Wikipedia talk:Verifiability

• See also: Wikipedia talk:Verifiability/Dates
• Archive 1
• Archive 2

Policy, guideline... both?

Why is this in both the official policy and guidelines categories, and have the guidelines header? Surely it's either one, or the other? This is at least confusing, if not out-and-out contradictory. Alai 06:18, 5 December 2005 (UTC)

I can't see that it is in both, Alai. Or perhaps it's already been dealt with. SlimVirgin ^(talk) 07:14, 5 December 2005 (UTC)

Indeed. The truly remarkable thing is that the edit to redirect the policy template was only just reverted... You'd think someone might have noticed we were apparently without (at least explicitly-labelled) policies for over two days... Much better now. Alai 07:44, 5 December 2005 (UTC)

Query

It's important to note that "verifiability" in this context does not mean that editors are expected to verify whether, for example, the contents of a New York Times article are true. In fact, editors are strongly discouraged from conducting this kind of research, because original research may not be published in Wikipedia. Surely OR would not include checking a seconday sources' primary sources ? Rich Farmbrough 16:25, 6 December 2005 (UTC)

• By the same token, knowingly publishing false information, that is verifiable from a reputable published source, is not considered "original research", despite willful failure to qualifying the source as "dubious". Wow. nobs 16:40, 6 December 2005 (UTC)
• There are at least two kinds of news article. A news summary is a simple compilation of information from other sources, and can be verified by checking those sources. (Indeed, this is exactly the sort of verification that editors do daily at Wikinews.) A news report, on the other hand, is what Wikinews terms original reporting, and is a firsthand account or interview, and primary source material. The only way to verify it would be to repeat the observations or interviews that the author performed. The difference is that between verification by reference to a publication and verification by repeating the experiment/data analysis/measurement/research. The latter is part of what peer review involves, and isn't what Wikipedia is organized or intended to perform. Uncle G 17:29, 6 December 2005 (UTC)
• I can't make head nor tail out of Nob01's response but if say a New York Times article cites other publications, it would be fine to look at those other publications. Should they differ, rather than engaging in a "search for truth" you should report that they differ. Fred Bauder 17:36, 6 December 2005 (UTC)

• I was trying to figure out how to respond, but Fred Bauder explained it well already. I just want to re-iterate the idea that it's not up to us to check whether sources are correct, it's just up to us to cite reputable sources. Friday (talk) 17:49, 6 December 2005 (UTC)

OK, so if a so-called "reputable source" is found to be dubious, nevertheless a provabely false insertion is made under the guise of "verifiability, not truth", that is still OK. nobs 18:17, 6 December 2005 (UTC)

Yes, that's okay. If the New York Times reports that the sun did not rise this morning, we write: "The New York Times reported on December 7, 2005 that the sun had not risen the previous morning." We might add: "This was cited by The Washington Post as one of a series of errors recently made by the Times." We do not add: "And one of our editors, Nobs, knew they were wrong because he saw the sun rise for himself." SlimVirgin ^(talk) 18:29, 6 December 2005 (UTC)

My question relates more to using the information at all, when a user knows it to be false. In otherwords, take out the portion about the Washington Post, yet nevertheless present the Times articles as "reputable" when the user inserting it knows it to be false. nobs 18:53, 6 December 2005 (UTC)

I don't know what your question is. As the page says, the criterion for entry into Wikipedia is verifiability, not truth, as counter-intuitive as that may seem. SlimVirgin ^(talk) 19:05, 6 December 2005 (UTC)

We do, of course, have the option to choose not to use a particular source if we think it's unsuitable. Maybe this helps address your concern? This is a bit of a judgement call - for example, we generally don't want people saying "We should never use the Times as a source, because they said >whatever< and it was wrong." On the other hand, we often choose not to use tabloids as sources due to their disreputability, but sometimes they may be an appropriate source. Friday (talk) 19:37, 6 December 2005 (UTC)

Let me give my own example: I do a search on Kiko Martinez, one result says he blew himself up in 1970, another makes reference to Kiko Martinez's widow, and other results cite Kiko Martinez attending rallies and lectures but does not give a date. So I write, "Kiko Martinez, who may or may not have blew himself up in 1970, attended rallies and lectures on unspecified dates", yet cannot attribute that to one particular source, would be considered original research. nobs 20:22, 6 December 2005 (UTC)

No you just say "According to X, Kiko Martinze blew himself up in 1970." As for Friday's comment: of course you personally can choose not to use a source you consider unsuitable. The real question is, what happens in someone else uses a source I consider unsuitable? We need policies and guidelines that can provide a common ground for me and the other person to discuss the appropriateness of the source, but ultimately, it is myself and the other person who have to seek some sort of compromise or accommodation. Slrubenstein | Talk 21:31, 6 December 2005 (UTC)

So I have a published source that says Kiko Martinez blew himself up in 1970; Fred Bauder says he sometimes chat's "with the same man at our local library", my inclusion is valid as WP:V, whereas Fred's is uncitable. nobs 21:48, 6 December 2005 (UTC)

What is important is not that you are making a claim about Kiko Martinez; you are making a claim about a verifiable source that makes a claim about Kiko Martinez. That is all we Wikipedians can do. Now, if Fred tells you "Hey, I talk to Kiko all the time" you should inform him that Kiko would do well to write to the New York Times or whatever paper reported the incident and ask them to retract it. If Kiko being blown up was worth reporting, then either (1) the fact that he is alive or (2) the fact that the source that reported his explosion is unreliable is also worth reporting. Ergo, you would be able to find a source that in one way or the other says that the first source was false. Slrubenstein | Talk 21:54, 6 December 2005 (UTC)

Nobs01 might check the Alamosa, CO phone book or DexOnline [1]. Fred Bauder 22:12, 6 December 2005 (UTC)

nobs did an extensive search of Martinez after Mr. Bauder's suggestion, and found a reference to his "widow", (note: not "x-wife"), and several other publicity bulletins about his attendence at various events, but I could never link any event to a specific date between 1970 and 2005. I believe this is about as far as Wikipedia allows without inserting mine, or Mr. Bauder's original research. nobs 22:37, 6 December 2005 (UTC)

I can't even find a WP article on Kiko Martinez - but it seems like a case of "are you talking about the same thing"? After all, there are at least two John Siegenthalers apparently of repute, and no doubt a handful more of no repute going by the same name. You need to be precise. For example, legal cases come with references - eg "941 P.2d 282, 283 (Colo. 1997)", and editions of books come with ISDN numbers - allowing this precision. Things can be confusing when you potentially have lots of things by the same or similar name, jguk 22:44, 6 December 2005 (UTC)

Well, what's happening is the issue is "verifiablity, not truth" on this page; whereas [2] and [3] the sourcing of proper and reputable citations are attributed to the poster as a "personal attack" against a priveleged expert. nobs 22:50, 6 December 2005 (UTC)

I'm really not sure what's going on here, but I'll go back to the question of verifiability and Fred's comment on his ArbCom candidacy page that reads:

Should you when informed that Francisco Kiko Martinez is alive and well, contrary to the information in your source, then investigate? After all, you can claim the falsehood is in your quoted source. Perhaps you should call and ask him at 1-719-589-6543. Fred Bauder 21:40, 6 December 2005 (UTC)

The answer from a WP:Verifiability perspective is "no". You should question how reliable your source is, and you should ask Fred to provide a published source (we, like other encyclopaedias, do not allow telephone numbers as references). There is also the question as to whether everyone is talking about the same Francisco Kiko Martinez (ie yourself, your source and Fred).

From a personal perspective, Fred may be right. You leave yourself open to being sued by libel for repeating and republishing an untruth, even though you have sourced it. The libelled doesn't have to restrict himself to going for your source.

From a Wikipedia perspective, Fred is the most influential member of the ArbCom and how he wishes WP:Verifiability to be interpreted is very relevant to Wikipedia. Ultimately what he says has loads more importance than what you say, regardless of the precise wording of the policy. Fred is one of those people who decides cases not on their precise merits, but rather on what side he wants to win, and he then picks an argument from the facts to get him there. There are loads of real-life judges who do exactly the same thing - I specialise in UK tax, and it wouldn't take too long to put together a list of UK tax cases that were decided more on whether the judges or Commissioners wanted the Revenue or wanted the taxpayer to pay. Personally, and I am a technically minded person, I prefer judges (and Arbitrators) who rule strictly on the merits of the case alone, without looking at wider issues - but unfortunately, in this world, we have to accept we don't always get them, jguk 23:17, 6 December 2005 (UTC)

I appreciate the comments very much, but let ask again, how does citing what a reputable, verifiable, publishied source, even if in error, constitute a personal attack by the user contributing the citation? This, presuambly, is what Talk pages are for, and this is why the material was raised in Talk, and not directly inserted, so it could be vetted. And the reference to Kiko's early demise is the only probable error I've discerned from an otherwise very credible source. nobs 23:26, 6 December 2005 (UTC)

That's a difficult question. You did not intend to make a personal attack, or to make a libel, but your edit may well have constituted a libel because you repeated a falsehood made by someone else. If you ask people the question "is a libel a personal attack?" they would say "yes". If you ask someone, if you clearly did not intend to make a personal attack in making a statement can it be a personal attack, the answer would be "no". The difference is in the intention. It appears from what you are saying that in answer to the question as to whether, in WP, you can break WP:No Personal Attacks despite you (even demonstrably) not having any intention to make a personal attack, the ArbCom has ruled "yes" - intention is irrelevant. I'm not sure I like that conclusion either, but that appears to be their conclusion - and, as noted above, it is consistent with how libel works, jguk 23:36, 6 December 2005 (UTC)

The falsehood (assuming the evidence) is minor relating to Kiko Martinez allegedly being alive, and does not directly relate to supposed libel. There has been no allegation of libel, and several issues relate there (1) being that it isn't libel if true, and (2) if proven libelous would only be applied to the party who originated the libel. But all that goes way beyond the scope of the current ArbCom case. Nonetheless, the Wikipeida Policy is verifiability, not truth, which if you read my first posting under this subhead [4], I expressed amazement for the intellectual dishonesty this invites. nobs 15:37, 7 December 2005 (UTC)

Nobs, you wrote above that because you have a published source that says Kiko Martinez blew himself up in 1970, including that in an article is valid under WP:V. Can you say what your source is, because that makes a difference? SlimVirgin ^(talk) 18:01, 7 December 2005 (UTC)

The source is (deja vu all over again) Wilcox, The Watchdogs: A Close Look at Anti-Racist "Watchdog" Groups, (Editorial Research Service, 1999), pp. 115-117. ISBN 0-933592-89-2 . nobs 18:12, 7 December 2005 (UTC)

Thanks, Nobs. Well, that provides an answer to your dilemma viz. your having a published source saying that X died in 1970 (meaning you can include it in an article in accordance with WP:V), while Fred says he has recently seen X alive. The answer is that Wilcox's book can't be used as a source about third parties because it's self-published. That puts it on the same level as personal websites and blogs. We can in certain circumstances use self-published material as a source about its author (e.g. we could use The Watchdogs as a source of information about Wilcox i.e. as a primary source), but we can't use it as a source of information about anyone else (as a secondary source). SlimVirgin ^(talk) 18:21, 7 December 2005 (UTC)

Thank you Slim. While still somewhat of a newbie after 10 months, I'm also somewhat of a slow learner (I used to ride the short bus), and will put those areas of Wikipedia policies & guidelines on my reading list regarding self publishing, etc., if you'd be kind enuff to direct my attention to where I can find them. Thank you. nobs 18:29, 7 December 2005 (UTC)

They're scattered around in several places: there's a reference in [WP:V]] to personal websites, there are others in WP:RS to blogs, personal websites and other less reliable sources, and there's a reference in WP:NOR about avoiding using sources that have no fact-checking procedures. I'm going to try to draw these together so that they're in one place and clearer, which might help to avoid this situation in future. SlimVirgin ^(talk) 18:33, 7 December 2005 (UTC)

<- Thank you. WP:RS#Reliability says, "exceptions may be when a well-known professional person or acknowledged expert in a relevant field has set up a personal website using his or her real name [5]. Even then, we should proceed with caution, because the information has been self-published, which means it has not been subject to any independent form of fact-checking;" which I read to say, it is not necessarily rejected out of hand, given the qualifiers; but this is exactly the sort of "fact-checking" necessary to determine if Kiko still has a pulse, etc. nobs 19:33, 7 December 2005 (UTC)

