evidence section contains the phrase challenged or likely to be challenged, without specifying WHO is making the challenge. I think most reasonable editors realize that this certainly includes challenges from other editors, but I have seen this used by an experienced ^{(at stonewalling)} editor who stonewalled for a long time by demanding V & RS from outside Wikipedia for the existence of a challenge or doubt about the included material.

To solve this problem, I suggest we include the words "by Wikipedia editors" at the end of that phrase. The whole idea here is to make sure that there is consensus among Wikipedia editors for the inclusion of material. When another editor challenges an edit, the one proposing inclusion should be able to provide a very precise explanation and sourcing to clear things up and to satisfy the challenging editor's doubts about whether the source actually says what is claimed. -- Fyslee / talk 16:01, 1 November 2008 (UTC)

I think you are talking about two types of "challenge"... the first (and the one that we are talking about in this policy) is when a statement is uncited, and someone asks for it to be cited. I think everyone is agreed that if this occurs, the information must be cited or it might be removed. The second is when a citation is given, but someone questions whether that citation actually backs what is stated. That is a different issue, and has a different solution. The first step there is to raise the issue on the talk page and discuss the problem.

In other words... this policy is talking about challenging uncited statements. It is not talking about challenging sources. Blueboar (talk) 16:45, 1 November 2008 (UTC)

I see your point (the difference between "before citation" and "after citation" challenges). Then we need to include this second issue here, which can be done in one sentence. Maybe including something like: "When the correct (or "proper") application of a source is challenged or likely to be challenged by any Wikipedia editor....." Doing this can save lots of discussion on talk pages. -- Fyslee / talk 16:56, 1 November 2008 (UTC)

Saving lots of discussion on talk pages is not within the goal of this policy. In fact, when there is a question as to whether a source is being improperly used, there should be lots of discussion on the relevant talk page. Also, there is no need to include the improper use of sources in this policy... as we already deal with that topic at WP:NOR (saying that a source supports a statement when it does not actually support is considered a form of Original Research). In other words... I don't really see a need to change things. Blueboar (talk) 17:24, 1 November 2008 (UTC)

(I have added a parenthetical phrase above.) When the existence of a clearly worded policy can be used to save a needless waste of time in a seemingly endless (months long) discussion, then that is a very positive side effect of the existence of such a clearly worded policy. Nothing wrong with that. The fact that such a discussion can occur indicates that the policy isn't clear enough. But,.... I think we aren't really understanding each other. My meaning is "the other way around." When an editor includes a source, and the way the source is being used is challenged by other editors, it is the obligation of the one including the source (IOW the editor being challenged) to justify their use of the source by pointing to the exact wording in the source. When that editor refuses to do so and just stonewalls by claiming the source isn't questioned outside of Wikipedia, and that it isn't being "challenged" outside of Wikipedia, and therefore other editors' objections have no weight, we end up in a pointless discussion. You see, it's hard to accuse that editor of OR if no one is being informed about which part of the source is being used to justify the inclusion and application, and the editor in question refuses to explain, except to point to this policy and claim that it refers only to outside-of-Wikipedia "challenges", and not to challenges from other editors. While I wouldn't exclude the possibility that the wording could apply to "outside" challenges, I believe it primarily refers to "inside" challenges from other editors during the editorial process. That is not clear from the presently ambiguous wording, at least ambiguous to that one editor.

Maybe that's as clear as mud to you (my fault), but will you at least admit that the inclusion of the three words ("by Wikipedia editors") won't hurt anything, won't weaken the policy, will make it unambiguous, and will increase the clarity, intent, and useability of this policy? Or do you believe that this only applies to challenges from sources outside of Wikipedia? The present wording shouldn't be ambiguous.

If this only refers to "before citation" challenges, then what is our "verifiability" (within the source itself) policy on "after citation" challenges? Wouldn't V apply in this situation in the absence of OR? -- Fyslee / talk 23:53, 1 November 2008 (UTC)

Fyslee: can you expand on what you mean by "after citation" challenges? Can you give some examples? — Carl (CBM · talk) 00:03, 2 November 2008 (UTC)

Well, I have always understood the wording to mean both "before citation" and "after citation" challenges. Blueboar states pretty clearly ("this policy is talking about challenging uncited statements. It is not talking about challenging sources") that the wording only applies to "before citation" challenges, IOW a demand that a given wording be backed up by a V & RS. That's obvious and we agree on that. I am seeking to ensure that we have wording here at V that covers verifiability within the provided source, IOW that the obligation still rests upon the editor providing the source to point out what wording in the source justifies the inclusion when they are "challenged" by other editors to do so. I believe that it is proper to challenge another editor to "verify" the application of the source. Is it really damaging to this policy to ensure unambiguity on this point when it can be done so easily, or has this matter already been discussed? -- Fyslee / talk 00:14, 2 November 2008 (UTC)

Agree with Fyslee, at least that he got some POV pushing. I don't think we can write policy to eliminate POV pushing, though, beyond a certain point. Perhaps he needed to ask for more help. I agree with this: "My meaning is "the other way around." When an editor includes a source, and the way the source is being used is challenged by other editors, it is the obligation of the one including the source (IOW the editor being challenged) to justify their use of the source by pointing to the exact wording in the source." We already have that here "When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested should be provided as a courtesy to substantiate the reference.", but perhaps it needs to be strengthened to a requirement. I have come across a similar situation to the one Fyslee describes. I don't see how the addition could hurt, and I do see how it could help. I suggest we modify the footnote to read "When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested must be provided to substantiate the reference, or the cited text may be removed." This seems to be a continuation of an older discussion. Now, where you're not going to eliminate POV pushing there is that an editor is going to ask for a detailed chemical analysis of the ink on the page. ——Martin^phi ☎ Ψ Φ—— 02:14, 2 November 2008 (UTC)

Fyslee, If you have not done so, please read WP:No original research. I think that policy covers what you are concerned about (disputes as to whether the article text is fully supported by the given source). Blueboar (talk) 13:01, 2 November 2008 (UTC)

Of course I have read it, many times at that, and it partially applies, since many of these policies are tangentially related to each other, but I'm here at WP:V because "verification" is the issue, and I believe that demands for verification (both "before citation" and "after citation") should be covered in this policy.

An example

Yes, NOR refers to this policy, and it's the "challenges" bit I am concerned about. Maybe we could have saved a lot of discussion here if I hadn't been relying on sketchy impressions about old and half forgotten incidents, instead of citing them directly. Maybe looking at such an incident might help. In the process we might even discover that this is indeed the wrong place for this discussion, but we will at least have learned something, which isn't a bad thing. I'm certainly open to learning more about how to apply policies, and the following dispute probably involves several of them, but my main point is that an editor (QG) refused to respond to the challenges from other editors because the "dispute" (notice that word) was at Wikipedia, not outside of Wikipedia. I'll refer to the archived locations for the discussion.

Here's the situation. We challenged QG regarding his claim that because there was no evidence of a dispute outside of Wikipedia being presented, that our editorial dispute at Wikipedia didn't count, and his edit should be accepted on his premises alone.

One version of the disputed text was this: A chiropractic article has described chiropractic as "a profession at the crossroads of mainstream and alternative medicine."^[3]

QG claimed that this was a fact, and others claimed that it was an opinion, and thus should be attributed (to the "chiropractic authors"). He claimed attribution was unnecessary because he considered it to be a fact. I too think it's a "fact" that chiropractic is "at the crossroads", but that's a personal opinion, and that the source represents the (correct) opinion of the chiropractic authors. To be on the safe side I considered attribution and precise quoting to be desirable. I never dreamed that it would be disputed that it was at least an opinion. That the factual nature of the opinion was being disputed by editors was also obvious from the discussions among editors, and I believe that dispute needed to be addressed, irregardless of whether the opinion was or wasn't being disputed outside of Wikipedia. I also know that there are differences of opinion on the factuality of that opinion outside of Wikipedia, but the fact that a dispute arose among editors should be sufficient for a " verification challenge" to be met by the editor making or defending the proposed addition. Copping out wasn't helpful. Here's a relevant section:

A bit from here:

For the second, this has nothing to do with "facts", but about what is clearly an opinion by two noted chiropractic authors, one of whom is also an MD. Other mainstream (and reform chiropractic) authors might - and do - have the opinion that chiropractic is mostly fringe nonsense, with only a pretense of mainstream veneer. That type of opinion would also need to be attributed. There is no question that the statement is an opinion, so don't meddle in what you don't understand, or don't wish to understand. What is opinion to one is fact to another, and vice versa, so both would need to be attributed and sourced.

Knowing your current running battle with Levine2112, I don't think an AGF is necessary here. Your comments here seem more like obstinate baiting to me, and I think you should leave him alone. -- Fyslee / talk 23:19, 20 September 2008 (UTC)

Attribution can sometimes be a necessary fact to include about an opinion. That chiropractic is at the crossroads of mainstream and alternative medicine is just an opinion, so we shouldn't assert that opinion as a fact. However it is a fact that so-and-so has the opinion that chiropractic is at the crossroads of mainstream and alternative medicine. -- Levine2112 ^discuss 21:36, 20 September 2008 (UTC)

Well put and quite correct. -- Fyslee / talk 23:19, 20 September 2008 (UTC)

You have not explained it to be necessary in this case. QuackGuru 21:39, 20 September 2008 (UTC)

It's necessary because it's an opinion, and one that can be controversial to many people. -- Fyslee / talk 23:19, 20 September 2008 (UTC)

Please provide evidence it's an oipinion, and that there is a seriuos dispute among reliable sources. It is a fact until evidence of a serious dispute is presented. QuackGuru 23:26, 20 September 2008 (UTC)

Sure I have. -- Levine2112 ^discuss 22:15, 20 September 2008 (UTC)

Based on WP:ASF, you have never explained any need for attribution in this particular case. QuackGuru 22:18, 20 September 2008 (UTC)

Yes I have. -- Levine2112 ^discuss 22:19, 20 September 2008 (UTC)

Levine2112 has not explained why attribution is necessary in this case based on WP:ASF. QuackGuru 22:28, 20 September 2008 (UTC)

QuackGuru, Levine2112 has explained very well. Since this is a very simple formulation, bringing up WP:ASF is just moving the bases and also irrelevant baiting, so please desist. -- Fyslee / talk 23:19, 20 September 2008 (UTC)

Please provide evidence there is a serious dispute among reliable sources in order for attribution to be necessary.

Bringing up WP:ASF is important part of discussion. Per WP:ASF, attribution is unecessary when there is no serious disagreement among reliable sources. QuackGuru 23:26, 20 September 2008 (UTC)

<-- QuackGuru, since you continue your baiting and are now moving the bases to "among reliable sources", I'm going to ask you to precisely quote what part you are referring to, because I really don't see your point. I'll even make it easy for you by copying the statement here so we are on the "same page":

Chiropractic authors have described the profession as being at the crossroads of mainstream and alternative medicine.^[3]

Please state precisely which of those words you are claiming are undisputed "facts". -- Fyslee / talk 23:48, 20 September 2008 (UTC)

It is an undisputed fact if no serious disagreement is presented. If an editor provides evidence it is an opinion (an opinion is when a serious dispute is presented) then it is not a fact. I am asking for evidence if a serious dispute exists. See WP:ASF. If the text is disputed then we can add attribution. If evidence is not presented then we can assert it as fact. QuackGuru 00:06, 21 September 2008 (UTC)

The stonewalling I referred to before this example is QG's constant repetition of "Please provide evidence it's an opinion...". That's a cop out. Instead QG should have dealt with the existing dispute among editors about whether it was or was not an opinion. Our dispute about the matter counts too.

Some from here, where the stonewalling continues:

There are still improvements to be made. The crossroads bit has to do more with a historical perspective and can be added to the history section and the combines aspects bit is an accurate description for the scope of practice section. There is also a discussion at the NPOV board about all the unecessary attribution added to this article. See Crossroads of Chiropractic. An archived discussion about asserting facts can be found here. QuackGuru 19:19, 28 September 2008 (UTC)

On this page you have people arguing that it is pseudoscience. . . I don't think the crossroads statement is a fact. . . I like it though because it is a far nicer description than the this place typically gives to chiropractic.TheDoctorIsIn (talk) 00:00, 30 September 2008 (UTC)

It is a sourced opinion about what I personally believe is a fact. Keating believed that the profession was a mixture of science, pseudoscience, and antiscience. -- Fyslee / talk 06:18, 30 September 2008 (UTC)
I believe it to be true, but I agree with you that it "is a sourced opinion", and as such it should be attributed. DigitalC (talk) 08:56, 30 September 2008 (UTC)

Per WP:ASF, please provide evidence it is an opinion. No serious disagreement exists. No evidence of an opinion has been presented.

Per WP:ASF, By value or opinion,^[4] on the other hand, we mean "a matter which is subject to dispute." This is Wikipedia's definition of an opinion. QuackGuru 21:11, 30 September 2008 (UTC)
And again, per WP:ASF, this is a statement is one that "clearly express values or opinions." DigitalC (talk) 08:50, 1 October 2008 (UTC)
Wikipedia's definition of a value or opinion is: "a matter which is subject to dispute." per WP:ASF.

No evidence of a dispute has been presented. That means we cannot claim it is disputed. When no dispute exists we are free to assert it. QuackGuru 17:44, 1 October 2008 (UTC)

I suspect that several policies and guidelines apply to the above dispute. Please share your thoughts about which policies apply here. -- Fyslee / talk 16:26, 2 November 2008 (UTC)

OK... to find out which policy applies I need to ask a few questions. First: I may have missed it, but where does the quote in question ("Chiropractic authors have described the profession as being at the crossroads of mainstream and alternative medicine") come from... is this a direct quote from a reliable published source, or an editors comment? Blueboar (talk) 17:39, 2 November 2008 (UTC)

Here's another version, indicating the editorial version, which quotes a source. I support both attribution and using an exact quote. I have added a functioning ref to that source. (I have also substituted this version above):

A chiropractic article has described chiropractic as "a profession at the crossroads of mainstream and alternative medicine."^[3]

Meeker and Haldeman are chiropractic's two top researchers. Meeker is also President of Palmer College of Chiropractic, West Campus, and Haldeman is also an MD. -- Fyslee / talk 18:13, 2 November 2008 (UTC)

Thanks, I would definitely say that the statement, as originally written, is verifiable (I will assume that Meeker-Haldeman article is not being challenged on the grounds that it is a non-reliable source). The statement that "Chiropractic authors have described the profession as being ...." is a factual statement that is backed by the source provided. We do indeed have some "Chiropractic authors" who have discribed the profession in that way.

That said, If Meeker and Haldeman are the only Chiropractic authors to say this, then we should probably directly attribute the satement (by changing it to: "Meeker and Haldeman, two Chiropractic authors, have discribed the profession as being...." )

We now come to the next level of question... on the basis of what Meeker and Haldeman say, can we bluntly state that: "The profession is at a crossroads..." That is a somewhat harder determination that depends on many factors. First, it depends on the reputation of Meeker and Haldeman... we need to look at what else they have written, and how their corpus of work has been recieved. Is the cited journal article itself cited in other journals?... more importantly, has anyone written a rebuttal of their work. In other words, how reliable are Meeker and Haldeman? What end of the reliability spectrum should they be placed (are they at the "unimpeachable expert" end or the "yea, they know what they are talking about" end). Without knowing the answers to these questions, no one here can advise you on this. I would, however, say that if there is any doubt (and since chiropracty is a controvercial medical topic I would assume there is), it is probably safer to directly attribute their statement, rather than accept it as blunt fact. Blueboar (talk) 19:40, 2 November 2008 (UTC)

Thanks for the very thorough and refreshing analysis. M&H are unquestionably top chiropractic researchers and leaders within the profession, and that article is cited widely. Since chiropractic ("chiropracty" isn't a real word) is indeed very controversial and has plenty of internal division, statements of any kind will be criticized by someone within the profession. ("For the consumer and third-party payer (that's insurance) this presents a dilemma that borders on a nightmare. In short, there is no standard of care in the profession. As Forest Gump might say, "A chiropractic office is like a box of chocolates: YOU NEVER KNOW WHAT YOU'RE GOING TO GET!" - John Badanes, DC, PharmD [3]) Ultra-straight chiropractors would find their statements to be concessions towards mainstream science and evidence-based medicine, which are anathema to the original straight wing of the profession. Because of this I hold to your opinion, that it is safer to attribute and quote them.

