Jump to content

Template talk:Unreferenced/Archive 10

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 5Archive 8Archive 9Archive 10Archive 11Archive 12

RFC: Should the template employ "unverifiable" or "unverified"?

This is a dispute over which word to use in the high-use template, {{unreferenced}}, "unverifiable" or "unverified" to properly invoke WP:BURDEN.--Fuhghettaboutit (talk) 15:24, 5 June 2009 (UTC)

  • Unverifiable mean not able to be sourced.
  • Unverified mean not yet sourced.

Verifiability is one of our core policies and its subsection, WP:BURDEN, provides that:

The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation... Any material lacking a reliable source may be removed... Do not leave unsourced or poorly sourced information that may damage the reputation of living persons or organizations in articles.

The disputed language in the template is the WP:BURDEN invoking: "Unverifiable material may be challenged and removed." That present construction, using unverifiable as opposed to unverified, functionally reverses the burden of WP:BURDEN. What it says can be restated as: "Material that cannot be sourced may be challenged and removed." By so stating, it requires the person challenging the material to show it can't be sourced. The policy says the opposite. It is the person wishing to add or restore material that has the sourcing burden. That's the dispute in a nutshell.

Upon changing the word to "Unverified", I was reverted. Discussion was had higher on this page and a bit more two sections up. I eventually settled on changing it to an equivalent of unverified—unsourced as at least one person objected to either verified or unverifiable. Note that for a time, approximately April - June, 2007 the template existed at various versions which properly invoked the burden, then stating "Any unsourced material that has been or is likely to be challenged may be removed at any time."

On June 26, 2007 "unsourced" was changed to the non-equivalent "unverifiable." From then until now it has persisted with versions of that language, with no one apparently commenting or showing they noticed the change.

In any event, WP:V is core policy and cannot be overruled. I believe this bowdlerizes it and I think getting this right is important—WP:BURDEN is important; thus stating correctly the burden when invoking it is important; this template is transcluded (as of this writing) in 127,664 articles, and it is more likely that a new user will see the burden through this template at the top of an article they created, than they are to actually visit the underlying policy page it invokes. I am being stubborn on this because I have seen much conclusory statements that unverifiable is better, but none providing any trenchant reasoning why it is better or why I am not correct, and I think this is a serious matter.--Fuhghettaboutit (talk) 15:24, 5 June 2009 (UTC)

This template has always been poorly worded, and this is just one aspect of that; of course it should be "unverified". In addition, I was howled down last year when I complained about the awkward wording "This article does not cite any references or sources", which would be neater and more formal as "This article cites no references or sources". A third, overarching problem, is that as soon as a single reference is added to a tagged section, the template becomes wrong ... plain wrong. That is why I've been advising editors not to touch this one with a bargepole, and instead to use {{refimprove}}. Tony (talk) 16:51, 5 June 2009 (UTC)
The template is not poorly worded. It reflects the current policy that material must be verifiable, not verified. This is not some poor diction, this is an important policy distinction. Gigs (talk) 18:36, 5 June 2009 (UTC)
  • To say there was a dispute between unverified/unverifiable is a mischaracterization. My objection was that it is problematic to prompt people to remove information that is "not yet sourced" from an article known not to have a single source, which is of course everything. However I don't support the template saying "unverifiable" either. We should be prompting people so search for sources on this particular template rather than prompting them with language which is about more fine-grained situations. --BirgitteSB 17:22, 5 June 2009 (UTC)
    • BirgitteSB My original post sought the change from one word to the other, only. That's all I sought. You were the first responder and objected. Maybe you misread it then?--Fuhghettaboutit (talk) 18:31, 5 June 2009 (UTC)
      • Disagreeing with your proposed change != supporting no change at all. The RFC summary describes this a choice between unverified vs unverifiable, when in fact there was significant discussion about removing the language from WP:BURDEN entirely. And there were proposals made that contained neither word.--BirgitteSB 19:11, 5 June 2009 (UTC)
        • That's true, but only because you objected to my change of the one word, and others commented, so I sought other ways to accommodate people. I wanted the change of one word; I still want the change of one word; it's all I ever wanted.--Fuhghettaboutit (talk) 19:29, 5 June 2009 (UTC)
        • I would support some of the versions above that don't mention WP:BURDEN. But I really think the template is OK right now. Gigs (talk) 19:17, 5 June 2009 (UTC)
  • Absolutely leave it as unverifiable the policy is named "verifiability" not "verification". As it says on my user page "The idea that people are not free to add unsourced knowledge goes strongly against the entire idea of a wiki. Unsourced is not the same thing as unverifiable." Leave the template alone, it's fine the way it is. Gigs (talk) 18:33, 5 June 2009 (UTC)
    • Gigs, instead of simply championing unverifiable, can you respond to the fact that the template invokes WP:BURDEN, and apparently changes the burden by its use of that word? That's the sharp issue. You seem to have a problem with WP:BURDEN by your statement. But that doesn't change the fact that this template invokes it on the one hand, but says something inimicable to it on the other.--Fuhghettaboutit (talk) 18:42, 5 June 2009 (UTC)
      • Your understanding of WP:BURDEN is incorrect. Material that is verifiable may not be removed, regardless of its current sourcing. Looking at the above conversation, this was pointed out to you multiple times by multiple people. This is a long standing and community wide consensus, and attempting to change it on the talk page for a template is not the right way to go. Gigs (talk) 19:02, 5 June 2009 (UTC)
        • No. Simple and objectively. WP:V states "The burden of evidence lies with the editor who adds or restores material. All quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation... Any material lacking a reliable source may be removed." If you want to change the policy, fine, but that discussion doesn't belong here.--Fuhghettaboutit (talk) 19:25, 5 June 2009 (UTC)
          • You can't take a single part of a single sentence out of context and present that as the entire policy.

            "Any material lacking a reliable source may be removed, but editors might object if you remove material without giving them sufficient time to provide references, and it has always been good practice, and expected behavior of Wikipedia editors (in line with our editing policy), to make reasonable efforts to find sources oneself that support such material, and cite them."

            That's a far cry from what you are implying. The only way you can possibly get your interpretation is to ignore every other part of the policy and only look at that sentence fragment. That doesn't reflect what the consensus is. The policy is verifiability, not verification. Gigs (talk) 22:05, 5 June 2009 (UTC)
            • I am not implying anything. I grow weary of this and this is all sidetrack which I shouldn;t have bit on. "The burden of evidence lies with the editor who adds or restores material." How could it be any clearer? You are a not a new user by registration but you only have 670 edit to articles so you probably don't see WP:BURDEN in practice very often for its uncontroversial, plain meaning. So you want to see consensus as to what it means? Okay. Here's multiple links to people applying it for its straightforward use. 1, 2‎, 3, 4, 5, 6, 7, 8, 9, 10. The burden is on the person who wishes to keep challenged material to find sources. Not the people challenging the material. The template's use of the one word over the other switches that burden.--Fuhghettaboutit (talk) 11:56, 6 June 2009 (UTC)
  • Upon reviewing WP:V in totality, I think Fuhghettaboutit is exactly correct about this. The meaning of what he quoted is not changed by the qualification cited by User:Gigs. That section essentially merely recommends reasonable efforts to warn editors and even try to find sources prior to removing unsourced material, but it does not require it. The absence of source is sufficient cause to remove material, period.
I see no substantiation for Gigs' assertion that: "Material that is verifiable may not be removed, regardless of its current sourcing." That is not stated or even implied in WP:V nor in any other WP policy, AFAIK. Even the part Gigs quoted contradicts this assertion: "Any material lacking a reliable source may be removed, but editors might object...". Sure, they might object if you remove it, especially without warning, but that does not at all mean it "may not be removed". Further, Gigs' assertion flies in the face of Jimmy Wales himself:

I can NOT emphasize this enough. There seems to be a terrible bias among some editors that some sort of random speculative 'I heard it somewhere' pseudo information is to be tagged with a 'needs a cite' tag. Wrong. It should be removed, aggressively, unless it can be sourced. This is true of all information, but it is particularly true of negative information about living persons.

Jimmy Wales [1]

--Born2cycle (talk) 21:17, 6 June 2009 (UTC)


(unindent) It should read "Unverifiable Unverified material may be challenged and then removed." That would better reflect the process in the context of a completly unsourced article. Of course, if it is a biography of a living person with no sources, just nominate the article for deletion.