I don't know what you mean about fact-checking. That doesn't apply to Wilcox because he's not a recognized expert in any field. I know you think he is, but he isn't. He could certainly be used if published by a proper publisher, but not when he self-publishes. SlimVirgin ^(talk) 19:35, 7 December 2005 (UTC)

The Military Law Review [6] says he is; I got multiple sources say similiar (incidentally, Chip Berlet "was a Mencken Awards finalist", Laird Wilcox actually won it [7]). nobs 19:42, 7 December 2005 (UTC)

Perhaps I'm misreading it, but that says he was a finalist. SlimVirgin ^(talk) 20:01, 7 December 2005 (UTC)

You've quoted the Military Law Review many times. But who other than them says he is? SlimVirgin ^(talk) 19:59, 7 December 2005 (UTC)

Nobs is neglecting to disclose that the personal attack involved was on Talk:Chip Berlet, see Wikipedia:Requests_for_arbitration/Nobs01_and_others/Proposed_decision#Disruption_by_Nobs01 (you will have to look in the history as the personal attacks have been removed). The problem is that there is only the most minimal nexus between Kiko Martinez and Chip Berlet. In the time frame that they were both in Denver Kiko was accused of a crime. Chip, a private investigator may have worked on the case. But Nobs contention is that since both were members of the National Lawyers Guild that Chip Berlet is somehow responsible for the alleged crimes of Kiko Martinez. That sort of stuff is why a one month (or one year) ban is proposed at Wikipedia:Requests_for_arbitration/Nobs01_and_others/Proposed_decision#Nobs01_banned_for_one_month Fred Bauder 20:02, 7 December 2005 (UTC)

Fred - just out of interest, do you really know either Chip Berlet or Kiko Martinez - and is it just a coincidence that Martinez lives near you or is that connected with Nobs01's enthusiasm for the matter? jguk 20:10, 7 December 2005 (UTC)

Neither Chip Berlet nor I are sure we met, but we might have. Kiko Martinez and I both practiced law in Alamosa Fred Bauder 20:16, 7 December 2005 (UTC)

nobs never made any personal contention regarding anything; nobs inserted a direct citation from Arleigh McCree, A Case For Self-Defense, Military Police (Summer 1981), quoted in Laird Wilcox, Political Research Associates: A Study in "Links & ties", subchapter Chip Berlet and the National Lawyers Guild, (Editorial Research Service, 1999). As SlimVirgin raised at the time, and now reviewing the applicable policy (WP:RS#Reliability) "we should proceed with caution, because the information has been self-published, which means it has not been subject to any independent form of fact-checking." This is one reason nobs used the Talk page (it now appears obvious, too much material, too fast). Now we have discovered an error in the original self-published source. Beleive it or not, the process works.

Let me add, the whole matter has now been carried farther then ever intended, and I have no plans to spend restless days and nights plotting to expose Chip Berlet for all sorts of things. This has been an intrusion into my ordinary work of researching matters from an earlier period, and I have no interest of writing American history from the 1960s & 1970s, which, Mr. Chip Berlet it appears, has played a part in shapping. nobs 20:24, 7 December 2005 (UTC)

Hopefully ArbCom will take note of that and allow you to move on without a ban, jguk 20:36, 7 December 2005 (UTC)

Well, yes, the process does work, but normally the process is to edit in accordance with our policies and guidelines, which caution against using self-published sources. If Nobs had paid attention to that, this entire thing could have been avoided. SlimVirgin is also wondering why Nobs keeps referring to himself in the third person. SlimVirgin ^(talk) 20:28, 7 December 2005 (UTC)

Thank you. This is a good lesson in why we have Admins. As to the 3rd person, I don't know, I guess it became a habit I picked up studying 19th Century German poetic texts. nobs 20:53, 7 December 2005 (UTC)

Verifiability does not mean "potentially verifiable"

This is already covered in the sentence

It is unreasonable to expect other editors to dig for sources to check your work, particularly when the initial content is questionable. The burden of evidence lies with the editor who has made the edit,

Nevertheless.

I would like to have specific language that makes it clear that "potentially verifiable" information is not good enough.

Specifically: "verifiability" requires information to be

accompanied by a source citation that is quoted, paraphrased, or summarized faithfully

and is not met by

An unsourced piece of information, even if there is no specific reason to doubt it, and even if one could potentially find a source for it by applying well-known research tools.

In other words:

MacShane wrote that John O'Hara's "Appointment with O'Hara" columns in Collier's Magazine were "garrulous and outspoken" and says neither "added much of importance to O'Hara's work."

MacShane, Frank (1980): The Life of John O'Hara. Dutton, New York

satisfies the verifiability policy (even if you do not have a copy of MacShane's book in front of you and even though what MacShane says is clearly opinion; the verifiable fact is that MacShane said so).

On the other hand:

The Saigon Cafe is a cafe in Duluth, Minnesota

does not satisfy the verifiability policy, even though one could potentially verify it by consulting an online Yellow Pages. Dpbsmith (talk) 19:24, 12 December 2005 (UTC)

Unless written as "The Saigon Cafe is located in Duluh, Minnesota^[8]

^ Yellow Pages, Minnesota 2005 edition, Yellow Book USA, Inc.

≈ jossi fresco ≈ t • @ 19:43, 12 December 2005 (UTC)

Sure. That works. Assuming good faith of course, i.e. assuming that the editor wouldn't put that in without having actually looked it up. (If a specific edition is being cited, BTW, I don't see why the editor shouldn't cite the page number). But that's true of any reference. Dpbsmith (talk) 19:46, 12 December 2005 (UTC)

This works a little better for me: [9]. The point is, the editor who inserts the information is expected to provide the source, particularly if challenged. Dpbsmith (talk) 19:48, 12 December 2005 (UTC)

Another example:

Adam Ant Has spoken openly on television about being bipolar

does not satisfy the verifiability policy, even if the editor saw and heard him talking about it on television. If the editor saw him on television, then it is potentially verifiable because there is almost certainly a recording of that program somewhere, and it may even have been described, but it does not satisfy the verifiability policy until the source citation is actually found. (To end the suspense: [10]).

Dpbsmith (talk) 19:45, 12 December 2005 (UTC)

Bleh. Perhaps I'm misunderstanding what you're getting at, but given the current state of Wikimarkup, such a requirement would turn this from an encyclopedia that anyone can edit into the unreadable and virtually uneditable (except by the most fanatical) encyclopedia. older≠wiser 20:41, 12 December 2005 (UTC)

Either you're misunderstanding me, or else I'm misunderstanding the current policy. I believe I'm talking about a "requirement" that already exists, not a new requirement. Dpbsmith (talk) 21:28, 12 December 2005 (UTC)

I value verifiability as much as anyone at Wikipedia. But I tend to agree with older≠wiser. "Potentially verifiable" is at best redundent, possibly just meaningless. Rather than say something like "You can't add information unless you could back it up in the future," I think it makes much more sense to say something like "If any editor feels that information in the article may be incorrect, they have the right to challenge the person who added that information to provide a source." I don't like the way I phrased either of these choices, but you get the idea. Slrubenstein | Talk 21:00, 12 December 2005 (UTC)

I obviously completely failed to express myself clearly. I've not infrequently seen people complain about "removing potentially verifiable" information. My understanding is that the standard is "verifiable," not "potentially verifiable." Once challenged, if a verifiable source citation isn't coming within a reasonable period of time, the information should be removed... probably placed on a Talk page... and kept there until someone, preferable the contributing editor, provides the citation. Dpbsmith (talk) 21:28, 12 December 2005 (UTC)

I agree Slrubenstein | Talk 05:42, 13 December 2005 (UTC)

Yes, I agree completely. I think I got confused by some of the interspersed comments about footnoting a phone directory, which IMO is not of much use unless the existence of said establishment is genuinely challenged. But even so, I'd really hate to see stuff like that footnoted. That would make many people even more inclined to completely ingore footnotes (that is, I think footnotes should be used to provide substantive elaboration that doesn't fit into the flow of the text rather than to provide citations for tedious trivialities). older≠wiser 21:47, 12 December 2005 (UTC)

I would like to see the policy made more explicit. I see too many editors in deletion discussions saying an article should be kept because it is "potentially verifiable", even after other editors have complained about not being able to find sources to verify the information in the article. I think it is the responsibility of every editor to cite sources for everything of substance they add to an article. This does not mean footnoting everything, it does means citing your sources. -- Dalbury^(Talk) 23:06, 12 December 2005 (UTC)

I'm going to disagree. Literally the majority of Wikipedia is accurate but lacks explicit citation. I think that being hardassed about this is an invitation to mass deletion and (because not everyone will accept this) edit warring. I'm probably as good as anyone here about citing, and still I'd venture that half of what I write lacks explicit citation.

Let me give a tangible example: do you think we would be better off without Thermidor, which as far as I can tell is entirely accurate, but lacks any explicit references? -- Jmabel | Talk 06:54, 13 December 2005 (UTC)

And how can you tell Thermidor is accurate? How can any any reader of Thermidor judge the reliability of the article? Given that anyone can add anything they want to WP, the authority of articles in WP has to come from the sources used in writing the article. If you write an article without checking sources, you are depending on your memory, which may be incomplete, or even wrong. And we cannot rely on all editors to add only facts. When I start or add to an article, I always cite references. -- Dalbury^(Talk) 11:25, 13 December 2005 (UTC)

• (Interjecting belatedly, because I presume the question was addressed to me.) Because I know the topics in question, and the only things there I'm not absolutely sure of from my own knowledge is the detail about th particular chef who may have created lobster Thermidor, and the remark about Monty Python. -- Jmabel | Talk 06:45, 14 December 2005 (UTC)

And how can I tell that the article on Thermidor in the Encyclopedia Britanica or Encarta is correct? I can't. There are no citations provided for specific facts (and even if there were, I probably would not look them up unless I happened to become interested in delving into a topic in more detail). There is an element of trust that these publishers have a process for vetting the accuracy of their articles. A persistent criticism of Wikipedia is that the information is not trustworthy because there are few controls over who can add/edit information. But that is in a way the very premise of Wikipedia, that over the long haul, having hundreds or thousands or millions of reader/editors providing input will eventually result in a comprehensive and high quality encyclopedia. Some people instinctively distrust this premise.

Back to point, the mere presence of a citation is no guarantee that the cited reference actually supports the fact. I think it is extremely misleading and dangerous to promote the illusion to general readers that the mere existence of citations = accuracy. What is needed is a more reliable process for vetting articles. IMO, a simplistic "requirement" that citations always be provided is not the answer. Checking citations may be a partial step in such a process, but the presence of citations alone are not sufficient. I'd fully support having a separate page for editors to provide verified citations of factual content (the talk pages partially serve such a function, but talk pages are frequently archived and past discussions become forgotten as old editors leave and new editors come along). But I very strongly oppose the approach of some to require that every fact be cited within the article. older≠wiser 14:04, 13 December 2005 (UTC)

At least with citations you have somewhere to start in verification. If an editor is not just simply pulling 'facts' out of his or her head, it should be easy enough to cite his or her sources. My feeliongs on this have grown out of seeing too many editors argue that an article up for AfD should be kept because it is "potentially verifiable", even though no sources are cited, and other editors have reported having problems finding verification. An article is "verifiable" only if there are sources that can be checked. If you can't cite publicly available sources for everything you say in an article, it's "original research". -- Dalbury^(Talk) 17:47, 13 December 2005 (UTC)

I'm not completely disagreeing with you. Especially with new articles, it may not be unreasonable to insist upon some verifiable source to warrant keeping the article. I just think we'd need to phrase such guidance carefully to avoid inciting a pogrom for the removal of any and all uncited facts. Also I'm not sure that we should always require that the original contributor provide the citation--cerainly it'd be nice if all contributors would do so. I mostly avoid AFD as it is largely a rather unpleasant aspect of Wikipedia, but I think "potentially verifiable" is a pretty lame reason for keeping an article--either an article is verifiable and someone does a little bit of digging or it is not (or no one cares about it enough to bother digging, which IMO amounts to the same thing). Perhaps there could be some sort of staging area for such "potential" articles and if no one bothers to improve them within a period of say a month, then they can be deleted. That way they'd have a chance beyond the relatively quick turnaround periond of AfD, but still have a limited shelf life. older≠wiser 18:32, 13 December 2005 (UTC)

I've not nominated any article for Afd, or voted for deletion, simply because it lacks cited sources. I have started tagging articles with {{Unreferenced}}, but that doesn't seem to actually accomplish anything other than warning readers the articles has no cited sources. I would be in favor of some process that raises the profile of these articles, but not necessarily AfD. The deletion process is already overwhelming, and doesn't need whole new classes of articles being nominated. -- Dalbury^(Talk) 18:43, 13 December 2005 (UTC)

I'd suggest that much more useful than tagging entire articles as {{unreferenced}} is tagging specific claims that you doubt with {{fact}}. Let's fish where the fish are: put the fact-checking work into what someone actually thinks is likely to be wrong. -- Jmabel | Talk 06:47, 14 December 2005 (UTC)

I agree about asking for citation for specific points rather than tagging the whole article. SlimVirgin ^(talk) 17:55, 15 December 2005 (UTC)

Responding to "the mere presence of a citation is no guarantee that the cited reference actually supports the fact", above: Yes, this is certainly a problem in academic writing. I have found many cases in my work where a statement in a journal article is followed by a citation. Too often, the cited work has only a loose association to the statement it supposedly supports. However, the statement looks more trustworthy with a citation. I feel that in a case like this, it would be preferable to have no citation at all, rather that creating this illusion for the reader. ike9898 14:58, 14 December 2005 (UTC)

Now is not the time

This is not the time to start trying to force editors to use footnotes. The Siegenthaler incident has highlighted Wikipedia's poor use of sources, and there's more interest than ever now in encouraging editors to provide sources for their edits, or at least for any that are even slightly contentious. What we have to concentrate on is persuading editors to provide any sources at all, rather than fussing about the precise form they have to come in, and especially not one as complicated as SEWilco's.