I felt that QG's refusal to deal with the issue in a proper manner was disturbing and disruptive. Was it proper to demand that he deal with the existing dispute among editors, or was it proper for him to demand sources for proof of some external dispute? -- Fyslee / talk 03:04, 3 November 2008 (UTC)

Certainly, a dispute among editors is a sufficient challenge for attribution. Do you now feel that the current policy is strong enough, or do you think we should modify the footnote or something? ——Martin^phi ☎ Ψ Φ—— 03:19, 3 November 2008 (UTC)

I obviously think that editorial disputes are sufficient to justify challenges for "after citation" explanations and evidence. Since this policy currently already covers challenges for "before citation" needs for evidence, we just need to cover the other situation, since it also deals with verifiability. -- Fyslee / talk 03:27, 3 November 2008 (UTC)

I suppose when it comes down to it, my take on the situation is that before citation challenges are within the proper scope of this policy, while after citation challenges are within the scope of WP:NOR. If a statement has been cited to a reliable source, this policy has been complied with. The statement and citation may still run afoul of other policies and guidelines... but this one has been met. I don't see the need to expand it. Blueboar (talk) 16:07, 4 November 2008 (UTC)

References

^ When content in Wikipedia requires direct substantiation, the established convention is to provide an inline citation to the supporting references. The rationale is that this provides the most direct means to verify whether the content is consistent with the references. Alternative conventions exist, and are acceptable if they provide clear and precise attribution for the article's assertions, but inline citations are considered 'best practice' under this rationale. For more details, please consult Wikipedia:Citing_sources#How to cite sources.
^ When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested should be provided as a courtesy to substantiate the reference.
^ ^a ^b ^c Meeker WC, Haldeman S (2002). "Chiropractic: a profession at the crossroads of mainstream and alternative medicine" (PDF). Ann Intern Med. 136 (3): 216–27. PMID 11827498.
^ Opinions involve both matters of fact and value; see fact-value distinction

Discussion

Fyslee, I'm not sure what you want from us here. Yes, that's an opinion -- a dominant opinion, but still an opinion. You could have short-circuited the above idiotic conversation with QG by presenting actual examples of divergent opinions: "Medical researcher John Smith says that chiropractic is complete bunkum and ought to be banned along with all forms of alternative medicine.^[1] Herbalist Mary Jones says that it's mainstream science and therefore sold its soul to the medical-industrial complex.^[2] Major Physicians' Club says that it's mainstream for acute pain and voodoo for chronic pain.^[3] Therefore there is a dispute among the reliable sources on exactly how to classify chiropractic with respect to conventional and alternative medicine approaches."

But I don't see any particular need to change this policy. Changing this policy is not going to make QG be less tendentious. WhatamIdoing (talk) 19:40, 3 November 2008 (UTC)

I'm not seeking some earth shattering change to the policy, but a clarification of it. Right now it seems to apply only to challenges to "before citation" matters. The idiotic discussion shouldn't have happened, because QG should have realized that the difference of opinions (his against the rest of us) did indicate it was likely an opinion (well, even the expression of facts are the opinions of the one expressing them), or at least that adding attribution would solve the dispute. Attribution is playing on the safe side. Rather one too many attributions than one too few.

BTW, this example is mentioned in a discussion at http://en.wikipedia.org/wiki/Wikipedia_talk:Disruptive_editing#It_is_time_to_make_this_a_policy . There Slrubinstein makes this comment:

"Here is a good example of what I mean: each individual edit is civil. The user appears to care about policy. But the pattern is trollish. I think we need to find more such examples and really come up with a good list of diagnostics of the pattern. I think repeating the same questions rather than responding to another user's answers may be one sign. It has a lot to do with not explaining one's own reasopnign AND a pattern of ignoring the reasoning provided by others. But I think there is more to it. I really hope other experienced eidtors will join in here!!"

QG should have replied properly to the challenges, instead of stonewalling. If "after citation" challenges were part of this policy, we could have avoided the above discussion. Disputes among editors must not be ignored. If a provided source is challenged as to its "fact" or "opinion" status, then attribution is the safest way to deal with it.

Attribution relates to verification of the authorship of the item in question, and also to WP:NPOV. This shows how our policies are interrelated. There the phrase "a matter which is subject to dispute" is used, and I think that editorial disputes qualify as challenges. We could tweak the NPOV policy on that point to include editorial challenges as proof that the "matter is subject to dispute." -- Fyslee / talk 15:03, 4 November 2008 (UTC)

There seems to be some confusion here. As I understand QG's point, it was based upon WP:ASF which is a portion of a different page (WP:NPOV). He contended that the statement made had the status of fact per that policy. In showing that this was instead an opinion requiring attribution, the burden was upon Fyslee and others to produce contrary statements, as ably done by WhatamIdoing above. QG was neither trolling nor stonewalling but simply sticking to his correct position, based upon a source. This is a different circumstance from that discussed on this page (WP:V). Generally speaking, editors should not expect their own personal opinions to trump a contrary position which is based upon a source as editors' own opinions have little standing. Colonel Warden (talk) 23:36, 6 November 2008 (UTC)

The confusion you perceive is because we have diverged from the original discussion, which is understood by the previous participants in this discussion. You have just entered at a late point and are excused. My original point was based on a faulty recollection and thus used a hypothetical situation that wasn't identical to the QG situation used as an example. That brought other policies into the discussion, but since they were related, they were OK for discussion. QG had only asserted that it was fact (I agree that it was fact, but still an opinion about a fact), and ignored all the other editors' contentions that it was an opinion, and he ignored those editor's contentions, IOW he demanded outside sources for something that was being proven before his very eyes right here. Editorial disputes by a large number of editors are, in and of themselves, sufficient proof of an existing dispute, and for editing to progress those concerns should be recognized, not ignored. -- Fyslee / talk 20:10, 9 November 2008 (UTC)

burden of evidence

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

Is this only saying that the editor adding material must provide evidence in the form of citations to reliable sources, or does it also mean that the burden of evidence for showing the source are reliable also lies with the editor adding content?

For example if an editor adds information based on an article from the LA Times, and another editor challenges it saying the LA Times is not a reliable source, does the first editor have to somehow proof that the LA Times is a reliable source? If so, what kinds of evidences are sufficient? Readin (talk) 18:43, 7 November 2008 (UTC)

I think we should clarify the article to make it clear what we mean. Readin (talk) 18:46, 7 November 2008 (UTC)

I would say that the burden of evidence shifts once the citation is provided. If there is a question as to the reliability of the source, I think that the person making the challenge should have to state why he/she feels the source is unreliable. On the other hand, I also think that people who think the source is reliable should have to defend the reliability of their source if asked to do so. In other words... it is up to both sides to enguage in discussion, and seek outside opinions if they can not settle the matter.

In the example you give... since the LA Times is the second largest mainstream newspaper in the country, it should be easy to defend it. If someone is seriously challenging the reliability of the LA Times, then you should seek a second opinion and get some support. That said, no source is always reliable... even the LA Times can be unreliable in specific situations (for example, if the paper states something as a fact on Monday, but it can be demonstrated that on Wednesday the paper issued a retraction, then the Monday edition is not reliable for that statement of fact.) A lot depends on the source, what it is being used to support, and how it is being used. Blueboar (talk) 23:22, 7 November 2008 (UTC)

Advocacy groups?

Recently I came across a section in AK-47 which cited a single source which was controlarms.org. I removed it per WP:V's barring of sources which are "promotional in nature". Someone accused me of "going out on a limb" by doing so. Where does WP stand on sources like that? Controlarms.org is an organization which advocates international arms trade regulation. This to me seems to indicate that they would be a biased or POV source. What are your thoughts? —Memotype::T 15:00, 9 November 2008 (UTC)

The fact that an organization may have a bias or POV does not affect whether we can use them as a source (see WP:NPOV). However, it can affect how we use them. If they are extremist in their POV, their use is very limited. If they represent a relatively mainstream POV (and it sounds like this is the case with controlarm.org) then there use is less limited. I would say that something like Controlarms.com can be used for a statement of opinion with attribution (as in: Controlarm.org, a political advocacy group that supports international arms trade regulation, has stated "blah blah blah"<cite to Controlarms.com>) however it should not be used to support a statement of fact. Blueboar (talk) 15:21, 9 November 2008 (UTC)

Ok, well it was being used as a statement of fact. Thanks for the clarification. —Memotype::T 19:55, 9 November 2008 (UTC)

When in doubt about whether an opinion is just an opinion, or whether it's an opinion about something generally recognized as an absolute fact, attribution will never hurt. Better one attribution too many than one too few. Stop and think about scientific research papers, where even statements of fact are often attributed as a means of giving proper recognition and avoiding charges of plagiarism. We can do the same here, regardless of the reasons, and thus ensure that we also comply with verifiability requirements. -- Fyslee / talk 20:29, 9 November 2008 (UTC)

You know... we should add something like "When in doubt, attribute" to the Policy. It would solve a lot of the questions we get here and at RSN. Blueboar (talk) 20:35, 9 November 2008 (UTC)

True enough. If there happens to be any error here, at least it's on the safe side. Attribution ensures that readers know that some editor isn't including their personal POV. I wonder how many would object to adding such a short sentence to the policy? -- Fyslee / talk 03:13, 10 November 2008 (UTC)

Verifying unpublished information

Why does this policy persist with excluding independent and unpublisised material and information? There are many significant, important, notable, influential and even popular, writings, musicians, films, etc that may not have the financial means to gain mas media attention to be publisised in order to comply with this policy and as such WP inevitably excluse vital information of encyclopedic importance. Nick carson (talk) 04:11, 7 November 2008 (UTC)

I note that you have posted similar querries at NOR, and at RS. You have recieved answers on those talk pages, so I will not repeat them here. All I can say is that I am sorry that you don't like the fundamental principles behind Wikipedia. Perhaps you would be happier editing at some other wiki that has different rules and philosophy. Blueboar (talk) 14:08, 7 November 2008 (UTC)

It would be apathetic of me to avoid such problems by leaving Wikipedia. I wouldn't be asking these questions if I didn't genuinely think there were better ways these policies could be structured. What I would like to see is some discussion regarding these questions rather than witness the acceptance of these policies as absolute. Surely they are subject to progression and amendment? Nick carson (talk) 10:20, 8 November 2008 (UTC)

Actually, no, in this case our policies are not open to "progression and amendment", at least not in the way you want. The idea that information in Wikipedia must be verifiable to reliable published sources is one of the fundamental core policies of the project. We can discuss how to improve the wording of these core policies (WP:V, WP:NPOV, and WP:NOR), but the basic concepts they impart are considered absolute. Blueboar (talk) 16:05, 8 November 2008 (UTC)

There's somewhat of a contradiction there. I'm not arguing against these "core principles", rather I'm requesting that they be amended. Altering their wording constitutes amendment, so you see where this contradiction lies. Nick carson (talk) 02:23, 9 November 2008 (UTC)

I would not call allowing unpublished (and thus unverifiable) sources an "amendendment"... I would call that a change of core principle. It changes the policy in a substantial way - to the extent that it would be a complete reversal of what it now says. Blueboar (talk) 04:00, 9 November 2008 (UTC)

I'm not sure how it could be achieved, but there is information that is notable as per WP notability policy that is excluded based on the V & RS policies. Nick carson (talk) 05:56, 9 November 2008 (UTC)

Can you give an example? SlimVirgin ^talk|edits 07:24, 9 November 2008 (UTC)

Indeed, an example would help. Even more so, a specifically worded proposed amendment would be even more helpful. NPOV is absolute, while other polices can be tweaked, but I would strongly advise you not to attempt to do so. Start with a new section (or a subsection here) and show us your proposed improvement. Then it can be discussed. -- Fyslee / talk 20:19, 9 November 2008 (UTC)

I'll have to give that a bit more of a think. Hasn't anyone ever raised the question of notable unpublished informaton in the past? If so, has this been adressed in the current policy? Nick carson (talk) 10:07, 10 November 2008 (UTC)

Lots of people have raised the question... usually because they "know" some fact is true, but have no source to support it. They were all given the same answer we have given you so far: Facts in Wikipedia need to be verifiable, and the only way for that to happen is to have the fact stated in a reliable published source. Blueboar (talk) 13:10, 10 November 2008 (UTC)

A discussion regarding the primary source or secondary source classification of highway maps

Wikipedia talk:No original research#Regarding maps being "primary sources" according to this policy --Rschen7754 (T C) 06:41, 16 November 2008 (UTC)

Self-published sources on themselves

Someone changed this section in two ways, [4] but it's not clear whether the change was deliberate. It was changed from:

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

to

(b) Self-published and questionable sources may only be used as sources about themselves, and then only if all of the following conditions are true:

The two changes are (i) that self-published sources may only be used in articles about themselves, which isn't true or sensible; and (ii) that they may be used, without further qualification, as sources about themselves in articles not about themselves, which would be a significant change, and would somewhat contradict the rest of the sps section of this policy. I believe it would be a welcome and much-needed change, but we'd need to add that reliable third-party sources must have identified that source as relevant to the subject matter. Otherwise, we're allowing any sps, or any fan of an sps, to decide that his views or actions are notable in relation to any subject.

I've changed it back until we decide whether change (ii) was deliberate and desirable. SlimVirgin ^talk|edits 19:06, 5 November 2008 (UTC)

~~I'm sorry ... but how are (i) and (ii) not contradictory?~~ Nevermind, I see the difference. Anyway, there's a lot of confusion regarding self-published sources, not the least because WP:SELFPUB redirects to the wrong place -- it should direct to the "Self-published sources" section, which provides an overview and introduction to the policy regarding self-published sources. The section you edited is more directed at sources which don't meet the bar of experts published in reliable publications on the issue at hand. And yes, I think (ii) is obviously quite reasonable, especially when you consider things like politicians or people under attack from someone. II | (t - c) 19:14, 5 November 2008 (UTC)

Regarding your last point -- yes, exactly. We would just need to build in a qualification that would disallow e.g. self-published astrologers from being used in Astronomy. SlimVirgin ^talk|edits 19:42, 5 November 2008 (UTC)

I have taken a crack at this. The grammar probably needs to be worked on if this sticks, but I have realized that policy edits have to be taken slowly. This addition should also help clarify that you can't simply pick any old website as a primary source (not understanding this caused me some trouble.) PSWG1920 (talk) 21:11, 5 November 2008 (UTC)

I went ahead and fixed some rather awkward wording, here is my cumulative change. PSWG1920 (talk) 01:34, 6 November 2008 (UTC)

I removed that the material shouldn't be contentious, because it's precisely for contentious material that we sometimes want to use an SPS for information about itself. For example, someone has accused X of something bad, and it's widely accepted that X did it. X responds on his website saying he didn't do it, and offers a convoluted, unconvincing defense that no other source has bothered to publish. We might want to cite that defense, even though we know it's contentious. SlimVirgin ^talk|edits 02:26, 6 November 2008 (UTC)

I agree with that removal. The one part of your edit that I'm not sure about is your addition of "reliable" here: "the source has been specifically mentioned in relation to the article's subject by an independent reliable source". I deliberately left that word out because of issues addressed by WP:PARITY. PSWG1920 (talk) 02:39, 6 November 2008 (UTC)

Also, a "convoluted, unconvincing defense that no other source has bothered to publish" could be seen as unduly self-serving, so perhaps that restriction should also be lifted or further qualified. PSWG1920 (talk) 02:44, 6 November 2008 (UTC)

I'm not sure about WP:PARITY. It's problematic, and it could end up allowing a self-published astrologer to comment on astronomy, so long as other astrologers had written about her in relaton to astronomy. Also, it's a guideline and this is a policy, so it should be guided by this page, not vice versa.

