Wikipedia talk:Verifiability

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Policy and Guidelines
WikiProject icon This page is within the scope of the Policy and Guidelines WikiProject, a collaborative effort to improve Wikipedia's policies and guidelines. If you would like to participate, you can visit the project page, where you can join the project and see a list of open tasks.

Citing objects rather than claims[edit]

I'd like to get some clarification on a topic that seems to come up for me again and again. I frequently see references included in articles that point to an object of the article itself rather than a claim about the object. Some recent examples:

  1. I worked at an article the topic of which had been the subject of numerous works of fine art. Claims were made along the lines of "The 1745 depiction by William Hogarth included pastoral flourishes." The citations were to jpg-file copies of the artworks hosted by reputable fine arts sites.
  2. I gave advice on a video game article where certain gameplay elements were the topic of discussion. Claims were made along the lines of "The main character can defend himself with a bioforce shield or by using a double jump to escape." The citations were to the games themselves.
  3. I reviewed an article about a performer whose act was described in detail. Claims were made along the lines of "In one show Mr. Knievel jumped over 10 schoolbusses laid end to end." The citations were to episodes of the performer's show.

I'm broad-minded about sources and per WP:PRIMARY and WP:SPS I believe that sources such as those in my example can often be used, however in in all three of the above cases I objected to their use due to the fact that in my view none of them make a claim. This is kind of similar to the copyright concept of the idea/expression dichotomy. The intangible idea that a painting is pastoral cannot be cited whereas the concretely expressed claim (i.e. the tangible claim fixed in a text or audio medium) can be cited. In the first example, it may be true that Hogarth's work is pastoral but to me this represents the interpretive claim of the Wikipedia editor. In the second example, it may be true that the character can use a certain shield or a particular evasive maneuver but the game isn't actually making a claim by including these elements in its substance. In the third example Knievel may indeed jump over 10 schoolbusses and the editor, the reader, and all others surveyed might also come to the same conclusion after watching the video but this conclusion is in fact nowhere concretized in a tangible medium.
I view references as borrowed authority. As an editor I view my job as presenting the claims of those who are in a position to make authoritative claims. My intention is for the reader to be able to verify claims with links to claims made by an authority. What I see in the above examples are invitations by the editor to the reader to perform her own analysis to verify the analysis of the editor interpreting primary source material. The counter-argument I tend to see is that these claims are obvious. ("Clearly Knievel jumps over 10 busses. Count them yourself. Here's a link to the YouTube video.") I consider this kind of argument to be an appeal via WP:BLUE/WP:MINREF. My position is then that these claims don't need citations at all. If they are obvious, non-controversial, unlikely-to-be-challenged claims then they don't need any reference at all. Well, community? Am I right about this or am I being a stickler for no good reason? -Thibbs (talk) 13:22, 8 October 2014 (UTC)

I think you're putting too much emphasis on the notion of "claim." While that word is used in V occasionally, the basic rule, BURDEN, is stated in terms of "content" and "material," not "claim". Also remember that "likely to be challenged" in BURDEN doesn't mean that it's likely to be challenged after viewing the cited source, but whether its likely to be challenged without regard to the source. Personally, I'd think that all three of those examples would be likely to be challenged and ought to be reliably sourced; whether the sources advanced are adequate is a different question: Of your three examples, I believe the first should be sourced to a reliable source other than the work itself because whether something is "pastoral" or a "flourish" or not requires an element of artistic analysis, definition, and judgment, not mere observation and common meaning. The third one I'm okay with (provided that the link to the YouTube video is not a copyvio, as it often is) and there's no visual ambiguity about the number of the buses or the fact that it's Knieval. The second one turns on just exactly what the game shows: Is the double jump a defense, per se, or is it just something the character can do which might be used in that way? Is the "shield" clearly named in the game as a "bioforce" shield or is it just a self-generated shield which might be telepathic, a mutant power, a magical power, etc.? Are there other possible defenses since this claim implies that these are the only two? Without actually seeing the game, I can't answer that but would AGF unless I was extremely sure that it didn't show it. Finally, remember that articles cannot be entirely or primarily built upon PRIMARY sources. That's my two cents. Regards, TransporterMan (TALK) 15:52, 8 October 2014 (UTC)
 An interesting question, well put. And I think that "claim" is proper here, as basically everything in an article can be taken as a claim of some sort. This also touches on original research (prohibiting "any new analysis or synthesis of published material that serves to reach or imply a conclusion not clearly stated by the sources themselves"), and particularly on routine calculations. While there could be some variance as to what constitutes a "pastoral flourish", counting what are clearly ten school buses would seem clear enough. (Or does it matter that, say, one of them is not really a school bus? Sigh.) This is perhaps no different in the essentials than counting the number of theories listed by an author, which (presumably) can be verified by any editor.
 But observations of game experiences seems rather less than obvious. E.g., how do we directly observe that the hero can defend himself without engaging in the game? And how do we know that this applies only for certain cases? To make a definite statement (lacking a clear statement by the source itself) would seem to require considerable experience with the game, perhaps across a wide span of players, which implies expertise beyond that of the typical editor. If an editor did have such expertise, his/her writing on the subject would be more in the mode of first-person journalism. ~ J. Johnson (JJ) (talk) 21:55, 8 October 2014 (UTC)