Yes. There is a pattern of the policies and discussions about sources being totally out of sync with the reality of how people source. I don't have facts or figures, but my observation is: People don't source. However, WP:Cite sources says "For the reasons listed above, if you add any information to an article, you must cite the source of your information." (emphasis in original). It's silly to have a rule respected by so few. Stevage 17:28, 16 December 2005 (UTC)

In any event, he is just wrong about embedded links providing less information than sources. When using a footnote system, you're meant to add the number after the sentence, then go to the Notes section to add the full citation information. When using embedded links, you add a numbered URL after the sentence, then go to the References section to add the full citation information. Therefore, both styles, when used correctly, provide exactly the same amount of information.

I find his edit warring about this very tiresome. It has been going on for months over multiples articles, policy pages, guidelines, the Village Pump, at least one RfC, and now RfAr. We are not going to stop editors from using embedded links. We're going to encourage their use, because we want to encourage any source citation at all. SlimVirgin ^(talk) 16:11, 15 December 2005 (UTC)

Wikipedia already accepts embedded links as a minimal source, just as a book title alone is accepted. It is expected that citations will be improved, just as it is expected that everything else in an article will be improved. (SEWilco 19:59, 15 December 2005 (UTC))

I have to agree that edit warring over format and requiring a certain format when there is no consensus is silly. But I would also have to say that going and changing a bare external link into a footnote with added title and author information is an obvious improvement. And SlimVirgin, if you don't wan't edit wars, why are you going around on articles with no controversy and removing that extra information that SEWilco's bot has added? I refuse to get dragged into this case/conflict and I would prefer we instead just separate out the issues and make common sense decisions. More information in citations is to be encouraged, though not required. Enough said. I don't agree with a lot of SEWilco's edits, but I am able to separate the person from the edits. When they make sense, we should support them. - Taxman ^Talk 20:21, 15 December 2005 (UTC)

Taxman, I'm not removing information. I'm restoring embedded links that he removes. If he wants to add citation information to the References section, that'd be most welcome and he'll never find me removing it. But that's not what he wants. He wants Wikipedia to outlaw embedded links, and he's been editing warring on policy pages and guidelines for months, misrepresenting what one says on another, adding misleading information then edit warring to keep it in. WP:CITE says pages shouldn't be changed from one style to the other without consensus, and I'm asking that he stick to that. SlimVirgin ^(talk) 21:51, 15 December 2005 (UTC)

Citation format poll

The poll has been moved to Wikipedia talk:Verifiability/Citation Poll.

/originally by SlimVirgin ^(talk) 16:26, 15 December 2005 (UTC) / comment edited for new location Mozzerati 20:15, 15 December 2005 (UTC)

You should wait until discussion is over before moving stuff. The fact it is "taking up the whole page" is not a clue to you, participant SlimVirgin, that it is an active discussion? (SEWilco 19:55, 15 December 2005 (UTC))

No, it was you who put it up. As you've asked so many questions, it's more appropriate on a subpage. All that matters is that people who want to take part know where to find it. SlimVirgin ^(talk) 21:43, 15 December 2005 (UTC)

Biographical material

I'm going to try to add something about the importance of using good sources when it comes to biographical material about living people, following the Siegenthaler incident. SlimVirgin ^(talk) 17:56, 15 December 2005 (UTC)

Not everything can be verifyalbe from "reputable sources"

Lets say I created an article on Zeliard, a video game from 1990. Under common WP conventions, video games that were developed, published and sold, usually are considered notalbe enough for an article.

During the description of the game, I say, "the top boss in the game is called Jashiin".

Now, this statement is:

A. True.

B. Verifiable. Anyone who owns and had played the game can verify the truth of the statement.

However, no notable newspaper has ever published an article about Zeliard, so there is NO WAY to get "solid references".

So we have a notable article, with a true and verifiable statement, which, however, was never referenced.

About 90% of the article's length (pretty much anything except the infobox containing data that probably can be verified somewhere, such as corporate press-releases), covers in-game plot, theme, and characters, which were never described in press.

Does that mean the article on Zeliard should be thrown out alltogether?

This question is critically important to the Verifiability policy. Its not just about Zeliard, its pretty much about ANY article statement which is true and verifiable, but is not documented in press.Elvarg 21:46, 15 December 2005 (UTC)

• These details are likely to be in the instructions, or in some other material published by the manufacturers. SlimVirgin ^(talk) 21:42, 15 December 2005 (UTC)
  • Maybe they are, maybe they arent. OK, lets say I say "the third level containst a forest-like map". This statement is also true and verifiable (and its truthfulness is so trivial and noncontroversial that it cannot be labelled "original research", yet it probably is not mentioned in the manual or anywhere else.) Now what?Elvarg 21:47, 15 December 2005 (UTC)
    • It almost certainly has been published somewhere by the manufacturers. If you're certain it's trivially true, add it to the article, but if someone challenges your edit, you'll have to find a source. If no one challenges it, and so long as you're certain it's something not likely to be challenged, there'll be no problem. SlimVirgin ^(talk) 21:53, 15 December 2005 (UTC)
      • This doesn't solve the problem. The statement is trivially true, and nobody has a reason to challenge it on grounds of having a different opinion. However, someone may still want to challenge it just for the sake of challenging it (anything can be challenged).Elvarg 21:56, 15 December 2005 (UTC)
      • If the game is notable, you should be able to find a third-party description of it, if only on the Internet. Superm401 | Talk 22:04, 15 December 2005 (UTC)
        • 1. Even famous games (especially older ones) may not have press releases about them. And not only games. Notability is a subjective meaning. Wikipedia is not a paper encyclopedia -- and as such, it can and should include articles that may not be notable enough for press releases, but that would be interesting to a sufficient number of on-line readers.
          • Press releases are unnecessary. Superm401 | Talk 22:14, 15 December 2005 (UTC)
        • 2. The Internet is not considered a solid referential source anyways. Elvarg 22:11, 15 December 2005 (UTC)
          • As I explain below, we accept all sources on a case-by-case basis. We do not and should not discard internet sources wholesale. Superm401 | Talk 22:14, 15 December 2005 (UTC)
            • What's to stop me from creating a web page and putting facts on it, then writing a WP article "referencing" facts from that page? Yes I know its against the rules, but its impossible to enforce. Or if you want to stay legal, pretty much any subject has SOME supporters, and its not hard to find a webpage to support pretty much anything... The point is, that by itself, a reference from a web page is not any more solid then a statement from wikipedia -- without credibility, they both are worth zilch. In fact, I'd argue that a wikipedia statement is usually more credible, since it is subject to peer review, unlike web pages. Elvarg 23:08, 15 December 2005 (UTC)
              • That's comparable to writing an article and putting it in a xeroexed newsletter you control, and then citing the "publication" as a source. I think very few editors would accept that as a "reliable" source. There will always be disagreements about whether a particular source is "reliable", and there will not always be consensus, but I can't think of a better system than the collective judgment of Wikipedians on the reliability of a source. -- Dalbury^(Talk) 23:28, 15 December 2005 (UTC)
                • Then we are back at square one, agreeing on a fact that accepted sources have to be reputable. And reputable sources may well not consider an article not notable enough for print, yet it IS notable enough for Wikipedia (refer to Not a Paper Encyclopedia). Elvarg 23:37, 15 December 2005 (UTC)
                  • We use Web sources that we judge to be reliable when there are no reliable printed sources readily available. We apply the same judgements of reliability to Web sources that we apply to paper sources. Of course, that still leaves holes, particularly for topics that are well-known to English speakers. As I see it, we need to make a good-faith best effort to document, from whatever source, the content in Wikipedia articles, always subject to challenge from other editors. If we can't point to a reasonable source for the content, we shouldn't put it in. -- Dalbury^(Talk) 00:02, 16 December 2005 (UTC)
• Aren't we forgetting that the game itself is a sort of "publication" -- it is copyrighted material. I don't see how this is much different from someone writing a plot summary for a novel -- either there is prima facie evidence to support the assertions or there is not. I know some people have a bugbear about citing primary texts, but when the subject is the "text" (or game, movie, TV show, etc.) itself, it is kinda hard to avoid. older≠wiser 22:00, 15 December 2005 (UTC)
  • Good point, Bkonrad. The game itself is a publication. SlimVirgin ^(talk) 22:16, 15 December 2005 (UTC)
    • Actually, this can be problematic. Yes, the game is a publication, but what if I say "the level 8 boss is named Alguien"? Its also true and verifiable, and the game is a publication, however its not mentioned in the manual, and getting to level 8 would take a considerable amount of effort and skill, hardly justified for the sole purpose of verification of a statement. It's incomparable to a book, where you can just open it at page 857 without reading the whole book first.Elvarg 22:24, 15 December 2005 (UTC)
  • Yes, but when summarizing a novel, I can always refer to the novel itself as a reference (refer to a page number, ISBN number, etc). I can't do it say in a video game.
  • Also, the discussion always seems to go about finding workarounds to the conditions stated. I'd rather deal with the concept. Maybe a video game, as a "publication", was not a perfect example. The question remains -- does Wikipedia allow facts which are true and verifiable but have never been documented by anything that can be considered a reference?Elvarg 22:04, 15 December 2005 (UTC)
    • If you are describing a valid concept, you should be able to find many valid examples. To answer your conceptual question, we don't allow what you just asked about. We permit information that is verifiable through secondary sources. Secondary sources do not necessarily need to be "academic" or from scholarly journals or the like. We judge on a case by case basis whether the secondary source is trustworthy. However, it is vital that you can verify it by a means other than "seeing for yourself". Just as we would not include information in China that could only be verified by visiting it, we do not accept information about publications or software that can only be verified by using them. Such information is original research, which you probably know is specifically forbidden. Superm401 | Talk 22:12, 15 December 2005 (UTC)
      • Please don't make it personal. I could spend several hours coming up with a dozen examples, but I'd rather focus on the concept. And now we came close to the source of the problem -- the acceptance of verifiable yet unfeferenced facts, which obviously also involved the definition of original research. Do you say that ANY statement which was never before published is considered original research? Elvarg 22:19, 15 December 2005 (UTC)
        • Do you say that ANY statement which was never before published is considered original research? Although this wasn't directed at me, I'll answer "no". There are two components to the phrase "original" and "research". To consider commonplace observations as "research" is IMO stretching the notion of "research" beyond the limits of sensibility. older≠wiser 22:30, 15 December 2005 (UTC)
          • This is my opinion as well. I think that "no original research" cannot be applied to trivial statements. For example, the statement "488837 + 399498 = 888335" is probably never stated in amy book exactly, and thus is NOT REFERENCED, yet it is NOT original research, as there is no research component to it (it is true, verifiable and trivial. Whether its notable or not is another matter). Elvarg 22:35, 15 December 2005 (UTC)
            • The problem is, any idiot may write nonsense in an article, and argue for its inclusion saying that it is trivial. We need to clarify "trivial". Borisblue 23:02, 15 December 2005 (UTC)
              • Whether its trivial is irrelevant; we shouldn't allow original research regardless of how trivial. The page WP:NOR, however, makes clear that the equation you list above would not be original research. It represents a synthesis of available published data (e.g. from a math textbook). We can publish such syntheses so long as they do not create a "novel narrative or historical interpretation." Clearly, this simple equation creates no such novel interpretation. Christopher Parham (talk) 23:04, 15 December 2005 (UTC)
              • Another point: in response to your example if the video game hasn't spawned a mention of the Boss' name in a strategy guide, manual or something, then my take on it is that the Boss is not notable enough to mention. Borisblue 23:08, 15 December 2005 (UTC)
                • "A Wikipedia entry (including any part of an article) counts as original research if it proposes ideas". The keyword is IDEAS. A statement of trivial fact is NOT AN IDEA, and hence cannot be original research (IMHO). Elvarg 23:13, 15 December 2005 (UTC)
                  • We have to include facts in that definition I'm afraid, otherwise we'll have nazis putting in articles that 6 rather than 6 million jews died in the holocaust. The statement "6 jews died in the holocaust" is a "fact". An unverifiable fact, but a fact nevertheless. WP loses nothing from omitting that boss, if he's not worty of a mention anywhere else. In fact, the first paragraph of this policy mentions "Wikipedia should only publish material that is verifiable and is not original research. One of the keys to writing good encyclopedia articles is to understand that they should refer only to facts, assertions, theories, ideas, claims, opinions, and arguments that have already been published by a reputable publisher." Borisblue 01:13, 16 December 2005 (UTC)
                    • I disagree. First of all, since 6 million jews were killed, 6 have also obviously been killed (it's the "if I have a million dollars, then I have a thosand dollars" argument), and hence its true, verifiable, and dubiously obvious. Secondly, even if there was a statement by nazis that "ONLY 6 jews have been killed", then its neither true nor verifiable (evidence of at least 7 killed would make the statement verifiably false, and there is more then sufficient references to back that up as well). You may have a point, but this argument does not validly convey it, since neither of your statements exibited a situation of a statement which is true and verifiable but not referenced, and hence did not deal with the issue at hand. Elvarg 18:31, 16 December 2005 (UTC)
  • On another note, you are free to cite a video game as a source for information, so long as it is a reliable source for that information. (e.g., [11] which has style information for citing software/video games) However, if the information appears in an article about the game itself, you usually don't actually need to cite the subject of the article as a source; if there are multiple editions or translations, you may want to cite the one you used. Christopher Parham (talk) 23:18, 15 December 2005 (UTC)
  • I've had this conversation a few times, but never in so much depth. It seems to me that although it's never really stated, there are cases of verifiable, true statements which are not original research, which can however, not be included, and this is one of them. IMHO, there is nothing "original research" about playing a computer game and documenting the names of each of the levels, including secret levels. I doubt the game's documentation will document the secret levels at all. There may very well be no reputable sources that have written about these levels. Most sites that go into this much detail are run by kids and are not "reputable". Magazines that do reviews rarely go as far as secret levels. And lastly, the information is "verifiable", but it could be really, really hard to verify. For example, in Sega Rally, if you beat a particular time for the world championship, a secret track is revealed. Now that the game has been around so long, you could find that information easily. But not 10 years ago. And not for tiny games which may otherwise be notable. Stevage 17:24, 16 December 2005 (UTC)
    • Case in point. While I (as a person who never played Sega Rally, could raise doubts about the truthfullness of your sentence, which you cannot backup through references, I CANNOT SAY your statement constitutes "original research". The point is true, verifiable yet unreferenced sources are not ALWAYS original research (are in some cases, not in others), and in those cases where they AREN'T, there needs to be another policy/guideline to determine the criteria by which they should be included or not. Elvarg 18:31, 16 December 2005 (UTC)

