Wikipedia talk:Verifiability

Replacements for VnT

Jehochman:

One standard for inclusion in Wikipedia is verifiability, not truth. Material in Wikipedia must have been published by a reliable source. Editors may not include content merely because they think it is true.

Becritical:

Verifiability is the ability to cite reliable sources that directly support the information in an article. All information in Wikipedia must be verifiable, and only reliable sources may influence the decision to include or exclude information. Whether editors personally believe the information is true or false should never determine Wikipedia content. Verifiability does not guarantee inclusion, because Wikipedia has other policies and guidelines that affect inclusion. If the verifiability of any text in Wikipedia has been challenged, or if it is likely to be challenged, the source must be stated in the form of an inline citations. Appropriate citations guarantee that the information is not original research, and allow readers and editors to check the source material for themselves. Any material that requires a source but does not have one may be removed without further discussion, and unsourced contentious material about about living people must be removed immediately. For help on adding citations, see Citing sources. This policy applies to all material in the mainspace.

Blueboar: (This is a re-working of the entire lede section)

Verifiability is one of the core concepts of Wikipedia. Our readers must be able to verify that the information presented in an article has been presented accurately. This is achieved by citing reliable sources that directly support the information in an article. All information in Wikipedia must be verifiable. Editors should not add unverifiable material, even if they are convinced that the material is true. In this context, the initial threshold for inclusion is Verifiability, not truth.

However, Verifiability is not the only threshold for inclusion. There are other policies and guidelines that influence what information may be included in an article. The fact that information is verifiability does not guarantee its inclusion.

Note that the policy requirement is for verifiability, not actual verification. It must be possible to attribute the information in a Wikipedia article to reliable, published sources that are appropriate for the content in question. However, in practice it is only necessary to provide inline citations for quotations and for any information that has been challenged or that is likely to be challenged (see below).

Verifiability, No original research and Neutral point of view are Wikipedia's core content policies. They work together to determine content, so editors should understand the key points of all three. Articles must also comply with the copyright policy.

Edit request on 27 March 2012

Change the first sentence from:

"Verifiability on Wikipedia is the reader's ability to check cited sources that directly support the information in an article."

To:

"Verifiability on Wikipedia is the reader's ability to check that cited sources directly support the information in an article."

Currently, the first sentence implies that 'cited sources that directly support the information in an article' definitely exist, and that 'Verifiability' is the ability to of a reader to 'check' them (dunno what that means, read and grade them?). This is a typo I believe. By moving the word 'that', the sentence takes on a more reasonable meaning, that verifiability is about the ability to check (verify) that cited sources directly support information in the article. LK (talk) 06:12, 27 March 2012 (UTC)

I think the former is what readers do and the latter is what editors do. Seems correct to me as is. — Bility (talk) 16:37, 27 March 2012 (UTC)

Like the change but oppose making it now. This section is the reason it was locked and it a subject of a mediation effort. North8000 (talk) 17:33, 27 March 2012 (UTC)

As North says... we are currently in the process of mediation over the opening paragraphs of this policy... so we should hold off on making any changes until that mediation is complete.

That said, the suggested change highlights a question that has been looming in the background and has never been adequately addressed... when we cite a source, what exactly are we verifying?

The current language implies that we are verifying the fact that sources supporting the information exist. This reflects the original intent of the policy... the requirement for verifiability on Wikipedia grew out of WP:NOR, and the original concept was essentially that we needed to verify that the material we add to our articles is not Original Research.

The suggested change would shift us away from that original intent... if adopted, we would be saying that we are verifying the information presented in the article. That may be the way we want to go... but it would be a fairly significant change from the original intent... and not a step we should take without a lot of thought and discussion. Blueboar (talk) 17:36, 27 March 2012 (UTC)

Actually, I thought that a piece of information is not supposed to be in an article in the first place, unless there is a source for it. (Except for information about which there is no dispute, or at least that's what this policy implies.) And hopefully, it is a source that says (or implies) that the information is true, otherwise it isn't worth much as a source. So I think that in a way, we are using the source to "verify" that the information presented in the article is true, in fact we are using the source to provide the information that we put in the article. Right? Neutron (talk) 22:59, 27 March 2012 (UTC)

The "cited sources" bit is wrong. We care whether (reliable) sources exist, i.e., have been WP:Published in the real world. Material must never be included unless "there is a source for it—in the real world." Material may be included without a source being WP:CITEd in the article (unless it is one of the four types of material listed at WP:MINREF, for which iWP:Inline citations are mandatory [only one of which has a deadline for providing the citation; the other three require an inline citation either before the heat death of the universe or whenever your fellow editors choose to deal with it).