My thanks to both of you for your thoughtful responses.

  • TransporterMan, I appreciate your comments on example 2 regarding whether the shield was "clearly named in the game as a "bioforce" shield". That's my preferred line of inquiry on that question as well. On reflection I think you're probably also correct in your interpretation of "likely to be challenged" from WP:MINREF. J. Johnson has probably found the exception that is most closely on point for counting objects, though it still leaves the question about simple analytical claims like:
    • "X has brown hair.<ref>X's Facebook profile picture</ref>" or
    • "Y eats apples.<ref><nowiki>Images from ''Le Monde'' depicting Y eating an apple.<nowiki></ref>" or
    • "Factory errors at the pressing plant resulted in some copies of The White Album with two A-sides.<ref>A copy of the erroneous pressing bearing official pressing plant identifiers.</ref>"
WP:BURDEN is not a bad point of reference, either, but I guess my main question is what is meant by "directly support" in WP:BURDEN? Is (A) a citation to an artwork/game/episode considered "direct support" for encyclopedic claims regarding the elements of the artwork/game/episode? Or does (B) "direct support" require tangible claims made by a claimant? In the B case there is a third party (i.e. not the Wikipedia editor and not the reader), even if it's only the author of the work, who serves as the authority supporting the encyclopedic claim. In the A case the editor is expressing what she hopes is common sense that the reader will agree with if verification takes place. I think J. Johnson may have hit on the precise policy at issue.
  • J. Johnson, that line from WP:OR is exactly what I was looking for. Specifically, the use of the word "stated" gets to my point about concrete/tangible/fixed expressions constituting claims. Clearly looking at a picture and determining that X has brown hair is not much of an analysis, but WP:OR pretty clearly bars "any analysis". This is where I start to feel extremely pedantic. I feel silly arguing for stronger sources than something clearly and easily analyzable which when analyzed could really only result in one conclusion. Perhaps this is one of those rare cases where WP:IAR should be used. But how would IAR manifest itself here? Would it be better to ignore WP:V and say that no sources are actually needed? Or would it be better to ignore WP:OR and ask for sources that don't actually state the claim but require a tiny bit or reader analysis to verify? Or should that simply be a case by case matter - say for example ignoring WP:V unless challenged and ignoring WP:OR if/when challenged?

I'm still interested in hearing from others as well. These kinds of questions can be quite irritating when they arise at FAR, GAN, DYK, etc. so the more opinions the better. -Thibbs (talk) 12:43, 9 October 2014 (UTC)