Two meanings for "verifiable" are being confused. A source is verifiable if it is easy to check the source and make sure that it really says what the person citing it says it says. If someone quotes "The New York Times" that's a verifiable citation. If someone quotes a personal communication, or something they saw themselves, that's not verifiable.

The essential point is that, unlike scientific papers and traditional encyclopedias, the accuracy of Wikipedia's material does not and cannot rest on the authority of its contributors. Therefore, it has to rest on someone else's authority. Therefore, in principle, everything in Wikipedia needs to be referenced to a source.

It is the source that needs to be verifiable, not the fact.

Due to the existence of public libraries, print materials are verifiable. Also, unlike web references, they are durable--the same book is likely to be available from libraries five years from now--and the expense of publishing in print constitutes some kind of mild sanity check. Due to the ease of checking web references, web references are verifiable.

Personal observation of the contents of a video game might be verifiable if one could show that there were many public libraries that had copies of that game available for circulation. But it is much better to find a description in a game magazine, because that is more verifiable than the game itself.

An unreliable source can be verifiable. What's important is to state the source. The reader can judge its reliability. Furthermore, in cases where a source is obviously non-neutral, it is not inappropriate to make a comment to that effect, provided of course that the comment is neutral and, if necessary, sourced.

For example, in a discussion of whether chocolate is healthy, it would be perfectly reasonable to cite Harvard researcher Norman Hollenberg to the effect that it contains flavanoids which may reduce high blood pressure, and equally reasonable to note in the citation that Hollenberg's studies have been funded by the American Cocoa Institute and by Mars, Inc. Dpbsmith (talk) 18:54, 16 December 2005 (UTC)

Reputable and authoritative sources

Possibly we should extented the discussion of dubious sources to recognise that certain sources can be authoritative for certain claims: here, we would not allow that the video game counts is generally reputable as a source, but the game provides an authoritative basis for claims about for the name of anotagonists encountered in it. Similarly, we may not trust the government of Zimbabwe to be generally truthful, but it is nonetheless, the ultimate authority for what the law of Zimbabwe is. --- Charles Stewart 19:07, 16 December 2005 (UTC)

• Exactly. The source has to be evaluated based on the information you are citing to it; any publication is likely to be a reliable source in some cases, and not reliable in other cases. The government of Zimbabwe might not be a reliable source for an assessment of the fairness of elections in Zimbabwe, but for other data, e.g. weather information for Zimbabwe's airports, it might be the most reliable source of all. Christopher Parham (talk) 21:52, 16 December 2005 (UTC)

another question

How does Verifiability work in this instance:

Russ Bellant of Political Research Associates writes an report. David Lee Preston of the Philadelphia Inquirer writes about the report. Chip Berlet of Political Research Associates quotes the Philadelphia Inquirer article about the Political Research Associates report. Mr. Berlet states,

"The Philadelphia Inquirer runs an article by David Lee Preston ... cites the Bellant report which describes how the Republican Party has been recruiting ethnic facists, racists and anti-Semites for over 20 years, through its Heritage Groups Council.

Is this considered "verifiable"? [12] nobs 23:12, 15 December 2005 (UTC)

yes Borisblue 01:14, 16 December 2005 (UTC)

That's how it works. SlimVirgin ^(talk) 01:22, 16 December 2005 (UTC)

The short answer is "Yes". A longer answer is that it is a perfectly valid citation, though if one doubted either Berlet's or the Inquirer's word, it would be natural to seek out the original report, and if it contradicted the report there are several ways to go depending on the topic at hand. One could just ignore Berlet and the Inquirer and cite Bellant, or one could overtly discuss where in the chain someone misquoted their source. Also, typically, in academia at least, if one got the information via Berlet and later went to the primary source, one would normally still acknowledge Berlet. Because our purpose is a bit different than an academic's—we are not specifically trying to show the originality of our research—some might argue that once the primary source has been checked, we could drop mention of the (in this case) tertiary source. I'd be more inclined to keep it and cite both: I think it is useful to indicate whose pathway we are following through an article. This is a lot like when we translate an article from a non-English Wikipedia, we overtly acknowledge that article as a source, as well as citing its sources. -- Jmabel | Talk 01:56, 16 December 2005 (UTC)

There are two questions here, (a) a self-publishing source (b) quoting a secondary source that is quoting the same self publishing source (may not even qualify as a tertiary source). Plus the axiom "exceptional claims require exceptional sources". The same paragraph even states,

"most major media drop the story. The charges in Bellant's report are not covered in the New York Times, Washington Post, Associated Press, or United Press International."

This may be a text book example of how defamatory information is manufactured and spread with phraseology like "linked in published accounts " [13], and how these "published accounts" proliferate. This deliberate abuse may not even be apparent to users acting in good faith.

My only suggestion is that the series of tests beginning with Authenticity and Provenance, and continuing through the rest of that article, be incorporated into WP:RS & WP:V to eliminate fraud like this. nobs 18:16, 16 December 2005 (UTC)

Tools to assist verifiability

I would like to propose that the MediaWiki interface be improved somehow to make it easier to cite the source of edits, and that the requirements of sources be relaxed to encourage people to give whatever source information they have. Information with shaky source information is better than no information, or information with no source at all. At least if there is a shaky source, someone can make the informed decision to remove it, or source it properly. So, I propose that when an edit is made, there be a mechanism to state the source, as one of: (from most valuable to least valuable)

1. Article or website with full citation
2. Article or website with just URL or name of work
3. Vague idea of source (eg, NY Times article in February, somewhere on Slashdot)
4. Presumed common knowledge backed up by direct observation
5. Presumed common knowledge
6. Believe to be true

Currently we ask people to cite their sources, and provide no assistance in doing so. Footnoting is *hard* for a beginner. The least we can do is give them a multiple choice question to work from.

Comments please! Stevage 17:35, 16 December 2005 (UTC)

I'd like to second this. It would be excellent if under the "submit" text window there was a form (just a few titled text entry boxes) asking for the title of the - book, authour, website, etc - and that way WikCodei could take care of the formatting depending on the info and even start a "References" section if there isn't one yet. Referencing as it stands is a pain and the guidlines are far from clear. I had to flip through three sections to find how to correctly cite web pages, for example. ThePromenader 09:57, 17 December 2005 (UTC)

The reason for the complexity is that there are different ways of citing sources, and editors can't agree on which is best, so all three styles have to be described. But the page also makes clear that formatting is less important than providing some information about where you found the material. SlimVirgin ^(talk) 22:26, 17 December 2005 (UTC)

Previous books now refuted by their own authors on their homepages

I noticed that this policy is quite dismissive of homepages, but what about homepages of authors that refute their earlier works? It would be strange if Wikipedia could use the published books by these authors as a reference, but not the homepages on which the very same authors under their real names refute their own books. See Talk:Sathya_Sai_Baba#Abuse_of_referenced_links Andries 11:19, 17 December 2005 (UTC)

I think the author would be a 'well-known person' or 'expert' in my reading of such. —Matthew Brown (T:C) 22:01, 17 December 2005 (UTC)

I have also asked the question at Wikipedia_talk:Reliable_sources which seems a more suitable place for it. Andries 22:26, 17 December 2005 (UTC)

Religioustolerance.org

Please see the discussion on Wikipedia talk:Verifiability/Religioustolerance.org and Wikipedia:Verifiability/Religioustolerance.org as to whether Wikipedia should cite the religioustolerance.org website, jguk 13:57, 17 December 2005 (UTC)

Now what are we going to do, following this example? Are we going to create Wikipedia:Verifiability/University_of_Virginia_religious_movements_homepage too? [14] This seems like a good reference but is somewhat dubious because its entries on indivual movements were mainly written by students and as a result have strongly varying degrees of quality. The website is used in several articles in Wikipedia. We could create some more of these subpages. (I do not oppose them). Andries

An invitation to edit warring

"Any edit lacking a source may be removed"? I think this is an invitation to edit warring. I also think it's a terrible policy, just in general. For example, United States strikes me as fundamentally accurate, but largely unreferenced. Are we really saying that it would be appropriate to delete nearly all of the text of the article?

This strikes me as a wild overreaction to certain recent problems we have had. It is one thing to demand sources when one is genuinely in doubt. It is another to remove material that is known or believed by you to be true, merely because it is uncited. People are very likely to do this because they find particular material politically inconvenient. -- Jmabel | Talk 19:38, 17 December 2005 (UTC)

I was going to agree with you, until your last sentence; "because they find particular material politically inconvenient." - Isn't this a form of being "genuinely in doubt"(remembering WP:AGF)? It's a bad thing to remove material that you believe is accurate, just becuase it's unsourced, but it doesn't seem like a bad thing to remove material you don't believe, since it's unsourced. If a fact is controversial (i.e. someone wants to remove it), it particularly needs a source. This is basic. I look forward to your response. JesseW, the juggling janitor 19:58, 17 December 2005 (UTC)

Jmabel, if a new "fact" is added by an edit and that fact is not referenced, it strikes me as entirely appropriate to remove it straightaway with a request to reference that fact if it is to be re-added. That isn't really an invite to an edit war - either the editor finds a reference and re-adds in the information, or doesn't find a reference, in which case the information is not added.