As for unduly self-serving, I've always liked that qualification. It's there to stop self-published sources from adding a lot of vanity stuff to their websites then adding it to their WP bio. It means we can consider using it, but have a good reason to turn it down if it's over the top. SlimVirgin ^talk|edits 02:55, 6 November 2008 (UTC)

I still think "reliable" is a bit too strong in the above context. For example, the original material on Quackwatch is self-published, and by my understanding is not considered reliable in a strict sense by Wikipedia (I think I remember hearing about an ArbCom case which decided that), but nonetheless is highly notable and presumably could establish the notability of another source. "Source independent of the subject" would prevent your example of a self-published astrologer from happening, but I suppose it still might not be strong enough. PSWG1920 (talk) 18:54, 6 November 2008 (UTC)

Given SlimVirgin's above reasoning, should "it is not contentious" be removed from here as well? PSWG1920 (talk) 03:58, 8 November 2008 (UTC)

Actually, if I remember correctly, it was the inclusion of the "it is not contentious" language at BLP that caused us to include it here. It was an attempt clear up confusion between two policies. Blueboar (talk) 16:04, 9 November 2008 (UTC)

Are we concerned about individuals or groups as "self-published sources"? For example, the material published by Amnesty International and Human Rights Watch is "self-published". Should it be excluded? Jayjg ^(talk) 02:36, 13 November 2008 (UTC)

Self-published always meant by one individual or a very small, informal group, with no professional oversight. Amnesty is a huge organization with lots of employees, including lawyers and professional writers, so it doesn't count as an SPS. SlimVirgin ^talk|edits 02:56, 13 November 2008 (UTC)

I'm not sure the policy makes that clear enough. Can you provide an example or two of a self-published individual we'd like to be able to cite in an article not about the individual? Jayjg ^(talk) 03:23, 13 November 2008 (UTC)

It's not just self-published sources, but questionable ones too. An example would be citing material from Mein Kampf in The Holocaust (not self-published, but definitely a questionable source). As the policy stood, we wouldn't have been able to do that, though of course people did where appropriate. SlimVirgin ^talk|edits 06:10, 13 November 2008 (UTC)

I've definitely had to explain the difference between selfpubs by experts and selfpubs by nonexperts. Selfpubs by nonexperts are generally limited to biographies of themselves. Selfpubs by experts can be cited in their area of expertise, but not BLPs and not to establish notability. And I'm of the opinion that publications by political watchdog groups, etc, fall in between secondary sources and primary sources; "self-published" to me implies by individuals. Squidfryerchef (talk) 04:28, 13 November 2008 (UTC)

I agree with the notion of distinguishing between recognized experts and non-experts. However, I disagree with the idea that self-published should only aply to individuals. Organizations can certainly self-publish. I would say that the reliability of organizational self-published sources rises along with the reputation of the organization, no different than for an individual. Blueboar (talk) 13:46, 13 November 2008 (UTC)

I'm concerned about using self-published experts; something an expert publishes in a peer-reviewed journal is quite different from something he/she publishes on his/her blog. Jayjg ^(talk) 02:19, 14 November 2008 (UTC)

Now that I read it again, WP:SPS is a bit unclear on this. "Self-published material may, in some circumstances, be acceptable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications." Was this originally intended to mean that such could be treated as reliable in a strict sense, or just that such a source could be used to cite opinions of said expert? If it's the latter, then that is now covered by WP:SELFANDQUEST, thus that part of WP:SPS can be removed. Or if we really can treat an expert's self-published work as being reliable for more than just his/her opinions, then that needs to be made more clear. PSWG1920 (talk) 03:24, 14 November 2008 (UTC)

Thinking about this some more, it seems like quite enough to say that so-and-so is an expert in this field (cite independent source), and here is what they have said in a self-published work. WP:SELFQUEST currently would cover that, just as it would cover a non-expert who is nonetheless notable in the field, the difference being that you note in-text the credentials or lack thereof. PSWG1920 (talk) 05:30, 14 November 2008 (UTC)

I find the change unacceptable. The older version was perfectly sensible. "Not contentious" cop-outs don't work either. We only look to policy when things are contentious anyway. --Relata refero (disp.) 14:28, 16 November 2008 (UTC)

WP:SELFPUB #7

The qualification reads: "the source in question has been mentioned specifically in relation to the article's subject by an independent, reliable source." This is a bit confusing/redundant when the sps is used in an article about the sps. Obviously independent, reliable sources have mentioned the sps in relation to the article's subject or there wouldn't be an article in the first place. It may make things clearer to differentiate between sps in articles specifically about themselves, and sps in an article about another subject. Ty 04:16, 17 November 2008 (UTC)

First, I don't really see the need at this point for the clause "including in articles about themselves" in the first sentence, and in fact I think that wording is potentially misleading. I have realized that even in an article about a Fringe theory, you can't simply choose whatever proponent of it you see fit to reference for their opinions, but until recently this was not made clear by any policy. What is now qualification #7 fixes that and makes clear that even if a questionable source is notable in itself, a connection to the article's subject (whatever that happens to be) must be recognized by an independent source. PSWG1920 (talk) 04:53, 17 November 2008 (UTC)

Updating for primary sources

11-Nov-2008: Again - For years, this policy page has omitted the use of primary sources, which are well-explained under WP:NOR. To avoid the typical instant condemnation of primary sources, I propose adding short text under "Reliable Sources" for clarity. A good balance about primary sources would state more than one short sentence, but give details beyond the WP:NOR wording. I suggest the following text be added instead:

(under "Reliable Sources")

"Primary sources can be used to support statements of simple description, such as definition of terms or measurements, but avoid subjective, broad claims, such as about actual effectiveness, marketplace rank or social impact. Any interpretation of primary source material requires a reliable secondary source for that interpretation.

For example, hurricane wind speeds and storm surge levels could be quoted directly from the U.S. National Hurricane Center (NHC) as a primary source using airplane or buoy readings; however, stating, "The storm became more dangerous" must be traced to that source or another reliable source, rather than stated as a conclusion of the Wikipedia writer. Opinions must be traced to reliable ~~independent~~ objective sources.

Articles should also use reliable, third-party published sources with a reputation for fact-checking and accuracy, for statements beyond mere definition about the subject."

(Text not added yet to WP:VERIFY 11Nov08)

In the balance, the combined goal is to state the general use of primary sources, as well as warn about some details (such as avoid marketplace rank: "our product is the industry leader") and include long-term examples, such as hurricane data using NHC as a primary source rather than wait until network news reports. It has been a confusing issue, to state "third-party" and also get people to understand use of primary sources; otherwise, there would not be so many debates about rejecting primary sources. Also, I agree, there is no need to also copy the wording from WP:NOR, since the policies will be used together, regardless of the wording of either. -Wikid77 (talk) 07:03, 11 November 2008 (UTC)

In the proposed wording, I struck "independent" and put "objective" as the more general term, allowing NHC comment about their hurricane data when acting objectively. -Wikid77 (talk) 07:16, 11 November 2008 (UTC)
- While I like that, I think independence and objectivity often go hand in hand. NHC is probably a good objective source for opinions regarding hurricanes (such as predictions), since they are independent of the hurricane. On the other hand, if the accounting practices of NHC were criticized, NHC would probably not be an objective source on that matter, since it is not independent. Of course, there are plenty of independent sources which are not objective, such as editorials, but I'm having a hard time thinking of someone or something that could comment objectively on something it's not independent from. Seraphimblade ^{Talk to me} 07:29, 11 November 2008 (UTC)

I agree that we should reitterate that primary sources are allowed, but I think it can be done in even simpler language. I suggest something short and sweet... along the lines of: "Primary sources may be referred to in Wikipedia, however there are limitations as to how they are used. Primary sources must be used with great caution so as to avoid Original Research." Blueboar (talk) 14:22, 11 November 2008 (UTC)

Wikid77, our choice of independent is deliberate: that's the legal standard that Wikipedia has to meet in defamation cases. So we must stick to independent (although we can add objective, particularly if you're willing to volunteer to mediate all the POV disputes that will then cite the perceived "non-objectiveness" of a reliable source that happens to disagree with them).

I think that the proposal is much too long for what we need at this particular policy. There is no value in duplicating NOR and always harm (cost of time for keeping things synchronized). Why not say something simple like, "While primary sources may be used in limited ways [link to relevant section of NOR], articles should rely upon reliable, third-party published sources with a reputation for fact-checking and accuracy." WhatamIdoing (talk) 19:37, 11 November 2008 (UTC)

I would not use the words "Primary and Secondary" in one part of the sentence and then "third-party" in another... as it confuses the issue and makes it seem as if they are related. They are not... Both Primary sources and Secondary sources can be third-party.21:11, 11 November 2008 (UTC)

I've always thought that this policy page was the most "natural" and intuitive location for any discussion of the dos and donts of primary sources. I don't think we ought to have the same material in more than one place, but maybe we ought to consider moving what is currently at NOR here, or better yet, moving it here and then replacing it with something less baroque and theoretical, like general "rules of thumb" followed by concrete examples along the lines discussed above. COGDEN 00:04, 12 November 2008 (UTC)

Can you give an example of a primary source that is a third-party document? Classic examples of primary sources include autobiographies, diaries, e-mail, letters -- none of these are third-party documents.

The 'and secondary' shouldn't have been in that (half of the) statement anyway. Would you be happier with this? "Primary sources may be used in limited ways, but articles should rely upon secondary sources published by reliable, third-party sources with a reputation for fact-checking and accuracy."

I agree that putting these in the same section introduces the possibility for confusion; that's one of the reasons that I oppose having WP:PSTS merged into this policy. WhatamIdoing (talk) 17:45, 12 November 2008 (UTC)

An autobiography, diary, email etc that is written by somone other than the subject of the article would be a primary source that is third party. Say... a published diary written by Mickey Mouse cited in the article on Donald Duck. In that article, Donald would be the first party... Mickey is a third party. However, Mickey's diary is still a Primary source. Blueboar (talk) 19:38, 12 November 2008 (UTC)

Reasonable enough, although that example still fails the "reputation for fact-checking and accuracy" aspect. We have to meet all of the conditions to defend against libel claims, not just some of them. WhatamIdoing (talk) 03:13, 13 November 2008 (UTC)

Granted, but that is a different issue... to give a better example: The Anglo-Saxon Chronicle is a primary source in English history, but it would also be considered a "third-party" source in the aritcle on Alfred the Great (since it was not written by Alfred). What I am getting at is that "primary source" and "first-party" are not synonyms ... there may be overlap, but the terms relate to different concepts. The same goes for "secondary source" and "third party" Blueboar (talk) 14:00, 13 November 2008 (UTC)

A WP:Guideline or an WP:essay spelling all of this out in readable detail would be useful. A wikilinked-mention of such an essay would be useful in tyhis WP:Policy page. I am not qualified to contribute this. -- Boracay Bill (talk) 00:24, 14 November 2008 (UTC)

Blueboar, the sources that an article rely upon must meet all of the criteria, not just some of them. For legal reasons, we need to rely on third-party sources. For scholarly reasons, we prefer secondary sources. This is a "both/and" situation, not an "either/or".

As for your example: The Anglo-Saxon Chronicles are not primary sources. Annals aren't diaries; they're much more like encyclopedias. You take information collected elsewhere (read Anglo-Saxon Chronicle#Sources.2C_reliability_and_dating) and summarize it in your "yearbook". I don't know enough about these particular documents to classify them as secondary or tertiary sources, but they really do not seem to be primary sources.

Do you have any objection to the sentence I proposed above? WhatamIdoing (talk) 02:05, 14 November 2008 (UTC)

Um... The Anglo-Saxon Chronicles is absolutely considered a primary source by historians. But to get back to the question at hand... I agree with your revised wording. Blueboar (talk) 03:27, 14 November 2008 (UTC)

we need to rely on third party sources for legal reasons only if matters like defamation or other similar controversial matters are an issue. we do not in the least need to rely on them for uncontroversial material about anything at all. Just what legal scenario do you have in mind? Anyway, how can a first party source from an individual defame him? It could defame someone else, but so could a source of any other origin or nature. (Note that this is a different issue from the sources being third party sources to demonstrate notability--there the reason is to avoid self-advertising) DGG (talk) 04:37, 14 November 2008 (UTC)

Blueboar, Wikipedia does not use the same definition of primary source that historians do. I'm not saying that our definition is the One True™ Definition -- just that we have one, and that according to Wikipedia's definition, any source that reports information taken from other sources isn't considered a primary source for Wikipedia. ASC took information from other publications (including other encyclopedias): it's therefore not a primary source.

DGG, we also rely on third-party sources to give us an idea of whether something is important -- not just whether the article should exist at all, but whether any given fact is generally considered important enough to get (that much) space in the article. WhatamIdoing (talk) 18:26, 14 November 2008 (UTC)

Very old documents are Primary sources

Sure it does... from WP:NOR: "Primary sources are sources very close to the origin of a particular topic or event. An eyewitness account of a traffic accident written or narrated by the eyewitness is an example of a primary source. Other examples include archeological artifacts; photographs; audio and video recordings; historical documents such as diaries, census results, ...." The Anglo-Saxon Chronicle would qualify under "historical documents such as diaries". Again, while there is overlap "first person" and "primary" are not exacltly the same thing. Blueboar (talk) 22:53, 14 November 2008 (UTC)

So when the original compiler of ACS copies something out of Bede's Ecclesiastical History, you think that makes ACS the primary source, simply because ACS is a lot closer to the original events than we are? Would you, a thousand years from now, consider today's Wikipedia a primary source?

Personally, I don't think that summarizing some other book makes the summary a primary source for what happened. WhatamIdoing (talk) 00:39, 16 November 2008 (UTC)

Well, first off, you can not really talk about the original compiler of the Anglo-Saxon Chronicles... the work was written over several hundred years, and had several chroniclers. Secondly, a large portion of the ASC pre-dates Bede, so I doubt there is much that is copied (and in fact, the ASC frequently disagrees with Bede, or vise versa... different POVs). As for Wikipedia... if, in a thousand years time, Wikipedia ended up being one of the few sources that existed, then yes... it would probably be considered a Primary Source by historians of that future time. In any case... we are getting side tracked. Blueboar (talk) 01:05, 16 November 2008 (UTC)

I don't think that any source that summarizes information from another source should be considered a primary source for the purposes of WP:PSTS. "Very close to the origin (compared to me)" is not the same as "very close to the origin (end of criteria)". If you tell me a story that someone else told you, and then I tell someone else, then you are not a primary source, even though you're closer to the origin of the story than I am. WhatamIdoing (talk) 02:04, 16 November 2008 (UTC)

I agree... except for the fact that the compilers of the ASC mixed information taken from others and information they knew first hand... thus it is primary. To move the discussion away from the ASC, I think you are taking an overly narrow view of Primary... Age does matter. For example, The Dead Sea Scrolls are considered a primary source, even though parts of them contain material that was copied from older texts. There comes a point when the age of the document moves it from being a mere copy to being a historical or even archeological artifact. Blueboar (talk) 14:26, 16 November 2008 (UTC)

Book Covers

Are book covers, reliable sources? If so, what section of policy does it fall under? If not, should they? (ie: a paperback book cover says that there are over one million copies in print.)--Jojhutton (talk) 01:34, 16 November 2008 (UTC)

I would be wary of statements taken from a book cover... they are essentially advertizing by the publisher. Thus, I would hesitate to call them reliable for a statement of fact (as in: "The book has sold more than one millian copies"). They are reliable, however, for a statement as to what is stated on the book cover (as in: "According to the cover of the paperback edition, the book has sold more than one million copies"). Blueboar (talk) 14:48, 16 November 2008 (UTC)

Merge sections?

It seems to me that a merge of WP:SPS, WP:QS, and WP:SELFQUEST would be logical and could make this policy simpler without changing the effective meaning. Especially since questionable sources are subject to the same restrictions as self-published sources per SELFQUEST. This could also solve the problem of WP:SELFPUB directing to the wrong section (which apparently can't simply be fixed since that would mess up links in discussions elsewhere.) PSWG1920 (talk) 17:48, 19 November 2008 (UTC)

That is how it used to be... and was changed only a few months ago. The objection to the old wording was that it implied that self-published sources were the same thing as questionable sources... toss in extremist sources and there was real confusion. While all are dealt with in similar ways, they are not necessarily the same thing. A questionable source is not always self-published. Blueboar (talk) 18:11, 19 November 2008 (UTC)

Yes, but that could be made clear even in one section. Bold self-published and questionable. PSWG1920 (talk) 18:26, 19 November 2008 (UTC)

OK... How is this for a solution... I have created a new level 3 header (and moved some of the shortcuts to it), while shifting the existing sections to level 4 subsections. This maintains the idea that the topics are distinct, but are dealt with in the same way. Revert if you don't approve. Blueboar (talk) 18:34, 19 November 2008 (UTC)

I do approve, but I think more could be done. I may edit it significantly and then self-revert and ask what people think. PSWG1920 (talk) 18:45, 19 November 2008 (UTC)

That works for me... then we can compare your ideas and the current text side by side. Blueboar (talk) 18:59, 19 November 2008 (UTC)

FYI... Having compared your ideas with the current text... I have no objections to your edit. But wait to see what others think before you go make the change. Blueboar (talk) 19:16, 19 November 2008 (UTC)

Reliable

As far as local news anchors go, are their biographies on the official station websites reliable? —Mythdon (talk • contribs) 23:05, 19 November 2008 (UTC)

Lets put it this way... barring any better source, yes. The next question, however, is this: Is a local news anchor really notable enough for mention in Wikipedia? If the only source you can find on one is the website of a local TV station, I am not sure that he or she really is notable. They would have to have won some sort of national award or something. Blueboar (talk) 02:08, 20 November 2008 (UTC)

These sorts of questions are best asked at the reliable sources noticeboard, which is dedicated to answering them. As always, no source is universally reliable: it depends entirely on how you use it. WhatamIdoing (talk) 03:29, 20 November 2008 (UTC)

Geriatric nursing

am trying to write a page on geratric nursing could you review and comment—Preceding unsigned comment added by user:199.8.81.101 (talk • contribs) 13:39, 21 November 2008

Could you give us a link?. —Mythdon (talk • contribs) 06:16, 22 November 2008 (UTC)

That'd be Geriatric nursing. Has some major issues to be addressed.LeadSongDog (talk) 06:53, 22 November 2008 (UTC)

Reference library category

In order to help facilitate easier location of potential sources of offline information to help verify the notability of article subjects and contents, I have created Category:WikiProject reference libraries and placed into it all of the reference library pages of which I am aware. Please add more project reference libraries to this category if you know of more. Additionally, feel free to create new reference library pages for any particular project as well. They can be very useful. ···日本穣^{? · Talk to Nihonjoe} 20:05, 21 November 2008 (UTC)

Jeremy Tucker (Article) =

I edited the reference/source section because 3 of the 4 links do not work. They are invalid. So I deleted 2 of them and I would like that article to be challenged because I do not feel that it follows the guidelines of verifiable content. Thank you if someone can look it over.

