Wikipedia talk:Verifiability

El directorio enciclopédico desde la Wikipedia.

The project page associated with this talk page is an official policy on Wikipedia. Policies have wide acceptance among editors and are considered a standard for all users to follow. Before editing this page, please make sure that your revision reflects consensus. When in doubt, discuss on the talk page. Always remember to keep cool when editing. Changes to this page do not immediately change policy anyway, so don't panic.

Shortcut: WT:V

This talk page is automatically archived by MiszaBot II.

Any sections older than 10 days are automatically archived. An archive index is available here.

[edit] "Warning" signs on bio. pages and nomination for deletion

Have added a large number of

This article does not cite any references or sources.
Please help improve this article by adding citations to reliable sources. Unverifiable material may be challenged and removed. (August 2008)

This biographical article needs additional citations for verification.
Please help this article by adding reliable references. Unsourced or poorly sourced material about living persons must be removed immediately, especially if potentially libelous or harmful. (August 2008)

to a large groupe of players in the swedish top league, aka Allsvenskan, i myself is a major fan of Halmstads BK, also a team in Allsvenskan, and have made major contributions to that teams and players articles, i am trying to improve all articles regarding to Swedish top fotball, Allsvenskan and Superettan mainly, and was hopping that some user and IP numbers would help if i added this tags, this however havent happend and i feel that i know to little about some players to writte theire bios on my own. I also added this signs to the articles since i dont feel like they reach the demands on Wikipedia regarding bios. So now i wonder how long i should wait before i put up a bio. page for deletion or if there is something else i should do instead? --> Halmstad, Talk to me 21:17, 17 August 2008 (UTC)

[edit] Self-fulfilling verifiability

I've jut been reading my latest copy of Private Eye in which it suggests that the Daily Mirror used an (essentially vandalised) copy of a page on a Cypriot football team called Omonia Nicosia with a joke supporters group inserted into it. Private Eye's assertion is that our verifiability policy is essentially daft, given that it requires simply attribution in a third-party reliable source, such as a mainstream newspaper, which would allow this "joke" to be included as factual. I'm uncertain on this specific case, but I've seen similar point brought up before. Whilst I'm happy that in general we may catch these problems, do we have any procedures in place to prevent this, or specific guidance? This is an issue, since it impinges on our reputation and the quality of the encyclopeadia as a whole. Any thoughts? Fritzpoll (talk) 10:05, 2 October 2008 (UTC)

If we're using either Private Eye or the Daily Mirror to support article claims, the joke's on us. Regardless, reputable publications with established reputations for accuracy and editorial oversight are quite unlikely to be regurgitating vandalized and/or inaccurate information from Wikipedia. The neutral point of view policy is a further safeguard, since it indicates we shouldn't be presenting claims only present in one or very few sources. Vassyana (talk) 10:38, 2 October 2008 (UTC)

I'll dig around in my archives, but I'm pretty certain that several reputable newspapers (The Times, Telegraph) have been regurgitating portions of our pages. Maybe this is only a problem with newspapers, and as such isn't going to affect quality on a wide scale. I'm not sure - I just felt it was worth discussing. Fritzpoll (talk) 10:42, 2 October 2008 (UTC)

I think a deference to higher quality sources and ensuring that our articles reflect the body fo reputable sources is the best systematic way to counter the concern. In other words, sticking to the principles we have in place. I believe that at worst this highlights concerns periodically raised about the news media, particularly those outlets with less than a sterling reputations. Vassyana (talk) 11:23, 2 October 2008 (UTC)

"Verifiability, not truth" doesn't mean we should include what we know to be untrue. --NE2 10:34, 2 October 2008 (UTC)

Absolutely, but I think I was thinking more along the lines of how we try to prevent this from happening, if indeed it is possible. For instance, in this example, someone could unwittingly read the Mirror article, and update the article with a valid source. The edit is in good faith, and it could be very diffcult to catch. I was wondering if some sort of additional criterion, such as the date of the verifying source having to precede the first inclusion of the information in an article would prevent this? I'm certain that someone else can come up with ways in which this wouldn't work though... Fritzpoll (talk) 10:39, 2 October 2008 (UTC)

If the Daily Mirror is using Wikipedia as a source, we should not consider it reliable in any case. Powers ^T 10:42, 2 October 2008 (UTC)

Yes, but isn't that difficult to enforce in practise? The Daily Mirror crops up as a source in several pages, and we can't be sure that we're not being used as a source for info by various publications, even if it is less obvious. I'm just a bit concerned about the possibility of us verifying our own information via a proxy third-party source, and that maybe the date of the source needs to be used to assert it usefulness in verification. Fritzpoll (talk) 10:46, 2 October 2008 (UTC)

I've encountered this a bit recently - when trying to reference long-standing articles, we may need to be careful that the source didn't rely on Wikipedia. I've taken your approach, and prefer to rely on sources that either predate the entry here, or which are clearly reliable (although "clearly" is tricky to judge). Not perfect, but it makes me feel more comfortable. - Bilby (talk) 14:06, 2 October 2008 (UTC)

I was mostly wondering if we should make an expicit note/suggestion to that effect on one of the policy or guideline pages. Just to make people more aware of it? Fritzpoll (talk) 14:08, 2 October 2008 (UTC)

I've been thinking about this issue a lot, lately. :) Yes, I think it is probably a good idea to mention the issue, although I have no idea how. Perhaps in terms of what consititues a reliable source? - Bilby (talk) 14:44, 2 October 2008 (UTC)