With information that's already there, it's probably best to ask for sources on the talk page before removing information, unless that information is so outrageous, or is potentialy libellous, in which case remove it first and then ask for justification. For a long article needing references, it's best to take things slowly.

JesseW is, of course, quite correct when he says "if a fact is controversial...it particularly needs a source". Indeed, continual failure to provide a source for something that is controversial makes people believe more strongly that the "fact" is false, jguk 20:18, 17 December 2005 (UTC)

Just as a note, I've started detailed referencing of the "obvious" things in United States in a subpage, United States/References; any comments, corrections, or even better, help, would be appreciated. JesseW, the juggling janitor 21:09, 17 December 2005 (UTC)

Sounds like a good idea. In the medium-term, once almost everything is referenced, it can be merged with the article with the references becoming footnotes. Incidentally, I recommend referencing every fact - both "obvious" and "not obvious". I may pop along, if I'm not too busy fighting another cause or referencing up History of the West Indian cricket team so that I can make it a featured article, jguk 22:17, 17 December 2005 (UTC)

"Politically inconvenient"

Rather than intersperse into the exchange above, I figured it was cleaner for me to start a subsection.

WP:AGF doesn't mean that, in fact, all edits are made in good faith. It means that we should try to give each edit the benefit of the doubt.

I have absolutely no problem at all when people fight, and fight hard, to remove uncited material that they believe is false, or even possibly false. What I am objecting to is to is when people remove material that they know or believe to be true, claiming as a grounds that it is uncited, but they only do this to material they find politically inconvenient, while accepting equally uncited material that they find politically convenient. That is not a form of being "genuinely in doubt". It is a form of being "genuinely in bad faith". -- Jmabel | Talk 02:27, 19 December 2005 (UTC)

I agree with that, but surely the benefit to be derived from allowing instant zapping of misinformation, that anyone visiting WP to inform themselves could come across and innocently accept as true, is greater than the risk of politically motivated bad faith deletions which could in any case be solved by providing attribution for the edit? I say this conscious of a certain hypocrisy as someone who has once or twice (but only in order to replace tendentious nonsense) made unsourced edits I knew to be accurate... Palmiro | Talk 02:41, 19 December 2005 (UTC)

Jmabel, I think we're talking at cross-purposes here, and actually agree... I am (trying) to say that removing material (onto the talk page, of course) as unsourced, thereby provoking other editors to find sources for it, is a good thing. You seem (to me) to be saying that people will use the excuse of "unsourced" to remove material they know is true but find politically objectionable. I agree that this would, and will (and does) happen. However, I don't see how it is harmful - if the material is easily verified, then a source can be easily found, the material re-inserted, and the 'pedia is better for it. If the material is difficult to verify, and is controversial(i.e. someone finds it politically objectionable) then isn't it wrong to have it in the article without a source - controversial claims are likely to be viewed suspiously by many readers, who would greatly value a source, and likely not trust the fact without one - so isn't it to the benefit of the 'pedia for such material to be put onto the talk page until a source is found? Do you really want controversial, un-sourced material in the articles? Is it really so critical that controversial material for which a source cannot be easily found be repeated by Wikipedia? Is this really our goal? Thanks for your thoughts. JesseW, the juggling janitor 06:00, 19 December 2005 (UTC)

Let me give an example, one that, I guess, would more likely be motivated by trolling or WP:POINT-type behavior rather than politics. In our article Mexican Revolution, the following has no citation:

The armed conflict began over alleged electoral fraud perpetrated by General Porfirio Díaz in 1910; Díaz had been president virtually uninterruptedly since 1876. While his presidency was characterized by promotion of industry and the pacification of the country, it came at the expense of the working and farmer/peasant classes, which generally suffered extreme exploitation. As a result, wealth, political power, and access to education was concentrated in just a handful of families with large estates as well as some companies of foreign origin (mostly from the United Kingdom, France, and the United States).

There are a lot of facts there, and even a few interpretations. None of it is cited. It's not an area where I'm expert, but I'm certainly reasonably knowledgable. To the best of my knowledge, the material is entirely accurate. Would someone really be doing Wikipedia a favor by cutting that paragraph to talk and leaving a gaping hole in an article until someone could cite for each of the following:

• Porfirio Díaz was president of Mexico virtually uninterruptedly from 1876 to 1910. (My own feeling is that things like that are so easy to verify that explicit citation is a waste of resources. I know that's not a popular view on this talk page, but it is how 90%+ of our contributors work.)
• Electoral fraud was alleged in 1910.
• Porfirio Díaz was the party accused of being the intellectual author of that electoral fraud.
• This was the issue over which armed conflict began.
• Díaz's presidency was characterized by promotion of industry and the pacification of the country.
• This came at the expense of the working and farmer/peasant classes, which generally suffered extreme exploitation. (Proper citation for this would have to be an historian who says so: technically, this statement is POV, though I have never heard of any mainstream historian who disagrees.)
• In Mexico under Porfirio Díaz wealth, political power, and access to education was concentrated in just a handful of families as some companies of foreign origin.
• These falilies were the same ones who held the large estates (almost self-evident, but probably very difficult to actually cite)
• The dominant companies in Mexico at this time were under United Kingdom, France, and the United States ownership.
• Porfirio Díaz's policies, specifically those of promotion of industry and the pacification of the country were causal to this. (Again, proper citation for this would have to be an historian who says so: technically, this statement is POV, though I have never heard of any mainstream historian who disagrees.)

Of course, they probably wouldn't take the time to enumerate the separate points, so anyone assuming good faith probably wouldn't bother to find a citation for every one of these points, and when a good-faith contributor came back with citations for most of it but not all of it, our Machiavellian colleague could comb the paragraph again for what is still not cited, and begin the game over.

It is a lot quicker to challenge such things than to answer such challenges, so a few trolls with a little time on their hands apparently would have a license to gnaw holes in the middle of articles, almost at will, claim that they were merely following policy, and require enormous effort from other editors to get things roughly back where they were. I believe that if a group of a dozen people with about 2 hours a day wished to effectively destroy the English-language Wikipedia, they could devote their efforts to applying this rule arbitrarily and would either tie up or drive away several hundred contributors, or force those contributors to be the ones violating policy by restoring uncited material to articles. -- Jmabel | Talk 00:09, 20 December 2005 (UTC)

Sounds hypothetical to me. No doubt abuses occur, but this is the sort of reasoning that says newspaper vending machines cannot work because there is no way to stop someone from paying for one newspaper and removing many.

RIght now, there are a lot more problems with material being unsourced than with bad-faith application of Wikipedia:verifiability.

I realize that right now Wikipedia consists mostly of unsourced placeholders based on "common knowledge" and personal knowledge of contributors, and that's OK for now, but we need to get serious about fixing that.

All scientific publications do this, although traditional encyclopedias do not. Many popular, even bestselling books—a good example is Laura Hillenbrand's Seabiscuit: An American Legend—have nearly a fifth of their page count devoted to references and source citations. While the text itself is literate and entertaining. That's the sort of example we should be emulating.

When do we start doing it right? Dpbsmith (talk) 03:15, 20 December 2005 (UTC)

I don't like belaboring a point, but I don't think you are seeing mine. If we (well, you: I'm certainly not going to try to enforce this) require a contributor who wants to write a paragraph like the one above to provide an explicit citation for each of these facts and for every statement that expresses an opinion, no matter how uncontroversial, probably one of three things is going to happen:

1. People will obey, and creation of articles is going to grind to a halt. (I consider this the least likely scenario.)
2. People will simply ignore this and life will go on as before.
3. Most people will try simply to ignore this, while some go around getting all self-righteous about a policy change determined, in fact, by a handful of people. Either the credibility of policy pages will be greatly reduced in the eyes of most Wikipedians, or a lot of people will flood in here to change it.

As a rule, a relatively small number of Wikipedians are actively involved in policy pages, and I think that is entirely healthy: most are here to work on articles. This works, as long as those of us who work on policy pages try to keep policies workable and to accurately represent the loose consensus of the community on what the policies should be. It will fail if we try to dictate.

The nub of the matter: I think there is a strong consensus that it is appropriate to challenge a statement and cut it to the talk page if you believe it is false or significantly uncertain. I do not believe there is anything approaching consensus that it is appropriate to challenge a statement and cut it to the talk page merely because it is uncited, and despite being certain that it is correct. I don't see how to prove my point without going out and violating WP:POINT. Rarely have I been so tempted to do so. -- Jmabel | Talk 06:19, 20 December 2005 (UTC)

OK. I think I was confused. I agree with the nub as you expressed it above. We should cite material, but there is neither a need nor a cause to remove it if we believe it to be true (there is a need to cite it, but removing it is not the way to do this). Right. Sorry for making you explain yourself repeatedly, but I do understand it now, and agree. Thanks. JesseW, the juggling janitor 08:36, 20 December 2005 (UTC)

If we believe it to be true: I would have hoped that common sense would cover that. Obviously it is theoretically possible that someone could engage in what I might call filibustering type behaviour of this kind. But if you take a look at an article like Bat Ye'or or Jewish-Arab conflict, how can we hope to proceed without the right to immediately remove or at least comment out tendentious and misleading material? Discussing things on talk is fine, but is it really necessary to leave unsourced material we know or are fairly sure is inaccurate and misleading on an article pending a discussion? Palmiro | Talk 13:41, 20 December 2005 (UTC)

I think pretty much I agree with you, though I might stretch "significantly uncertain" farther than you'd like. Let me take a specific example. This is a general problem with lists, but the one I'm watching particularly is List of people believed to have been affected by bipolar disorder. People are constantly adding names with no citations whatsoever. Many of them are plausible. If I can find a citation instantly via Google, I'll add it. But if I can't, I move it to the talk page. I intend to continue to do this, even with names for which I think a source can likely be found. It's up to the person adding the name to provide a source. It doesn't have to be a good source, just a source. (It can't be just someone saying "it's been widely reported in the press" or "he said on TV..." In practice, probably more than half of the names I've snipped have, in fact, been reinserted with what I consider to be perfectly good sources.

People take lists as a challenge, and if they think they know a name that isn't on a list, they'll just add it. Lists have a tendency to become utter garbage, lists of "sure-I've-reads" and "thought-I've-heards." But if every name has a source, and an explanation that allows the reader to see why the name is on the list (whether they would personally agree with the categorization or not), the list becomes reasonable.

Also, while I think it is acceptable to allow plausible, not-seriously-disputed assertions to remain for a long time as placeholders, it should be a non-lip-service goal to cite them... with the same level of density and quality as Lauren Hildebrandt's book on Seabiscuit.

An alternative is to allow contributors to cite their real-world name, credentials, and contact information in lieu of a reference... but I wouldn't like that and I don't think other Wikipedians would.

All the information in Wikipedia needs to be traceable to sources. Dpbsmith (talk) 14:00, 20 December 2005 (UTC)

I think we are all pretty much in agreement on the substantive matter, then. "Any edit lacking a source may be removed," is an oversimplication. The following is not well worded, and I'm on my way out the door, but something like, "Material lacking an appropriate source may be removed, as described below in the section on Checking content; this is not a license to arbitrarily remove material that is clearly accurate, merely because it is uncited. Of course it is good to seek citation for such material, but such material should be left in the article while you look." Jmabel | Talk 04:00, 21 December 2005 (UTC)

That makes sense, but don't wait for ever. I would say that if there is no response within 5 days to a request for references, the material in question can be removed, and that re-introducing the material into the article without providing sources could be regarded as vandalism (i.e., subject to immediate removal again). Also, I feel under no obligation to spend my time searching for possible citable sources for material someone else has added to an article. I'll challenge and give time for someone with more of an interest in the article to do something. -- Dalbury^(Talk) 11:28, 21 December 2005 (UTC)

Yes, something along those lines seems reasonable, particularly if the action at the end of the five days or so is to copy the material onto the Talk page where it can easily be retrieved (we really need a feature for searching histories!). I can imagine some good-faith contributors who don't necessarily check Wikipedia as often as once a week, so it's not always easy to "engage." However, regardless of why the contributor doesn't respond, if they don't respond the material should eventually be removed.