I realise the process is different when someone adds a silly claim to an otherwise decent article; those can and should be deleted on sight. But that is not the context this template is used in. --Jc3s5h (talk) 15:48, 6 June 2009 (UTC)

As for the 1st word, agreed, tho "Unsourced" or "Unreferenced" would be better - someone can "verify" something without actually sourcing it if they're not used to the WP jargon. I don't agree with "then", since removing unreferenced nonsense is itself the challenge, and some editors don't consider their unreferenced silly edits to be nonsense. The burden should be on the people trying to turn the encyclopedia into Usenet or a messageboard. -- Jeandré (talk), 2009-06-06t16:05z
The purpose of the template is to alert editors that there is a problem with the article, to wit the article needs verification. It warns them that if no verification is fourthcoming information might be deleted. The template must be seen as a step in a process. At the point in that process when this template is added, the unsourced information is in a sort of "Shrodeger's cat" state... it might be removed and it might be kept, depending on whether it is eventually sourced.
Now, what happens after the teplate is applied is a different issue. Ideally, the person who adds the template attempts to fix the problem and find sources... however, he/she is not required to do so. The burden falls on those who wish to add or retain information to source it. On the other hand, simple courtesy demands that we give our fellow editors a reasonable amount of time to do so.
This brings us to the "unverified" vs. "unverifiable" debate. Our policy states that unverifiable information may be removed. That means that, if there is a reasonable belief that something can be verified, we leave it. However, the contrary argument is also true... if there is a reasonable belief that something can not be verified, we may remove it. If, after tagging the article or statement, and giving other editors a reasonable amount of time to provide sources, no sources are forthcomming, then it is reasonable to reach the the conclusion that no sources exist... and that the information is unverifiable. We may at that point remove it.
In other words, we maintain a balance between immediate removal and everlasting retention of unsourced information. With this ballance in mind, "unverifiable" is the better wording. Blueboar (talk) 16:59, 6 June 2009 (UTC)
Blueboar wrote "if, after tagging the article or statement, and giving other editors a reasonable amount of time to provide sources, no sources are forthcomming, then it is reasonable to reach the the conclusion that no sources exist... and that the information is unverifiable. We may at that point remove it." Sorry, but I disagree. It might be apparent that potential sources exist, but that examining them would be more work than any volunteer editor is inclined to undertake. For example, it might require examining the laws of every state of the U.S., traveling to a library to examine a manuscript, or reading a source in a foreign language. So it isn't necessarily unverifiable, it just requires more effort or expense than any of the participating editors care to undertake. Of course, the material may still be deleted after a reasonable period of time; if the adding editor didn't see fit to brag about all the wonderful source-based research he/she did to justify the claim, it is reasonable to assume that he/she didn't do any source-based research. --Jc3s5h (talk) 17:38, 6 June 2009 (UTC)
  • BURDEN goes both ways. The majority of text on Wikipedia is inadequately sourced, but nobody complains or removes it. Once some text has raised suspicions, it is incumbent on the uploader to provide verification, but this goes outside the scope of the template. The template is for unsourced articles, but the removal of the text is not because it is unsourced, but because the remover feels that it is untrue. This means that the remover's WP:BURDEN is to have done some research to see if it cannot be immediately verified. The template should read "Unverifiable material may be challenged and removed." Joey the Mango (talk) 18:14, 6 June 2009 (UTC)
If potential sources exist (such as laws, rare manuscripts, or foreign language sources) then all it takes is a note on the talk page to that effect halt the process. However, once a challenge has been issued (ie the template has been placed on the article or a fact tag has been placed by a statement), there must be a good faith attempt to find sources. The burden of evidence is on those wishing to keep information, not on those challenging it. The only question is how much time is given for that good faith attempt (I would be very generous, as long as someone is engaging me on the talk page and requesting more time). That said... please note that throughout my comment above, I make use of the term "reasonable". What is reasonable will differ from article to article depending on the nature of the unsourced info. My point is that there is a ballance between inclusion and exclusion that is best achieved by using "unverifiable" instead of "unverified" in the template. Blueboar (talk) 18:34, 6 June 2009 (UTC)
To say unverifiable is to make the template useless. You cannot show anything is unverifiable, even in our sense, without examining every speck of printed matter in the world. The balance is real; but it is implicit in may, which we could do worse than emphasize. Septentrionalis PMAnderson 18:58, 6 June 2009 (UTC)
How is the template useless? The template notes that the article unreferenced and needs to be fixed. The comment to the effect that "unverifiable information may be removed" is simply cautionary... warning editors that stuff might end up being removed unless it is fixed (in this context I take the word "may" to indicate a potentiality rather than permission). Blueboar (talk) 20:03, 6 June 2009 (UTC)
Well, it arguably does make it useless as a warning that material might be removed, given that "unverifiable" is technically unprovable. It takes the teeth out of it (which may be why some favor it). --Born2cycle (talk) 20:18, 6 June 2009 (UTC)

Great discussion folks. A lot of good points. The problematic sentence in question currently says:

"Unverifiable material may be challenged and removed."

I understand Gigs' point about what the policy is, but in the context of this template there is no reason to reinforce that.

What if it states something like this instead:

"Unsourced material that is challenged and remains unsourced may be removed."

Yes, in some cases, unsourced material may be immediately removed, but this template has no application in those cases. --Born2cycle (talk) 19:56, 6 June 2009 (UTC)

Thinking on my last comment (that in the context of the template, we want to express potentialiality and not permission), and Born2cycle's good suggetion above... perhaps it would better reflect what mean if we were to say:
  • "Unsourced material that is challenged and remains unsourced might be removed."
This would accurately warn editors of the potential result of inaction, but leaves everyone some flexibility of action. Blueboar (talk) 20:08, 6 June 2009 (UTC)
Agreed. --Born2cycle (talk) 20:18, 6 June 2009 (UTC)
As the RFCer, to coin a verb, I want to note that I for one am very flexible as to a solution. I had previously suggested getting rid of the invocation of WP:BURDEN entirely as a solution, as well as replacing it with something which simply flags the verifiability policy but not its burden subsection: ("Verifiability of information is one of Wikipedia's core content policies."). What I just can't stomach is the fact that it very specifically invokes the subsection of the policy dealing with the burden but reverses it. Blueboar's suggestion of "might" seems like a decent contender because the word takes the emphasis off of any present conflict with the unsourced material (it hasn't yet been challenged).--Fuhghettaboutit (talk) 20:34, 6 June 2009 (UTC)

Yes, leaving "unverifiable" in there is untenable. It must go.

Another candidate:

"Unsourced material that is challenged and remains unsourced may and might be removed."

Note that this says nothing about unsourced material that is not challenged. --Born2cycle (talk) 21:17, 6 June 2009 (UTC)

The important information is in the first sentence and like {{Refimprove}}, I think the second sentence should be <small>, because the first sentence is a warning to readers, while the second is maintenance aimed at editors and as such could just as easily be on the talk page. I don't mind if it remains as it is or if the second sentence of {{Refimprove}} is used. --PBS (talk) 21:42, 6 June 2009 (UTC)