I think I'll move this off this page and over to WP:RS, where I'll propose adding a bulletpoint to Wikipedia:Rs#News_organizations - Join me there? Fritzpoll (talk) 14:53, 2 October 2008 (UTC)

I think if a publication sources it's articles from Wikipedia, and doesn't fact check, then it wouldn't be a 'Reliable Source', and so the supposed fact cannot be supported by an article from such a publication. lk (talk) 10:42, 2 October 2008 (UTC)

But how do you know the publication is sourcing in this way? I only saw this one because someone else happened to be checking it out. Fritzpoll (talk) 10:49, 2 October 2008 (UTC)

It's not just tabloids like the Daily Mirror - a dubious "fact" from Wikipedia found its way into Ronnie Hazlehurst's obituries in several far more reputable newspapers, including The Times (see Talk:Ronnie_Hazlehurst#The_SClub_7_Hoax). The Times retracted it fairly quickly, but similar instances could easily have fallen through the net. Since otherwise good sources may well do this from time to time, I don't see a way to exclude the possibility of this sort of feedback loop completely, though it might well help to suggest that for unlikely facts, sources which pre-date the Wikipedia article are preferred. Iain99^{Balderdash and piffle} 14:30, 2 October 2008 (UTC)

I'd suspected as much. I guess, given Wikipedia's prominence, this was bound to happen - we'll catch some of the problems with people noticing similarities. In the meantime, I would like to propose an addition to either WP:V or WP:RS to make a note of this possibility. Any suggestions as to which or if this is even a good idea? Fritzpoll (talk) 14:38, 2 October 2008 (UTC)

By chance I just posted about a specific example of exactly this issue today at at the noticeboard! Times are hard in the news rooms! Carol Moore 16:07, 2 October 2008 (UTC)Carolmooredc

Any organsation that makes extensive use of Wikipedia as a source is not a credible source to be use by Wikipedia. I.e., if one-in-100 articles in newspaper x can be seen to have used Wikipedia-originated material, then maybe no articles from newspaper x should ever be considered citable as sources within Wikipedia articles. Meowy 17:00, 2 October 2008 (UTC)

I don't think it's as easy as that. What if the copying is non-obvious? I think an advisement of caution within the guidelines might be better than trying to organise overview of statistical information gathering on the use of Wikipedia articles by the press. I've proposed a wording addition of sorts at WT:RS, and I hope people can gather to comment there on whether the change is needed, and to help with the wording, since I am not an eloquent chap Fritzpoll (talk) 17:07, 2 October 2008 (UTC)

I agree especially if the information is actually coming from one of the many unofficial mirrors of Wikipedia some of which don't update as often as Wikipedia does. So you have the risk of someone using an old version of a Wikipedia article that may not match the current one and therefor be missed as actually originating from Wikipedia. This is an issue if when the mirror was created a piece of information was wrong or misstated and that article is then used to reinstate the old inaccurate information. Also even the best papers may use wikipedia; in the link Carolmooredc provides it was stated that "The New York Times has links to wikipedia all the time, that isn't a reason to remove their blog posts or news articles." Remember that mainstream newspapers is last in the hierarchy of reliable sources which implies that greater care must be used than with sources higher up the hierarchy.--BruceGrubb (talk) 06:51, 3 October 2008 (UTC)

A bullet point has been added to WP:RS to help with this issue Fritzpoll (talk) 10:35, 9 October 2008 (UTC)

[edit] How exactly are foreign language sources verifiable?

I've recently been into heavy discussions at FAC's etc about the use of foreign sources. I am not against them in any way, but I like to enforce the fact that as per the guidelines, they should only be used if there is no alternative in English. Very recently, this article passed FAC even though it uses German sources predominantly. I have bought this up with the reviewers and the actual mod who decides upon promoting or failing FA's, and he told me to bring this discussion here. Now long story short, I myself have been working on an article that uses a fair bit of foreign language sources. this alternative article uses a few Croatian sources here and there, but notice how I have included relevant translations of the source. I think all/most articles should do such, deepening on how often they use references in languages other than English. At FAC, the nominator refused to commit to such, pulling out insufficient cop-outs as to why his German sources shouldn't need to be translated. Basically, I asked the predominant question; "How can this be a Featured Article when it cannot even be verified in English". Rightfully so. The German sources are constant, and without translations cannot be verified. I don't see why an article should put the reader through further trouble by making them use online translators etc (which are usually difficult and inaccurate). Furthermore, its a Featured Article, which needs to "exemplify Wikipedia's best work".

I think the guidelines as to using foreign-language sources should be changed strictly. Unless in some extreme circumstances where exceptions can be made, most sources should be translated into English so readers can easily verify the information. For all we know, the majority of the German sources have nothing related to what they are supposed to backup in the article. Domiy (talk) 03:14, 3 October 2008 (UTC)

The policies regarding foreign-language sources should be more lenient, not more strict. When writing about topics pertaining to countries outside the Anglosphere, editors must often rely on non-English sources. Stricter guidelines would worsen systemic bias on Wikipedia. --J.L.W.S. The Special One (talk) 12:58, 3 October 2008 (UTC)

What systemic bias? The bias towards verifiablity and reliable sources? When readers can't read the language, how are they to know if the source is being reported accurately and is a reliable source? Doug Weller (talk) 13:07, 3 October 2008 (UTC)

Are you suggesting Wikipedia should only have articles about American pop culture and not Chinese cities, because most of the sources for information about Chinese cities (such as mainstram Chinese newspapers) are naturally in Chinese? --J.L.W.S. The Special One (talk) 13:36, 3 October 2008 (UTC)

There's also the question of verifying translations. Translations by editors would be frowned on, I'm sure there's a WP:rule against it somewhere. And who's going to pay for a certified professional translation? Do we have any pro translators who are editors / admins / whatever?