I'd prefer to see some vague wording that recommends common sense and etiquette, rather than an exact set of rules (instruction creep). "Any edit lacking a source may be removed eventually. Before removing an edit, the contributor of that edit should be invited to supply a source and given a reasonable length of time to respond." Dpbsmith (talk) 13:36, 21 December 2005 (UTC)

I'd prefer JMabel's version, as I think it's important that an editor who sees an uncited statement that is patently false should feel free to remove it on the spot. Palmiro | Talk 14:18, 21 December 2005 (UTC)

Shouldn't it be stronger, so that an uncited statement that is *probably* false, should also be removed, preferably to the talk page? Stevage 14:37, 21 December 2005 (UTC)

Whatever. The challenge is to find some wording that endorses the responsible removal of dubious material without sanctioning the abuse that JMabel is concerned about. I don't see the strength of Palmiro's objection. The issue is not patently false material, but dubious material. Anyone can remove patently false material at any time, not because it's unsourced, but because it's patently false. Dpbsmith (talk) 14:58, 21 December 2005 (UTC)

I think the material on checking content reasonably covers this problem. It essentially says you have to have a good reason to believe the material is problematic before you remove it. If you remove uncited material and can't justify why you feel it is incorrect, then you are violating this policy as I understand from what's written right now. We could make that point explicit in the checking content just to fully close the loophole if you think it needs it. The checking content section could use a little reworking, but I think the basic point is sound. - Taxman ^Talk 16:21, 21 December 2005 (UTC)

"Reasonably believed to be incorrect"

Taxman, I don't agree with the language you added:

Any edit lacking a source and reasonably believed to be incorrect may be removed.

Saying that the edit is "reasonably believed to be incorrect" is once again putting the onus on the person removing the material. It's too strong. I've been using the word "dubious," which means "in doubt."

Let's take a hypothetical example. Suppose someone adds a sentence to the article on Arnold Bennett saying "Arnold Bennett owned a Borzoi named Caesar who was trained to fetch Bennett's slippers," with no sources. This isn't common knowledge, and it's not easily checked.

It's not by any means "reasonably believed to be incorrect." If I saw a sentence like that, I'd figure there's about a 20% chance it was just made up, and about an 80% chance that the contributor had read this in some obscure biography of Bennett or some article, but didn't remember which one... or didn't bother to cite it... or remembered a professor saying so... or saw it in some article in a dog magazine about famous writers' dogs... or something. Whatever. It doesn't Google, even in books.google.com.

One could go a to a library and spend half a day researching Bennett, but even so, it's hard to prove a negative. If one came up with a book that mentioned Bennett's Borzoi Caesar, you'd have reasonable belief that it was correct. But if you failed to find it, you would have no reason to believe it was incorrect.

This is not "a statement that is reasonably believed to be incorrect."

It is a statement that is a) unsourced and b) dubious. It's dubious because it's not a well-known fact and you have no reason to believe that it is correct.

In this case I think it would be eminently proper to ask the contributor for the source and remove it after a few days if no source were provided. Dpbsmith (talk) 19:05, 21 December 2005 (UTC)

Ok, you have a point, I've reverted myself for now. But we need some language to avoid supporting a troublemaker that just wants to rip out material they don't agree with. Then again maybe it really is the time to focus more on accuracy than just any content and if it can't be verified and put back in so be it. The thing is, for information that you reasonably believe is incorrect, you should be able to yank it out immediately and challenge the editor to provide a source. For material that is plausible, but not verified, that wouldn't necessarily be the best path, and instead your method of pointing it out and waiting a couple days is better. So the timing of the allowed solution should be different in different cases perhaps. The problem there is instruction creep, so I'm a bit at a loss I guess. - Taxman ^Talk 19:29, 21 December 2005 (UTC)

Remember, we are not supposed to be figuring out whether information in an article is accurate; doing so would be "original research". All we are supposed to be doing is making sure that the information in articles is properly sourced. Therefore, I will request sources be referenced for an article or any substantive points in an article, and if verifiable references are not supplied in a reasonable time, I will remove the unsourced material, or nominate the article for deletion, depending on circumstances. If I think information in an article is wrong, but does have a reliable, verifiable source, I will not remove that information, but will find another source for what I think is correct, and add it to the article. It's that simple, in my opinion. -- Dalbury^(Talk) 21:57, 21 December 2005 (UTC)

I don't believe it is that simple. Any serious contributor to Wikipedia is continually evaluating sources. -- Jmabel | Talk 05:56, 22 December 2005 (UTC)

There are a lot of not-so-serious contributors. By requesting references for an article, I'm assuming that any serious editors interested in the article will do something. If nothing happens within a week or so, I think I can assume that it is unlikely any editor will provide references. -- Dalbury^(Talk) 10:59, 22 December 2005 (UTC)

Another example

Economic spectrum has now survived AfD, but I don't see a single cited (or at all obviously citable) statement in the article. But I doubt it would be acceptable to cut the entire content of an entire article that just survived AfD. Or am I still missing something? Or is it just that the people who voted on AfD implicitly violated policy? - Jmabel | Talk 07:42, 21 December 2005 (UTC)

I can't speak to the specifics of the vote on Economic spectrum, but far too many editors who participate in deletion dicussions seem to not care what the policies say, or have (to my mind) very strange interpretations of policy. -- Dalbury^(Talk) 11:17, 21 December 2005 (UTC)

Remember that AfD is primarily a vote on whether the article title should exist, i.e. is it a valid topic. IMHO too many AfD voters take a rather "Wiki first, -pedia second" attitude and vote keep as long as the article text vaguely roughs out something that could potentially become an article someday.

I think this is an attitude that has way outlived its usefulness. Once upon a time, a low-quality stub on Beethoven would have been better than nothing on Beethoven, and the assumption that it would be improved was more realistic than it is for the average bad article today. But that's just my opinion.

But in any case, an AfD is not necessarily a vote on the merits of the actual article text as it exists at the time of the AfD. Thus, as things stand at present, it is not a violation of policy to vote "keep" on an article full of unsourced material. However, I believe it is a violation of current policy to allow the article to remain in that state for more than a short period of time. Dpbsmith (talk) 13:47, 21 December 2005 (UTC)

Taxman's edit

Taxman, I don't want to nitpick, but I foresee some problems with your edit (in bold): "Any edit lacking a source and reasonably believed to be incorrect may be removed."

Suppose someone adds: "American Marines's dress uniforms have silver, not gold, buttons." I do not believe this to be incorrect, reasonably or otherwise. But I have no reason to believe it is correct either. Requesting a source is entirely reasonable, as is moving it to talk or deleting it until a source is provided. SlimVirgin ^(talk) 19:36, 21 December 2005 (UTC)

Yup, check above, I already reverted myself. The problem is in finding a way to avoid abuse for ex. from POV pushing that could come from selectively removing material. Then again, I'm torn because maybe allowing people to remove any unsourced material would be the thing that finally works to get people to reallize they need to do actual research from reliable references. - Taxman ^Talk 20:45, 21 December 2005 (UTC)

I agree that people should be allowed to remove any unsourced material, as the current version says. I know it could be abused, but realistically any qualification of it will be more prone to abuse. I've only seen the current version abused once: an editor was being asked for sources a lot, so he went to another page of interest to his accuser, which contained no sources, and blanked the whole thing with the edit summary "removing unsourced material." ;-D SlimVirgin ^(talk) 20:55, 21 December 2005 (UTC)

Lots of things work well when only a few people know about them. It remains to be seen if this will work when it becomes a commonly known policy, but it's worth trying. - Taxman ^Talk 21:10, 21 December 2005 (UTC)

I don't think you were wrong to tweak the wording, I just don't think you got quite the right tweak yet. Dpbsmith (talk) 21:47, 21 December 2005 (UTC)

• What about: Any edit lacking a source that the editor reasonably believes to be incorrect, or correctness of which the editor reasonably doubts, may be removed. Also, any statement lacking a source for which a source is requested on the talk page, may be removed if no source is provided within a reasoanble time. This handles the removal of unsourced but plainly correct content for PoV reasons, making it clar that such edits are not within the policy and are in bad faith, while preserving the principle that unsourced content cannot remain if challanged, and that dubious unsourced content may go at once. DES ^(talk) 21:58, 21 December 2005 (UTC)

Absolutely not. This is an invitation for POV axe-grinders to "dispute" any source they don't like. Firebug 22:57, 21 December 2005 (UTC)

Nothing in the above says that a souce may be "disputed" it says that statemets with no cited sources at all may be removed under some circumstances. DES ^(talk) 22:04, 22 December 2005 (UTC)

I don't think editors are obliged to respond to requests - many people pass through here once and never come back. Why not keep it simple: "If you doubt the truthfulness of an unsourced statement, remove it to the talk page. Otherwise, just request a source." Stevage 22:06, 21 December 2005 (UTC)

Editors are not, of course, obliged to respond to any request. But anyone who fails to source a statemetn and fails to chaek back runs the risk of the statement being deleted, or removed to teh talk page. Ideally an editor will source all significant statements without being asked, but we are a long way from that. And of course a request for a source should always be aimed (at least in part) at the entire community, as well as at the editor who made the edit. Your wording is not bad, but soemwhere it should be made clear that if soemone does "just request a source" and none s forthcomming, then the statement can and should be removed even if it is thought to be accurate. Also i do think that "doubt" should be "reasonably doubt", and I would prefer "accuracy" or "correctness" to "truthfulness" -- "doubting truthfulness" sounds like accusing another editor of lying, and while I'm sure you didn't intend it in that way, we don't want anyoen to read such implications into a policy or guideline text. DES ^(talk) 22:17, 21 December 2005 (UTC)

DES, you're talking about a culture change here - to encourage everyone to source statements that they add. I have a number of points:

• (1) I think nowadays the problem is not as bad as you'd imagine - the quality of sourcing has improved considerably in Wikipedia in the last few months.
• (2) Good editors don't mind quoting sources when they forget them. If you revert someone with the comment that you're only reverting it because it is unreferenced and they should only re-add it if they reference it, and link to Wikipedia:Verifiability - you'll find that in 80% or so of cases it works. Either the information is not re-added or it is re-added but with references also being added.
• (3) Come the end of the day, the precise wording of Wikipedia:Verifiability doesn't matter - either editors source their statements, or they don't. Good editors who source their statements should encourage those who don't source their statements: (i) by example; and (ii) by telling them to source their statements.
• (4) In short - to make things better, leave the guideline alone and (i) make sure you source every new fact you add (and we all make mistakes on this from time to time); and (ii) encourage others to source every new fact they add. And yes, this can be by politely reverting new edits with a request for references.

Remember, we can't legislate for commonsense, but we can encourage good practice by our actions. This is more important than the minutiae of a policy, which quite frankly will rarely be read, jguk 22:47, 21 December 2005 (UTC)

I agree with Jguk. SlimVirgin ^(talk)

Me too. And I like Stevage's proposed wording. - Jmabel | Talk 06:16, 22 December 2005 (UTC)

Joe, I reverted your edit because it throws up too many problems. If I have no idea whether something is true or false, it should still be removed if unsourced. What is wrong with what the page currently says? SlimVirgin ^(talk) 06:40, 22 December 2005 (UTC)

Gosh, I thought I'd already written far too much about what is wrong with it. The short version is that it provides too easy a cover for trolls and POV warriors. Have a look at my comments on this page over the last few days; I don't think I'd be doing you or anyone else a favor by rehashing them. -- Jmabel | Talk 08:48, 22 December 2005 (UTC)

I want to see JMabel's concerns addressed. I also want to see WP:V strengthened in practice. It's been policy in theory forever. People need to learn not to be offended by legitimate, civil, valid requests for sources. Dpbsmith (talk) 11:03, 22 December 2005 (UTC)

Plus, it was not compatible with this line from the second paragraph: "Articles should contain only material that has been published by reputable or credible sources, regardless of whether individual editors view that material as true or false. As counter-intuitive as it may seem, the threshold for inclusion in Wikipedia is verifiability, not truth." -- Dalbury^(Talk) 11:07, 22 December 2005 (UTC)

I also like Stevage's wording. And SlimVirgin, would you stop reverting everything jsut because you feel like it? There were already a few people agreeing with the wording, and in any case, reverts are considered harmful. Don't do it unless the edit is really damaging. - Taxman ^Talk 13:25, 22 December 2005 (UTC)

Reputable confirmed printed sources not available online

Recently, a number of administrators removed information from a group of related articles stating an objection to the book references provided, since they could not find the book's online version (which does not exist). Yet other editors confirmed the contents of the books as valid. WP:V seems to conclusively support the view that reputable published books can be used as verifiable sources. However, given that books are harder to get a hold of than printed sources (e.g. you have to go to your library or book store to check the references), is there a specific policy that applies to these? Can published reputable books ever be correctly stated as being "original research"? Zordrac (talk) Wishy Washy Darwikinian Eventualist 13:21, 22 December 2005 (UTC)