We don't get to dictate to the creators of reliable sources how they should write. It is not appropriate to disqualify a source that contains the information we are interested in just because it isn't stated in the manner some silly Wikipedia guideline wants it to be stated. Jc3s5h (talk) 13:51, 9 October 2014 (UTC)
While WP:OR does not allow editor to engage in analysis or interpretation of objects... it does allow us to give basic description of objects (which can, if needed, be cited to the object itself). The question is often whether the description is basic enough. For example...
  • The statement "The west end of the cathedral is dominated by a large rose window" is a basic description that can be verified by anyone who goes to the cathedral and looks at the west end itself.
  • The statement "The west end of the cathedral is dominated by a large rose window depicting the martyrdom of St. Eggfroth" is iffy... there is a secondary question as to how we know that the window actually depicts St. Eggfroth. Does the window contain the Latin caption "Hic Est Eggfroth" so the viewer knows who is depicted?... Or is the statement based on a wikipedia editor's semi-expert knowledge of medieval iconography (for example, knowing that you can recognize St. Eggfroth because he is usually depicted holding a flower pot)?... if the latter, then it is OR... since the average person would not be expected to know this fact.
  • The statement "the west end of the cathedral is dominated by a rose window in the English Perpendicular Style" would definitely be OR without a secondary source... the average person would not be able to verify the statement by simply going to the cathedral and looking at the window... recognizing the English Perpendicular Style requires some degree of expertise.
In other words, OR often depends on exactly what is written in the article. Blueboar (talk) 16:15, 9 October 2014 (UTC)
Thanks for your thoughts, Blueboar. I'm not so much talking about blatantly interpretive OR as I am about the kind of common sense claim that references a source in which no such claim is stated. So the first example statement ("The west end of the cathedral is dominated by a large rose window.") is directly relevant. Here we have a basic descriptive claim generated from first principles by an editor whose novel statements give voice to unstated elements of primary source material in a manner that comports with basic common sense. The reader will deduce the same from an independent analysis of the pictures and everyone who examines the matter in good faith will agree that the claim is verifiably accurate. You suggest above that such a claim "can, if needed, be cited to the object itself." Does this mean that the default should be to leave a basic description like this unsourced until and unless challenged? That would in fact be my gut feeling. These questions often arise in the context of a quality review (FAR, GAN, DYK, etc.), however, where rigorous sourcing is demanded. Do you think these quality reviews should demand sourcing for basic descriptive claims like this or would it be ok to promote an article despite its containing an unsourced subsection like "Basic description of the cathedral and grounds" that arose from the editor's novel common sense description? -Thibbs (talk) 11:34, 10 October 2014 (UTC)
Thibbs, first let me note that (as usual) I agree with what Blueboar has said above. Next, I just have a moment and want to respond to your new three examples by saying that much is in the context and detail: Brown hair — If this was important, I'd probably challenge it; brown is pretty subjective (is it brown or auburn) and can be temporary, if it was just a minor or passing detail and wasn't just pretty clearly contentious or wrong, I'd leave it. "X eats apples" — insufficient source, especially if made as a statement of preference or habitual practice or if the fruit in the picture is not clearly an apple, but in any event a person can pose for a picture in which it appears that they're eating it when they're not: X could be an actor and this a screencap from a movie when, in fact, X detests apples. White Album — insufficient source, could be fake or dummy, could be that it has two A-side labels when the tracks are actually A and B, could be a single, unique error such that "some" is incorrect. In general, whenever you refer to a thing itself as a PRIMARY source, it must be absolutely unambiguous, and, indeed WP:PRIMARY says, as policy,

Any interpretation of primary source material requires a reliable secondary source for that interpretation. A primary source may only be used on Wikipedia to make straightforward, descriptive statements of facts that can be verified by any educated person with access to the primary source but without further, specialized knowledge. ... Do not analyze, synthesize, interpret, or evaluate material found in a primary source yourself; instead, refer to reliable secondary sources that do so. Do not base an entire article on primary sources, and be cautious about basing large passages on them.

(Emphasis as in original.) Come to think of it, is that perhaps the clarification that you were seeking? Regards, TransporterMan (TALK) 15:47, 10 October 2014 (UTC)
Re Thibbs's question: "Does this mean that the default should be to leave a basic description like this unsourced until and unless challenged?" ... well, I would make the argument that a basic descriptive statement like "The west end of the cathedral is dominated by a rose window" is in fact sourced (to the cathedral itself)... the question is really whether the source needs to be cited. That's a judgement call. There is no default in such cases. You can cite the cathedral itself if you want to... or you can wait until someone challenges the statement. It's your choice. That choice goes away if someone actually does challenge it. In that case you would need to format a citation to the cathedral itself (well... I suppose you could attempt to convince the challenger to drop the challenge... but that is usually a long, difficult and frustrating battle, and not really worth the effort... it takes but a few seconds to format a citation, while it can take weeks to convince someone that a citation isn't needed). Blueboar (talk)
Thanks again to both of you.
TransporterMan I appreciate what you're saying and I agree that context is usually everything in determining reliability in specific cases. But its your comments on the general case that go directly to the heart of my question. And I do agree that PRIMARY serves as about the best direct policy support for Blueboar's comments from yesterday re "basic description".
Blueboar: "Basic descriptive statement like "The west end of the cathedral is dominated by a rose window" is in fact sourced (to the cathedral itself)..." - Now that's nuance! I like it a lot. :) And point well taken about sources versus cites.
Unless there are others who wish to raise opposing views, then, I think I well grasp the community view now. This discussion has been very helpful for me. -Thibbs (talk) 21:24, 10 October 2014 (UTC)