OTOH I agree that we should not worsen systemic bias on Wikipedia.-- Philcha (talk) 13:05, 3 October 2008 (UTC)

Read carefully what I said. It is indeed acceptable to use foreign language sources by all means, as long as it becomes a guideline for them to be translated into English for verifiability. If an editor has used a German source heavily throughout an article as he is very reliant on it, then he has every right to; as long as he provides a translation into English so others can verify it except him. Domiy (talk) 13:08, 3 October 2008 (UTC)

The policy currently requires to quote the source, in its original language, in the footnotes. Native speakers of the foreign language can thus check the accuracy of the translation. --J.L.W.S. The Special One (talk) 13:36, 3 October 2008 (UTC)

Yes but what about those who read the article but don't understand the foreign language? Why should they be doomed to not being able to verify the information? Everybody should have the same ability to verify facts. If an editor has used a foreign source then I don't see why they can't go through the trouble to merely include their own translation of the source. The fact that numerous editors refuse to do this is a heavy ground for suspicion. Perhaps they have said something in the article and backed it up with a foreign source so most people (including reviewers) will not be able to suspect anything. Well I do, and it is very suspicious that so many editors are against this. Domiy (talk) 22:09, 3 October 2008 (UTC)

A choice has to be made between allowing only content in published translations or alternately sourcing the original and providing a rough translation which polyglots can verify. The latter frequently would be a copyvio, creating something of a double-bind. An English precis of a fair-use quotation from the original (provided in the citation) is often the best available option.LeadSongDog (talk) 22:47, 3 October 2008 (UTC)

<Shrug> I can read German. For me it's verifiable.

I can't read Croatian. But I can ask someone from the Croatian wikipedia to help me. :-) So once again, I have no problem there either. I'm not seeing it? --Kim Bruning (talk) 22:50, 3 October 2008 (UTC)

This is not the first time this has come up... so perhaps we need to make it clearer... people often mistake being able to verify information with being able to verify it right now. Wikipeida requires that information be verifiable... it does not require that it be verifiable right this instant. Blueboar (talk) 23:05, 3 October 2008 (UTC)

*nod* --Kim Bruning (talk) 23:15, 3 October 2008 (UTC)

Well I don't see what the process of waiting will offer. If it's not verifiable "right now" then why would it be verifiable later in a few days/weeks time? What I'm trying to say is that Wikipedia, at its best, should be a primary tool of research on a subject (hence the purpose of the Featured Article process). Making the reader go through all the trouble, navigating them away from this website etc really defeats the entire purpose of such. You say, for example, you cannot read Croatian. Personally I will tell you that you will have an equal amount of trouble trying to ask for help on the Croatian Wikipedia. And I don't see how providing translations yourself can be a copyright violation (?) I think it is actually more accurate as the editor who used the source for the information can translate it into what he has interpreted the information as saying. This would be much clearer to the reader or anyone else who tries to verify the source. At the very least, there has to be something that can distinguish a normal translation to a translation that has been accurately checked and verified, hence is a reliable source. Domiy (talk) 01:47, 4 October 2008 (UTC)

I agree with J.L.W.S. that "The policies regarding foreign-language sources should be more lenient, not more strict."

And I disagree with Domiy's comments: "Everybody should have the same ability to verify facts" and "Well I don't see what the process of waiting will offer", both for the same reasons. That makes it sounds as though any and all books should be forbidden, just because all readers don't happen to have them at home. We do, and we should do, allow papers and books which are not online, or which require a subscription, to be used as sources. Checking them out can require some digging, whether they are in English or not.

I was recently involved in a BLP AfD, not to !vote one way or the other, but because someone asked my opinion of several sources in a language I know. I evaluated the publications and articles, but did not do any translating. That required perhaps an hour of my time, while full translations would have required more than a day or two.

In most cases, AssumingGoodFaith (as Domiy's bolded comment above does not do) is sufficient. In a few cases, more digging and 2nd opinions may be necessary. --Hordaland (talk) 08:47, 4 October 2008 (UTC)

(((The preceding comment redacted with sincere apologies to LeadSongDog for having put words in her/his mouth! An honest but nasty mistake on my part.))) --Hordaland (talk) 17:12, 4 October 2008 (UTC)

Fair enough. I see where you are coming from. However, I still think something has to be done. Perhaps the foreign sources should require, instead of a relevant quote, a relevant part/paragraph/line of the source which identifies where the source is backed up. For example, if a German source is used to back up a statement which can be found in the 1st paragraph, 3rd line of the source, then this should be specifically and directly stated so readers may know what exactly needs to be translated/verified. Domiy (talk) 10:06, 4 October 2008 (UTC)

I agree that a very simple statement of fact can be quoted and/or translated in the ref itself. That's easier than paragraph#, line#, and should be fair use. One problem is, we're often referencing statements, opinions, debates and facts which cannot be stated in 5 words or fewer. Another is that we may need to establish the reliability of the author or publication. 'Tweren't meant to be easy. --Hordaland (talk) 13:46, 4 October 2008 (UTC)