Agree about <small>, and I think there is something to be said for using identical wording in both templates (so only the first "big" sentences are different). --Born2cycle (talk) 21:54, 6 June 2009 (UTC)
  • Leave as unverifiable. Hello? Miles of problems calling, apparently your can of worms is here. Unverifiable means no source exists or could exist, etc. With mountains of books and news media, scholarly works and other texts locked away presently we have a major information gap that defies that certain subjects are inherently notable and source-able and don't need to be removed simply because the book isn't in front of you. Sources do exist they jsut haven't been plastered up and down an article. The present sourcing expectations have changed considerably since even a few years ago. Encourage more sourcing not more deletions and edit-warring. -- Banjeboi 11:16, 7 June 2009 (UTC)
    • If something is labeled as "unverifiable", that strikes me as a declaration that the tagged statement can never be verified -- even ignoring the impossibility of proving such a negative, I should think that any content which will never be verified should be removed as a matter of course. Simply labeling something as "unverified" is a much simpler charge, with fewer such implications. The easiest way to demonstrate a statement's verifiability is and will always be, of course, to verify it, so why not invite that? – Luna Santin (talk) 00:38, 8 June 2009 (UTC)
      • The practical issue is how such statements are unfortunately used in contentious manner. Similar to an editor adding dozens of fact tags to content they wish to be deleted, similar to editors deleting entire paragraphs because each sentence didn't have a ref on it. Despite general references, which is still a common practice, being on the article. From being on the other end of this where someone wikilawyers to delete material for whatever reasons forcing others to stop what they were working on and repairing the damage I can assure you that logic only plays a bit role here. The moment you open the door that something that is unverified can be deleted ... it will. It won't be pretty. It will be mass slicing through content with little regard to prose and what serves our readers, it will be chop-shop writing that resembles a committee that each wrote a sentence with a ref because only a sentence with a ref will stand a chance to survive. It will sadly be a degrading effect where someone on a mission will point to a word and says "this says i can and nothing says I can't so chop, slice, delete," this really foretells many contentious problems ahead I'm afraid. -- Banjeboi 11:14, 9 June 2009 (UTC)
        • My attitude is that a template that contains a condition that is impossible to fulfill is null and void, and I am free to ignore it and/or delete any instance of it that appears in an article. Since it is impossible to search all the publications in the world to determine if a claim can be verified, the template is null and void; it is the template that should be deleted, not the material in the article. --Jc3s5h (talk) 12:44, 9 June 2009 (UTC)
  • I have had editors say ridiculous things like "I challenge every statement in the article, therefore every sentence must have a reliable source" WP:BURDEN should not be taken as a license to change the verifiability policy into a verification policy that requires every statement in every article to have a cited source. This template should not reinforce that idea either. Gigs (talk) 16:41, 7 June 2009 (UTC)
    • What's ridiculous? It's de-facto standard, love or hate it. And if a sentence combines info backed by different sources, it goes deeper. Sure, if a whole paragraph is lifted off a single book it's simple, but it's not always so simple.NVO (talk) 02:51, 8 June 2009 (UTC)
  • I think both of the following statements will work: Unreferenced material may be challenged or removed or Unverifiable material should be challenged or removed. They both mirror what happens here. On the one hand material that is impossible to verify shouldn't be here at all and it is every editor's duty to remove such material. On the other hand, regarding information that could be verified yet currently isn't; editors still can challenge this and noboby should reinsert the material until backed up by a source. I prefer "unreferenced" myself as an article here is only as good as the sources currently in it, not the sources stacked away in libraries that none of the article's writers has bothered to consult. ThemFromSpace 23:10, 7 June 2009 (UTC)
The first suggestion works just fine. The second suffers from exactly the same problem the current one does. By saying Unverifiable material should be be challenged or removed" you are saying: "material for which sources cannot be found should be be challenged or removed". This leads to only two possibilities. 1) We treat "cannot be found" as a good faith attempt by the disputer to find sources, in which case we are reversing the burden in exactly the same way the present statement does, i.e, the person who disputes the content, regardless of what the material is, must first attempt to find sources and only after failing, may they then state their failure to find sources somewhere, then challenge, then remove. 2) The second possibility, as noted by a few users, is that if we treat the technical definition: cannot be found strictly, using unverifiable states an impossibility, thus eviscerating any meaning from the sentence. Unsourced/unverified/unreferenced/uncited all work. Unverifiable does not.--Fuhghettaboutit (talk) 03:27, 8 June 2009 (UTC)
Then go get the policy changed, since that's what you seem to have the actual problem with. 206.248.204.121 (talk) 12:18, 8 June 2009 (UTC)
That's a non sequitur. Not betraying the existing policy is what we're here about. Have you actually read anything above?--Fuhghettaboutit (talk) 12:46, 8 June 2009 (UTC)
  • I didn't read most of this, but "unverifiable" means that the reader is not able to verify it, not necessarily that it is impossible. Citations don't verify, readers do. Adding a citation doesn't verify the line. Unverified implies that we haven't gotten around to verifying information, which is something we don't do in the first place. As for text, try "This article requires additional references to ensure verifiability". Everyone will read a few words in a box, everyone will ignore pretty much anything over one sentence.   M   17:33, 8 June 2009 (UTC)
  • Go for Unverified. The statement that a statement is unverifiable is itself unverifiable (it can be falsified, but not proved), so that statement must itself be removed, leading to a logical short circuit. --Alvestrand (talk) 18:02, 8 June 2009 (UTC)
  • Use Unreferenced or Unsourced. I think people are getting confused over the context of "unverifiable" in the wording of this template and the verifiability policy. If you read the first sentence of WP:V it states:
"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." (some emphasis/bolding removed and some added by myself)
As a casual reader would see it, if an article is unreferenced then it is unverifiable (by the reader) as there is nothing in the actual article to support the claims it makes. You can argue that somewhere there will be a reference or source in a book or anywhere else, but for the casual wikipedia reader who has no interest in searching google or the library to check that the information is true, it is unverifiable to them. Unverifiable/Unverified is too subjective as the discussion above proves, whilst unreferenced and unsourced are objective: It's either sourced/referenced or it's not. ascidian | talk-to-me 23:52, 8 June 2009 (UTC)
  • Use unreferenced or unsourced for all of the reasons mentioned above. However, if I have to choose between "unverified" and "unverifiable", It should definitely be "unverified". It makes much more sense and is less confusing, especially to new readers. hmwithτ 20:53, 9 June 2009 (UTC)
  • Oh! Thanks to User:M, I get it now. "Verifiable" means that the article has sources cited which a reader can use to verify the veracity of the material in the article. If there are no sourced cited, the article is not "verifiable"... it is "unverifiable". However, as much as that makes sense now, I still think it's problematic because it's so easy to misunderstand. --Born2cycle (talk) 21:44, 9 June 2009 (UTC)

Incorrect premises

Going back to the top of this discussion, there are a couple of incorrect premises:

  1. Unverifiable mean not able to be sourced.
  2. "Unverifiable material may be challenged and removed." That present construction, using unverifiable as opposed to unverified, functionally reverses the burden of WP:BURDEN.

These premises are incorrect because:

  1. "Unverifiable" means the editors have not cited sources, leaving the reader with no way to verify that sources support the material in the article.
  2. The construction does not reverse the burden of WP:BURDEN. A simple absence of citations to sources indicates that the article is "unverifiable". The burden remains squarely on the editors to provide those citations so that the reader can use them to verify that the sources cited by them support the material in the article.

With that clarified, I don't see a problem with the current wording, except that it might be misunderstood. That can be fixed by changing it to say:

Unverifiable material (material that lacks citations for a reader to use to verify that the sources cited support the material) may be challenged and removed.

I'm sure better wordsmiths than I can say the same thing more succinctly. --Born2cycle (talk) 22:08, 9 June 2009 (UTC)

Where verifiability means "whether readers are able to check that material added to Wikipedia has already been published by a reliable source", a pure language definition, using normal interpretation of the change from an "ability" ("capacity") ending to "un" ("not") and "able" ("capable"), unverifiable means "it is impossible (there is no capability) for readers to check that material added to Wikipedia has already been published by a reliable source". You conclude from this that "impossible to check" means unsourced but that doesn't follow. Unerifiable means it is not possible to take the steps needed to allow the ability, i.e., it is not possible to source the material to allow the checking to occur. Using the same pure language interpretation, unverified means "readers are not yet able to check that material added to Wikipedia has already been published by a reliable source". The only reason they are not yet able to do so is because it is unsourced. No matter how you work it, unverifiable should be replaced. Probably the greatest reason to do so is that the actual policy section it invokes, does not use the word unverifiable and we can see that it is very problematic to do so because so many people find it slippery to define. The actual policy says (I am not paraphrasing here, the direct quotes are) "...must be attributed to a reliable, published source using an inline citation" and "Any material lacking a reliable source may be removed". Given that we're referring to this section of the policy, why not use the form of language it actually uses to craft its invocation. Unsourced/unreferenced/uncited all do a yeoman's job. Unverifiable seems to be a sinkhole.--Fuhghettaboutit (talk) 12:43, 10 June 2009 (UTC)
(To Born2Cycle) Clarified? Unsourced is not the same thing as unverifiable. The presence or absence of cited sources has absolutely no bearing on whether information is verifiable or not. Gigs (talk) 12:49, 10 June 2009 (UTC)
I don't disagree with you in an abstract sense, but unverifiable, like many words, can mean different things based on the context. In this Wikipedia context an article is "unverifiable" to a reader if the editors did not provide citations for the reader to verify that the cited sources actually support the material in question. Adding citations to sources that support the material is what makes an article change from unverifiable to verifiable. Any thing that is "unverifiable" in the strict/absolute sense shouldn't be in there at all, and should be removed, not cited with a tag. --Born2cycle (talk) 14:45, 10 June 2009 (UTC)
I don't believe that view is the consensus. You seem to be the only one here asserting that unsourced==unverifiable (in any context). Even Fuhghettaboutit is not asserting that. We should keep in mind that we aren't out to redefine the policy here, this template needs to reflect consensus interpretation of the policy, not put a new spin on it. That was the original concern anyway, that it was putting an inappropriate spin on the policy. Spinning it the other way and implying a stricter standard would be just as bad, and we'd be back here with a mess of defenders of WP:BOLD complaining that we were discouraging edits. It looks like the straw poll is showing a rough consensus at this point anyway, so we might be able to just agree to disagree and move on. Gigs (talk) 15:41, 10 June 2009 (UTC)
This "dispute" is bogus. The spirit of the policy is that content added to WP should be able to be cited properly. Arguing over the words used is not the point. WP is not a court of law. Please abide by the spirit of policy, not the letter of the "law". --rannṗáirtí anaiṫnid (coṁrá) 22:07, 14 June 2009 (UTC)