-David Scalzo

DScalzo23 (talk) 01:08, 29 November 2008 (UTC)

Wikipedia itself and about.com as sources

I'd never consider Wikipedia a reliable source and had long believed that it explicitly ruled that it wasn't a reliable source for itself. And that about.com was no better. (Of course, both sites have good stuff, but the domain name alone doesn't start to assure you that you can rely on its quality.) Which is why I yesterday removed a whopping amount of Wikipedia/about.com sourcing from a certain article (edit summary neither answers.com nor wikipedia.org is an acceptable source). Seemingly unimpressed by either this or the short message I left on his/her talk page, the article's editor promptly reverted my edit. (Summary: Most all citations are checked against third party resouces. Material requires cites.)

Time to quote policy, I thought. However, I see nothing about this in WP:V (or the guideline WP:RS). It's hard to search for "Wikipedia" but a search for "about.com" in the "Wikipedia" and "Wikipedia talk" spaces brings zero hits.

The author of the article seems pretty scrupulous, quoting and translating from the Czech and German-language Wikipedias; I know nothing about the former but have read that the latter has higher standards than has the English-language one. So if anybody is going to cite Wikipedia (or about.com), then this is the way. But does WP policy now permit the citing of Wikipedia and such dodgy sites as about.com? If so, I'm amazed (and horrified); if not, I wonder where this is stated and why it isn't clearly stated in WP:V. -- Hoary (talk) 00:55, 2 December 2008 (UTC)

No, one Wikipeida article is not considered reliable for citation in another Wikipeia article... the excepting being in an article about wikipedia. This is explained in WP:SPS which reads: ... self-published books, newsletters, personal websites, open wikis, blogs, knols, forum postings, and similar sources are largely not acceptable to cite in Wikipedia. (my bolding, for emphysis). About.com is also not a reliable source, as it makes such heavy use of Wikipedia. Blueboar (talk) 01:27, 2 December 2008 (UTC)

No shinola. Within the very same section: Articles and posts on Wikipedia, or other websites that mirror Wikipedia content, may not be used as sources. This is what happens when, as today, I get up an hour earlier than usual. -- Hoary (talk) 01:41, 2 December 2008 (UTC)

"may object" v. "may wish"

I am hoping there is a way to rephrase "Any material lacking a reliable source may be removed, but editors may object if you remove material without giving them sufficient time to provide references." to give more guidance for editors. The statement gives clear authority to an editor removing unsourced material, but then throws in a vague phrase that others "may object" - without providing any guidance as to how an editor is supposed to object, what a proper forum for the objection may be, what happens to the material in question while a consensus is worked out about sourcing the material. I feel that a phrasing similar to "Any material lacking a reliable source may be removed; however, because Wikipedia is a collaborative project and someone thought the information was important to include, you may wish to allow time for editors to provide references." would help clear up a contentious point in this policy by underscoring Wikipedia principles. -- The Red Pen of Doom 19:28, 3 December 2008 (UTC)

I think people understand that editors who "object" usually do so on a talk page, either the article's or the user's. I don't see that this point is contentious as currently phrased. UnitedStatesian (talk) 19:33, 3 December 2008 (UTC)

The suggested phrasing places the subjectivity/options/promotion of consensus building with the editor wishing to remove the unsourced material rather than the current phrasing which gives the remover full authority and places the "objector" in an opposing stance rather than encouraging collaboration. -- The Red Pen of Doom 19:41, 3 December 2008 (UTC)

Actually, I do understand RedPen's confusion... it is the dual use of the word "may". When we say unsourced material "may be removed" in the first part of the sentence, we are clearly intending it as being permission to remove. But when we say "editors may object" the intent is less clear. Are we granting a counter permission to object? (in which case we probably should lay out what that means an how to do so) Or, are we cautioning potential removers that other editors might object?

I think the intent is more of a caution... to say: Go ahead and remove if you really think the unsourced material is egregious. However, since others might object to your summarily removing it, you should explain what you are doing and why (in either a comment on the talk page or in the edit summary, and give people a chance to respond. Blueboar (talk) 22:42, 3 December 2008 (UTC)

Per my above comment... are there any objections to changing the sentence to "Any material lacking a reliable source may be removed, but editors might object if you remove material without giving them sufficient time to provide references"? I think this would clear up the confusion. Blueboar (talk) 17:07, 4 December 2008 (UTC)

It's OK with me, but I think a purely descriptive phrasing is even better: "editors sometimes object if you remove material without giving them sufficient time to provide references". This is all splitting hairs, though. — Carl (CBM · talk) 17:34, 4 December 2008 (UTC)

I like CBM's descriptive wording. UnitedStatesian (talk) 17:57, 4 December 2008 (UTC)

I support CBM's version as well. -- The Red Pen of Doom 19:17, 4 December 2008 (UTC)

This passage has long read (without significant complaint in at least the last year or so): "Any material lacking a reliable source may be removed, but editors may object if you remove material without giving them sufficient time to provide references. " [emphasis mine] AFAIK, the grammatical mood was not only indicative but also imperative. That is, other users may (have permission to) object if they believe the removal of material is unreasonable or unduly hasty. As TheRedPenofDoom observed in her/his edit to the "Burden of proof" yesterday, WP is an interactive process, and this sentence expresses one manifestation of this process. The sentence needn't say explicitly what WP's general type of interactive consensus process is, because this is dealt with many other places in WP. But of course it should reflect community understanding of how that process is to be implemented in the context of Verifiability#Burden_of_proof. If the sentence is intended to be merely descriptive (indicative grammatical mood), there may not be any need for the clause at all. No need, on a purely descriptive basis, to mention other editors objecting. The sentence can just as easily read: "Any material lacking a reliable source may be removed," ... Kenosis (talk) 18:29, 4 December 2008 (UTC)

"inclusionists" might not be in favor of such a blank check handed to "deletionists".... -- The Red Pen of Doom 19:39, 4 December 2008 (UTC)

The main point I was making is that the word "may" gives permission to object, as well as implying that users often do object. There are valid policy grounds for objecting on the basis of factors other than merely the amount of time permitted to provide citations prior to removal of the material. The rules, long established, are, essentially, "if it ain't cited, it can be removed, period." Such a removal, if there's any further disagreement, engages broader discussion process to (hopefully) arrive at a consensus. A user may, for instance, argue back at the person doing the removing that "no way, it's common everyday knowledge", or "it's already cited farther down in the article and WP:LEAD allows us to summarize it in the lead without citing it". Etc. etc. Thus the word "may" provides a more explicit policy basis for objecting to an outright removal of unsourced material. All this could, of course, reasonably be written in many different ways. ... Kenosis (talk) 19:56, 4 December 2008 (UTC)

FYI: WP:LEAD the last time I was there no longer contained the exception that if material is sourced within the body of the article it did not need a citation in the lede. -- The Red Pen of Doom 17:42, 11 December 2008 (UTC)

Sure, it's a frequently shifting landscape, though most of the very basics are fairly stable at this stage of the wiki's growth. Presently, WP:LEAD reads, in part: "editors should balance the desire to avoid redundant citations in the lead with the desire to aid readers in locating sources for challengeable material." As you've already observed, it's an interactive process, sometimes seemingly endlessly. ;-) ... Kenosis (talk) 18:22, 11 December 2008 (UTC)

I;'d like to find a different word, because "may" has proven ambiguous. I certainly interpret it as a warning: if you do not source,someone might object. But I don't think we can say that "deletionists" are stricter in general about sourcing than "inclusionists" -- they are two different things. V is about content, not inclusion, except to the extent that if there is no verifiable content at all, an article can not be written. DGG (talk) 04:32, 11 December 2008 (UTC)

Understood-- totally reasonable, DGG. As you may have noticed, there's also a balance or tension on these policy pages between casual language on the one hand, and more explicit, legalistic, strictly analytic language on the other. Whatever balance is chosen, at present i imagine we agree there's permission to remove unsourced material, without prior warning and with only a statement that it's unsourced, and also that there's permission to replace the material or object to the removal, stating the policy grounds for the objection. This, of course, is expected to engage a discussion process seeking to arrive at a consensus about how best to apply the relevant WP content policies. As we've seen, this is where "inclusionists" and "deletionists" commonly wrangle with one another. ... Kenosis (talk) 17:31, 11 December 2008 (UTC)

Woulda, coulda, shoulda

WP:VER requires claims to be verifiable. When someone says "well I know it is true..." If it seems doubtful, or, especially if there is a source that says the contrary, we tell them: "please provide a source," and refer then to WP:VER, which is pretty clear:

This page in a nutshell: "Material challenged or likely to be challenged, and all quotations, must be attributed to a reliable, published source."

Note the use of "must" in the policy summary. However, in the section on "Burden of evidence," it had said that material challenged, "should be attributed..." I changed it to "must be attributed" to be consistent. Generally the difference between policy and guidelines is that policy tells you what to do, guidelines say what you should do. Sunray (talk) 08:12, 8 December 2008 (UTC)

Conflicting sources

If there are several websites, all of equitable reputation, which have conflicting information, how should this be handled? For example: A video game is going to be released in the near future. The developer's website has a release date of "Early 2009", while many reputable gaming website state "March 2009", "Q1 2009" (fiscal quarter, which begins in March), or "1 March 2009". Should all viewpoints be expressed in the article or should one source be considered superior to all others? Thanks! (EhJJ)^TALK 21:48, 10 December 2008 (UTC)

Um... Isn't "March 2009" and "Early 2009", essentially saying the same thing? I don't see that these sources actually are conflicting. But to answer your question... if you have several sources of equal reliability that state conflicting things, you should mention them all, as per WP:NPOV. Blueboar (talk) 15:31, 14 December 2008 (UTC)

patents and patent applications

Regardless of what their standing may be legally, issued patents are acceptable sources. In the sciences, they zare cited just as published papers are cited, and abstracted and indexed in all the major disciplinary indexes,such as Chemical abstracts. (In fact, CA lists preliminary patents also, as the first & sometimes only published form, and chemists normally cite these also.) I am therefor reverting the addition of patents from the list of questionable sources; to ensure the widest consensus, I have not howeveryet added them to the list of reliable sources. DGG (talk) 04:28, 11 December 2008 (UTC)

Alas, not all patents are created equal.. take for instance, this PCT Patent Application, which is 100% Spam.. or any one of the patents found here.. Indeed, patents can be easily misused to explain wild fringe theories. Patents can really only be considered as reliable sources that patents of that nature exist.. not of whether or not the content/subject of the patent is legitimate. --Versa geek 05:17, 11 December 2008 (UTC)

This is all I'm going to say.(At the moment) Someguy1221 (talk) 08:59, 11 December 2008 (UTC)

As usual, I agree with DGG. The fact that we must concoct some standards for reliable sources on wikipedia does not change the fact that in the real world there are no such things. Everybody makes mistakes, and no source, no journal is without them. One can equally impeach any source by finding errors which could be misused by diabolical fringepushing editors. Not all university press books, not all articles in the best journals, etc are created equal. Most patents aren't science fiction, there's some checking, so they should not be called questionable sources. Not always the best, not able to defeat better sources if they're one of these crazy patents, but not unacceptable in general. We should try to emulate standard academic / real world practice as DGG described, not throw the baby out with the bath-water.John Z (talk) 09:43, 11 December 2008 (UTC)

The difference between patents and journals is that these few examples are nothing like the bad journal articles you might find. A bad article in a good journal is something that slipped through the cracks. A crazy patent is something that fell through a giant hole, since being based on good science is no requirement for getting one. And isn't that the very essence of a self-published source: there is no form of revew for it. A patent is certainly reviewed, but not for compliance with mainstream scientific thought. Like any blog, they should held in question, by default, without some good reason to do otherwise. I don't think the fact that some scientists cite them is any real reason to allow us to broadly cite them. News sources often quote blogs, but that does not evict the entire blogosphere from the realm of questionable sources. Someguy1221 (talk) 09:52, 11 December 2008 (UTC)

The standard of review for journal articles is frustratingly high: patents less so. The U.S. government does not test the claims made by everyone who files a patent. As such, patents should be considered self-published in terms of WP:RS. RJC ^Talk_Contribs 14:15, 11 December 2008 (UTC)

While a patent doesn't get the same level of scrutiny as a major journal, that's not the point. There are many newspapers and books less vetted than patents that we accept. RS isn't an absolute standard but a relative one and should be used that way. A patent verifies the idea existed at a certain date and that the applicant was aware of the idea, even if has not been realized as they call it. It does not verify the idea (or the theory behind it) works. We should use them with caution in the understanding of that character.LeadSongDog (talk) 14:42, 11 December 2008 (UTC)

John Z (09:43, 11 December 2008) and LeadSongDog (14:42, 11 December 2008) have got it right - there is no a such thing as a 100% reliable source, not even in peer-reviewed science journals (I won't bore you with details, but can easily do so). Patents and patent applications have weaknesses, but as a coarse filter they require applicants to put their money where their mouth is. So I'd accept one provisionally, unless / until a better source refutes it. --Philcha (talk) 15:02, 11 December 2008 (UTC)

DGG, a published patent that asserts that Silver cures AIDS. (Do an internet search on "tetrasilver tetroxide".) Shall we re-write AIDS on the basis of this supposedly non-questionable and non-self-published source?

Can you, in fact, think of any fact in a patent that you cannot find published elsewhere and that you would accept as being (1) a proper secondary source for the supported scientific/engineering fact and (2) something your professional conscience would allow you to describe as a responsible choice for the supported scientific/engineering fact? WhatamIdoing (talk) 06:19, 12 December 2008 (UTC)

100% reliability isn't the issue: basic reliability is. The applications aren't vetted at all for what we're looking for, and the argument that they filter some claims out by forcing the applicant to put their money where their mouth is does not get around the problem that they are still self-published: vanity presses do the same thing, but don't count as a third-party appraisal of the work. RJC ^Talk_Contribs 18:16, 12 December 2008 (UTC)

Philcha, you seem to be focusing on the claims, whereas I usually see people citing the specification instead. The specification does not get changed during the patent review process. (Note that the small entity filing fee in the US is $165, so it's cheaper to file a patent than it is to print books on the vanity press.) WhatamIdoing (talk) 20:49, 12 December 2008 (UTC)

Implying, I take it, that patents are no more and no less reliable than vanity-press self-published books, and that patents should be required to meet WP:V#Self-published sources criteria. -- Boracay Bill (talk) 01:27, 14 December 2008 (UTC)

First are we talking a patent application or a patent that has been granted? They are not the same thing. I would agree that a patent application is similar to a self published source... a patent that has been approved is an official government document, and thus not self published. Blueboar (talk) 15:37, 14 December 2008 (UTC)

Blueboar, if by "published" you mean "willing to take any sort of responsibility for the contents", then you are mistaken. The fact that the patent on curing AIDS with tetrasilver tetraoxide is available through the USPTO does not mean that the United States government makes any sort of claim about the accuracy of the contents, from whether or not the inventor actually fed silver to people in Honduras, whether or not those people actually had AIDS (something the inventor did not bother to test for), to whether or not the claimed invention actually worked. They have no choice but to assume that the inventor has met his statutory requirement to tell the truth about his investigations. WhatamIdoing (talk) 22:49, 14 December 2008 (UTC)

That was not really what I meant by "published"... what I mean is that there is an official notice of the existance of the patent, placed in the government archives, that we can look up. In other words, we can verify that the patent existed and we can verify the fact that the patent makes certain claims. What a patent does not verify is whether the claims made are valid or not... a patent does not tell you that the thing being patented actually worked. This is why I would treat it in a similar manner to a self-published source... they are of limited reliability, but not outright unreliable. Blueboar (talk) 04:10, 17 December 2008 (UTC)

Yes, and... it's as reliable as its author - no more and no less - a self-published primary source. If the author is a notable published expert, then the patent could be treated as reliable in similar fashion as a blog post of a noted expert. But no added reliability of the content is conveyed by the acceptance of the patent application, other than to verify the fact that the patent was filed and granted, and to verify that the text of the patent is not a fake. The patent is published by its author; it just happens to be mirrored by the government's hosting of the document in their system. --Jack-A-Roe (talk) 04:26, 17 December 2008 (UTC)

When something can't be cited

I'm currently having an issue in I Will Follow You into the Dark and its GAN. This policy says the threshold is verifiability, not truth. A section on the article isn't cited and pretty much can't be cited because of lack of third-party reliable sources. In this case, what should be done? DiverseMentality 19:33, 16 December 2008 (UTC)

To clarify for DM, the section in question is "Release", which I feel completely consists of non-controversial information that I don't feel policy requires me to cite and is tricky to cite if I had to (it takes five semi-unreliable source citations to do so).--Remurmur (talk) 02:34, 17 December 2008 (UTC)

Actually, the material in question can be verified... by looking at the item being discribed (ie the primary source). The section simply contains a discription of each release (that there were two releases in 7" colored vinyl, and a CD single, with differing songs on the on the "B side" of each). This is all information that can be verified by anyone just by looking at the items themselves. If someone is silly enough to challenge the information and demands that you cite a source for it, just cite the item as if it were a primary source (I assume you have already looked to see if you could verify the information on a distributor's website, such as Amazon.uk?)Blueboar (talk) 03:51, 17 December 2008 (UTC)

What if ?