A different scenario: Lets say there is a German Feature Article. That means it has passed all the FA criteria, including having reliable sources, in German. Now someone comes along and translates the entire article as it stands into English. Surely the only reason such an article could be denied a high rating is if the translation is shown to be incorrect? Roger (talk) 16:01, 9 October 2008 (UTC)

The reasoning here is faulty on numerous counts. First, that an article is featured on the German wiki means: Nothing. We don't know that each source was checked in their review process. Wiki is not a reliable source, and we don't rely on non-reliable sources; editors are supposed to say where they got the material, and if you say you got material from a Wiki, well, you shouldn't be adding it at all. Second, Wiki articles are dynamic and can change after gaining featured status. Third, the editor entering text into en.Wiki is responsible for verifying the original source, not a Wiki, which is not a reliable source. SandyGeorgia (Talk) 06:34, 12 October 2008 (UTC)

Forgive me, but I think this whole question is ridiculous. Suppose someone sources from an obscure scholarly work, one that you can only find in the libraries of major universities. Are we going to argue that because 99% of us don't have access to the library of a major university, that source is invalid? Are we going to limit sources to only those articles that exist on the internet? Just because 99% of us don't read Chinese, or German or whatever, doesn't mean that a source is not verifiable. LK (talk) 17:54, 9 October 2008 (UTC)

Yes - and this has been discussed already, with the same conclusion (a translation may be a problem for accessability, but is not an obstacle as long as verifiability is concerned). Michelet-密是力-Me laisser un message 18:24, 9 October 2008 (UTC)

And similarly, the sources to articles like C-minimal theory are completely incomprehensible to at least 99% even of mathematicians. Nevertheless it is established consensus among the scientists who edit Wikipedia that articles on such specialised topics are notable if enough has been published about them (which is the case here).

Typically, when there are only non-English sources the topic of the article is actually most interesting for the native speakers of the language in which sources exist, and typically there will be enough of them around at the article. E.g. the initial example of German language sources for an article on German football seems completely pointless to me. The purpose of the verifiability policy is to ensure that people can rely on the correctness of what is reported here. It is most definitely not to protect the speakers of English from information that is only available (yet) in another language. --Hans Adler (talk) 18:26, 9 October 2008 (UTC)

You know, I begin to think that the reason that this is even mentioned is not about supporting facts, but about supporting direct quotations.

Think about it: if you quote something by, say, Martin Luther or Adolf Hitler from their own writings, it's going to be in German. But we don't want a block of German text in the middle of an English article. So you put up a translation, and put the original in the footnote so that others can check your translation work. This is a pretty reasonable approach.

Compare that to quoting, "Sports team ______ won the game," which seems unnecesssary to the point of being stupid. WhatamIdoing (talk) 19:35, 9 October 2008 (UTC)

That is certainly part of it... but not the entire reason. The other half of the equation does have to do with supporting facts...or at least the interpretation of facts. Often, the English words you use translate something written in a foreign source can effect its meaning in English. If a foreign source says "ABC", does it correctly translate as "XYZ" or as "something close to XYZ, but with a subtle difference"? We require both the original language and a translation so that people who know the language can see the original context... so that they can judge whether the original material is being translated in a way that is both a clear translation, and accurately conveys the contextural meaning of the original. Blueboar (talk) 20:04, 9 October 2008 (UTC)

On 9 October 2008, Margaret Atwood was interviewed on CBC Radio One about her upcoming Massey Lecture and her new book "Payback: Debt and the Shadow Side of Wealth". In the course of the discussion, she revealed that, as a child, she was intrigued by the fact that different versions of the Lord's Prayer use the words "debts" and "trespasses" in the same place. This lead her to an adult exploration of the topic and the ultimate conclusion that both English words have the same Aramaic source-word because in that society the concepts were interchangable. Two teams of accomplished translators arrived at different, widely published translations. An eminently notable author comments on the translations in a verifiable medium. Even so, a footnote explaining the source text and not just listing the authoritative translations would still contribute to the quality of the discussion.LeadSongDog (talk) 21:11, 9 October 2008 (UTC)

Just out of curiosity, how do we view this as a source (and as a reference):

Д. M. Проэктор, "Агрессия и катастрофа. Высшее военное руководство фашистской Германии во второй мировой войне", Глава восьмая. "Катастрофа", М.: Наука, 1972.

Note: No English at all is used. Not even a transliteration of the author's name. --ROGER DAVIES ^talk 04:58, 10 October 2008 (UTC)

See fr:Wikipédia:Prise de décision/Utilisation de texte dans une langue autre que le français for a rule about foreign quotations (verifiable, but in french... ;o)

"Il est possible d'inclure du texte dans une autre langue que le français, mais la traduction est obligatoire. Celle-ci doit se situer dans le corps du texte, et la source de la référence doit être donnée." = It is possible to embed text in a language other than English, but translation is required. It must be in the text, and source of reference must be given.