Straw poll

  • Unverifiable material may be challenged and removed.
Support:
  1. 2nd choice - I still think this verbiage is OK as well. Gigs (talk) 12:57, 10 June 2009 (UTC)
  2. 1st choice - to emphasize that it is everyone's role to try to source material if the editor who introduces it neglects to. DGG (talk) 23:27, 12 June 2009 (UTC)
  • Unverified material might be challenged and removed.
Support:
  1. 2nd choice. SilkTork *YES! 23:32, 9 June 2009 (UTC)
  2. 2nd choice. hmwithτ 13:06, 10 June 2009 (UTC)
  3. 2nd choice.--Fuhghettaboutit (talk) 18:06, 10 June 2009 (UTC)
  4. 2nd choice. ascidian | talk-to-me 08:16, 11 June 2009 (UTC)
  5. 2nd choice. Unclear who does the verification, editors or readers. --Jc3s5h (talk) 13:43, 11 June 2009 (UTC)
  • Unsourced material might be challenged and removed.
Support:
  1. 1st choice. SilkTork *YES! 23:32, 9 June 2009 (UTC)
  2. 2nd choice. --Born2cycle (talk) 00:31, 10 June 2009 (UTC)
  3. 1st choice. This is OK. It dodges the verifiability question in a way that's acceptable, allowing that unsourced information may be verifiable, but the burden is on the adder. I think this reflects wider consensus. Gigs (talk) 12:57, 10 June 2009 (UTC)
  4. 1st choice. hmwithτ 13:06, 10 June 2009 (UTC)
  5. 1st choice and what I had previously changed it to (other than the use of "might") and was reverted; this is the parallel language to that in {{refimprove}}, used since 2007). Note though I would have set up this poll differently as 1) keep current form and use unverifiable 2) Keep current form and use unverified 3) Keep current form and use unsourced 4) Regardless of what first word we use, should we change "may" to "might" 5) Should we get rid of all language invoking WP:BURDEN and instead simply flag the existence of the verifiabiity policy by a replacement statement such as "verifiability of information is one of Wikipedia's core polices".--Fuhghettaboutit (talk) 18:04, 10 June 2009 (UTC)
  6. 1st choice. JackSchmidt (talk) 19:43, 10 June 2009 (UTC)
  7. 1st choice. ascidian | talk-to-me 08:16, 11 June 2009 (UTC)
  8. 1st choice. This not only makes it clear challenged material must be sourced, but "sourced" makes it clear who neglected to do something: the editors. --Jc3s5h (talk) 13:43, 11 June 2009 (UTC)
  9. acceptable, because of the use of "might". it does not say it will be removed, it says it might, which is perfectly true. It gives by implication the proper advice. DGG (talk) 23:27, 12 June 2009 (UTC)
  10. Although I prefer "may" to "might" as any editor should remove such material that infringes one of our core policies (WP:V). ThemFromSpace 22:38, 26 June 2009 (UTC)
  11. Prefer 'may' instead, otherwise ok. PhilKnight (talk) 13:21, 28 June 2009 (UTC)
  • Unverifiable material (lacks citations for a reader to use to verify sources) might be challenged and removed.
Support:
  1. 1st choice --Born2cycle (talk) 00:31, 10 June 2009 (UTC)
  • Unreferenced material may be challenged and removed (just throwing this out there)
Support:

A modified approach

The proposals above are tinkering too much with one sentence (or even one word), and therefore demanding that it carry too much of the weight. In my opinion, we should consider the language of both {{Refimprove}}

and {{Unreferenced}}

to arrive at the best solution. I propose taking the best from both to rewrite {{Unreferenced}} as follows:


Points to consider:

  1. Adding "for verification" to the end of the first sentence puts the Verifiability policy where it belongs. The reason for requiring source citations is for verification, that is, to make it easier for readers and editors to verify the article's content.
  2. The second sentence of {{Unreferenced}} is clearer that the second sentence of {{Refimprove}}.
  3. Unverifiable should not be the first word of the last sentence because it is not impossible to verify content without a citation; a trip to the library will do it in most cases. We require citations to make verification by others easier.
  4. Unverified should not be the first word of the last sentence because a citation does not make content verified. The citation just makes it easier for anyone else to verify that the content is supported by the cited source.
  5. Unsourced, with all that precedes it, sufficiently conveys what we mean. A wordier alternative would be Material without citation of supporting sources may be challenged or removed.
  6. In the last sentence, may is preferable to might; may correctly conveys both what an editor is permitted to do and that it might happen.
  7. In the last sentence, I would change and to or because (a) no one editor would to both at the same time and (b) or sufficiently conveys the range of consequences. Finell (Talk) 20:38, 14 June 2009 (UTC)

Support:

  1. Finell (Talk) 20:38, 14 June 2009 (UTC) (my proposal)
  2. Looks good too. JackSchmidt (talk) 21:37, 14 June 2009 (UTC)
  3. --Jc3s5h (talk) 21:53, 14 June 2009 (UTC)
  4. Brilliance/perfection. hmwithτ 22:15, 14 June 2009 (UTC)
  5. I'm ok with it. Gigs (talk) 14:54, 15 June 2009 (UTC)
  6. Great idea, takes care of all the problems. -- Oldlaptop321 (talk·contribs) 02:32, 16 June 2009 (UTC)
  7. I am not sure why we need two separate templates (rather than redirecting this one to refimprove), but at least they should be uniform. This is a step in the right direction, and the text is also one of the best variants so far. (I agree with some below that "and" would be slightly better than "or", but that's a WP:BIKESHED problem.) --Hans Adler (talk) 11:53, 16 June 2009 (UTC)
  8. Great! --Born2cycle (talk) 22:23, 19 June 2009 (UTC)
  9. This fixes all the problems. Very nice. Werson (talk) 19:23, 20 June 2009 (UTC)
  10. Very nice. ThemFromSpace 22:39, 26 June 2009 (UTC)

Comments:

Finell wrote "The reason for requiring source citations is for verification, that is, to make it easier for readers and editors to verify the article's content." That is not the only reason. Common courtesy, and in some cases copyright license agreements, require attributing direct quotes to their source. Another reason, which does not rise to the level of a requirement, is to guide readers to additional information beyond what is in the article. --Jc3s5h (talk) 21:12, 14 June 2009 (UTC)

I agree with all your points. WP guidelines reqire a specific citation, with page number if any, for each direct quotation, for all the above reasons (courtesy, copyright, and verifiability). I would use {{Fact}} for an un-cited quotation, and would not rely on {{Refimprove}} or {{Unreferenced}} for that purpose. Citations also help to police copyright violations or plagiarism (the two aren't the same) for material that does not purport to be a direct quotation, if too much text was borrowed from a source. I also thought about the "further reading" function of all citations, but it is an added benefit of satisfying WP:V. And I didn't want to be wordier than I already was. Finell (Talk) 22:42, 14 June 2009 (UTC)
I am very happy we are no longer stuck on unverifiable, and I like the integration of refimprove's first line. However the change from "and" to "or" does not work. It robs the intended meaning from the framed language to set forth an order of events (challenge, then remove if no sources are found) and worse, strongly implies something incorrect. This does not work in the same way that the following statements, also having one thing that must happen before another, are almost meaningless and do not work with a change from "and" to "or":
  • "the instructions say to prime or paint"
  • "If you're caught you will be tried or convicted"
By using "challenged or removed", we are by normal language construction saying that the challenging, and the removal, are two alternatives things that can be done where sources are lacking, and not necessarily connected things at all. The sentence no longer parses as stating the process has an order. "Challenged OR removed" means material can just be removed, without a challenge first, or, separately, that material may simply be challenged. Retaining "and" works, as does something more pointedly instructive, such as "Unsourced material may be challenged and thereafter removed if sources are not provided."--Fuhghettaboutit (talk) 20:39, 15 June 2009 (UTC)
I don't like your last proposal. It implies that you need to discuss before removing instead of being bold. Gigs (talk) 22:28, 15 June 2009 (UTC)

(indent)Upon further thought, I think "or" is valid, and "and" is also valid:

  • Challenge - You can discuss before removing dubious info that someone might be able to source
  • Remove - You can be bold and remove info you think is unlikely to be verifiable
  • Challenge and remove - You can stick an ugly citation needed tag on, and then come back and remove it later, or remove statements with stale fact tags, or remove information to the talk page... etc.

All of these courses of action are allowed and encouraged, depending on editor preference and the situation at hand. If anything "and/or" would work, but I think leaving it as "or" is fine; in common English, "or" is often not exclusive (when lacking "either" at least). Gigs (talk) 22:39, 15 June 2009 (UTC)

First, it's normally a challenge before removal operation, with the policy strongly recommending that order. That's gotten across by and, and not be or. Second, by separating them you are making it into two statements: unsourced material may be challenged" and "unsourced material may be removed". The second one relates to the policy invoked. What does just challenging have to do with anything? You say: "you can challenge before removal." Yeah but it no longer says that. It just says you can challenge. The coupling between them is lost. No, making it "or" is confusing and no longer captures the policy invoked.--Fuhghettaboutit (talk) 09:23, 16 June 2009 (UTC)
If you think people need permission or prior discussion to remove material, then you are just wrong. This is a wiki, not a "design by committee" encyclopedia. Gigs (talk) 11:55, 16 June 2009 (UTC)
Well, I don't think that, and haven't a clue what your post is addressed to but very apparently it's a strawman, and I really wish you'd read what I've said with a bit more focus, because you responding to something else than what I've said. or entirely missing what I'm saying is happening with some frequency. Talking about the effect of language and parsing it can be slippery but we seem to be miscommunicating a great deal --Fuhghettaboutit (talk) 12:54, 16 June 2009 (UTC)
Maybe so. In any case I do think challenging alone is a valid course of action (as well as the other 2 possibilities), so "inclusive or" is perfectly fine to leave in the template, and probably should be changed in refimprove as well. Gigs (talk) 13:18, 16 June 2009 (UTC)
The term and/or is what we mean, but it is an abomination to the English language. The word or is presumptively inclusive unless qualified, and therefore embraces both and and or. Therefore, or is the best choice here and for all similarly phrased templates. Finell (talk) 20:25, 19 June 2009 (UTC)
You speak English so you use or as an exact disjunctive every day for the exact opposite of what you are saying (the wikt:exclusive or. I can even state the choice we are faced with using the exclusive or, or not. That last sentence is actually a disjunctive use and the promised choice before us is: to use "or" OR "and". That sentence has a natural meaning and it is that we can use one OR the other (and this sentence is yet another usage). Try to substitute and/or in any of those preceding sentences and the result is an absurdity. There are times when or can be used as the inclusive or, but we understand whether something is the exclusive or, or the inclusive or, through context, which new users seeing the template do not have.