I generally do not support changes to the lead during the mediation, but this is actually an error. The simplest solution is to remove the word "cited" entirely. WhatamIdoing (talk) 00:43, 28 March 2012 (UTC)

Now that you mention it, "cited" clearly is an error. Possibly a non-controversial fix? Note that this is not the original edit request. North8000 (talk) 00:57, 28 March 2012 (UTC)

As is inevitable with any one sentence pseudo-summary/pseudo-introduction, it is imprecise. And, upon rigorous dissection, "imprecise" means "wrong", so we should not give it a rigorous dissection. In Wikipedia, the definition of wp:verifiability is "complies with the wp:verifiability policy". And that's not using a word to define itself; the first is an attribute, and the second is a policy. Sincerely, North8000 (talk) 23:28, 27 March 2012 (UTC)

In any case, this issue is likely to be fixed by what ever we come up with at the Mediation... so there is really no point in discussing it too much now. Blueboar (talk) 03:23, 28 March 2012 (UTC)

To remove the erroneous word "cited" from the first sentence, per comments above. (The link to WP:RS should be kept.) WhatamIdoing (talk) 04:34, 5 April 2012 (UTC)

I disagree with the change because I do not believe that "cited" is an error. A reader has the practical ability to verify the content of an article only if the sources for this content are actually cited in the article, and not if they merely exist somewhere but are not cited. Therefore, verifiability does require that sources are actually cited in the article in conjunction with the material that is thereby made verifiable. I also do not support the initially proposed wording change for this reason.  Sandstein  17:07, 11 April 2012 (UTC)

"All information"

Could we swap the second sentence "All information in Wikipedia..." to "All contested or potentially controversial information" (or something like that) in compliance with the second paragraph and WP:When to cite? Brandmeister^talk 22:16, 28 March 2012 (UTC)

As with the discussion thread immediately above, this is one of the things being looked at in the current mediation process. However, in my opinion, the second sentence is really referring to all information, not only that which is contested. --Tryptofish (talk) 22:22, 28 March 2012 (UTC)

This is correct as it stands. NB that we make a distinction between what is verifiable and what is WP:CITEd. All material must be verifiable (possible for someone to verify, assuming he's willing to go to sufficient amount of work to find the sources himself). Only contested material must be cited (provided with the name of a source that supports the material, so that it's easier [but not necessarily easy] for other people to verify the material).

This is, BTW, addressed in the FAQ, currently third question from the bottom (and will probably be fourth from the bottom before long). WhatamIdoing (talk) 19:17, 29 March 2012 (UTC)

I agree I think its grammatically correct and fits perfectly, see no reason for a change. Maxasher (talk) 03:24, 5 April 2012 (UTC)

Kindle books don't have page numbers; how do we cite them?

I'm not sure where I should ask this, but it's directly related to WP:V so I'll ask here. I just discovered that apparently Kindle eBooks don't contain page numbers. Instead, it gives a location. In my particular case, the cite I wanted to add was at location 1970 out of 6998. I brought this up at the article talk page and also at the Computing Reference desk. So, I guess I have two questions:

1. Is there currently a citation template that allows us to cite a Kindle book?
2. If not, how do we go about creating a new citation template to handle this situations?