Books are of course considered very good references (depending on their quality of course), and we should assume good faith that it says what the editor says it does unless an editor has been shown to misquote sources before. If that hasn't happened, then the burden does fall on the people who don't believe the reference says what it is quoted as. The only caveat could be for dubious claims from very difficult to obtain sources. If no libraries carry the book for ex. Otherwise removing information because you can't find the online version (if your characterization of the situation above is correct) is in clear violation of this policy. - Taxman ^Talk 13:30, 22 December 2005 (UTC)

That's what I would have thought. And as for what is considered to be "reputable", can we suggest that if Wikipedia has an article about the reference book in question, and the Wikipedia article suggests that it is a reputable source, then it can be considered to be reputable? But then, what of the issue that some editors suggest that the book's contents do not say what they think it says? Does it become a popularity contest? Do some people's views take precedent? Is there a burden on the part of the person that says that the book says that to prove it somehow? I would have thought that insertion of a tag such as {{fact}} would be appropriate in such a case, rather than a deletion of the information. Zordrac (talk) Wishy Washy Darwikinian Eventualist 13:45, 22 December 2005 (UTC)

Not checking the specific elements of the debate in question, that would be generally reasonable to consider that reliable, but it depends on the topic. Pauling is considered authoritative on quantum chemistry for ex, but not on his vitamin C theories. And yes, like Dpbsmith says requesting the person using the reference provide a page number and exact quotes is very reasonable. After that, like I said, the burden of proof falls on the challenger to disprove that is what the book says. If someone is demonstrated to have lied about what a reference says, that is of course a severe violation of this policy and then it is no longer reasonable to take their word for what a source says in the future. - Taxman ^Talk 15:28, 22 December 2005 (UTC)

As of 2005, printed books are better references than online sources. They are durable and stable. Their links don't break. The vast majority of books mentioned in Wikipedia are available through interlibrary loan via most public libraries in a matter of days. And recently I've been finding that quite a lot of books, though not available online, are searchable online via a www.a9.com book search and/or books.google.com.

It is reasonable to insist that a book citation include an ISBN number, or, if older than ISBN, other identifying number.

It is reasonable to insist that an editor providing a reference to a book, once challenged, should provide a page number and an actual quotation verbatim quotation. You can't say "Kryptonite has a molecular weight of 165" and merely cite "The Handbook of Chemistry and Physics."

As always, the issue is not "is the book true." The issue is whether the book exists and says more or less what the citing editor says it says. If a book is disreputable, it is perfectly reasonable to cite sources that call it disreputable. "At the Earth's Core," by Edgar Rice Burroughs, is a fine reference for Hollow Earth, as is McBride's Symmes' Theory of Concentric Spheres, despite the fact that the world is not hollow, despite the fact that the first is a work of fiction, and despite the fact that McBride's book is fairly obscure.

If a book is self-published, it should be cited only with a note to that effect.

In an ideal world, we'd have some Wikipedians with access to good university libraries systematically checking the book references.

There's a good case for removing a book whose existence can't be verified online (between Amazon, used-book sites like abebooks, and many, many big libraries that have their catalogs online, this really should be possible).

There's also a good case for removing as unverifiable a citation to a book that can't be gotten from a major research library, either directly or by interlibrary loan, by someone who's seriously trying. Dpbsmith (talk) 14:03, 22 December 2005 (UTC)

P. S. I would think that it's reasonable to assume, as a first approximation, that a publisher's blurb is a fair representation of the contents of the book. That is, until someone has the time to check the book itself, the Amazon blurb of The Bell Curve, which says

But despite decades of fashionable denial, the overriding and insistent truth about intellectual ability is that it is endowed unequally.... Charles Murray explores the ways that low intelligence, independent of social, economic, or ethnic background, lies at the root of many of our social problems. He also discusses another taboo subject: that intelligence levels differ among ethnic groups. According to the authors, only by facing up to these differences can we accurately assess the nation's problems and make realistic plans to address them.

is good evidence for a statement that "The Bell Curve says that intelligence levels differ among ethnic groups."" While it's not impossible that an author might disown the presentation of his thesis in a blurb, it's unlikely. Dpbsmith (talk) 14:09, 22 December 2005 (UTC)

Okay. I'll just take the case in point. A book is sourced, is available on Amazon, is referenced as being the most notable possible source for the subject, has its own Wikipedia article with references to that effect, and, after challenged, exact quotes are made with reference to page numbers, edition number etc. Is it reasonable for that book to still be removed as a reference purely on the basis that the book's contents are not available online? Is it also reasonable for an editor to remove the citations based on their inability to find its contents online? I would have thought not, but would like clarification. Zordrac (talk) Wishy Washy Darwikinian Eventualist 21:10, 22 December 2005 (UTC)

Assuming your summery to be complete and accurate, IMO this is compeltely unreasonable. Somehow I suspect that the otehr parties involved would summerize the events differently, but without knowing what articel is involved, i have no way to asses the merits of the actual case. Note that if the book is on amazon, the "search inside the book" feature, if activated for the particular book (it isn't for all books) may provide an online version that can assist in such cases. DES ^(talk) 21:17, 22 December 2005 (UTC)

Hi Zordrac, in the example you've outlined, the offline source sounds as though it would be acceptable as a source. The only exception would be if the editor offering it as a source was not trusted for some reason e.g. because he had caused problems with sources in the past, or where his political views were very extreme, perhaps. I've only seen this done once, where a fully cited reference from an old newspaper was turned down by one editor who didn't trust the editor who was offering it. The latter therefore scanned the page onto his website, so the first editor could read the quote in context, and that worked: the citation was accepted. But generally speaking, there's no need for a reference work to be online. SlimVirgin ^(talk) 21:42, 22 December 2005 (UTC)

I certainly agree that it is inappropriate to remove a source solely because it is not online. Of course, that does not mean that all "sources" that are not online are suitable for references on WP - they still have to be considered on their own merits, jguk 06:41, 23 December 2005 (UTC)

Material from WP:CITE

Dan, I deleted the material you copied over from CITE, because I couldn't see anything it added that wasn't already on the page in some form. SlimVirgin ^(talk) 22:44, 22 December 2005 (UTC)

You should actually have checked :-). As Uncle G notes in his edit summary when rv'ing you, I moved the material. Personally I think it's the only logical solution if WP:CITE is "only a style guide", and personally I'd prefer to have to only refer people to this page for the complete run down on why to cite sources! Dan100 (Talk) 09:09, 23 December 2005 (UTC)

I've reverted because it seems repetitive. What does the material copied from CITE say that isn't already on the page? I'm happy to have it if it says something new, but it seems pointless just cramming more stuff in, not to mention that this wasn't a particularly well written or illuminating section of CITE. SlimVirgin ^(talk) 10:05, 23 December 2005 (UTC)

Why don't you revise it rather than reverting then? BTW, you're the queen of adding text, so I thought you'd like it :-) Dan100 (Talk) 14:16, 23 December 2005 (UTC)

No, Dan, actually that's you. You've done this several times before: mass deletions in one policy, copying it over to another. Why won't you answer the question? What does the material you've copied from CITE add to this text that wasn't already there? SlimVirgin ^(talk) 17:23, 23 December 2005 (UTC)

I would ask people in this dispute to carlify for me what's being disputed. SlimVirgin, are you merely saying you find inclusion of this to be redundant. Or, more fundamentally, do you object to this being part of a policy (not a mere style guideline in WP:CITE). I strongly support this which makes clear citing sources when contributing is mandatory, not just something we want, but something we demand. But obviously if its not policy, it shouldn't be here, no matter how good it is. Policy needs consensus, so I'm trying to figure out a) What is current policy and b) What, if any, changes is anybody seeking in current policy. Sorry for my confusion, but I'm really unclear what's being debated here, and I find it frustrating to see a core policy page being changed frequently. --Rob 18:51, 23 December 2005 (UTC)

Hi Rob, I'm also frustrated at the frequent changes. Dan has done this elsewhere with policy, practically blanking them and moving material elsewhere, I think under the "be bold" principle, which I'd argue shouldn't apply to policy pages. But regardless of that, my objection is that the material, so far as I can tell, is repetitive. I've asked Dan a couple of times what he feels it adds that isn't already there in some form, but he won't say, just keeps reverting. Also, this material was the weakest part of the cite-sources page, inherited from a long time ago I think, which is why no one changed it, so it's a shame to see it transferred wholesale into a policy page that already makes the same points.

Can you see anything new in the additions that isn't already on the page? I'm happy to keep anything that really is new.

Regarding whether citing sources is mandatory, the policy is that people ought to (ideally) cite sources when they make an edit, but it's not mandatory. That is, if someone makes an edit and doesn't cite a source, you can't take them to the arbcom for a policy violation. However, if the edit is challenged, a source must be provided, or the edit may be deleted. Exactly when it's deleted depends on the context and on common sense. If it's a potentially defamatory or in some other way dangerous claim, for example, it should be deleted immediately. If it's a harmless claim, it could be moved to the talk page, or left in the article for a few days. SlimVirgin ^(talk) 21:08, 23 December 2005 (UTC)

I can absolutely see something new in the additions. It's clearly stating the sources *must* be included with any addition. That's apparently new to *this* *policy* page. WP:CITE is just a guideline. So, really I don't understand the focus on the discussion of redunancy. I consider redunancy a trivial "problem", and think its often good to be repetitive of critical points (like this). If this is policy, I absolutely want it on this page, and wholeheartedly support its inclusion. If it's not policy, keep it off the page, and make it clear everywhere that's its just a guideline. Then, let us start-up a policy discussion in the appropriate place (as we would then have a real problem). I read the edittools message which says "Content must ... be based on verifiable sources..." and I take both the literal and implied meanings very seriously. Basing on sources doesn't mean you write the content first, and try to find sources later. We seem to have a problem where different things are being said in different places. --Rob 21:27, 23 December 2005 (UTC)

We shouldn't ever delete a mass of text from a guideline and just dump it onto a policy page, because the guidelines make suggestions that the policies may not mandate, so it causes confusion exactly like this. This page already is repetitive on the main points, Rob, so to produce yet more of the same is pointless if it's identical, and problematic if it's adding new material. No, it isn't policy that sources must be included with any new edit. At least it wasn't before Dan inserted it without discussion onto the page (if that's what it says exactly). We can't make that policy in part because it's unenforceable, in part because we'd have to say exactly what kinds of edits sources had to accompany (do we need sources for "2+2=4" and "the sun rose this morning"?). The policy as it stood said that any edit that is challenged and has no source may be removed, and I don't feel we can go further than this, particularly as some editors (e.g. Joe Mabel above) feel that may be going a bit too far already. I completely support the idea that using good sources is crucial, but I also don't think we should produce a draconian policy; and anyway, if we do, people will just ignore it. The policy should be reflective of community consensus, and based on how people edit, no one thinks that adding a source for every edit ought to be mandatory. SlimVirgin ^(talk) 21:49, 23 December 2005 (UTC)

I read the edittools message which says "Content must ... be based on verifiable sources..." and I take both the literal and implied meanings very seriously. Basing on sources doesn't mean you write the content first, and try to find sources later. We seem to have a problem where different things are being said in different places - very good point Rob, and I'm going to edit that text to reflect that. Dan100 (Talk) 10:08, 24 December 2005 (UTC)

Why I've removed the material moved from WP:CITE

I've deleted the material from WP:CITE again because it's mostly repetitive (please read the page before making additions), and also because Rob thinks one part of it may have changed the policy. Rob, you were talking about being confused about what the two pages say regarding sources, so here's a rundown:

• WP:V (a policy and therefore mandatory) currently says, and has said for some time, that WP articles should contain only material "already ... published by a reputable publisher" and that "[t]he burden of evidence lies with the editor who has made the edit. Editors should therefore provide references. Any edit lacking a source may be removed."

That's the part JMabel thinks should be qualified a little (see below).

• WP:CITE (just a guideline and therefore not mandatory) said for some time: " ... if you add any information to an article, you must cite the source of your information," and "Providing sources for your edits, in particular for edits that are challenged by other editors, is mandated by Wikipedia:No original research and Wikipedia:Verifiability, which are policy." [15] The phrase "in particular for edits that are challenged by other editors ..." diluted the sentence sufficiently so that it was clear not every single edit had to be accompanied by a source, but without saying as much so as not to give POV warriors and lazy editors an explicit loophole.

• The qualifying phrase was removed on November 30, [16] leaving the sentence "Providing sources for your edits is mandated by Wikipedia:No original research and Wikipedia:Verifiability, which are policy," which was false because the policy doesn't exactly say that. Unfortunately, none of the regular editors on the page noticed the change and so it wasn't reverted.