I defy you to think of any use of "challenged and removed" that doesn't convey what we are trying to say, that material can be challenged and, material can be removed, which has a secondary language meaning of implying the order of operations which is de facto, how we use it in the context of WP:BURDEN. Specifically, Gigs says "we do just challenge"—yes, on Wikipedia, speaking generally of things that are possible, people do just challenge if they drop a drive-by fact tag, but that is not what we're talking about in this specific context. In the context of WP:BURDEN, challenge is in relation to removal. We can just remove, yes, but we only challenge here as a prelude to removal if sources are not provided. And, thus, is inclusive of all operations we want to convey. You must concede (because you speak English) that "challenge or remove" can, as one of its possible parsing interpretations, mean "material can be challenged or material can be removed, not both"; as well, that is can mean: "material can be challenged, alternatively, material can be removed", both of which constructions are not what we mean in this context. What, then, does using or bring to the table over and that is not a liability?--Fuhghettaboutit (talk) 02:46, 20 June 2009 (UTC)

If you try to run across eight lanes of freeway you may be seriously injured or be killed. The or implies any of the following may occur
  • serious injury without death
  • death without injury (instantaneous death)
  • serious injury leading to a coma and eventual death a week later
. Similarly, in the proposed wording with or, all of the following possibilities are included:
  • challenge without removal
  • removal without challenge
  • challenge followed by removal a week later
Or in both of these contexts is correct, and I think the meaning should be clear to most. --Born2cycle (talk) 03:12, 20 June 2009 (UTC)
Yes, you are using or in its inclusive form in this example; the context makes anyone reading it immediately aware of that. What you are not responding to is that or also has an exclusive form. Or in the context here, can be taken as the exclusive or that the context of your example usage above does not allow. And suffers from no such problems. You say "you think the meaning should be clear to most." Even if that's true, you leave in the possibility of the meaning not being clear. Using and raises no such confusion possibilities. So, even if you say that or will probably be understood by most, why does that impact on whether or is superior to and?--Fuhghettaboutit (talk) 03:24, 20 June 2009 (UTC)
And is misleading. While the material may be challenged and (necessarily) removed, it may also be challenged, fixed, and not removed. "May be challenged and removed" does not parse to mean that - that is, the "may be" modifies "challenged and removed" together, not each word independently.
Also, the exclusive or meaning of or has so little application in standard English that it is not even mentioned in the article. For example, even in an apparent exclusive or situation such as when a server asks whether you would like coffee or tea, the possibility of having coffee and tea is not precluded.
Further, even in logic and computer science the inclusive form is the default meaning. Using or may not be perfect wording, but I do think it's the best, certainly better than the misleading and. --Born2cycle (talk) 16:58, 20 June 2009 (UTC)
See also what User:Werson says below under Conclusions - while or is ambiguous, and least it's not wrong like and is. --Born2cycle (talk) 05:24, 21 June 2009 (UTC)

Styling

Why has the line break and small text crept back in? It doesn't make the template any smaller, unbalances it on wider displays and makes the instructional text less legible. Chris Cunningham (not at work) - talk 10:53, 16 June 2009 (UTC)

Presumably to make its appearance consistent with the refimprove template. Both variants are the same to me, as long as they are both the same. --Hans Adler (talk) 13:13, 16 June 2009 (UTC)
Chris Cunningham (not at work) we have already had a discussion on this (see Template talk:Refimprove#Style tweaks, so why are you asking this question, (when you posted your last comment to that section on the 11th) and phrasing it the way you have? But to summarise for those who do not want to read "Style tweaks": The first part is of use to readers and editors of an article, the second half is only of interest to editors as a maintenance comment; and as such having the second part in small, emphasises the important part, (as the second part could be placed on the talk page as it is of no importance to readers). --PBS (talk) 09:35, 17 June 2009 (UTC)
Again, on wide or high resolution displays the line break and small text adversely affects the appearance and readability of the template. Using them to marginalise part of the text because of a personal distaste for cleanup text on articles is inappropriate, as was PBS's use of his admin bit to revert the change in question after it had been proposed and performed. Chris Cunningham (not at work) - talk 09:45, 17 June 2009 (UTC)
I've got 3800 horizontal pixels between my two monitors, and it looks fine to me. The second part of the text is indeed less important. Gigs (talk) 11:48, 17 June 2009 (UTC)
Here's a comparison. On high-resolution, high-DPI screens there's a clear win here. I'm still struggling to see what the advantage to the two-line format is. Chris Cunningham (not at work) - talk 13:17, 17 June 2009 (UTC)
The two line format is easier for me to read. It shows that the template is making two comments about the same subject. One is a warning to readers, and another is instruction about how the problem might be addressed through editing. Gigs (talk) 18:25, 17 June 2009 (UTC)
Gigs is right. Maintenance templates should be as succinct as possible so readers (who generally don't care about wiki maintenance) aren't distracted from actually reading the article (which is usually their goal!). Leading with a big, short sentence gets all the meaning across and lets the readers continue onto the article if it's something they're not concerned with. Putting it all on one line forces them to read the whole list of wiki cruft. It's just bad usability, regardless of simple aesthetics. Werson (talk) 19:21, 20 June 2009 (UTC)

Conclusions?

So, it seems that the discussion has died out. This one seemed to be OK with most people (minus one objection for the two line format, and one objection to "or" instead of "and"):



Are we ready to call it a day for now? Gigs (talk) 18:59, 18 June 2009 (UTC)

Premature. Four of the people who participated in the straw poll have not commented on the later proposal section, 3 people expressed a preference for "might" over "may" that haven't commented in the section. One of those that commented above agreed that "and" is better and that issue hadn't even been broached at the time the others signed (and or is wrong). At this point I see more participation and more support for the original straw poll proposed language than the above later addition. RFCs, though they have no mandated time period, are removed automatically at the 30 days (we're at day 13), so no, let it run a proper period. --Fuhghettaboutit (talk) 00:41, 19 June 2009 (UTC)
Support. This wording clarifies the meaning of verifiability – as seen from the original discussion, even experienced editors didn't understand what the term meant in that context. It makes much more sense here. As for the and/or debate, I don't think there's any succinct way to accurately get the meaning across. There are three possibilities for unsourced material:
  • It may be challenged and kept (if a source is provided in response to the challenge)
  • It may be removed without a challenge (while challenging is polite, it's certainly not required by any policy, and this template should not falsely imply that it is)
  • It may be challenged, and later removed
The word and falsely implies only the third possibility, while the word or falsely implies only the first two. The word or can be inclusive, but it simply doesn't sound inclusive in the above sentence.
Therefore, I propose a different wording entirely, which gets the point across as clearly as possible: "Unsourced material may be removed at any time." It's better for the template to be too direct than too equivocal. If that's not an option, then I support Finell's wording. The word or is better than and here, as or is ambiguous while and is simply wrong. Werson (talk) 19:10, 20 June 2009 (UTC)
Still opposed. The arguments for the two-line format are bogus, especially Werson's argument which says that making the second line twice as difficult to read is a usability win. Chris Cunningham (not at work) - talk 17:20, 21 June 2009 (UTC)
How about the following wording? Would that satisfy everyone?
Unsourced material may be removed with or without prior challenge.
Hope that helps... --Chris Jefferies (talk) 22:09, 24 June 2009 (UTC)
I think that's a good compromise Chris. It should just be and but that language does take care of the problem. I could even have lived with or over and if I had to, but that is better than using or in the current formulation. The original problem of unverifiable is behind us, which is was far more important than any of this. What really rankles is that all this was so unnecessary. 50,000 words spilled to change the obviously incorrect WP:BURDEN wording, but for the switch from and to "or", to exactly what I changed it to word-for-word more than 3 months ago.--Fuhghettaboutit (talk) 22:52, 24 June 2009 (UTC)
I think this new proposal is fine as well. Fuhghettaboutit, you may think all the discussion was unnecessary, but remember it was you who often raised objections to developing consensus. I think this discussion is important, as unnecessarily discouraging casual users from adding unsourced information would be disastrous. We must carefully communicate our policies in a way that avoids doing that. I am glad we were able to reach a consensus here (save Chris). Gigs (talk) 15:21, 26 June 2009 (UTC)
I support Chris Jefferies wording proposal too, although I have to agree with Chris Cunningham that on my monitor (widescreen at 1920*1200) the <small> text is a bit too small? ascidian | talk-to-me 22:26, 26 June 2009 (UTC)
As this template is frequently replaced by footnote citations, which are often presented using {{reflist}} and so are in a small font how do you manage to read them? --PBS (talk) 13:51, 28 June 2009 (UTC)
@Gigs: I continue to be gobsmacked by your indiscriminate posts; your niggling insinuations; your off-the-wall misinterpretations. As uncivil as this post is, this is me biting my tongue. It is better we don't interact.--Fuhghettaboutit (talk) 04:26, 27 June 2009 (UTC)

A conclusion?