If verifiability is the criterion, and not "truth", what happens when you can find RS that say contrary things? This happens all the time in contentious areas. Do you count the references on each "side" or attempt to determine (ie by a consensus of sources) factual accuracy (ie verifiability ?) in this way, or does one attempt to determine which RS are the most/least reliable? Or are all RS sources counted equal to one another? Thanks Tundrabuggy (talk) 02:04, 17 December 2008 (UTC)

When the sources are contradictory, pretend you are a journalist and present both sides. When the subject is less controversial, and you have some knowledge in the field, you can try to weigh the sources to give each one due weight in the article. The more controversial a subject is, the more difficult that can be. — Carl (CBM · talk) 03:57, 17 December 2008 (UTC)

Read WP:Neutral point of view, which will give you more details on dealing with this. Blueboar (talk) 04:00, 17 December 2008 (UTC)

How to handle an uncited statement. Delete or citation-needed template?

Are there any guidelines on what to do if a statement has not been supplied with a verifiable source. What is the preferred approach: query the claim with a template like {{cite}} or delete the text? WP:VER mentions using a template but doesn't seem to recommend either way.

In my particular case, I started an article and before getting it off the ground have repeatedly had chunks of text deleted. It's been quite a frustrating experience. Personally I think it would have been more constructive to add a citation-needed template rather than delete. This would have given editors the opportunity to find sources. Advice and opinions please. pgr94 (talk) 09:45, 18 December 2008 (UTC)

I would place a "cite needed" tag on it (unless it is obvious "vandalistic" nonsense). Deletion is justified if there already is a "cite needed" tag on it and the tag has been there for a long time without anyone responding to it. Depending on the nature of the article it could take up to a few months for someone to come up with a good cite - so give it time. Roger (talk) 10:33, 18 December 2008 (UTC)

It really depends on how "damaging" the unsourced statement is, and what type of article you are dealing with. If the article is a Biography of a Living Person (BLP), then we take a hard line on unsourced information. WP:BLP tells us to delete imediately. In other types of articles, it is more of a judgement call. If you think it likely that the information might be accurate, but is simply unsourced, then the first step is to see if you can find a source. If you can't, just leave a {{fact}} tag and hope that sooner or later someone else will find a source. If you think the information is questionable, tag it and explain what is wrong on the talk page. Then wait a while, and if no one provides a citation after a reasonable time, you may remove the uncited information. If the information seems to be really off the wall, and you are positive that it is incorrect, you are allowed to remove it as a first step (but if someone returns it, don't edit war over it... slap a tag on it, and discuss the issue on the talk page). In other words, use some common sense and courtesy. Blueboar (talk) 14:34, 18 December 2008 (UTC)

Thanks. I completely agree with both of you Dodger69 & Blueboar. pgr94 (talk) 16:54, 18 December 2008 (UTC)

I feel your pain, Pgr94, but there's a simple solution. Put the citation in first and the statement after. Then you never have unref'd statements to delete.LeadSongDog (talk) 16:16, 18 December 2008 (UTC)

Ok. I see your point. Nonetheless, if we were to delete everything that is not cited Wikipedia would be much smaller and it might cause quite a stir. Shouldn't WP:V provide a clear position on how unreferenced material be handled? A position along the lines of that described Dodger/Blueboar above seem a good starting point. pgr94 (talk) 16:54, 18 December 2008 (UTC)

The point is that WP:V can't (and shouldn't) provide a clear position... whether to delete or tag is (and needs to be) determined on an article by article, and even statement by statement basis. Sometimes it is better to delete, sometimes it is better to tag. The important thing in your case (being the author of the material being deleted) is that you need to find a source for the information you want to add. Blueboar (talk) 18:37, 18 December 2008 (UTC)

Agreeing with (most of?) the above. Today I replaced a CitationNeeded which had been sitting there doing no harm since September 2007. So what? The fact looked reasonable, so it had remained. Perhaps the press release I added as a ref isn't good enough; in that case someone will be along to replace or supplement it some day. The project is a work-in-progress, and unless an article is totally peppered with such tags, I think they generally are a good thing. - Hordaland (talk) 19:02, 18 December 2008 (UTC)

I see that I oversimplified my response above. Proactive citation-before-the-fact will help prevent your edits suffering deletion. Other's edits you can only address after-the-fact. As others have said, there are judgement calls to make. If harmful, delete. If probable, tag {{cn}} or (better) find and cite a source. If improbable, tag {{dubious}} or {{cn}}. If the article is rife with omitted citations, tag {{refimprove}} at the top. The main thing is to try and ensure an innocent reader doesn't mistakenly trust an unsupported assertion.LeadSongDog (talk) 06:00, 19 December 2008 (UTC)

Thank you all for your suggestions. I am still left wondering why the above advice doesn't belong in the guidelines, but I'll defer to you on this one. pgr94 (talk) 09:37, 19 December 2008 (UTC)

We should have a section with precedents

We should have a section that deals with decisions from the community but most importantly the arbitration commity. --CyclePat (talk) 07:30, 20 December 2008 (UTC)

This new "#7" under WP:V#SELF

I had reason to look at WP:SPS today, and I was surprised that a major new restriction was made to the policy in the addition of this point #7: "the source in question has been mentioned specifically in relation to the article's subject by an independent, reliable source." Looking at the archives of this talk page, that addition seems to have been made by only two editors, with some discussion on another aspect of the change by a few others. It seems to me the process was about like this:

It was decided to finally change WP:SPS from "SPS may only be used in articles about the author" to "SPS may only be used to support statements about themselves/their authors". IMO, this is a good change and reflects current practice where this is useful.
Someone grew concerned that J. Random Crackpot could self-publish his views on a subject, and then insist on including those views in the article on the topic because his SPS is now an allowable source for the statement that he holds that view.
Instead of realizing that this sort of thing is covered by WP:NPOV, WP:WEIGHT, WP:FRINGE, and so on, the reaction was to add this additional criterion. Oddly, the criterion is applied to the source being specifically connected to the subject rather than the author, which makes it far more restrictive.

I'm not sure how many SPS this additional criterion actually leaves usable; I suspect it eliminates the majority of them completely. I propose that the statement in question be removed, and replaced with a note such as the following: "Just because a source exists for a fact does not mean the fact must or even should be included in an article; WP:NPOV, WP:WEIGHT, WP:FRINGE, and other content policies and guidelines still apply. For example, the opinion of an individual on a particular topic is unlikely to be worth mentioning in the article on that topic, unless reliable, third-party sources specifically connect the individual with the topic." This should have the same effect of closing J. Random Crackpot's loophole without being overly restrictive on legitimate uses, and it keeps content inclusion decisions where they belong. (For a parallel example, we could source the birth weight and length of many people by looking for the birth announcement in their hometown newspaper. We don't usually bother to include that information because such minutia is normally unimportant to an encyclopedic biography. But we also don't add a policy that birth announcements are somehow "not acceptable" as sources to prevent someone from doing that, we just point out that such trivia is not worth including.) Anomie ⚔ 06:03, 19 December 2008 (UTC)

I too saw the problematic wording, was puzzled by this addition and spent some time trying to understand it. As for #3 above, "source" must be read as "author" (as it often enough is) because otherwise it would create a new and bizarre restriction on SPS's in articles about themselves, clearly not the intent. Your proposed new wording is reasonable, and not terribly different from the present #7 with the necessary "author" for "source." John Z (talk) 09:07, 20 December 2008 (UTC)

I think point 7 is overly restrictive. For example, if an author or artist has an official website including biographical details, such as where they were born, where they grew up, went to school, etc., I think we should be allowed to use that as a source in their BLP, subject to conditions 1 to 6 and the WP:BLP section on using self-published sources, even if the bio sketch on his site has not been mentioned in the media (why would anyone mention it?). Jayen 466 12:47, 20 December 2008 (UTC)

As per above, the intent at the time section 7 was added was to expand usage of SPS's, but the natural interpretation of the current wording has this unfortunate and unintended consequence, ruling out sources like the bio sketch that have been considered acceptable for a long time. However, it seems to have been (rightly) been ignored so far, and I'd be happy enough with Anomie's clearer proposed slight further liberalization.John Z (talk) 04:30, 21 December 2008 (UTC)

the proposed wording remains unacceptable. It would, for example, prevent the use of people's posted CVs, which are the source of a great many of the biographical articles for such things as birthdates and education and positions. Birth announcements in local newspapers are only marginally more reliable thqn if posted on the parent's blog, and very much harder to find. If you want to broaden it, do it as an alternative. Self published sources may be used even when not about the subject, if they are mentioned in RSs (and meet the other criteria). DGG (talk) 20:12, 21 December 2008 (UTC)

Agree with DGG and John Z, the wording is too restrictive and would eliminate quite a bit of sourcing on WP:BLP's. I've removed it until a consensus version can be found. Dreadstar † 21:46, 21 December 2008 (UTC)

I agree that we want to limit crackpots... but this is not the way to do it. We should remove the recently added restriction, or at least seriously re-word it. Blueboar (talk) 21:59, 21 December 2008 (UTC)

Podcasts etc.

I think the change undone here made sense. Just an extension of SPS principles to other media. Any views? Jayen 466 12:50, 20 December 2008 (UTC)

As the first person to remove this change (here), I am actually coming around to supporting it. This change in opinion is based upon discussions at WT:RS, and Jossi's edit summary reverting my initial removal. What is important is the intent... In adding podcasts, Jossi does not seem to be saying that we should apply the restrictions of SPS to all podcasts... those would be applied only to those podcasts that are self-published. This makes sense. A self-published podcast is no different than a self-published website, self-published book, or any other SPS. So (for example) a podcast created by Joe Blow would be treated as an SPS (with all the limitations thereof), while a podcast created by a major news outlet would be treated as a fully reliable source.

In other words... We can use podcasts to verify information, but self-published podcasts are limited. Blueboar (talk) 14:44, 20 December 2008 (UTC)

Yes, these issues are all covered by existing policies, and it is not a change but a clarification. ≈ jossi ≈ (talk) 17:00, 20 December 2008 (UTC)

I have asked this at WT:RS... but I will ask it here as well... When someone is interviewed on a podcast, who is the "publisher"? Blueboar (talk) 17:24, 20 December 2008 (UTC)

It might make sense to consider it "self"-published with respect to both the interviewer and the interviewee. A similar situation obtains if I have my own private website and a friend writes a text and allows me to put this (otherwise unpublished) text up on my website. Neither I nor my friend can claim that the hosted text is anything other than self-published. Jayen 466 17:42, 20 December 2008 (UTC)

The existing text is already extensively qualified: it says that because self-published stuff is not acceptable, a bunch of things (including blogs and patents) are largely not acceptable, and then there's a big footnote discussing different cases of "blogs". There isn't even any implication that the things listed are the only things that are likely to be unacceptable due to self-publication, and the footnote is likewise not absolute. Making the list of examples structurally more complicated in order to redundantly qualify it is not only unnecessary, but IMO would actually be confusing. Consequently, I am (so far)

opposed to this change.

I could see adding some things to the list, changing "largely" to "usually", and perhaps even changing the ordering of the entire sentence so that the list goes after the statement that these things are only usually unacceptable rather than always unacceptable. Pi zero (talk) 18:46, 20 December 2008 (UTC)

How about this alternative phrasing of the sentence, which simply reorders it, changes "largely" to "usually", and adds podcasts and vcasts?

For this reason, it is usually not acceptable in Wikipedia to cite self-published books, newsletters, personal websites, open wikis, blogs, knols, podcasts, vcasts, patents, patent applications, forum postings, and similar sources.^[4]

Pi zero (talk) 19:02, 20 December 2008 (UTC)

We should include something to the effect that, like self-published sources, audio and video interviews are reliable primary sources for the views of the interviewees. ·:· Will Beback ·:· 01:41, 21 December 2008 (UTC)

No way, unless the caveat established in WP:SPS is addressed. I really do not understand the need to dilute that portion of the policy which has well establishged. Interviews, are primary sources, sure. But the issue of prior publication on a reputable source stands. Who cares is if it an interview, an essay, a blog post, or a speech. No secondary sources referring to it, no deal: we don't use it without considerable caution and within the caveats expressed at WP:SPS ≈ jossi ≈ (talk) 04:10, 21 December 2008 (UTC)

Yes, that's exactly what I'm suggesting: treating audio or video interviews as reliable self-published primary sources, with all of the caveats appropriate for such a source. ·:· Will Beback ·:· 20:17, 21 December 2008 (UTC)

Self-published by WHOM? It seems that here we are saying that podcast interviews are published by the person being interviewed, while at the discussions currently going on at WT:RS we are saying they are published by the person who owns the website where the podcast is found. Can we have some consistency please? Blueboar (talk) 21:18, 21 December 2008 (UTC)

(unindent) Sounds reasonable to me. We clearly specify that all the kinds of source described in SPS can be used for information about their author. I also like one other aspect of jossi's phrasing that hasn't been mentioned: it clarifies that there are some blogs that aren't self-published, which I've noticed being an issue before: some editors just don't seem willing to accept that if a blog is published under editorial supervision (e.g. by a newspaper) it may be considered reliable. JulesH (talk) 09:57, 21 December 2008 (UTC)

Once the word "usually" is at the front of the sentence before the list, I'm comfortable leaving to the footnote the specific mention of supervised blogs. Pi zero (talk) 12:26, 21 December 2008 (UTC)

Lists

Are lists exempt from the verifiability policy?--John (talk) 05:32, 17 December 2008 (UTC)

No. They must be sourced as well. --MASEM 05:42, 17 December 2008 (UTC)

The list inclusion criteria are particularly important as relates to the verifiability of the listed items. Here are a couple pages that may be helpful: Wikipedia:Lists & Wikipedia:Stand-alone lists . --Jack-A-Roe (talk) 06:05, 17 December 2008 (UTC)

If you're talking about List of Irish people, one thing to consider is whether the articles for the people in the list state that the person is Irish and have a source. --Pixelface (talk) 06:54, 17 December 2008 (UTC)

Exactly. Such as Jim Morrison, Paul McCartney, John Lennon.... --John (talk) 07:22, 17 December 2008 (UTC)

I had a similar difficult which resulted in much backlash when I requested references for the disambiguation page at MS. As I click to view the MS disambiguation page I reallise that this is an ongoing problem. For example, there is a link to Mats Sundin (at the top) but at no point does that article mention that his name is abreviated as MS, perhaps it's M.S., (which would make more sense), but even so, there are no references within the main article to support the dissambiguation page. Which is why I will/have removed it. (Depending on the circumstances (such as this example) people will most like be contentious towards any removal!) Keep up the good work. --CyclePat (talk) 18:40, 17 December 2008 (UTC)

p.s.: Just look at all the edits I did to clean up MS. Good luck! --CyclePat (talk) 19:05, 17 December 2008 (UTC)