A Quest For Knowledge (talk) 22:33, 5 April 2012 (UTC)

Kindle books do actually normally (but not always) have page numbers. I'm not sure if there's a quicker way, but if you do a text search for a distinctive phrase in the part you want to cite, you will normally get a page number with the result. That's using a Kindle Keyboard. FormerIP (talk) 22:42, 5 April 2012 (UTC)

I just picked up my Kindle and noticed that you can get the page number just by clicking "menu". FormerIP (talk) 22:44, 5 April 2012 (UTC)

Do we know if the page numbers on the Kindle version are the same as they are in paper? If so, just cite it as if it were paper. Blueboar (talk) 22:54, 5 April 2012 (UTC)

I hope someone will investigate the potential variations in Kindle citations. Do they vary depending on any possible device settings? Do they vary from one type of Kindle to another? Whatever factors cause a change in pagination will have to be included in the citation. But there is no reason not to cite a Kindle page number if it does not match the page in a paper edition, just as there is no reason not to cite a page number in a paperback edition just because it does not patch the page in a hardcover edition. Jc3s5h (talk) 22:59, 5 April 2012 (UTC)

I don't have a Kindle device. I have the Kindle app installed on my iPad. I don't see a way to view a page number. I also tried the Kindle app for Windows and the Kindle web reader. None of them seem to have a way to see a page number.

The location was consistent across the iPad, Windows and web reader, so I don't think that this is an issue. A Quest For Knowledge (talk) 23:04, 5 April 2012 (UTC)

There is certainly no WP:V issue with citing a kindle version... It is just a question of how to format it. I would suggest raising the issue at WP:CITE... it is more in the scope of that guideline than a policy issue for this page.

(ec) I'm assuming the page numbers match up to the paper editions. But there is also a location number, which I think goes by screens in a typical Kindle display. If you can't get page numbers, maybe just do an oldschool non-templated citation. FormerIP (talk) 23:11, 5 April 2012 (UTC)

Glancing at some information from the Amazon site, it seems that the PC version lets you display the page number, but the i-whatever blurb does not brag about that capability. It seems to all be the same thing that you buy and view on several different devices, according to your whim of the moment. So I suppose I would cite it with |edition=Kindle and the page number. If you discover page numbers are not consistent (when present) across all devices, I would not give a page number and instead give a quote that can be searched for. Jc3s5h (talk) 23:24, 5 April 2012 (UTC)

The Windows app doesn't show page number for me. Maybe it's the particular book I'm reading? The Amazon web site seems to leave some wiggle room with the statement that "This feature is available with thousands of books in the Kindle Store" as if it's not available for all books. According to this page, not all Kindle devices support page numbers. A Quest For Knowledge (talk) 23:33, 5 April 2012 (UTC)

The answer's already in CITE. See WP:EBOOK. WhatamIdoing (talk) 21:17, 10 April 2012 (UTC)

Ah, thank you. :) A Quest For Knowledge (talk) 21:19, 10 April 2012 (UTC)

Fall and Decline of the Roman Empire by Edward Gibbon

Please "verify" quote and material under the sub-title: "Christianity as a Contributor to the Fall and to Stability". I have a three volume set of "The Decline and Fall of the Roman Empire" by Edward Gibbon; The Heritage Press, NY; The special contents of the edition are copyright, 1946, by THE GEORGE MACY COMPANIES,INC., and Chapter XXXIX (chap. 39) and your quoted material is not here as cited. In Chapter 39, the fall of the Roman Empire in the West has already occured and this chapter is about the invasion and conquest of Italy and the Gothic king, THEODORIC the Ostrogoth. — Preceding unsigned comment added by Sandrameyer (talk • contribs) 18:40, 8 April 2012 (UTC)

This page is for discussing the verifiability policy, not for checking individual sources. You could raise your question, instead, at the talk page of the article where the material occurs, or, if you need further help, at WP:RSN. --Tryptofish (talk) 19:35, 8 April 2012 (UTC)

Accessibility

Compare this sentence:

"The principle of verifiability implies nothing about ease of access to sources"

with this one:

"Because this is the English Wikipedia, English-language sources are preferred over non-English ones, assuming English sources of equal quality and relevance are available."

In the first one, and certainty in the linked essay - Wikipedia:Reliable sources/Cost, the implication is that accessibility is not regarded as a high priority; while in the second one it is clear that accessibility is a high priority. As we are not writing doctorates or other scholarly papers, but general interest articles for a general encyclopaedia for the general public, who are usually looking for quick and reliable information rather than an in depth scholarly analysis or a degree course in the subject matter, then accessible sources should be the priority. We should prioritise a decent and recent text which has been scanned on Google Books, over a 100 year old text only found in an out of print bookstore in Greenwich Village or in the vaults of the British Library. We should be directing readers to decent accessible sources, not making matters difficult because we have no clear policy on it.