So, we have this tag that says that unsourced material may be removed. So, what if someone removes it? And what is left is speedy-able or clearly fits inside the reasons for deletion - do we as a community have the resolution to carry forward and not permit restoration of the challenged material without source and go ahead and delete the remaining stump of an article? And, the logical extension, is that blanking of wholly unsourced articles is permitted and deletion will occur in short order. If we are unwilling to do this, then this template has no teeth and will just be a long-term place holder rather than a real means to improve the encyclopedia. Carlossuarez46 (talk) 15:46, 30 June 2009 (UTC)

What's under discussion above is the wording of the template where it invokes the long-existing subsection of the verifiability policy at WP:BURDEN. This template, as well as {{refimprove}}, have contained wording addressed to that for ages, and we're here about the correct way to say it. The underlying policy itself is what I think your post is addressed to so, unless I misread you, I think Wikipedia talk:Verifiability is a more natural fit for your post. However, I would note that having seen WP:BURDEN used in many situations, it is rarely closely connected with any form of article deletion. It is usually employed with regard to specific material that a person challenges, most often in a already fairly well-developed article that already has some references, but none or unreliable ones for a challenged sentence or section, and is not used to remove most of the material from an article to leave it in a "speedyable" state. Many users would revert the wholesale blanking of most of a non-BLP article while invoking WP:BURDEN as indiscriminate; a failure to capture the spirit of the policy, which invites a challenge, a look for sources yourself first, some time for the majority contributors to rectify the problem and so on.--Fuhghettaboutit (talk) 23:55, 30 June 2009 (UTC)
There are cases were the use of such a template is the forerunner of a AfD. For example see Wikipedia:Articles for deletion/Axis plans for invasion of the United States during WWII I added a {{refimprove}} along with {{fact}}s for specific paragraphs not already tagged with {{fact}} on 07:06, 16 September 2007. On the 2 October 2007 as no additional citations had been added, I put the page up for an AfD. During the week of the AfD, all paragraphs with a {{fact}} were removed, There was one sourced paragraph on a World War I plan, and two other sentences (1)"Hitler declared war on the United States of America on 11 December 1941", (2) "The Hawaiian Islands could be used for future landings in the continental U.S., or to conduct aerial surface, naval, or submarine attacks against U.S. coasts." from a source called "What If? The World's Foremost Military Historians Imagine What Might Have Been". But this is not what normally happens, instead the warning template alerts a reader that not all/any of the information in the article had been verified with the use of citations, so therefor it may not be completely accurate (no way to judge), and over time citations get added until an editor makes a judgment call that the article now has enough citations to warrant the removal of the {{refimprove}} template. --PBS (talk) 09:06, 1 July 2009 (UTC)
@Fuhghettaboutit If we are bothering to rephrase it, then say what means and mean what it says. The old tag was basically handed down with a pedigree but not necessarily embued by consensus - now, if we have a consensus on what it should say - and therefore, what it should mean - what's wrong for expecting people to act upon it? In the early days when WP was desparate for content, there was a leniency that we'd rather have an unref'd article than no article at all - I thinkk that mentality has shifted over time, and now we may expect more from our articles per WP:BURDEN, WP:RS, and WP:V and esp. WP:BLP for those articles. So when people bent over backward to keep an unreferenced one-liner "Joe Blow was a famous Fooian singer/writer/scientist, whatever." three years ago, that won't pass muster today and may even be speedied. Again the thinking is we needed more articles on Fooians. Well, I'm not saying there aren't any famous Fooians we don't have yet, there are fewer and fewer as time progresses and we're less likely to make a leap of faith that Joe Blow is notable or a short article about him without references ought to be kept. Carlossuarez46 (talk) 23:34, 3 July 2009 (UTC)
People act upon WP:BURDEN every day but not normally to remove all content such that deletion is in the offing. The language of the policy does not speak to unfettered removal. It contains all types of caveats and recommendations before removal that results in people using it only for specific material they object to, or think sounds dubious, or controversial, rather than using it as a license to remove any unsourced material. When it arises, I see it typically being applied to articles that already have references, and the removal is as to those parts of an article that do not. As I intimated previously, I think you need to go to the source. The verifiability policy does not have the teeth it should or, at the very least, the way we apply is with kid gloves. I think we have dug ourselves a very, very deep hole that we are going to have a hard time climbing out of with hundreds of thousands of entirely unsourced articles sitting around. I think it's a systemic cancer on Wikipedia. But, for better or worse, there is a very deep and wide vein of editors who have shot down every proposal that has ever been made to seriously enforce verifiability; to apply it with any teeth. If you think I'm one of them, that's not the case. I am reporting what I see, not what I believe. If it was up to me, we would start systematically removing unsourced material from the encyclopedia. What we need is something like Wikipedia:Requests for verification, which was shot down. This template acts to flag that unsourced material is a problem. And it acts to also tell people that unsourced material can be removed. But this template and its progeny are side issues. The underlying problem can only be addressed by a change in the culture, where we don't just require the hypothetical ability to be sourced, but actual sourcing on some type of a time frame, or deletion will follow. Verifiability as presently written and interpreted can only take us so far. It doesn't require sourcing. It requires the ability to source. The WP:BURDEN subsection doesn't get us to a place where we require sourcing either (unfortunately).--Fuhghettaboutit (talk) 05:19, 4 July 2009 (UTC)

{{editprotected}} Please, add the cs interwiki [[cs:Šablona:Bez zdrojů]] as already added to the doc page. I hope I follow the process properly. ;) Franta Oashi (talk) 08:53, 7 July 2009 (UTC)

interwikis enhanced: bg, da, de, fr, hsb, ru. The same, to be put on the template page. Franta Oashi (talk) 09:54, 7 July 2009 (UTC)
If they're on the doc page, then they don't need to be added to the template itself. — Martin (MSGJ · talk) 11:24, 7 July 2009 (UTC)
It is not clear to me: Once I added the interwiki to the doc page (because the tamplate page is protected), is it automatically put to the templet too? So, how does the combination of "protection & doc" work? Franta Oashi (talk) 22:34, 8 July 2009 (UTC)

Fact template is now named citation needed

{{editprotected}} In this template, the word "challenged" is a piped wikilink to {{fact}}. The target template is now formally named {{citation needed}}, in an effort to improve clarity (or so I'm told ;-) ). We should prolly change this template to match. advTHANKSance! —DragonHawk (talk|hist) 02:02, 21 July 2009 (UTC)

 Done Let me know if there are any problems. Plastikspork (talk) 04:45, 21 July 2009 (UTC)

Accessibility improvement

{{editprotected}} For WP:ACCESSIBILITY by visually impaired readers, purely decorative images such as the one used by this template should not have links, so that they don't bog down screen readers; see WP:ALT. Please install this obvious sandbox change to remove the link. Thanks. Eubulides (talk) 12:40, 26 August 2009 (UTC)

 Done — Martin (MSGJ · talk) 13:24, 26 August 2009 (UTC)

Date on Unreferenced template

Should newly added Unreferenced templates ("This article does not cite any references or sources") be backdated to the time of the articles' creation, or should they reflect the date that the template was added to the articles? (Adding Request For Comments tag to existing discussion.) Mudwater (Talk) 12:41, 10 November 2009 (UTC)


Note: The discussion below has been copied from User talk:Rich Farmbrough#Date on Unreferenced template. Other editors are encouraged to join the discussion here on the template talk page.