• Dan removed this sentence and a lot of other material, without discussion, from WP:CITE [17] and inserted it, also without discussion, into WP:V, thereby arguably changing the policy, and reverted when I deleted it.

• I've therefore restored this material to WP:CITE [18] and added a tighter qualifier: "Providing sources for edits is mandated by Wikipedia:No original research and Wikipedia:Verifiability, which are policy. What this means is that any edit that is challenged and has no source may be removed by any editor."

• I've also restored WP:V to the status quo ante.

What we should decide now is where we want the emphasis to lie. Do we want to:

1. Weaken the current policy slightly (per Joe's suggestion) to say: "Any edit lacking a source may be removed. If you doubt the truthfulness of an unsourced statement, remove it to the talk page. Otherwise, just request a source." (This implies that you should not remove an unsourced statement unless you doubt its truthfulness; that is, being unsure about an edit would not be sufficient grounds to remove it.)

2. Strengthen it slightly (per Dan's additions) to say: " ... if you add any information to an article, you must cite the source of your information," which arguably changes the emphasis of the policy from "any editor may remove an unsourced edit" to "every edit must be accompanied by a source when first added." As this is a policy page, and not just a guideline, in theory an editor could be taken to the arbcom for not providing a source with an edit.

3. Leave the policy as it currently is: "Any edit lacking a source may be removed."

My preference is to leave it as it is, because I foresee problems with weakening or strengthening it. However, I'm prepared to go along with consensus, but we should reach an agreement first rather than adding material that not everyone agrees with, and which may make this page internally inconsistent or inconsistent with other policies. SlimVirgin ^(talk) 04:06, 24 December 2005 (UTC)

Again, I'm trying to resist WP:POINT, so I'm keeping this to a "though experiment". I just went and hit "random article" a few times. I got Hits (Phil Collins album), AFC East, and Eugene Paul Bennett. The first two have no references at all. Are you saying that it would be appropriate for me to cut the entire article to its talk page? The third gives three references, but no statement in the article is tied to any reference. Same story?

Slim, do you actually do things like this, or is this all theory? -- Jmabel | Talk 08:39, 24 December 2005 (UTC)

Hi Joe, no I never do it, and I agree with you in practice. I don't want to see vandals/POV pushers roaming through articles randomly deleting sentences that aren't sourced. It's just that I foresee problems with any words we use to dilute what's already on the page. I suppose at some point we need to rely on common sense. I've never been asked for a source for something obvious that I couldn't find. Yes, it's annoying to be asked if you know the thing is true, and obviously so, but it only takes minutes to find one. SlimVirgin ^(talk) 08:56, 24 December 2005 (UTC)

If it "only takes minutes" we must work on some very different topics. Try finding a citable online source for the first name of a minor figure in the French Revolution or to document that a particular bar was a pioneering locale in Seattle fringe theater. -- Jmabel | Talk 10:44, 24 December 2005 (UTC)

Sources don't have to be online, if that's the issue, Joe. However you yourself know that X is the case, whether from a book or from an article on a website, it's probably citable. SlimVirgin ^(talk) 10:48, 24 December 2005 (UTC)

• Slim, thank you for explaining your reverts. Why did you not do it sooner, but instead wait until you were on the brink of breaking the 3RR? Now you've explained it, it makes sense. You could've done that a lot sooner... I note that you've clarified the intro at WP:CITE too, so some good of this has come of it. Dan100 (Talk) 10:07, 24 December 2005 (UTC)

For now, I think 3. Leave the policy as it currently is: "Any edit lacking a source may be removed." seems to make sense. I strongly support the right to remove uncited stuff on site. Perhaps a good rule-of-thumb to encourage (but not make policy), is people should feel extremely free and encouraged to remove *new* uncited material prompty, but be more cautious of older well established material. It's essential RC Patrollers remove anything that might be sneaky vandalism quickly. If somebody sees something they recently added is removed due to lack of sources, they'll likely re-add it with a source (as its fresh in their memory, and still on their watch list). But if somethings been sitting around for a year, there's little harm for taking extra time before removing it. --Rob 10:51, 24 December 2005 (UTC)

That's a good point. I think there's some value in treating older material a little differently, and at a minimum requesting a source before removing. Newly added material without a source however can be shot on sight. But I also agree we shouldn't overlegislate that. As for the emphasis, I believe there is clear consensus for encouraging citing sources for added material and for removing unsourced material (which both encourages sourcing in a roundabout way and enforces it). But SlimVirgin is right that we're not requiring citing sources in the sense that it's generally a blockable offense not to. If someone is repeatedly adding unsourced material or replacing sourced material with unsourced, that could meet the disruption policy, but that's different. So I think as long as we make it clear it's not a blockable offense unless it rises to disruption, we can say citing reliable sources to add material is required. - Taxman ^Talk 16:31, 27 December 2005 (UTC)

I would prefer 2. Strengthen it slightly ..., but I'm not a fanatic about it. I do want to see more done to encourage references. I have started marking articles {{unreferenced}} and am seeing a response in a few articles. I'm keeping the articles I mark on a suspense page, but I'm not sure yet how long I will wait to take further action, or what that action will be. -- Dalbury^(Talk) 12:51, 24 December 2005 (UTC)

Verifiability is...

Non-verifiable! --Knucmo2 14:20, 27 December 2005 (UTC)

"reputable publisher" and press ???

Main press

press don't always take the information from press agency or people working for them... sometimes they takes the information from another parution,for example a local newspaper that lead to problems

• for example if someone publish a hoax on internet and this is taken by a not so credible newspaper and itself taken by a more and more credible one...then you have unverified content into "reputable publisher"

there are known example but at this moment i have none of them in mind

• another example is april fools

a french magazine "science et vie junior" that is a well known scintific magazine targeted at young people(teenagers) has a fake article on the april parution...and you don't know wich article is fake...until the next mounth parution wich reveals it so an important newspaper told that cell phones were making "pousser des poils dans les oreilles(don't know how to traduce it in english)"(to be verified)

so even with reputable publisher you need to be carefull

about scientific articles and magazine

exept the "science et vie junior thing" normaly scientific parution in magazines such as nature are verified before the parution... but i what is the exact process of verification(mabe they check also the protocol of experiencies...that is crucial in science because you have studies that says something and others that claim the contrary)

Normaly the people publish result of studies on specialised location on the internet(don't remember where) then it goes into specialised publications such as neuroscience magazines for neuroscience studies and then it goes into scientific magazine such as nature,then into magazines of scientific vulgarisation(such as science et vie(french magazine...please tell us if there are more well known and international ones)) and then into the press And so if the parutions on magazines such as nature are verified they are credible but sometimes there are errors that can pass trough this for example an error about the universal constant...they had made an error calculating their variation trough the time and thoat that they had vary much more than the reality There is also a carefull interpretation to — Preceding unsigned comment added by 81.241.31.173 (talk • contribs)

Since the standard for Wikipedia, as currently stated in the policy, is verfiability, not truth, the first consideration in whether to use a source is whether it is reliable. Now I would not use a source, even if it were normally reliable, if I knew that other reliable sources contradicted it. But determining whether an item in a normally reliable source is true, or an April Fool's joke, or something else, is original reasearch. If you are uncomfortable with information that appears in a reliable source, look around to see what other reliable sources have. If reliable sources disagree on something, you may have to note that in the article. -- Dalbury^(Talk) 12:27, 28 December 2005 (UTC)

I say the hell with verifiability if the reference is not true, or has falsehoods. OIn the Bigfoot page, skeptics are quoting very bad authors who make errors and pass on really, really

false and bad stuff.I feel the skeptics here __choose_ those bad authors to support their own case.

As for NPOV, here I am, on Wiki, a resident Bigfoot expert, and I am not allowed to post my experiences because there ARE NO JOURNALS THAT DISCUSS BIGFOOT. And all books are two years or more behind the times.

Let us not forget -- this is Not Brittanica. It is a big website. Nothing more.

beckjordBeckjord 07:43, 29 December 2005 (UTC)

No, this is an encyclopedia, which is why we don't allow original research. If you can get your research published by a credible third party, we can use it as a source. SlimVirgin ^(talk) 10:08, 29 December 2005 (UTC)

I understand your frustration, but we have no way of vetting editors, so we have to require published sources. You can, however, question the reliability of a source per Wikipedia:Verifiability#Dubious sources. -- Dalbury^(Talk) 12:58, 29 December 2005 (UTC)

Well, to be precise, you can include verifiable information that helps the reader assess the quality of the source. For example, if other reputable sources have criticized the source in question (the author, press, or publication) one can add "However, according to x, ...." or "However, x has claimed that ..." One can also add verifiable information about the publication, for example, whether or not it is peer-reviewed; what is its circulation; is the journal associated or affiliated with an organization (for example, I would like to know whether a publication is put out by a branch of the AMA, or a pharmaceutical company, or an independent body of scientists, or an organization promoting homeopathic or natural remedies). An editor need not directly say "this is a good source" or "this is a bad source" or "this is a qustionable source." An editor can, however, include a range of views not just about the topic but bout the sources, and provide enough contextual information about the sources so that a reader can identify any biases in the source. I think all of this can be done in ways that comply with this policy as well as NPOV. Slrubenstein | Talk 17:57, 29 December 2005 (UTC)

Proposed rewrite

At the moment the page is long (16kb) and discursive. It is also, no doubt, little read. I'd like to see it much shorter and to the point - and something that assumes WPians will use common sense in interpreting it, rather than something that tries to cover everything. I have put up a draft on Wikipedia:Verifiability/temp and would be grateful for comments. In essence, I think we should keep the rules simple and straightforward - and have them written in a way where we can hit troublemakers over the head rather than have them hit us over the head, jguk 19:27, 29 December 2005 (UTC)

I made a slight change on foreign language sources. Otherwise I think it's better, but I'm strongly of the opinion that we have to try to make a wider consultation about this. Right now I think that changes are being made with no consensus from the rest of the community, but without them caring enough to oppose those changes. We should try to stimulate them into discussion. Mozzerati 20:46, 29 December 2005 (UTC)

I appreciate Jguk creating a temp file. As with all articles here, this page certainloy can be improved upon. But, as Mozzerati points out, major changes should only follow broad discussion; this is especially important concerning policy pages. While we can and should improve policy pages, we should give precedent more weight than on other pages, and involve more general discussion concerning major changes. I do have one specific comment about the policy — this is in Jguk's proposed version, but he adopted it from this version. I am referring to the statement that sources should be unimpeachable. I agree that the policy should convey to editors the importance of highly reputable sources. However, when it comes to contentious topics, some sources will not be "unimpeachable," indeed, the sources themselves may be objects of contention. I do not think such sources should be eliminated. Rather, in compliance with NPOV, we should make it clear what the source of contention is, and provide some context about the parties who contend over the un/impeachability of the source. The source's verifiability of course must be unimpeachable by which I mean that the provenance must be clear and accurate (i.e. verifiability means, if we say "this is the source," i.e. provide information to help people find the source (so they can read it for themselves), this information must be incontrovertably accurate. Slrubenstein | Talk 21:03, 29 December 2005 (UTC)

I would like to separate out the big bold box (in "The Policy" section) and use that, as it seems like the one great part of the suggested change. --Rob 21:13, 29 December 2005 (UTC)

I would agree with the above two voices regarding wider consultation. The rewrite is certanly an improvement, although I have some concerns with the purple policy box, which seems to state the policy stronger than the previous page. Steve block talk 21:16, 29 December 2005 (UTC)

By all means, I agree we'd need wider consultation before adopting a rewrite. But to begin with, let's keep it to the regulars to this page until it's improved a bit:) jguk 21:31, 29 December 2005 (UTC)

I think jguk's temp version is very good; in fact, I think it's basically what the policy says now, only more clearly. Information added to pages should be verifiable, and properly cited from reputable sources. If something is added which is neither, then it certainly can be summarily removed - in fact, in many cases it absolutely should be summarily removed. In practice, depending on how nonsensical the information is, one could deal with it in a number of ways. If it is particlarly horrendous, or libellous, it should just be removed. If it is almost certainly false (or wildly POV), but might contain a grain of truth, it could be brought to the talk page. If it might be true, but seems dubious, then a {{fact}} template can be placed beside it. But while all of these should be suggested as possibilities, the fact remains that policy requires information to be cited and verifiable, and added information which is neither can simply be removed. Jayjg ^(talk) 22:32, 29 December 2005 (UTC)