When reading a Wikipedia article I tend to check the sources, and if the sources are print only (even if easy to obtain), then I am frustrated. If I need to go to my library to get reliable information, then the ease and much of the value of using Wikipedia is lost (and if I have to go to my library, then I will first get whatever texts my library has - regardless of what texts have been used in building the Wikipedia article). I have done more than a handful of Good Article reviews where on checking sources I find the information in the source does not quite match what is said in the article, and even highly respected and experienced editors have made mistakes in copying over/summarizing information. I have, over time, learned not to completely accept sources in good faith - indeed the actual point of sources is that they are there for me to check, and additionally for me to learn more from. The less accessible the source, the less verifiable and reliable the article becomes, and the less I am able to learn. I would regard accessibility to be a high priority, and if the same information is contained in a closed source and an accessible source, then - at the least - the accessible source should also be included, and, if appropriate, preferred.

A suggested alternative wording:

The principle of verifiability implies that information is able to be verified, so consideration should be given to citing reliable sources which are the most accessible. Ideally a search should be done to see if there is a freely accessible online version of a text on a website such as Project Gutenberg, Google Books, Feedbooks, or Wikisource, and citing to the scanned page so that a reader is able to immediately verify the information in context. Occasionally information may only be found on online sources requiring payment, or in print sources only available in research libraries; such restricted sources are encouraged if no accessible reliable or comparable sources are easily found. WikiProject Resource Exchange may be able to assist in obtaining source material.

Does my wording go too far in the direction of accessibility? Could it be better balanced? Thoughts? SilkTork ^{✔Tea time} 19:51, 10 April 2012 (UTC)

• There's a tension between making it easy to check articles and making it easy to write them. I think this proposed change goes too far in the wrong direction: we still need to write material. For example, we should clearly have something on Flora of France, but when I finally get around to writing it, I don't want to take any grief for using the foreign-language print sources that are on my bookshelf. I do not wish to trawl around the snippet view in google books looking for a way to satisfy suspicious and hostile editors of my integrity. Therefore, I am opposed to this change.—S Marshall T/C 20:08, 10 April 2012 (UTC)

• Oppose. Your proposal is flawed for several reasons. First, we are not just writing "general interest articles for a general encyclopaedia for the general public". By its breath and being a digital encyclopedia we have lots of specialist articles. You may want to read up on pillar 1. Quoting: "[Wikipedia] incorporates elements of general and specialized encyclopedias, almanacs, and gazetteers." Second, we are aiming for the best sources on a topic, not for its most recent and most accessible one. If the best books are in hidden in a library then you should consider entering one. Third, inline sources are first and foremost provided to comply with our verifability policy; at the same time, they should indeed help you to learn about a topic. But you cannot seriously expect that all the best books are flying around on the net for you to download for free. If you really want to learn something, going deep into a topic, you should certainly consider buying some good books or textbooks on it. Fourth, nothing is more annoying that someone with no knowledge on a difficult topic, coming along with some Google Books searched references and claiming that these are to be put into the article, when they are clearly poor and inappropriate. But on finding a better balance: certainly, when you've got two equal-quality sources, and one is free and the other one is not, you should go for the free one. But preferring a worse source for a paywalled one, no. Btw, there is always WP:WikiProject Resource Exchange/Resource Request. Nageh (talk) 20:47, 10 April 2012 (UTC)
• Oppose. Hear, hear! To write a high quality article, you have to draw on the best available material. And for some years to come, these will be on paper. Researching the article on Robert Oppenheimer, I read several high quality books from cover to cover, one of which won a Pullitzer Prize (the subject is quite well covered); but few of these are are available online, and some may not be for decades. Reviewing another article recently, I travelled to a nearby library to obtain the books necessary to perform the spot checks. Someone will be able to repeat this process in a few years; but this may not be true of online sources, which come and go due to link rot. Hawkeye7 (talk) 21:53, 10 April 2012 (UTC)
• Oppose as this would discourage the use of almost any post-1923 academic journal material in citations. Content must be verifiable in principle, not necessarily in every single case. In a research paper, for example, or a nonfiction book, references/sources are often present but that doesn't mean there exist for every reader to be able to access. Sources should be accessible to at least some number of interested readers, but not all of them. Yes, English and free sources are preferred because it makes verifying easier, but we should not have to spend extra time finding similarly relevant sources if using a nonfree or non-English one suffices. /ƒETCHCOMMS/ 23:43, 10 April 2012 (UTC)
• Oppose - This is exactly backwards of where this project needs to be going. There is much too much sourced out to quasi-semi-dubious web sources because that stuff is "easy"... What's needed is for people to get their asses down to the library and start making use of university press monographs and journal articles. The BEST sources are often times less accessible or FAR less accessible than the web gunk. Nor should we be kissing Google Books' behind, since that is getting to be more and more of a for-profit enterprise, as expected. Carrite (talk) 01:29, 11 April 2012 (UTC)