Actually, the translation can be given through an external source, or can be made by a wikipedian. If the translation is problematic, the NPOV policy means that the problem must be explained as such, and Wikipedia can't choose for the reader, that's all. —Preceding unsigned comment added by Michelet (talk • contribs) 06:00, 10 October 2008 (UTC)

Roger, do equally good, or better, English-language sources this subject (the German government) exist? If so, we use them. Otherwise, this is fine. Non-English sources are permitted. WhatamIdoing (talk) 18:45, 10 October 2008 (UTC)

Thanks for your thoughts. With that though comes an assumption that any source is reliable. In this case, not even the title (roughly, Aggression and catastrophe: The Nazi High Command in World War II) is accessible to the average editor so it's difficult to judge. --ROGER DAVIES ^talk 06:22, 12 October 2008 (UTC)

[edit] Burden of evidence

The source cited must directly support the information as it is presented in the article (When there is dispute about whether the article text is fully supported by the given source, direct quotes from the source and any other details requested should be provided as a courtesy to substantiate the reference.).

The current wording could be misunderstood if a reliable sources cites another (reliable) source to support the information as it is presented in the article. This practice is very common in peer reviewed publications. We have a section in the citation guideline to cover this (see WP:CITE#SAYWHEREYOUGOTIT). I would suggest removing the word "directly" from the sentence.--Philip Baird Shearer (talk) 09:32, 3 October 2008 (UTC)

Good point. Or the word 'directly' in that first line could be replaced with 'fully', if that is what is meant. --Hordaland (talk) 12:54, 3 October 2008 (UTC)

Hmmm... perhaps we do need to think of a way to rephrase this, because similar wording is being questioned at WP:NOR as well. The reason for the words "directly support" was to stop editors from misusing sources... taking something a source says out of context and using it to support a statement that the source does not actually agree with. I think we would all agree that this is something we do not want... but perhaps the wording needs to be rethought. Blueboar (talk) 23:25, 3 October 2008 (UTC)

How about saying that the "source cited must clearly support the information..." That deals with the indirection of someone citing someone else's work and also the NOR problem of taking statements out of context (e.g., to quote a relatively famous example, the assertion that "Psalm 14 says, 'there is no God'", without noting that the entire sentence reads, The fool hath said in his heart, "There is no God.") WhatamIdoing (talk) 00:53, 4 October 2008 (UTC)

My favorite "out of context" line is: "Congress shall pass no law"! (First Amendment to the US Constitution. (sometimes I wish it did end there...) Blueboar (talk) 13:00, 5 October 2008 (UTC)

I agree, I think "must clearly support the information as it is presented" is good. lk (talk) 05:29, 5 October 2008 (UTC)

As no one has done so, I'm going to be bold and change it myself. LK (talk) 09:18, 8 October 2008 (UTC)

I prefer "unambiguously" to "clearly". Is that a dealbreaker for you? --ROGER DAVIES ^talk 09:23, 8 October 2008 (UTC)

Looking at the new text, I'd like to suggest a further change to prune and tighten up:

from: "the source cited must clearly support the information as it is presented in the article"
to: "the source cited must support unambiguously the information presented in the article"

Any objections? --ROGER DAVIES ^talk 09:37, 8 October 2008 (UTC)

I don't know what unambiguously means in this situation, and how it would mean anything different from clearly. May be you could explain why the change? LK (talk) 14:37, 8 October 2008 (UTC)

Sure :) The "source cited must clearly support the information" puts the onus on the source to be clear. The source statement may not be particularly clearly written but otherwise be accurate, hence, "unambiguously". Now I think about it more, the whole emphasis can be placed on the article content by inverting the sentence structure and recasting, thus:

"the information presented must be (unambiguously) confirmed by the source cited".

Does this help? --ROGER DAVIES ^talk 11:06, 9 October 2008 (UTC)

Thanks for that. I've been thinking unambiguously since clearly was suggested, but I couldn't (clearly) explain why. --Hordaland (talk) 12:22, 9 October 2008 (UTC)

[edit] Is a publication from a fringe group a Questionable Source?

I'm being reverted on this edit where I add the word 'fringe' to the section on Questionable sources. I believe it's uncontroversial, but Vision thing (who apparently thinks some Fringe sources are reliable sources) insists that I get consensus first. So, I would like to ask, 'is it ok to describe Fringe sources as Questionable sources'? By 'Fringe', I mean as defined by WP:Fringe and WP:PSCI. lk (talk) 15:07, 3 October 2008 (UTC)

I think it would be better to refer to this as pseudoscientific instead of fringe. Count Iblis (talk) 15:37, 3 October 2008 (UTC)

I think lk's edit was well within both the spirit and the intent of the Questionable Sources section. Fringe sources are questionable sources... as such they can be reliable, in certain limited circumstances... and the limitations are the same as for any other questionable source. Blueboar (talk) 23:15, 3 October 2008 (UTC)

Citing WP:Fringe on this doesn't seem entirely applicable here, as that page is about fringe theories, not "fringe sources." --Hordaland (talk) 08:57, 4 October 2008 (UTC)

You are correct to a point... WP:Fringe is not completely applicable, but it isn't completely unapplicable either... Fringe theories are often citable to Fringe sources. I think lk's point is that a Fringe source (say a publication by a fringe religious sect) would fall into the "questionable" category. Such sources can be reliable in certain limited situations, but generally they are not reliable. Blueboar (talk) 13:12, 5 October 2008 (UTC)