I don't think there is clear community consensus on citing lists. Some feel that as long as the information is sourced in one article, it is verifiable in any other article that links to the one with the source (ie it is verifiable by going to another page, even if it isn't actually verified on the one you are looking at). Others think that information should be both verifiable and verified directly in any article the information appears in (including lists). My feeling is something of a compromise... When it comes to lists, the determination as to whether to require citations (or not) should be left to the consensus of the editors working on the list... with the strong recommendation that if inclusion in the list has a reasonable potential for controversy, then the inclusion criteria should require references in the list. Blueboar (talk) 19:09, 17 December 2008 (UTC)

That makes sense. It seems to me that an article which includes living people and deals with ethnicity definitely needs citations per WP:BLP. The problem is that we have a long list which contains much that is probably true and much that is false. I have no interest in cleaning the article up; I have better things to do and this article duplicates a category in any case. I therefore posted a request for citations on the article talk a month ago. When none were forthcoming I made the article a redirect but this was undone. We thus have an article on Wikipedia which currently (or at least the last time I checked) asserts that Jim Morrison, Paul McCartney, John Lennon (among many other spurious entries) are or were Irish. I am not sufficiently uninvolved to take administrative action on this matter but maybe somebody could back me up, or else do the work to trim out the crap from the article and cite the rest? At the moment it is a disgrace to Wikipedia and the sort of thing that gets us (rightly) laughed at by some. --John (talk) 19:28, 17 December 2008 (UTC)

It's fine if a list duplicates a category. Some people prefer using categories, some people prefer using lists. Lists and categories complement each other. List of Irish people states that it's a "list of famous Irish people" and that it's not just about ethnicity. It says it covers "People who were born on the island of Ireland and/or who have lived there for most of their lives" and people who "have been raised as Irish, and/or have adopted Irish citizenship." If some names don't meet the criteria for inclusion in the list, you can remove them. Jim Morrison is in Category:Irish-Americans and Category:Irish-American musicians. The Jim Morrison article says "He was of Scottish and Irish ethnic heritage" and cites The Biography Channel. So that citation can be added to the list if someone disputes Jim Morrison's inclusion on the list. If there's still a dispute, talk it out on the list's talk page. Jim Morrison is dead, so BLP doesn't apply, but I suppose if you wanted to be really strict you could remove every name that's in Category:Living people. --Pixelface (talk) 05:02, 19 December 2008 (UTC)

That doesn't make sense though Pixelface. Read what you wrote again. There is a mismatch between the criteria of the list and the inclusion of people who were not "...born on the island of Ireland and/or who have lived there for most of their lives" or "...raised as Irish, and/or have adopted Irish citizenship." --John (talk) 19:39, 23 December 2008 (UTC)

There's a clear community consensus that all articles (or "encyclopedic content") in main namespace are (is) subject to WP:V. There's a clear community consensus that stand-alone lists in main namespace (except disambiguation pages: "Disambiguation pages ... are non-article pages, in the article namespace,...") are a type of articles or "encyclopedic content".

This resolves to: there's a clear community consensus that all lists in main namespace except for disambiguation pages are subject to WP:V.

CyclePat's example (of a disambiguation page) is off-topic (or: an example of the accepted exception) regarding the general question.

Blueboar's comment is equally missing the point: there is a clear community consensus (etc).

See WP:lists#Listed items

Verifiability-by-main-article-of-the-listed-item applies, for example, for categories (but we're no longer in main namespace then, but in category namespace). Similar example: lists in Portal namespace. --Francis Schonken (talk) 19:52, 17 December 2008 (UTC)

Thanks all, that was really helpful. Unless there is a volunteer to sift through and remove the unverified and verify the rest I will redirect again tomorrow per WP:BLP. There are a whole bunch of ethnic lists like this one so it may be a bigger issue. --John (talk) 04:48, 18 December 2008 (UTC)

Podcasts and WP:SPS

I assume that the article issues that started our recent discussions over podcasts must have been resolved, as discussion here has stoped... unfortunately, there are some unfinished questions that never got answered... namely, when someone is interviewed on a podcast, who is the publisher of the interviewee's remarks? This has baring on determining whether WP:SPS applies, or not, and whether the remarks are reliably published or not.

Looking at the conversations here and at WT:RS, some seem to feel that the interviewee is the publisher, the rational being that when someone speaks they are the author of their own words and are thus self-publishing them. Others seem to feel that the publisher is the the website that hosts the podcast (or rather the group or person that runs the website). The rational here being that a podcast interview should be considered similar to a print interview... and, when someone is interviewed by a magazine, we treat the magazine as the publisher.

Now, many times these two views will converge ... a person creates a podcast of themselves being interviewed and posts it on their website (or the website of an organization that they are directly connected to). But there are situations situations where the interviewee does not run the website (or does not have any direct connection to the group or person who runs the website). In this situation, is the interviewee self-publishing his remarks (ie WP:SPS applies), or is the group or individual publishing the remarks (ie WP:SPS does not apply)?

I see several potential issues that arise here... involving both WP:SPS and WP:RS ... Let us say that a recognized expert on stamp collecting is interveiwed by an amature stamp collector, who posts the podcast on his personal web-page. The expert is reliable, but the amature is not. If we say that the intervewee is the self-publisher, then his or her remarks can be cited as a reliable source, however that reliability would be limited by SPS. If the website is the publisher, then the remarks should not be considered reliable as they are not reliably published (the personal page of an amature stamp collector is not a reliable source). We could not use the remarks, even though they come from an expert...and even if we wanted to use them purely to express an attributed view of the expert.

Now let us slightly change senario, same amature conducting the interview, but instead of posting it on his personal website, he gives it to a national Philatelic Society to which he belongs, and they post it on their website. That website is a reliable source. If the website is the publisher, then the restrictions of WP:SPS do not apply... and if the expert says something controvercial (say an accusation of fraud against another expert) we can discuss the allegation and cite the podcast in our articles. If the interviewee is the self-publisher, then the allegation is restricted under WP:SPS.

As you can see... these are potentially puzzling issues... We need to form a consensus here. Blueboar (talk) 17:41, 22 December 2008 (UTC)

I think there are multiple cases to consider:

If the interviewee is the one publishing the interview, the interviewee's statements can certainly be treated as self-published. This is no different than posting the same statements in blog.
If a reliable source publishes the interview (this could be the interviewer or a third party), then we can treat it as a "normal" source. This is no different than a reliable source picking up a blog post.
- If the reliable source disclaims editorial oversight (e.g. "The views expressed in this interview are solely those of X and do not necessarily represent the views of the publisher."), we can trust that it really is X but we should treat it as self-published. This is no different than a reliable newspaper publishing an opinion piece.
If an unreliable source publishes it, IMO the interviewee's statements can be considered self-published if we trust that it really is the claimed person and we trust that any editing hasn't misrepresented the actual interview. If there is any reasonable doubt that it is faked or mis-edited, don't use it, and don't use anything that comes close to a WP:BLP violation.
- If the interviewee reliably "claims" it, for example by linking it from their website, then it's no different than the interviewee having a vanity press publish it for him and thus we can treat it as self-published.
- We could also use it as a self-published source for the publisher, which might be useful in extremely limited cases (e.g. "X claims A said Z,^[1] but A denies this.^[2]" might be a useful example in a section about X's faked interviews).

To address your example specifically, the first case falls under my point #3: it depends on whether the expert "claims" the interview and/or whether there is any doubt it really is the expert being interviewed. In your second example, since you stipulated that the Philatelic Society is reliable then it falls under #2. Anomie ⚔ 18:46, 22 December 2008 (UTC)

OK, so the website is the publisher, while the interviewee would essentially be the "author" of his/her remarks. And both publisher and author have to be considered as aspects of "the source" when looking at RS issues. That makes sense to me. Blueboar (talk) 20:20, 22 December 2008 (UTC)

That is a great way of putting it! I'll have to remember that. Anomie ⚔ 21:47, 22 December 2008 (UTC)

Straw poll on 'trial' implementation of FlaggedRevisions

The discussion on the implementation of a 'trial' configuration of FlaggedRevisions on en.wiki has now reached the 'straw poll' stage. All editors are invited to read the proposal and discussion and to participate in the straw poll. Happy‑melon 18:08, 2 January 2009 (UTC)

RFC at WP:NOR-notice

A concern was raised that the clause, "a primary source may be used only to make descriptive claims, the accuracy of which is verifiable by any reasonable, educated person without specialist knowledge" conflicts with WP:NPOV by placing a higher duty of care with primary sourced claims than secondary or tertiary sourced claims. An RFC has been initiated to stimulate wider input on the issue. Professor marginalia (talk) 06:31, 3 January 2009 (UTC)

Blogs associated with published periodicals other than newspapers

Footnote 5 says:

"Blogs" in this context refers to personal and group blogs. Some newspapers host interactive columns that they call blogs, and these may be acceptable as sources so long as the writers are professionals and the blog is subject to the newspaper's full editorial control. Where a news organization publishes the opinions of a professional but claims no responsibility for the opinions, the writer of the cited piece should be attributed (e.g., "Jane Smith has suggested...").

I assume his is applicable to blogs associated with other published periodicals, e.g., magazines, correct? For example, Information Week, a reliable source, publishes a number of blogs by its writers. TJRC (talk) 19:26, 22 December 2008 (UTC)

Correct. Blog collumns hosted on the website of any media outlet (newspaper, magazine, TV network, etc) should be treated exactly as if they were in the printed/aired version of the media outlet. Generally this means that they would be considered "opinion" pieces... reliable if attributed as being the opinion of the collumnist. Blueboar (talk) 20:07, 22 December 2008 (UTC)

Actually, Blueboar, the hosting address is not enough. Take for example this blog (Spanish language) by the head of Scientology in Spain, hosted in the "la comunidad" section of the website of the El País newspaper. I thought at first the facts reported there were good to cite with all the considerable authority of El País, until after some digging around I realised that anyone can open a blog account on the elpais.com website ([5]). The la comunidad section of the website is basically a feature comparable to blogger. We have to verify in each case who the author is, and whether they are a staff member of the paper or not. Jayen 466 13:51, 4 January 2009 (UTC)

Interesting. Given this, I have to agree with Jayen's caveat. Blueboar (talk) 15:16, 4 January 2009 (UTC)

WP:NONENG is one of the most stupid policy in English Wikipedia

It assumes all human knowledge is described in English. But this is not true on this Earth. Python eggs (talk) 23:54, 9 January 2009 (UTC)

The policy doesn't require that sources be written in English, only that sources in English are preferred. I think that makes sense, since the content of this Wikipedia is also in English. --Megaboz (talk) 00:02, 10 January 2009 (UTC)

I believe guidelines also say that a translation should be provided upon request. This is reasonable because part of "verifiability" is that other editors can verify - and most editors know English and don't know the language of a document in whatever other language is being used. How can you verify something you can't read? Of course there is still the danger of deliberately false translations. Readin (talk) 00:06, 10 January 2009 (UTC)

And what about NONENG "assumes all human knowledge is described in English?" Please read NONENG and you will see what it actually says. Slrubenstein | Talk 00:33, 10 January 2009 (UTC)

You might have the wrong policy, Python eggs, as there is nothing in NONENG that makes the assumption that all works are in English. In fact, this policy assumes the opposite, and addresses that assumption: that there will be times when sources are not in English. --KP Botany (talk) 01:41, 10 January 2009 (UTC)

Then I think this section should have more words to clarify the policy. See this discussion, someone thinks English is "required" not "preferred" as most of you do. Python eggs (talk) 09:13, 10 January 2009 (UTC)

I agree with the others who have commented here. Reading the policy, it is clear that it is a preference and not a requirement. The policy also spells out that it does not apply when no English source is available, nor when one is available but a foreign-language source is better for some reason. When sources both in English and in a foreign-language are available but the foreign-language one is superior, my personal feeling is however that as a convenience to those readers who do not know the foreign language, both should be given. Joeldl (talk) 09:27, 10 January 2009 (UTC)

I suggest the following rewording, on the grounds that the current wording may be confusing for readers whose native language isn't English.

	change	assuming the availability of an English-language source of equal quality,
	to	when an English-language source of equal quality is available,

Pi zero (talk) 13:53, 10 January 2009 (UTC)

I have nothing against the proposed change, but I doubt it will solve the problem at High Speed Rail - unfortunately, it is probable that at one point or another there will always be some editors who misread policy or apply it with poor judgment, which is why edit conflicts are best resolve on article talk pages and not here. Slrubenstein | Talk 14:34, 10 January 2009 (UTC)

By the way, if one looks at the discussion at High Speed Rail, one would see that Python misread the policy there, and also others were objecting to his sources not on the fact that they were not in English but that they are of poorer quality than the English sources and fail to meet other standards. Confusion over NONENG is not the source of the content conflict as Python suggests. This seems like a red-herring. Slrubenstein | Talk 14:37, 10 January 2009 (UTC)

the practical rule is to use both 1/ the best English language sources available (if there are any) AND the best sources available, regardless of language; Supplemented by, in discussing non-Anglophone area topics, the best sources will usually be non-English, in the language relevant to the topic being discussed. Though not all editors can identify them, those who can, should. DGG (talk) 01:16, 12 January 2009 (UTC)

"challenged or likely to be challenged"

The words "challenged or likely to be challenged" are in bold in this policy, and yet there appears to be no guidance on what actually constitutes a good-faith challenge. While it is understandable that the burden of evidence lies on the editor adding or restoring content when it is challenged, a common but mistaken interpretation of this policy is that there is absolutely no burden on removing content. This policy is often used to justify removal of content that no one really disputes for the sole reason of it being unsourced, or to justify the mass removal of large swaths of content just because it is unsourced, or the person removing it doesn't like the sources being used. But if this was the way this policy was intended to be used, the bolded words would be unnecessary; it would be equivalent for it to simply say "All material must be attributed to a reliable, published source using an inline citation." It is a huge waste of editors' time to force them to find sources for things which are undisputed common knowledge; or to force them to continually "prove" to the remover that the sources justify the content, when the person removing the content doesn't want to bother making any effort to find and look at the actual sources. It is my opinion that this policy, or at least a guideline, needs to clarify that material cannot be challenged for the sole reason that is is unsourced, or by an unsubstantiated opinion that the source is not reliable or relevant; it should be required that the person removing content needs to demonstrate a good-faith belief that the information is false, unverifiable (not just unverified), or harmful. And that once a source is cited in good faith, the burden should then fall upon those challenging the content to actually try to find and look at the source themselves, and to provide a good-faith explanation of why they think the source doesn't exist or is unlocatable or is unreliable or doesn't say what it is claimed that it says. Any thoughts? DHowell (talk) 06:16, 23 December 2008 (UTC)

Essentially this is an aspect of the never ending inclusionist vs. exclusionist debate. Another way of phrasing this aspect is Verifiability vs. Verification. I think the wording of the policy is a nice compromise between these two positions... To add information it simply must be verifiable (meaning that it can be verified... not necessarily in a given Wikipedia article, but somewhere.) However, if you do not provide actual Verification (ie a citation) that information can be removed. In other words... The threshold for inclusion is Verifiability; the threshold for removal is Verification.

Remember, just because you think a statement is obvious or is common knowlege, that does not mean everyone else will agree. My feeling is that if someone has bothered to delete material because it is unsourced, that constitutes a good-faith challenge. Sure, there may be more polite ways to handle the situation than summarily removing the material from the get go, but summary removal can be justified under WP:BOLD. And ultimately, a simple edit summary of "Unsourced" is enough to make challenging unsourced information legitimate. To return the material you must either cite a source (which should be easy to do if the material is obvious or common knowledge), or convince the challenging editor that a source is not needed.