And here's an example of what I'm talking about (work in progress, don't go all ONESOURCE on me!)... Compare the very accessibly-sourced, lightweight, and shitty A versus the correctly sourced out B. Carrite (talk) 01:35, 11 April 2012 (UTC)

See my comment below, where I indicate that the book you are using is available for snippet views - [1]. SilkTork ^{✔Tea time} 09:29, 11 April 2012 (UTC)

• Oppose I suggest you manage your frustrations on your futon, and leave encyclopaedism to people willing to audit the full scope of available sources. Fifelfoo (talk) 01:38, 11 April 2012 (UTC)
• Unfortunately Oppose. For many subjects, the best sources are still relatively inaccessible. Perhaps we should revisit the issue in 4 or 5 years and see where things are at. Kaldari (talk) 04:23, 11 April 2012 (UTC)
• Oppose, as exactly how not to do it. This is supposed to be an encyclopaedia, compiled from the best sources we can find, not a collection of trawlings from the net. As for Google books, it should never be cited as a source, because it isn't one. The source is the book itself, and if you don't have access to it, you shouldn't be citing it... AndyTheGrump (talk) 04:50, 11 April 2012 (UTC)

I'm not actually arguing any point as I can see there is currently no consensus - I just want to clear up a misunderstanding. Google books is not a source, it is a resource - it contains sources. It may help to think of it as a library where texts may easily be found. So if someone is using a text, it seems worth checking to see if the text is available on Google Books. It's a quick check, simply copy and paste the name into Google Books, and see what is available. For example, Carrite is using James N. Giglio's H.M. Daughterty and the Politics of Expediency for the Harry M. Daugherty article. His first cite is to support that Daugherty's father, John H. Daugherty, was the son of Scotch-Irish immigrants and worked as a farmer and tailor. Well, the book is not available fully scanned, but it does have a snippet view that supports the info. It's not much effort to link to that page, but can aid the reader. While a snippet view is not ideal, it's a bit better than nothing at all. SilkTork ^{✔Tea time} 09:23, 11 April 2012 (UTC)

And who tells you that just because a snippet view is available this is really a good source on the topic? Maybe it is really poor one otherwise, and you are misled into further looking up this source (wanting to learn more about it as you claim)? It really is at the discretion of the editor to decide which sources to use, at his/her best knowledge and in accordance with our guidelines. Nageh (talk) 11:02, 11 April 2012 (UTC)

• Oppose per the above. While we don't get points on Wikipedia for tracking down obscure theses or editiones princepes we shouldn't discourage the use of print sources or paywalled scholarly databases where they are the best sources, as they very often are. A sentence pointing out the desirability of using (or at least linking) freely available source material might be crafted, but this is not it. Eluchil404 (talk) 07:13, 11 April 2012 (UTC)
• Oppose per the above. Further, Google links should be viewed as transitory when pages cited have been checked, read and the book is properly cited per WP:RS standards. Kierzek (talk) 15:35, 11 April 2012 (UTC)

Need uninvolved editors to examine my analysis of a source

At Wikipedia:Arbitration/Requests/Enforcement#ZScarpia_and_WLRoss, there's an issue over the proper use of a source. According to my own analysis, the use of this source fails criteria #2 of WP:SPS but no one else is agreeing with me. Can some uninvolved editors examine this source and let me know what they think? A Quest For Knowledge (talk) 03:31, 12 April 2012 (UTC)