Regarding the point about theory vs source, I remember being a bit surprised when reading the sources section here. Because it starts off defining reliability in terms of (reputation for) fact-checking, scrutiny etc. Fine. Then in the questionable sources subsection it's talking about sources that rely instead on rumors and personal opinions, promotional etc. Ok. But then: "or express views that are widely acknowledged as extremist". I was surprised because it switches from the issue of the editorial quality of the source, to the issue of the 'viewpoint being expressed therein. In other words, it seems to be implying that extremist viewpoints in sources means there can't have been fact-checking etc... EverSince (talk) 17:13, 5 October 2008 (UTC)

I think it's justified to characterize organizations that espouse fringe, pseudoscience or extremist views as Questionable Sources. Since by espousing such views, these organizations have given prima facie evidence that they do not care much for fact checking or the factual accuracy of their views. LK (talk) 05:32, 6 October 2008 (UTC)

I took that subsection to be talking about sources in the sense of documents rather than organizations/authors. And about how much scrutiny/fact-checking there's been of the document (and it's explictly not talking about self-publishing). I think "pseudoscience" is different, being more inherently about the nature of the facts. EverSince (talk) 20:24, 6 October 2008 (UTC)

[edit] Order of words in "Questionable sources" Section

[Note: This section used to the part of the section above, '... fringe group a questionable source?']

Having heard one of the editors involved in these changes misquote WP:V and WP:RS essentially to assert that a certain political/economic class of widely quoted sources are basically all fringe wackos who aren't reliable at wikipedia, you'll have to pardon me if I grow a bit suspicious when the editor makes/supports changes that make pages with "opinons" be the first most questionable category!! The relevant changes are in the second sentence.

Original Sept 30 version: Questionable sources include websites and publications that express views that are widely acknowledged as extremist, are promotional in nature, or rely heavily on rumors and personal opinions.

New October 10th Version: Questionable sources include websites and publications that rely heavily on rumors and personal opinions, are promotional in nature, or express views that are widely acknowledged as extremist or pseudoscience.

Because sites that rely "heavily on rumors and personal opinions" has been moved up first it casts suspicion on many sites (left, right and libertarian) that offer opinions promoting allegedly minority views, often from both lay people and experts, and infers that even if a renowned academic expert published there with footnotes, it's too "questionable" a source to use the article! I think it is best to go back to the original so the emphasis starts with extremist and promotional. Also "personal opinions" might be modified to "lay person personal opinions" just be 100% clear. Carol Moore 15:25, 10 October 2008 (UTC)Carolmooredc

The sources are listed in order from most reliable to least reliable – which makes sense, as the WP:SOURCES section also presents reliable sources in order from most to least reliable. Publications that rely heavily on rumors and personal opinions can be reliable in certain circumstances, promotional sources are highly suspect (I wouldn't usually source from an advertisement). And extremist sources are prima facie unreliable. This ordering makes sense to me. LK (talk) 15:49, 10 October 2008 (UTC)

First this section header doesn't really address the change that was made; and better late than never there is a relevant explanation for the change that was made.

As I've said before, and today replying to Blueboar on WP:RS/talk spectrum language needs to be in there and there needs to be a way to make it clear that a notable person who footnotes his material can be quoted for fact but a layperson even with footnotes cannot. And where does, for example, a Nobel winning economist publishing an article on economics in a large conservative paper fit? Only useable for opinion? Or useable for fact?? Or do people have to keep running to WP:RS/noticeboard to clear up something that better guidelines might make clearer?? Carol Moore 17:02, 10 October 2008 (UTC)Carolmooredc

I don't appreciate the personal attacks, so let me address your concerns one at a time. 1) This section isn't for discussing the reordering of the sentence, it is discussing the insertion of the word fringe into the paragraph. 2) A reordering of the examples of questionable sources does does not change the meaning of the sentence and is trivial. Your comment that my explanation of why this ordering reads better is "better late than never", is petty. If you had checked at WP:RS you'll find that this is also the ordering used there. 3) WP:RS addresses to some extent the differences between most questionable sources and extremist sources – that are prima facie unreliable. LK (talk) 08:59, 11 October 2008 (UTC)

Don't take common colloquial expressions personally. Thought provoking alternative language to deal with my concerns was posted on WP:RS and I'll quote it here:

I think part of the issue here is that the word 'source' is used interchangeably to refer to venues and people (thus Carol's objection that a reliable person can publish something in an unreliable venue, and less commonly vice-versa. maybe expanding the description like so (rough draft): "Questionable sources include websites and publications that rely heavily on rumors and personal opinions or are promotional in nature, or organization or individuals who express views that are widely acknowledged as extremist or pseudoscientific. It may be necessary to evaluate the venue and the author independently." does that move in the right direction? --Ludwigs2

What do people think? Any problem with that or a similar change?

Also note that I have these concerns not just because of specific current and past problems in editing articles, but because I know these issues come up over and over on WP:RS/noticeboard because I put together this selection of most of the discussions of important general and political oriented sources in WP:RS/noticeboard archives. Carol Moore 14:29, 11 October 2008 (UTC)Carolmooredc

[edit] arXiv papers, self-published?