If, on the other hand, someone summarily removes material that is sourced, that is a different matter. Remember that the challenge may have been made in good-faith (a typical example of good-faith removal of sourced information is when an entire paragraph or section can be cited to one source... the citation may be stuck on at the end, and the challenger may not realize that it applies to the entire paragraph or section.) In any case, if the material is sourced, you have a legitimate reason to contest and revert the removal (noting in the edit summary that the removed material was sourced, and requesting further discussion on the talk page). Blueboar (talk) 19:04, 23 December 2008 (UTC)

DHowell, is it safe to assume that you're in a current dispute? Would you like to tell us what article(s) are involved? WhatamIdoing (talk) 22:52, 23 December 2008 (UTC)

A challenge is a challenge. If an idea is not verified and it is challenged it should be removed. The burden in removing content is a burden of the content not being verified and the challenger being dubious or concerned. No more is really needed. The idea is to err on the side of not spreading false information. If someone is challenging things to make a point or to be disruptive then that is something to be handled through behavioural policies. Chillum 23:01, 23 December 2008 (UTC)

Thanks for your responses. This isn't really tied to any specific dispute, it is more of a general feeling based on observations of various disputes, especially calls to delete articles on the basis on them being "unsourced", even when sources have been shown to exist. Something that bothers me is the excessive removal of content without a proportionate amount of actually looking for and using sources to improve content. It far easier to remove something than to look for a source and verify it; and thus perhaps for every 10 editors removing content, there might be one actually looking for sources (yes, that's just a wild-ass guess based on personal perceptions), and at that rate a lot of valuable content will be simply lost to attrition. One of the problems is there is nothing in policies or guidelines to help us distinguish between real good-faith challenges and "challenging things to make a point or to be disruptive"; it seems to take something to level of Wikipedia:Requests for arbitration/Episodes and characters 2 for any sanctions to even occur, while mass removal of content less egregious, and yet still disruptive and pointy I believe may be happening on a regular basis. But since any removal of content can theoretically be justified by WP:BURDEN, this makes such disruptiveness and pointiness difficult to determine and defend against. One has only so much time to spend looking for sources. DHowell (talk) 08:22, 24 December 2008 (UTC)

In most cases, AfD will not delete an article simply for lack of sources, especially if people can demonstrate that sources do in fact exist. However, it is then up to those who want to keep the article to fix the problem... ie to actually add the sources. If no one cares enough about an article to actually fix the problem and provide sources, after an AfD has noted the problem... the topic is probably not notable enough. Blueboar (talk) 16:53, 24 December 2008 (UTC)

Blueboar I would respectfully but strongly beg to differ with your characterization of people "[not caring] enough ...[to]... provide sources". For several months I worked on trying to provide sources for two articles (Joseph Murphy (author) and Church of Divine Science). To suggest that my failure to provide sources is because I didn't "care enough" is erroneous at best and insulting at worst. Such an attitude is based on a faulty assumption that any needed reliable sources are easily accessible online. Take for example the following text from the currently {OR} tagged WP article on Murphy:

"Dr. Murphy was Minister-Director of the Church of Divine Science in Los Angeles for 28 years, where his lectures were attended by over a thousand people almost every Sunday."

Certainly a claim of such large attendance requires a valid cite. I know from personal presence at some of these lectures that this is a true statement BUT my personal observation is original research and not verifiable to future Wikipedians, so I went looking for a WP:RS. The problem in my case is that because Dr. Murphy was popular but not controversial (controversy is what sells publications) little was written about him and I came to realize that the most potential RS available were from various Los Angeles area newspaper archives from between 30-50 years ago. These archives are not online (some of the newspapers are defunct) and the material by its nature was never going to be headline stuff so the physical search would naturally be very time consuming. I simply do not have the physical or financial ability to spend endless hours in archives located 100+ miles from my home and work. I also requested help from other Wikipedians via WP:REX who have access to such resources but to date only one person with limited resources has discovered anything. Thank you droptone! Unfortunately neither of these sources is adequate. Common sense tells us that citations must exist and that this author is certainly worthy of an encyclopedic entry but in this case WP:RS and WP:N policies demand citing resources that few people have time or money to access. Low Sea (talk) 22:21, 3 January 2009 (UTC)

All because of WP:NOR? another case of policy in need of amendment. Nick carson (talk) 12:03, 14 January 2009 (UTC)

I understand sympathize with what you are saying Low Sea ... but this is not what I was talking about when I made my "no one cares" comment. I was referring purely to situations where an article has been nominated for AfD, kept due several editors stating that sources exist, and then no one bothering to actually add the sources they said existed during the AfD. I take that as an indication that no one really cares. In cases such as your's, where there is a good faith effort to locate sources, I have no problem. Blueboar (talk) 23:24, 3 January 2009 (UTC)

I agree with Blueboar's comments on the good faith removal of unsourced material and 'no one really cares', a trap (if that's the right word and it probably isn't) I almost fell into when I posted a source for a current AfD and failed to add it to the article (but did when I realised what I'd done). — Preceding unsigned comment added by Dougweller (talk • contribs) 07:41, 4 January 2009 (UTC)

The thing is, "no one really cares" and "no one's working on it" are not valid reasons to delete an article. When someone cites a source to support information, then they should be deemed to have met the burden required of this policy, no matter where they have put that citation (whether it is in an edit summary, a talk page, an AfD, or the article itself). Anything beyond that (i.e. putting in the inline citation) is an editing issue, and our editing policy says we should strive to preserve information, and our deletion guideline says in bold "when in doubt, don't delete". The idea that editors should go around removing and deleting information willy-nilly just because it is unsourced, with out a good and explicit reason to challenge it, is not in conformance with those policies, even if it seems to be supported by this policy. My point is that if that idea were to be supported by this policy, the bold words would be redundant and unnecessary. Most editors do not spend all of their time on Wikipedia and many may not know how to properly cite sources in articles; so just because an article is not perfectly up to standards does not mean no one cares or no one wants to work on it. Personally, there are far, far more articles that I would like to work on than I have the time to work on, so to assume no one cares because they aren't putting forth the effort that you think they should is just wrong. If I cite a source in an AfD, and don't put it in the article, perhaps that's all I had the time to do; it is wrong to then delete the article because I supposedly didn't care enough to add the source to the article. Wikipedia works by many editors making incremental changes to articles (with a few making substanital changes); deleting perfectly verifiable articles on the assumption that no one cares hinders this process greatly. If I don't have the time to make an incremental change, I'm certainly not going to have the time to write the article from scratch; and this has zero to do with whether I care about the article's existence. Just because it is easier to delete than to add the source to the article does not mean it is the right thing to do. WP:BEFORE puts a burden on a deletion nominator to look for sources as well, but for some reason this burden appears not to be taken as seriously as the one mentioned in this policy. DHowell (talk) 05:30, 7 January 2009 (UTC)

There's a reasonable middle ground possible. Rather than deleting we could just comment out until a usable cite is found.LeadSongDog (talk) 06:14, 7 January 2009 (UTC)

LeadSongDog, you could also move the text to the article's talk page.

DHowell (and others), you appear to have confused deleting a single fact -- perhaps a sentence or a single word -- with deleting entire articles. WP:V says that if you think a statement is inaccurate, and it's unsourced, then you should delete the statement. This has nothing to do with whether the entire article is appropriate for the encyclopedia. We won't delete George W. Bush just because a single word is unsourced and probably wrong. We should delete the inaccurate word. WhatamIdoing (talk) 20:49, 7 January 2009 (UTC)

Actually, article deletion was only a part of my concern (and Blueboar chose to expand on that aspect, and I was just continuing that line of discussion). In my opinion WP:V should not be used as a deletion reason unless the entire article was unsourced and no sources were found after a reasonable and judicious attempt to find sources were made (and this is exactly what deletion policy says). But I also object to mass removal of content that is unsourced unless the same efforts are made. What I'm trying to get to is that there should be a clear distinction between challenging content and simply removing content; and that removing content should not just be interpreted as a challenge unless there is a clear reason for the challenge, and that reason should be something more than just "it's unsourced!". I'm not concerned with the occasional removal of sentences that people find dubious, I don't have a problem with that (and in those cases usually a good edit summary is sufficient to explain the reason for the challenge); but I do object to the wholesale removal of entire sections or articles (whether through AfD or by turning them into redirects without merging any content) without making a good-faith challenge and allowing for reasonable discussion with involved editors before doing the removal, and allowing them a reasonable amount of time to find sources proportionate to the amount of content being challenged. And five days is often not enough time, which is why I object to AfD in most of these types of cases. DHowell (talk) 03:24, 9 January 2009 (UTC)

So you'd like to require volunteers(!) to first invest time and energy in finding a reliable source before removing unsourced information.

While I have a little bit of sympathy for this, it's simply not practical. I also think that, with the exception of the occasional disruptive editor, the deletion of important accurate information simply for lack of of a source is pretty rare.

Also, when I'm patrolling pages, I find that an explanation of "unsourced" is a politer way of saying "probable nonsense" or "entirely off-topic" or "self-promoting garbage". If other editors happen to do the same thing (and I have no idea whether anyone else does), then it's possible that you're consequently seeing a bigger problem than actually exists. WhatamIdoing (talk) 05:32, 9 January 2009 (UTC)

But those removing information are "requiring volunteers(!) to ... invest time and energy in finding a reliable source before" adding the information! This is not practical when the amount of information being removed exhausts the resources of the editors trying to keep it in. I'm just asking for some balance. Investing time and energy in removing unsourced information simply because it is unsourced does not improve the encyclopedia. Putting a greater burden on others to improve the encyclopedia than one puts on oneself does not improve the encyclopedia. Every editor should be looking for reliable sources, those adding information and those removing information. In my opinion, information should only be removed if the information is likely to be unverifiable, or is harmful in some way, not simply because it hasn't been verified yet. And you can't determine "likely to be unverifiable" unless you invest some time looking to see what sources actually say. By the way, I wouldn't consider it impolite to say in the edit summary "unsourced; probably misinformation" or "entirely off-topic" (which is certainly not in any way a synonym for "unsourced") or "unsourced self-promotion". At least that would be a reasonable challenge, if it is an accurate representation of the info being removed. DHowell (talk) 06:58, 13 January 2009 (UTC)

The basic assumption that the would-be deleters as acting in good faith is not always true, sometimes they're POV-pushers - yes, I've seen actual cases. Hence I agree with DHowell's initial comments (06:16, 23 December 2008 ).

AFAIK the norm at WP is that the last ref supports all preceding content until the preceding ref or the start of a para. Otherwise artciles would be unreadable because of blue pox.

If a ref is given the would-be deleter should thoroughly read and understand it. If someone deleted a passage I'd written without explanation or without clear evidence of having thoroughly read and understood the cited sources, I'd re-instate the passage and tell them to go read the sources - and if the deleter persisted without providing evidence I'd be considering a reference to WP:ANI for WP:DE.

That doesn't mean I go looking for excuses to jump down throats - in fact if someone asks a question on a Talk page, however naive, I try to answer politely and in as simple language as the subject allows, and have been thanked for it several times. --Philcha (talk) 09:50, 13 January 2009 (UTC)

This is essentially the same practice that I have. — Carl (CBM · talk) 13:42, 13 January 2009 (UTC)

Yes, DHowell, you understand my position correctly. See WP:BURDEN: the person that wants to include the information is the person that has the duty to demonstrate that it's correct. WhatamIdoing (talk) 05:59, 14 January 2009 (UTC)

Of course, the best way to avoid annoying deletions and challenges is to provide citations when you add information in the first place. Instead of writing articles and then looking for sources to support what you wrote, best practice is to have your sources lined up and ready to go before you write. Blueboar (talk) 15:46, 14 January 2009 (UTC)

Tag all articles with {{Refimprove}}?

Hi, after working on quite a few articles lately and finding a lot of unreferenced information I decided I would make a suggestion here for it to be discussed and noticed. My suggestion is that the {{Refimprove}} tag should be added to new, current and old (collectively all) articles on Wikipedia. Let me explain my reasoning. If this tag was on all articles I believe it would let readers know that the article they are viewing may have some material that is not true, accurate or correct, particularly the articles that have practically no references. I believe that without references anyone could add information. Furthermore, it would also provide a message to all willing readers that they could contribute to articles they had information and references for. Thereby, potentially increasing the number of contributors.

How it would work? Well with every article tagged it would be the job of the contributors to add material to the article until it reached a state of 'completion' where as much work as possible was referenced and verifiable. Of course, articles already substantially referenced and verifiable would move to the next stage. (i) I believe it could then be the job of an administrator/moderator to remove the tag from the page once they have been notified that the article has reached completion (this could also relate to [ii]) or, alternatively, (ii) when someone feels that the article has reached completion with references and all then this someone begins a discussion on the talk page arguing that the tag be removed and if the consensus agrees then it is removed.

What do you think? I know it may not work and I know it may be impractical but I believe this could help Wikipedia strive towards becoming 'Encyclopedic' and a legitimate source of information.- Jack ^(Talk) 21:53, 14 January 2009 (UTC)

I would object to tagging all articles with refimprove tags... I think you are assuming a universal problem, and suggesting a universal solution, based upon a limited set of examples. While many articles do indeed need better sourcing, other articles (the best ones) are very well sourced right from the start. Blueboar (talk) 15:58, 14 January 2009 (UTC)

Agree with Blueboar. For a start, all articles that have passed GA or FA review in the last year, possibly 2 years, have had their refs scrutinised. There are reassessment projects for older GAs and FAs, e.g. I got Spider through a GA reassessment a few months ago. I also know of non-GA-/-FA articles that are adequately referenced - for example I've created a few stubs (some of which have grown) to explain specific concepts, and these have adequate refs. --Philcha (talk) 16:24, 14 January 2009 (UTC)

I agree that this is a well-intended but dreadful idea in practice. It would likely create an enormous uproar, edit wars, and likely result in people mass-removing all such tags, even when they were appropriate. WhatamIdoing (talk) 21:30, 14 January 2009 (UTC)

Fair enough, I guess I will just have to continue tagging nearly all new articles I visit since it seems every new one I go to lacks the necessary number of references.- Jack ^(Talk) 21:53, 14 January 2009 (UTC)

You might think about whether tagging articles based on the number of references is really the right choice. Two truly excellent refs could be perfectly adequate for an article, whereas a dozen blogs is insufficient for just about anything. Also, assuming that you're not willing to make any effort to actually locate sources, you might suggest on the talk page exactly what you find insufficiently referenced. "The article" isn't very helpful. "This section" or "these three statements" can be very helpful. WhatamIdoing (talk) 03:59, 15 January 2009 (UTC)

You make valid points 'WhatamIdoing' (funny name by the way, I like it). Anyways, I guess I'm still attached to a reference system I've used since the beginning of my schooling. I had to almost always reference anything I wrote whether it be interpreted or direct because in reality no one can think of something new/informative without having to get it from somewhere. Also, I didn't add the reference tag purely on the number of references at the bottom I looked at the citations throughout the information too. I guess I'm still having to get use to Wikipedia's 'allowed practice' of only needing one reference for a whole paragraph, compared to when I was at school and had to provide citations for each line of text. By the way, if this is the proper 'practice' I think the citation at the end of the paragraph should be of a different type to other citations to let readers know. Thanks for the advice WhatamIdoing, I'll apply it whenever I can or I remember.- Jack ^(Talk) 04:33, 15 January 2009 (UTC)

Problematic line

In the self-published section, it says "if the information in question is really worth reporting, someone else is likely to have done so."

I'd like to remove that, as it's frankly silly - there are plenty of notable, significant fields where the primary means of discussion is blogging by notable sources. There are certainly cases where silly trivia is discussed by experts in blogs. But there's also lots of cases where it's not, and I think the warning ends up being dreadfully old-media focused in a kind of silly way. Perhaps just a note about caution, and a reminder that sources that are not self-published are always preferred when they are available? Phil Sandifer (talk) 19:59, 10 January 2009 (UTC)

I do not think it is silly. Blogging is all too often an easy way to self-publish, and is seldom guided by any kind of peer or editorial review. You may as well say Wikipedians can use their own user pages as sources for whatever topic they have written a WP or sandbox essay on. The issue here is not watering down the proscription of using self-published material as appropriate sources. What we need is a discussion of different kinds of blogs and distinguish between those that are, like most blogs, just a form of self-publishing, and other blogs in which there is an adequate form of editorial or peer review to make them meaningful. I do not think there are enough and I don't think we should make it any easier to use blogs as a way to side-step this point about self-published material, as an appropriate source for encyclopedic content. Slrubenstein | Talk 20:15, 10 January 2009 (UTC)

Well, and the third category of self-published sources, cases where the source is significant enough that the medium is irrelevant. My usual example here is J. Michael Straczynski, who has written extensively on Usenet and in self-published collections of scripts on the production process for Babylon 5. There's no question of the relevance or the reliability of that. My concern is that there's a genuine shift in the nature of publishing - truly important stuff happens in blogs now. We should still make them not our first choice, but the "if it really matters it'll be somewhere else" thing just isn't true, particularly for contemporary and recent topics. Phil Sandifer (talk) 20:24, 10 January 2009 (UTC)

We already allow for blogs written by "acknowledged experts" in a given field of study. As the show's creator, Straczynski surely qualifies as an expert on Babylon 5... thus, his blog would be considered a reliable source in that article. However, let us say that he comments on string theory in his blog... this comment would not be a reliable source for use in the String Theory article... as Straczynski is not an acknowledged expert in string theory. Blueboar (talk) 03:34, 13 January 2009 (UTC)

Yes, but I think Phil's point is that the sentence in question is applied even when the author is an expert. I understand it's purpose: it's meant to stop us from taking announcements of (e.g.) new discoveries on scientists' blogs and using those, without waiting for peer review, etc. But as a warning that does not always apply, I don't feel it belongs in a policy. It should be at WP:RS, not here. JulesH (talk) 09:58, 17 January 2009 (UTC)

First, the whole SPS thing is in the wrong place, it should be in WP:RS.