Greetings, Rich Farmbrough. I noticed that in some articles you have replaced the category "articles lacking sources" with an Unreferenced template. Fine, but, I believe the Unreferenced template should not be backdated. If you put the current date on the template, other editors will then see how long the article has been tagged. Take a look at Template:Unreferenced#Usage, where it says, "The date parameter is used to indicate when the template was added to a page." (If you reply here I will see what you say.) Mudwater (Talk) 01:36, 9 November 2009 (UTC)

Yes, except it's not "articles lacking sources" - it's Articles lacking sources (Erik9Bot) which is about 100,000 strong and not dated. Essentially the Erik9Bot articles would swamp the November 2009 section. If there were just a few hundred I would simply stick them in there. I added a footnote to the{{Unreferenced}} page. Rich Farmbrough, 01:59, 9 November 2009 (UTC).
I see what you mean, but I think it's confusing to backdate the Unreferenced template. Other editors will get the impression that an article has been tagged for a long time -- more than three years in this example -- when it's only been tagged for a short time. If the tagging is partly based on the actions of a bot, and results in many articles being tagged for the current month, I don't see that as a problem. Mudwater (Talk) 02:10, 9 November 2009 (UTC)
Yes I understand. When we set up most of the dated categories we did go back to previous database dumps to get the dating as accurate as possible, we would date that way now if it were effective use of resource. Rich Farmbrough, 02:15, 9 November 2009 (UTC).
It's not a question of the date being "accurate", it's a question of what the date means. The general usage on this and many other tags on Wikipedia is that it shows when the tag was added. Maybe the Unreferenced template should be enhanced to have an additional parameter showing how long the article has been unreferenced, but that's not how most editors are going to interpret the current tag. Mudwater (Talk) 02:43, 9 November 2009 (UTC)

Dating the tags when they were added (or when the date was added) has always been a compromise. It was one I was aware of when I created the parameter and categories. I am also aware , because I deal with them on a daily basis, that articles move from category to category, and all sorts of strange things go on. It doesn't matter that some articles are "late" - even if it weren't a waste of human time to date the tag at all, people would make so many errors digging back (I get probably a couple hundred misdated tags a day as it is) that adding the current date or better still no date at all is the preferred option for human editors. The important thing is that the articles get references. But having said that I am reluctant to re-create old dated categories (though it happens often enough due to article reversion). Rich Farmbrough, 03:04, 9 November 2009 (UTC).

Recreate old dated categories? I think my last post has muddied the waters, so let me restate my message. By common usage of the Wikpedia tagging system, and as stated in the documentation for the Unreferenced tag, the date in a tag indicates when a tag was added to an article. Backdating the tags should not be done, because (1) it will give most editors the erroneous impression that the article has been tagged for much longer than it actually has been -- in the example cited above, three years instead of two hours -- and (2) there is not a significant advantage to backdating the tags. I'm going to copy this discussion section Template talk:Unreferenced so please feel free to reply there. Mudwater (Talk) 04:02, 9 November 2009 (UTC)

Note: The discussion above has been copied from User talk:Rich Farmbrough#Date on Unreferenced template. Other editors are encouraged to join the discussion here on the template talk page.


It may interest, amuse and entertain to know that all the articles in the "October 2006" category and earlier are backdated since the date parameter wasn't introduced until November 2006. Template_talk:Unreferenced/Archive_1#This_should_be_a_deletion_template Rich Farmbrough, 05:04, 9 November 2009 (UTC).

The reason they are tagged with {{Unreferenced}} is because of that first tag, so of course they should be backdated. Debresser (talk) 06:52, 9 November 2009 (UTC)
What first tag? "Category:Articles lacking sources (Erik9bot)"? In the example cited above, the category was added in September 2009, but the Unreferenced tag was backdated to October 2006. Also, it appears that User:Erik9bot has been blocked indefinitely. Mudwater (Talk) 09:06, 9 November 2009 (UTC)
That Erik9bot is blocked is part of the reason we are superseding the category. While the principle was reasonable the category requires maintenance which Erik9bot is not here to perform. There are other flaws in the implementation too. Rich Farmbrough, 12:35, 9 November 2009 (UTC).
  • It really makes little difference which date is put on the tags (short of recreating cleared categories). Both methods are a valid answer to the question of what date is appropriate. Why don't you each use the method that you prefer in your own edits. This is not an issue where there is any value in forcing others to conform to a one's personal preference.--BirgitteSB 13:18, 9 November 2009 (UTC)
    Indeed. I'm failing to see what actual problem is being caused here. It's not as if people are backdating these tags to 1970 or anything. Chris Cunningham (not at work) - talk 08:54, 10 November 2009 (UTC)

Here's a summary of what I understand so far. A few months ago, a bot that was subsequently banned as a sockpuppet found thousands of articles that did not have references. One or more editors is removing the bot-added category and adding an Unreferenced tag to the articles. If the current month is added to the new Unreferenced tags, the category "Category:Articles lacking sources from November 2009" will be "swamped". The problem there being what? That the category will correctly show which articles had the Unreferenced tag added to them this month? That would seem to be a good thing, not a bad thing. So, instead of that, the tags are being backdated, approximately but not exactly to the month of the articles' creation. As a result of the backdating, the vast majority of editors, and readers, will get the erroneous impression that the tags were added a long time before they really were added -- several years before, in many cases. Thus any value that the date in the Unreferenced tag may have is being completely undermined. Am I missing something here? Because I feel like I am. I'd like it if more editors were involved in this discussion, so I'm going to add a Request for Comment to this talk page section. Mudwater (Talk) 12:31, 10 November 2009 (UTC)

I was under the impression that the date added now, was the same date when the banned bot added its category. Is that not so? Debresser (talk) 12:58, 10 November 2009 (UTC)
Definitely not. I gave one example earlier in this talk page section: "the category was added in September 2009, but the Unreferenced tag was backdated to October 2006." The bot added the category in the last few months but the tags are being backdated approximately to the articles' creation, which is often several years earlier. Mudwater (Talk) 13:27, 10 November 2009 (UTC)
That is not good. Debresser (talk) 14:53, 10 November 2009 (UTC)
Just to clarify here: AFAIK it was Erik9 (talk · contribs) who was banned; Erik9bot (talk · contribs) was just collateral damage. This may not change anything, but just to ensure that we're not going assuming malice on behalf of an approved bot. Chris Cunningham (not at work) - talk 16:01, 10 November 2009 (UTC)

Here's another point of interest. Category:Articles lacking sources (Erik9bot) is a hidden category. While that might be appropriate for a maintenance category, it means that someone who's reading an article won't see the category, even if they scroll down to the bottom of the article and read the categories. The only ways I know to see that an article is in that category is to edit the article and go to the bottom, to see if it's in the markup, or to start looking through the category itself. My point is that backdating the Unreferenced tags to when the hidden category was added would also be a bad idea, because most people wouldn't have noticed that the article is in that category. So, once again, Unreferenced templates should not be backdated, but should reflect the date that the template itself was added. Mudwater (Talk) 14:43, 13 November 2009 (UTC)

Findsources

{{Editprotected}} We should add the {{Findsources}} template to this template. The edited version is available in the sandbox.
V = I * R (talk to Ω) 23:45, 4 December 2009 (UTC)

Not done: People seem to be very much opposed to a similar idea on Template talk:Refimprove. Either way, I think the RfC should be allowed to conclude before adding such links to other maintenance templates. — Martin (MSGJ · talk) 00:02, 5 December 2009 (UTC)
...there's an RfC? O_o
V = I * R (talk to Ω) 00:10, 5 December 2009 (UTC)

When this template is wrong

This template is often used on articles which have one or more sources. It would be much more useful if it said "sufficient" sources instead of "any." As it is, it is usually wrong, and should be either changed or removed from most of the articles where it exists. Becritical (talk) 01:20, 15 December 2009 (UTC)

I agree that this template should not be used in articles that have any references. I would suggest that in such cases you feel free to replace {{Unreferenced}} with {{Refimprove}}. Mudwater (Talk) 01:26, 15 December 2009 (UTC)
Well, I do (: but the thing is that that is a nuisance, and it could be fixed by changing one tiny word in the template. Becritical (talk) 00:20, 16 December 2009 (UTC)
I know there is a way for a template to test to see if there are any <ref>...</ref>s on the entire page. It wouldn't be able to test for other methods of citing. Plastikspork ―Œ(talk) 01:23, 16 December 2009 (UTC)
That would be cool, I did not know that was technically possible. Becritical (talk) 03:13, 16 December 2009 (UTC)

Category

Couldn't the addition of this template add the article to a category for unreferenced articles? That way it would be easier to find which articles have been recognised as needing references and they can be improved quicker. A similar category could be introduced for the {{refimprove}} tag. Wikiwoohoo (talk) 21:25, 8 January 2010 (UTC)

Requested merge/redirect

The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the move request was: No action. Not the proper venue, as almost everyone here agreed. TFD may be the best option. Ucucha 13:41, 25 January 2010 (UTC)



Template:Unreferencedtemplate:Citations missing — Sorry for the misuse of the move request template, since I didn't find "Requested redirect" mechanism.