Are research papers posted on the arXiv archive considered self-published? Thank you. --Phenylalanine (talk) 00:32, 6 October 2008 (UTC)

Yes, they are self-published. There's no editorial control or peer review apart from minimal checks to make sure papers are on topic. — Carl (CBM · talk) 01:28, 6 October 2008 (UTC)

I agree. Although there is a moderation process, so that sheer nonsense is excluded, the papers should probably be styled as self-published. However many of the papers also appear as working papers published by the author's university or other affiliated institution. These working papers should generally not be considered self-published. LK (talk) 07:18, 6 October 2008 (UTC)

I also agree. The quality is generally very good, and most are formally published later. But initially everybody with an email account at a university could get an account. This led to a small amount of fringe stuff being put on the server, and they have subsequently tightened the requirements, but only a bit. An arXiv article is generally as reliable as an article available from the author's home page. --Hans Adler (talk) 11:41, 6 October 2008 (UTC)

The arXiv model of retrospective open peer review (as opposed to the traditional preemptive or prospective peer review shouldn't be itself be an issue except perhaps for recent papers. Possible problems:

If an article hasn't been trashed in a year after publication, can we accept that it has had a de facto open review?
Are there any concerns that the papers now published on arXiv are not correctly attributed to their real authors, or are they adequately attributable for purposes of copyright? Does their system provide non-repudiation?

LeadSongDog (talk) 14:10, 6 October 2008 (UTC)

Of course we can't take a lack of negative reviews as an indication of quality. I agree with Hans - arxiv papers should be treated the same as papers on the person's personal website. There's no reason to think that arxiv papers are more or less accurate than any other preprint. Being put on arxiv doesn't give them any extra credibility. — Carl (CBM · talk) 16:29, 6 October 2008 (UTC)

Perhaps an examination of these withdrawals, "The arXiv moderation system" and "The arXiv endorsement system" will help clarify the discussion.LeadSongDog (talk) 16:46, 6 October 2008 (UTC)

I am somewhat familiar with the arXiv processes: [1]. The arXiv is a sophisticated preprint server that guarantees that papers remain citable in the foreseeable future. It tries to keep the amount of crap (crackpot papers and non-scientific stuff) low to save resources and keep its reputation intact. This requires a very small amount of checking, some of which is automated. My educated guess is that they phrase everything as if they were a bit stricter than the really are, to reduce the amount of arguing they have to do with the crackpot authors. But I know one former Wikipedia author who still has 12 papers on arXiv (some of which he tried to push onto Wikipedia). I have evaluated one of them, and it was incredibly stupid and would never pass a serious peer review in the appropriate field. He actually managed to get some of them published (in journals such as "Speculations in Science and Technology"), but that's another story.

The withdrawals that you cited were a widely reported (e.g. in Nature (journal)), and hopefully singular, event. They had nothing to do with accuracy (which would not be an issue, as far as arXiv is concerned), but were because of "excessive reuse of text from articles by other authors".

Scientists don't usually put random crap on their academic homepage, because it would be very bad for their career. This self-control is also expedient when publishing with arXiv. Even more so, because you can't let a paper disappear if you find out later how embarrassing it is. But all of this is already provided for by "Self-published material may, in some circumstances, be acceptable when produced by an established expert on the topic of the article whose work in the relevant field has previously been published by reliable third-party publications." --Hans Adler (talk) 22:06, 6 October 2008 (UTC)

Thanks to all! I am working on this article, where I cite this arXiv paper (mentioned here and here), and this arXiv paper (mentioned here), which concludes that the previous paper is bunk. Both papers are cited here and here. Am I allowed to cite the first arXiv paper? (the author subsequently updated his paper with a reply to the critics in the Appendix)? Cheers, Phenylalanine (talk) 23:18, 6 October 2008 (UTC)

You can definitely cite them for purely descriptive claims ("This person claims that X might happen. This other person disagrees.") If you go out of your way to explicitly attribute the claims to the authors of the articles, and and cite as many things to second-party sources as possible, you'll probably be OK. — Carl (CBM · talk) 13:05, 7 October 2008 (UTC)

That provides an excellent case in point. Note that the first paper's abstract is followed by "Comments: Now contains a reply to a comment on this paper (0808.4087) as appendix. Comments and criticism are still highly welcome". If the arXiv process makes these comment linkages independent of the preferences of the original paper's authors, then it greatly enhances the objectivity of the process and reduces the likelyhood of crap remaining uncriticised. LeadSongDog (talk) 16:52, 7 October 2008 (UTC)

I am not sure if you are reading something into this that isn't there. What happened is that someone criticised the original paper, so the author uploaded a new version with an appendix that replies to the criticism. Of course he was under no obligation to do this, and most crackpot authors would not do it. This is exactly the same situation as with self-publishing on one's homepage.

What is potentially interesting, though, are the two trackback links from the abstract page of the paper. Unfortunately it turns out that they are not useful in this case, and we can't rely on enough scientists using blogs for criticising papers anyway. --Hans Adler (talk) 20:50, 7 October 2008 (UTC)

I did say "If the arXiv process..." precisely because I didn't know what that process does. If we compare the traditional journal publishers, they provide a commentary process, whether it be "letters", "comment in" or whatever. It would be very disappointing to learn that arxiv doesn't do this.LeadSongDog (talk) 05:22, 8 October 2008 (UTC)

[edit] The phrase "verifiability, not truth"

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

This worries me greatly. I mean if Wikipedia is supposed to be the "sum of all human knowledge" then how can we only allow verifiability?

There must be some give-and-take involved here. I'm sure that this is an age old question, but please humour me. I've seen many maths additions be removed because of lack of verifiability, even though the things written are quite clearly true!