The big problem is the extreme variability of blogs and other SPS. Most are self-indulgent twaddle, but some are very thoughtful pieces written by very smart people. In "academic" subjects there are plenty of journal articles; one has to be careful with books, it mostly depends on the authors. In "non-academic" subjects the situation is much less clear-cut: peer-reviewed journal articles are virtually non-existent and the university presses do not publish books on these subjects; with books you have to know the strengths and weaknesses of authors, for example in chess Reuben Fine is reliable on the quality of play in a game but unreliable on history; and in many "non-academic" fields there isn't the money to support printing & distribution costs or management overheads, so a lot of the good material is published on the Web and in many cases self-published. To take chess as an example, Edward Winter is known to be one of the most thorough chess historians currently active, and most of his materail appears at his web site; Taylor Kingston, Tim Harding and Jeremy Spinrad also do good, thoughtful analyses, many published at ChessCafe.com. In video games, the printed books are mostly strategy guides and are poor, because they are published in a rush and based on beta versions of the games (see Decline of Guides by Alan Emrich). Big-name computer game mags are no better, as they mainly contain reviews of individual games, are written to tight deadlines, and there's evidence of reviews being slanted as a result of commercial pressure from advertisers. I've seen factual errors in reviews by people I know are "good guys", and can only attribute these to time presssure. OTOH the best self-published sources are very knowledgeable. For example Master of Orion II Online is the place for expertise about Master of Orion II: "Lord Brazen" has produced a series of patches that resolve bugs and present additional game options; "Siron" has produced a range of mods to improve game balance; because all of these run only in the MS-DOS version of the game, they know all there is to know about running the game on Win XP and Vista; and you'd better believe "Cybersaber"'s strategy guide, because everybody there is in awe of his playing skill.

Our rules about reliable sources make a big deal of "reputation for fact-checking and accuracy", but it's mostly fluff. Who says publisher X has a "reputation for fact-checking and accuracy", and who reliable is the source of that claim, etc., etc., etc.? (if you think this ian infinite regress, you're right) Where is there a good description of any publisher's QA processes (how would you verify it?), and an independent audit of how faithfully they are followed (where they exist)? Claims about "reputation for fact-checking and accuracy" are themselves almost entirely SPS - or fabrications by WP editors, and therefore violations of WP:V!. --Philcha (talk) 11:21, 17 January 2009 (UTC)

Verifiability vs. Truth

OK ... New guy in the classroom, be nice. I come here from the Larry Sanger page, where the whole "The threshold for inclusion in Wikipedia is verifiability, not truth" is misquoted. I read back through a couple of this articles talk archives to see if there was a suggestion exactly like the one I'm about to make. While it is an old dead horse I might be beating, and I know that "truth" is a subjective thing, I can't help but wonder if the following might be an alternative:

The threshold for inclusion in Wikipedia is truth must be supported by verifiability -

To be honest I was shocked by the statement, until I read through the remainder of the sentence. Unfortunately, many people don't read through the remainder of the sentence, and eventually twist things into "Wikipedia is interested in consensus not truth". Well ~~that's~~ my coming here and saying that is my "Be Bold" moment for today. Not that I don't try to edit articles as well, I just try to keep in mind that there is a difference between being bold and being foolish. (boy I hope that one doesn't come back to bite me). Kind Regards Ched (talk) 07:02, 17 January 2009 (UTC) typo fix Ched (talk) 07:03, 17 January 2009 (UTC)

Actually, Wikipedia is an encyclopedic project, and the real threshold for inclusion is knowledge - not truth. Knowledge is objective, and means that when you learn something you can rely on it and do something relevant with it - as the saying goes, "the proof of the pudding is in the eating".

This implies that knowledge must be verifiable, or, to put it the other way around, if something is by nature not verifiable (like "Cleopatra dreamt of Ceasar before dying") it is an opinion, not real knowledge. You may thing that "Cleopatra dreamt of Ceasar before dying" and you will obviously never be proven wrong, you may guide your life on such a perfectly respectable opinion, but it is not knowledge (by nature) because it does not pass the test of Verifiability (touchstone) - that's all, so such an assertion has nothing to do within an encyclopedia.

This is why "The threshold for inclusion in Wikipedia is verifiability, not truth", the reason being that "the touchstone for knowledge is verifiability" - and "the purpose of an encyclopedia is to give access to the sum of all human knowledge - not truth".

But, OK, the Wikipedia:Five pillars page dosen't say much about that, and the "Wikipedia is an encyclopedia" links to Wikipedia:What Wikipedia is not, which is obviously not enough : there should be somewhere a link to Wikipedia:Wikipedia is an encyclopedia, which should be more specific on this subject ... Michelet-密是力-Me laisser un message 08:45, 17 January 2009 (UTC)

My immediate response to "truth must be supported by verifiability" was "I like it!" Then I started thinking. The problem with "truth" is that, in a really contentious subject where emotions permanently run high, (a lot of politics & religion, much worse if combined), both / all sides believe they know "The Truth". Including "truth" in the rules for WP:V would probably lead to edit wars where each side deletes the other's content, saying "I don't care what the academics say, that is not The Truth". --Philcha (talk) 09:23, 17 January 2009 (UTC)

In that context, you can't verify anything, since there is no "good faith / common sence" consensus. The only solution left is, in that case, to describe the debate (X says X, and Y says Y) and abandon any hope to give any knowledge on the subject itself :( but this is only true for polemic subjects, not the general case. Michelet-密是力-Me laisser un message 19:14, 17 January 2009 (UTC)

The proposed change goes against the whole philosophy underlying Wikipedia. Sanger's point is that people will not agree to "the truth" and in his analysis as a philosopher there is no basis or means by which people will agree to the truth, so truth should be abandoned as a goal. And I think this is an important and useful point, because in my experience the worst POV pushers always defended themselves by claiming they were simply adding the "truth." Since it was true (they proposed), their arguments for why it was true could never be disproven. This just leads to useless edit-wars. Sanger's point is pragmatic. People cannot find a universal or objective standard for the truth (any who claims to is wrong, since someone else will invariably disagree with them) ... but they CAN come up with mutually acceptable principals of verifiability. And we HAVE come up with those principles, and they have worked really well for us! Slrubenstein | Talk 23:46, 17 January 2009 (UTC)

Well, I appreciate you hearing me out and responding. The "not truth" is still a sticking point for me, but I'll try to wrap my head around the subtleties as I go along. my best to all Ched (talk) 16:14, 18 January 2009 (UTC)

That which is verifiable/verified would usually end up being the truth anyway. (The value of "usually" is of course a moving target.) Roger (talk) 17:25, 18 January 2009 (UTC)

I'm not sure I agree. I always assumed that it's verifiability not truth because we generally deal with points of view. Sometimes those points of view are technical and sometimes they are not. I mean this, most biologists don't have a problem with evolution, there's a lot of evidence to support it. But the processes and mechanisms are still debated, and there have been some really heated debates, especially between so called Neo-Darwinism and Punctuated equilibrium. We should be giving both these points of view because they are both relevant, but whether either, or both, or neither is the "truth" is irrelevant. What is relevant is that we can verify that reputable scientists have published reliable documents that we can use as sources to support both ideas. Likewise with Michelet's example about Cleopatra. We shouldn't be interested in whether it's "true" that Cleopatra dreamed of Ceasar. We're interested in whether this is a "point of view" that can be verified from a reliable source. For example we can say that "academic x says in paper y that according to historical figure z, Cleopatra dreamed of Ceasar." So I don't think it's about things that are verifiable necessarily being more likely to be "true", just that things that are verifiable at least have (or have had in the past) the support of a recognised authority/reliable source. Alun (talk) 06:41, 20 January 2009 (UTC)

It is not our role to choose between competing theories and declare one or the other the winner. We should just be reporting the existence and content of the various theories/opinions/points of view. To use your Cleopatra example: The "verifiable truth" is the fact that Academic X said what he said <ref>Here is the article he wrote.</ref>. The "truth" or verifiability of the content of what he said is the responsibilty of the editors of the journal that published his paper, not the editors of Wikipedia. We deal in "meta-truth" (to coin a word). Roger (talk) 10:35, 20 January 2009 (UTC)

Where exactly did I claim that it was the role of Wikipedia editors to choose between competing theories and declare one the winner? Indeed I said the opposite of that. What I said was that we include all points of view that are notable, and verify those points of view from reliable sources. There is no "truth", even in the sources we cite, those are points of view and not truths. Ultimately it's about belief. One academic may believe that his point of view is "the truth", and another may believe that this point of view is fallacious. We provide both these points of view, and use reliable sources to verify them. We leave it to people who read the article to form their own judgment regarding these various points of view. Did you actually read what I wrote, because you seem to think I said the opposite of what I did say. Alun (talk) 13:27, 20 January 2009 (UTC)

Where exactly did I claim that you claimed that it was the role of Wikipedia editors to choose....? If you read my post properly you will see that I am actually agreeing with you. However I am putting forward a concept that could be a solution to the "conflict" between "truth" and "verifiability". The "truth" that we deal with is to report (with cites) the objective fact (truth) that opinions A, B and C exist and then describe the content of those opinions. The "truth" or otherwise of the content of the opinions is irrelevant. That is what I mean by "meta-truth". Some truths are of course absolute: 1+1=2 will always be objectively and immutably true, it is not suceptible to opinion or dispute. Roger (talk) 20:37, 20 January 2009 (UTC)

When there are differing points of view, we have to gage the prominence of each one, and report them according to their prominence. For example, we would not have an article that simply said: "Smith (1999) says the Earth is flat. Jones (2000) says the Earth is round. Davis (2001) says the Earth is shaped like a giant ice cream cone." The actual Earth article has just one sentence, "In the past there were varying levels of belief in a flat Earth, but this was displaced by the concept of a Spherical Earth due to observation and circumnavigation." This article does not present any sources in favor of the flat earth theory.

What people usually mean when they use the catchphrase "verifiability, not truth", is that just thinking something is true is not enough to include it – we have to have sources as well. However, the way in which we report on conflicting sources is much more subtle than simply listing them all and letting the reader decide. That's a cheap way out. We occasionally compromise to it in particularly controversial areas simply as a way of moving discussion forward, but for the vast majority of articles all editors agree both that the material is verifiable and that it is "correct" in the common language sense. — Carl (CBM · talk) 13:48, 20 January 2009 (UTC)

Which is why I said that points of view have to be from reliable sources, or are you suggesting that a reliable source would claim the world is flat? Furthermore I am perfectly well aware about our neutrality policy says about things like undue weight. But that is a different story. This discussion was specifically about why we say verifiability not truth. We say it because "truth" doesn't come into it, we cite reliable sources to support points of view. We don't care if those points of view are "the truth", only that they come from reliable sources. We don't give undue weight to fringe theories, but then fringe theories are unlikely to come from reliable sources anyway. When two or more points of view exist that are prominent, then we include them. Obviously the prominence and/or importance of a point of view can be relatively subjective, which is why we have talk pages, no? Alun (talk) 14:02, 20 January 2009 (UTC)

I, and many others, care both that claims in our articles can be sourced to reliable sources and that the claims made by our articles are correct. It's a false choice to see these as competing interests instead of complimentary goals, and naive to think that either one can be ignored as long as the other is accounted for somehow. — Carl (CBM · talk) 18:25, 20 January 2009 (UTC)

It might help this conversation if we distinguish between "Truth" (which is subjective), and "Factual Accuracy" (which is not). We can all agree that information in Wikipedia should be "Factually Accurate", while understanding that there will be debates as to whether a given piece of information is "True" or not. Blueboar (talk) 19:21, 20 January 2009 (UTC)

CBM, the claims made by our articles should be true/accurate (e.g. "John Smith argues that X is also Y," but the claims made by the article's sources needn't be (that X is also Y could be false). This is why we don't "do" truth on Wikipedia, because the only truth that matters for us is whether we are accurately reporting the claims of reliable sources. We make sure our articles don't contain absurdities by making sure that our sources are good ones. SlimVirgin ^{talk|contribs} 19:32, 20 January 2009 (UTC)

I agree with the last sentence. But of course the way that we decide which sources are good ones is by using our editorial judgment and knowledge of the topic. For example, if the editors of an article agree that some reference is simply wrong, or misleading, or idiosyncratic, they may decide not to use it or mention it at all. Sources are "good" to the extent that they are helpful for writing an article.

Blueboar's comment cuts to the core of what I'm saying. The thing to keep in mind when people ask about the "verifiability vs. truth" slogan is that we do care about correctness, and we don't want to give new editors the opinion that correctness is unimportant or that it is somehow in conflict with referencing. The point of "verifiability, not truth" is just that we don't include information merely because editors believe it to be correct when we cannot find a reference for it. "Verifiability, not truth" does not mean that we must include claims in our articles which are factually inaccurate merely because the inaccuracy has been repeated by a source somewhere. — Carl (CBM · talk) 20:47, 20 January 2009 (UTC)

Carl, I think everyone here agrees that sources and the views they contain have to be represented accurately. I have never seen an inaccurate account of a source or view survive the wiki-edit process, I do not think a conflict over this has ever even reached ArbCom. I bet there are plenty of articles with inaccurate accounts of views or sources and virtually everyone would support your fixing them as long as it was clear your edits made it more accurate. I do not think this is an issue at Wikipedia. The issue that I think concerns everyone else here is NPOV, which is non-negotiable. Slrubenstein | Talk 21:05, 20 January 2009 (UTC)

Yes, I agree completely. This is why I was saying it's important, when explaining "verifiability, not truth", to make sure we don't imply that inaccuracy is actually OK. Our fundamental goal is to write an encyclopedia, and that carries with it certain assumptions of correctness. — Carl (CBM · talk) 21:41, 20 January 2009 (UTC)

I'm not sure that's right, though. If we see something that an RS has written that's clearly an error on his part — say, he refers to the title of a book incorrectly, or misquotes someone — then we can ignore it. But if it's anything less clear cut, and the source is a reliable one, we don't really have grounds to keep it out. To give an example, we had a situation at Animal testing, where someone added the view of an RS — published, as I recall, in a peer-reviewed journal — that animals aren't really conscious, which is clearly absurd, but it was nevertheless used to uphold the claim that this is a view, albeit a minority one, that's held within the research community. The point of "verifiability, not truth" is precisely to stop someone like me from jumping up and down screaming, "False! Get rid of it!" SlimVirgin ^{talk|contribs} 02:48, 21 January 2009 (UTC)

This is an issue of due weight, from NPOV, that is not particularly related to verifiability. We decide on issues of due weight using our editorial judgment, and it's perfectly possible that we decide that even though something is verifiable the due weight for it is zero. As NPOV says, "Views that are held by a tiny minority should not be represented except in articles devoted to those views." So verifiability is not on its own a sufficient threshold for inclusion. — Carl (CBM · talk) 03:01, 21 January 2009 (UTC)

Re "possible that we decide that even though something is verifiable the due weight for it is zero", sadly that is occasionally necessary, because occasionally some real rubbish gets published. ---Philcha (talk) 08:23, 21 January 2009 (UTC)

I'd say not occasionally, but often. Lots of Wikipedia articles (more so historical and humanities topics) are worrisome because of sundry weaknesses in the mainstream sources they've been built from. Gwen Gale (talk) 08:27, 21 January 2009 (UTC)

^ [ref]
^ [ref]
^ [ref]
^ "Blogs" in this context refers to personal and group blogs. Some newspapers host interactive columns that they call blogs, and these may be acceptable as sources so long as the writers are professionals and the blog is subject to the newspaper's full editorial control. Where a news organization publishes the opinions of a professional but claims no responsibility for the opinions, the writer of the cited piece should be attributed (e.g., "Jane Smith has suggested ..."). Posts left by readers may never be used as sources.

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

[2] When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested should be provided as a courtesy to substantiate the reference.

[Meeker-Haldeman-3] Meeker WC, Haldeman S (2002). "Chiropractic: a profession at the crossroads of mainstream and alternative medicine" (PDF). Ann Intern Med. 136 (3): 216–27. PMID 11827498.

[4] Opinions involve both matters of fact and value; see fact-value distinction

[5] [ref]

[6] [ref]

[7] [ref]

[8] "Blogs" in this context refers to personal and group blogs. Some newspapers host interactive columns that they call blogs, and these may be acceptable as sources so long as the writers are professionals and the blog is subject to the newspaper's full editorial control. Where a news organization publishes the opinions of a professional but claims no responsibility for the opinions, the writer of the cited piece should be attributed (e.g., "Jane Smith has suggested ..."). Posts left by readers may never be used as sources.

[1]

[2]

[3]

[4]

[1]

[2]

[3]

[4]