I've pointed to this page, but an editor who disagrees is following me around the Project, reverting my edits as "vandalism" and the like, etc. I've had a long conversation with him. And already given him many warnings. Any suggestions? It's becoming difficult. Thanks. --Epeefleche (talk) 17:44, 12 October 2014 (UTC)

could you link some of the articles? Blueboar (talk) 22:33, 12 October 2014 (UTC)
@Epeefleche: I don't think that you should be deleting content unless you genuinely believe the material is wrong or unverifiable. Please note that this policy states that content must be verifiable, not verified. For example, if an article stated that "Paris is the capitol of France." without a cite, can anyone honestly saying that they don't believe that Paris is the capital of France or that it's impossible to find a source? If that's what's happening, deleting such content is POINTy in my opinion. You asked for suggestions so I would suggest using the {{fact}} tag instead. A Quest For Knowledge (talk) 23:24, 12 October 2014 (UTC)
I of course agree that "the sky is blue" comments don't require a ref, and therefore should not be deleted. If you look at the reverts in this string, that's not what we are talking about here. Epeefleche (talk) 19:24, 14 October 2014 (UTC)
@Blueboar. Sure:
  1. This began when I took a position contrary to that of the other editor at this AfD.
  2. That was followed by the talkpage discussion relating to wp:v and wp:burden reflected here.
  3. And by the editor restoring uncited material here.
  4. And then by the editor following me to an article I had just edited, and reverting me, restoring uncited material (he later submitted a ref for 1 sentence, but restored 2 wholly uncited sentences without providing a ref) here.
  5. The editor then wrote: "It seems as if I am going to have to watch all of your edits in the future".
  6. He then followed me to an article I had just created, and deleted a cat I had just added to the article, under the incorrect assertion that the article did not relate to the cat.
  7. He then wrote to me: "I said I will continue to take a look at those of your edits that consist of deletions of content."
  8. These are just some of the instances of the many articles which he followed me to. Through all of this, I discussed wp:burden and wp:v with him multiple times, and both requested and warned him multiple times to stop hounding me. Epeefleche (talk) 23:29, 12 October 2014 (UTC)
What A Quest For Knowledge means is the WP:Preserve policy. That stated, some content should be deleted even if it's not wrong and is WP:Verifiable. Flyer22 (talk) 04:30, 13 October 2014 (UTC)
Yup... verifiability does not guarantee inclusion. And WP:Preserve is balanced by WP:CANTFIX. The key word in there is balance. Blueboar (talk) 13:02, 13 October 2014 (UTC)
The editor makes a good point; WP:BURDEN is not a free pass to remove content that lacks citations, as was discussed at length not too long ago. If you delete content merely because it's unsourced, you're being disruptive; WP:BURDEN requires a good faith assertion that you believe the content can't be sourced at all. Once you have made that good faith assertion that the article can't be fixed by adding a reference, the burden is on the other party to provide references; but you're expected to be somewhat knowledgeable of the topic to challenge the content, if only because you've tried to reference it yourself and failed. As with all policies regarding inclusionism vs deletionism, WP:BURDEN is balanced with WP:CHALLENGE. If you don't want to bother with checking references and try to source the content yourself, you can always tag the content with [citation needed] and let others do the work. Diego (talk) 13:36, 13 October 2014 (UTC)
  • If someone writes "Sasquatches are aliens from outer space" in an article, I'll remove it without bothering to check for sources.—S Marshall T/C 10:47, 14 October 2014 (UTC)
@Diego - "If you delete content merely because it's unsourced, you're being disruptive"... it's more nuanced that that... yes, going around on a pointy deletion crusade can be disruptive (being pointy about anything is disruptive... even enforcing "the rules"). However... a focused, article specific removal of unsourced content is not disruptive. There is almost always a rational behind it (even if that rational is not expressed). Blueboar (talk) 11:18, 14 October 2014 (UTC)
I fully concur with Blueboar. I'd also note that it is my opinion the assertion provision is not mandatory — note the "please" — and that, in any event and opinion aside, it does not require "belief" but only "concern", a much lower standard. Though everything that we do here must be done (and, absent evidence to the contrary, must be presumed to have been done) in good faith, there is no specific or special requirement that the assertion be made in good faith. Regards, TransporterMan (TALK) 17:17, 14 October 2014 (UTC)
  • I can't seem to convince the editor to stop following me around as described above, to corners of the project, as described most recently here, despite that exchange and this one. Epeefleche (talk) 19:24, 14 October 2014 (UTC)
    • Focusing on deleting article content is not a good idea, and someone needs to check what is being done and revert if desirable. It would be fine to pick a small group of related articles and actually work on them—that would involve content additions as well as deletions. However, a habit of drive-by deletions is not desirable. Johnuniq (talk) 00:56, 15 October 2014 (UTC)
Just for the record... I do not think Epeefleche is engaging in "drive by" deletions... nor "focusing on deleting content". Blueboar (talk) 12:19, 15 October 2014 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Dismissing some edits as "drive-by deletions" is I think too simplistic.