If I were to have enough money I could self-publish a book full of a whole range of falsehoods and then come onto Wikipedia and post the stuff because it's verifiable by the definition of these guidelines. (I had previously argued against self-published work but found myself against a more vocal minority.) For example: I could cite many verifiable sources that say that NASA never went to the moon, and (disgustingly) that the Nazi holocaust never happened.

If I'm talking a load of rubbish then please let me know nicely!

Δεκλαν Δαφισ (talk) 19:34, 6 October 2008 (UTC)

Don't despair. Although the debate about what "verifiability v. truth" means has gone on and on, our fundamental goal is to write an encyclopedia, and that carries with it the obligation that our content is "correct" in a certain sense. You can read "verifiability, not truth" as "verifiably in agreement with the published literature on the topic"; then the extent to which the articles are "correct" will reflect the extent to which the published literature is "correct". The word "true" is too loaded with meaning, and there are too many armchair postmodernists, for discussions on Wikipedia about "truth" to get very far. The moral, though, is that we do try to make our articles correct in the sense that the average reader would expect. — Carl (CBM · talk) 19:52, 6 October 2008 (UTC)

The reason for it is the fact that there is no such thing as 'the truth', it's all subjective, every editor has an idea of the what is truth so we rely on sources that have a reputation. Any other way would be impossible. Also note that common sense always applies and obvious things do not necessarily need sourcing and vanity press releases are not usually acceptable as reliable sources. Also the examples you cite are usually considered fringe theories and are represented as a opinions and views not as truth. --neon white talk 20:34, 6 October 2008 (UTC)

"Verifiable, not True™" is usually invoked when an editor wants to include something that s/he "knows" is "true", because -- oh, my mother told me that's how it is. Or because I'm pretty sure that I saw a magazine article about that a while ago. Or maybe it was someone's blog somewhere. Anyway, I'm sure I remember it correctly, and I sincerely believe it, so that's good enough, right?

This rule does not excuse the addition of false information; it merely limits the information to that which is verifiable in a reliable source. The contents of the encyclopedia would ideally be both verifiable and accurate. WhatamIdoing (talk) 03:13, 7 October 2008 (UTC)

Declan Davis, your example about being able to cite "many verifiable sources that say that NASA never went to the moon, and (disgustingly) that the Nazi holocaust never happened." is exactly why we set the threshold at verifiability and not truth. We would not be able to write serious article about the Apollo moon landing hoax accusations or Holocaust denial without this statement. To do justice to any topic, you have to be able to discuss what different people say about a subject... even if we think what they are saying is not true. We have all sorts of policies and guidelines that limit how we discuss differing views. The key for me is to attribute the views to those who holds them... in other words express views as statements of opinion, and not as statements of fact. Blueboar (talk) 15:21, 8 October 2008 (UTC)

[edit] A little help?

I've been writing aviation-related articles for about a decade now. During that time I've come to rely on Joe Baugher's extensive list of American Military Aircraft. This list consists mostly of articles created by scraping other sources, essentially identical to the process used to create articles for the Wikipedia. They tend to be much longer than suitable for a Wikipedia article, but the creation process is similar, and the results excellent.

On several occasions I have been able to check his references after the fact. I have yet to find a single error that isn't in the original source. The simple fact that he includes his references in the articles is a good indicator of their quality, IMHO.

So now with that out of the way, I would like to gather some opinions on whether or not this is a suitable reference source. It is claimed that it is not, because it is a self-published source. However, the criterion right above SPS, "Reliable sources", states "Articles should rely on reliable, third-party published sources with a reputation for fact-checking and accuracy.", a label that his works fit to a T.

Soooo, which is it? In my opinion, Reliable trumps SPS. Further, IMHO, that SPS's primary concern is weeding out low-quality sources like moon-hoax conspiracy sites, which are generally SPS, and not to weed out high-quality sources just because they are SPS.

Am I interpreting this correctly?

Maury Markowitz (talk) 21:00, 9 October 2008 (UTC)

It's worth observing that Joe Baugher is notable in his own right as an established author on the topic in mainstream press. That's one criterion used to establish such a reputation. But the discussion should really be at Wikipedia talk:Reliable sources, not here.LeadSongDog (talk) 21:30, 9 October 2008 (UTC)

Actually, I'd take it to WP:RSN, the reliable sources noticeboard, which exists for the sole purpose of dealing with these kinds of questions. WhatamIdoing (talk) 02:38, 10 October 2008 (UTC)

Done! Maury Markowitz (talk) 17:53, 10 October 2008 (UTC)

[edit] When is an 'English verifiable' source available?

I read the complete section on foreign languages and discussion on it. The question I have is that if there is an English source on an issue that says, for example, value of X is 7, and there's a non-English source that says, no in fact X is 17, since a source in English is available that has the 'value of X', does non-English source become inadmissible? Or since the viewpoint is different, both are admissible as contradicting point of views?

I tend to believe latter, but I want to get some expert advice before going back on the article and pushing for the updates. Thanks. Omer (talk) 08:57, 11 October 2008 (UTC)

Assuming that the two sources are of equal reliability, you would be correct to mention both views. All that the foreign language section indicates is that if there are multiple sources for the same information, some in English and some in another language, the ones in English are preferred, as they are more easily verifiable by readers of this version of Wikipedia. This section is not intended to "over-rule" WP:NPOV. Blueboar (talk) 16:22, 11 October 2008 (UTC)

Página espejo de la Wikipedia

Directorio de Enlaces Directorio dmoz Directorio espejo dmoz Pedro Bernardo