There is a book called 1066 and All That full title is 1066 and All That: A Memorable History of England, comprising all the parts you can remember, including 103 Good Things, 5 Bad Kings and 2 Genuine Dates. One of the notable things about Wikiepdia is that it is aimed at a global audience of all ages. What is common knowledge in one group of people may not be to another. I doubt that there is a person schooled in England who does not know that the Battle of Hastings was fought in 1066. However should we assume it is known to all that fact for all readers of this encyclopaedia?

The problem becomes more complicated when one considers the recognisability clause in WP:AT "Recognizability – The title is a name or description of the subject that someone familiar with, although not necessarily an expert in, the subject area will recognize." Once a person get involved into more than superficial editing an article they become familiar with the subject of the article. Once that happens there are lots of basic information that expert books on the subject assume that a person reading the book already knows, and so a source for the basic information may not be readily available in the sources used to write the article, but which the editors take for granted, because of their knowledge around the subject (they can not see the wood for the trees). In these cases it often takes an outside passing interest to note that a well known fact is not well known to everyone.

Sometimes articles can be improved by the editors who work on them. For example if one compares the Battle of Waterloo at the start and end of 2007, the text of the article had not changed much but by the end of 2007 it was just about fully cited. I was involved in that exercise and finding sources for things that I knew for a fact was not as straight forward as it would at first seem (because one forgets where one has read the fact one remembers so well, or one does not have access to the source). Although the text has not changed much, because of the citations the article, casual readers could be reassured that most of the content was based on third party reliable sources and could identify those parts which were not -- making it a much better article.

However not all articles are subject to that sort of internal upgrade (often because of the "familiar with the subject of the article" syndrome). For example I know that the citations Battle of Worcester article (over the attack on the west side of the city) need to be improved, but the text is roughly accurate, so I'll go to the effort of fixing it tomorrow (when I have more time). Two years later it sill has not been done!

In situation like Worcester, and lots of articles other where the editors think either something does not need an in-line citation because it is common knowledge, or it is covered by the general references at the bottom of the article, a tag with fact needed and then deletion some months down the line if no citations have been provided, can ginger up the the editing of a page and in the end nearly always the page is much better for it. As an example see:

Between Mid November and Middle January a collaborative effort had taken an article that did not even have a general references section to to one that had over 30 inline citation (including multiple citation to the same sources). But to get there there was an acrimonious debate talk page, and without WP:BURDEN being as clear cut as it is, it is unlikely that (a) the edit war would have ended as quickly as it did and (b) the article would not have been improved as quickly as it was.

-- PBS (talk) 12:03, 15 October 2014 (UTC)

If I correctly take your point, let me summarize: "Depend upon it, sir, when a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully." Dr. Samuel Johnson, Life of Samuel Johnson, James Boswell. 'Tis FSM's own truth. Regards, TransporterMan (TALK) 13:35, 15 October 2014 (UTC)
Yes Face-smile.svg -- PBS (talk) 22:03, 16 October 2014 (UTC)
Oh god, not Synchronous motor again...I still have nightmares about that...(shudder) :p DonIago (talk) 16:19, 15 October 2014 (UTC)

Link to Wikipedia:External links/Perennial websites[edit]

An editor added a link to the "the external links policy" but it's an essay, not a policy.[1] I thought about simply changing the verbiage to say "the external links essay on perennial web sites", but perhaps a link to the External links guideline might make more sense. I'm not sure. So, I figured I would start a discussion on the talk page to see what others thought. At the bare minimum, we shouldn't call an essay a policy. But beyond that, I don't have any strong opinions either way. A Quest For Knowledge (talk) 15:40, 19 October 2014 (UTC)

  • I added the {{redirect}} template chiefly because I think it's currently confusing that WP:TWITTER and WP:Twitter point to different places. I don't mind you changing "policy" to "essay" in the second parameter, but I do think that having a {{redirect}} of some kind there is going to be useful for editors who get tripped up by the current redirect situation. It Is Me Here t / c 12:27, 21 October 2014 (UTC)