Hereby I request to consider deprecating {{unreferenced}} and redirecting it to {{citations missing}}. The "unreferenced" template unnecessarily splits hair about missing references. This has especially become evident recently with the proliferation of AWB and other automated edits. I have seen multiple times when the {{unreferenced}} tag was removed just because someone added a single ref or an extlink, which was far cry from really referencing article. —- Altenmann >t 00:51, 17 January 2010 (UTC)

I don't think this is a good idea exactly as proposed. The current {{citations missing}} template is vague, as its documentation notes. It could refer to just a lack of inline citations/footnotes (which isn't always a problem), or a total lack of any stated sources. We don't want to make things worse by forcing {{unreferenced}} into that. • Now, perhaps it would be a good idea for {{unreferenced}} to adopt the wording currently at {{refimprove}}, so that adding a single citation to an otherwise unsourced article doesn't invalidate the template's presence. • Can anyone think of a benefit to having a distinction between "no sources" and "insufficient sources"? —DragonHawk (talk|hist) 01:58, 17 January 2010 (UTC)
Whatever you say. My point (you seem to agree with) that adding a single source should not invalidate the template. BTW I was always wondering but never actually looked into what is the distinction between {{refimprove}} and {{citations missing}}. May be they may be merged into one as well? - Altenmann >t 03:18, 17 January 2010 (UTC)
Another suggestion is to introduce an universal template, with severity gradation, something like {{uw-t1}}, {{uw-t2}}, etc. - Altenmann >t 03:23, 17 January 2010 (UTC)
  • Speedy close This is not a move request, it should be proffered in another forum for comments, like RfC or TfD (TfD since this appears to be akin to a merger request, and they handle deprecation anyways) 76.66.197.17 (talk) 04:42, 17 January 2010 (UTC)
    • Wikipedia is not a bureaucracy. If it gets the job done, who cares what tag gets put on it? —DragonHawk (talk|hist) 23:37, 17 January 2010 (UTC)
      • The forums exist for a reason, if we just use any old forum for any old discussion, why bother having different forums at all? Next people will send articles for deletion at WP:MFD or request renames at WP:IFD, and chaos will ensue. The forum that is specialized for the particular type of discussion will have active and regular participants who are more likely to be interested in it, or who will have specialized knowledge of the trouble the request can bring. 76.66.197.17 (talk) 04:45, 18 January 2010 (UTC)
      • RFC is too arbitrary and does'nt attarct proper people in proper numbers, TfD is inappropriate: I am not suggesting to delete anything. Lacking good standard mechanism, I have chosen what I thought best. "speedy closing" for bureaucratic reason is not helpful. - Altenmann >t 06:27, 18 January 2010 (UTC)
        • Wikipedia:Templates for discussion is designed for exactly this purpose (namely, the stated "requested redirect" reason here is a specifically handled component of the TFD process). It shares a similar initialism, but the process itself is entirely distinct from AFD. You're correct regarding "attracting the proper people", but while (mis)using WP:RM for this may attract the numbers desired it's not likely to draw the participants which are really desired.
          V = I * R (talk to Ohms law) 07:28, 18 January 2010 (UTC)
  • Oppose "unreferenced" is for articles with no references at all, "citations missing" is used on articles that have references. 76.66.197.17 (talk) 04:42, 17 January 2010 (UTC)
    • My question is, is there any benefit to making a distinction between articles with no references and articles with insufficient references? Both have the same basic problem (WP:V) and both need the same solution (add references). But maybe there is a benefit that I'm not realizing. —DragonHawk (talk|hist) 23:37, 17 January 2010 (UTC)
      • Yes, since they don't have the same basic problem with WP:V, since a completely unreferenced work is wholly problematic, while a partially referenced work has some parts that can be verified, and if necessary, everything else could be excised. 76.66.197.17 (talk) 04:47, 18 January 2010 (UTC)

Thanks you brought them into one place. Here are they expanded.

You probably know this kind of puzzle: "find 4 differences in the pictures." I say half of the above are useless duplicating. If fact, we need only three warnings:

  1. insufficient or no references
  2. there are some refs, but not all of them are bound to text with footnotes
  3. citation style and format is inconsistent.

Warnins of second and third type must have link to talk page with explanations what must be done, because the instructio of kind "citation style may be unclear. The references used may be made clearer" is a no-instruction. - Altenmann >t 06:19, 18 January 2010 (UTC)

  • Opposed - In addition to the process issues discussed above, since {{Citations missing}} is currently a separate entity, this requested move isn't actually possible regardless. On the merits of the proposal, setting aside the process issues, the fact is that this template doesn't address the use case which the Citatiosn missing template does. As 76.66.197.17 pointed out, the two template names represent different things to most editors.
    V = I * R (talk to Ohms law) 07:32, 18 January 2010 (UTC)
    You didn't read carefully. No move, but merge/redirect. - Altenmann >t 07:54, 18 January 2010 (UTC)
    Which is completely out of process for this request, which is why I essentially "refactored" the request to fit the process. I read quite carefully, actually. See also my reply above re: TFD.
    V = I * R (talk to Ohms law) 08:30, 18 January 2010 (UTC)
The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Tiny stubs question

The sentence "Consider not adding this template to extremely short articles," was added to the documentation in response to Template talk:Unreferenced/Archive 7#Tiny stubs. I'm currently in discussion with another editor over whether this applies to stubs in general. If it doesn't the perhaps the wording could be made more specific.--RDBury (talk) 04:52, 19 January 2010 (UTC)

There has been a week with no response so I'll try a different forum.--RDBury (talk) 00:55, 27 January 2010 (UTC)
I would suggest Wikipedia talk:Citing sources. Template talk pages (and the people watching them) tend to be focused more on the template itself (i.e., formatting, style, technical issues), rather than the policy/guideline the template is documenting. —DragonHawk (talk|hist) 03:34, 27 January 2010 (UTC)
I am going to be bold and remove the line. I think the idea behind it was to encourage editors to try to fix the problem (ie find sources) rather than just tagging it. But that has nothing to do with the size of the article. Even stubs should be referenced. Blueboar (talk) 21:54, 27 January 2010 (UTC)
There was a time when stub articles were more common, the idea being a sentence or two on a placeholder page was better than a redlink. That was in the early days, when the wiki web was still being built. Plus, ther were lots of basic pages that might not exist, like water or whatever. In that situation, one didn't expect references. Plus the emphasis on citation has increased over time. So I think it's more of a historical thing. But anyway, I think it's fair to say that these days, even stubs should be referenced. So agree. —DragonHawk (talk|hist) 00:36, 28 January 2010 (UTC)
I re-inserted the sentence. Yes, stubs should have sources added. But this is a consequence of a deeper issue: stubs should be expanded into proper articles. Once an article is tagged with a "stub" tag, that is enough to let the reader know that the article is extremely incomplete. There is no need for an additional, reader-visible template saying to add sources. If the article is no longer tagged as a stub, but somehow is still unreferenced, then add the unreferenced tag.
One reason that I think it is useful to have this line here is that we don't want people mass-tagging existing stubs on obviously notable subjects as "unreferenced" when the articles are already tagged as "stubs". Such mass tagging would not achieve anything (and, in any case, there is an {{unreferenced stub}} template that is better because it can be hidden).
So the issue is not whether stubs need to be verifiable, it is whether to overwhelm the tiny amount of material in the stub with numerous templates that all come down to the same thing: the stub article needs to be expanded. — Carl (CBM · talk) 00:46, 28 January 2010 (UTC)
I agree with Carl, that was the view that was held when the line was first discussed in 20006 (see /Archive 1#This template should not be used with stubs.) and I think it is the best procedure to follow. -- PBS (talk) 02:35, 28 January 2010 (UTC)
In the discussion when this line was originally added, see Archive 7, the editor said it was because the "tag isn't particularly helpful on two-sentence-long dicdefs-with-possibilities." The meaning is being extended to stubs in general though and this was not the original intent. Yes, the stub tag does indicate that an article needs to be worked on as well, but the fundamental difference here is that stubs are allowed in Wikipedia and unreferenced articles aren't. Addressing articles that have no references (either adding refs or deleting as OR) is a much higher priority than expanding stubs in general. There are thousands, if not millions, of stub level articles about obscure topics that may never be expanded, and there needs to be a way of identifying those that have high priority issues such as lack of references. I realize that the noreferences tag is not ideal because of the clutter issue, but to me clutter is relatively minor compared to there being no way of verifying the information. I will try to rephrase the line to be more in keeping with the original intent.--RDBury (talk) 16:40, 28 January 2010 (UTC)
FYI: The stubs line has been there since as of 16:27, 25 November 2005 after this discussion-- PBS (talk) 20:41, 28 January 2010 (UTC)
The edit I was looking at was this. Going further back in the history there is a different version of the statement that was removed.--RDBury (talk) 02:25, 29 January 2010 (UTC)
Was there discussion and a consensus to remove it? -- PBS (talk) 10:36, 31 January 2010 (UTC)
When you say "Addressing articles that have no references (either adding refs or deleting as OR) is a much higher priority than expanding stubs in general." – whose priority do you mean? There are tens of thousands of articles that are tagged as unreferenced, and many have been tagged as unreferenced for years. The situation on the ground is that the unreferenced template does not make articles a high priority.
Also, unreferenced articles are permitted on Wikipedia, in the sense that (1) they are not immediately deleted and (2) WP:V only requires inline citations for quotes and material likely to be challenged. We all agree that stubs should eventually have references provided, just as they should eventually be expanded into full articles, but there is not a policy that the references must already be in the article when it is just a stub. — Carl (CBM · talk) 17:31, 28 January 2010 (UTC)
I think the current wording takes your concerns into account, while still summarizing the policy according to WP:V. We can hopefully agree to disagree on the priority issue.--RDBury (talk) 02:39, 29 January 2010 (UTC)
I don't think it does. This template should not be used on small stubs as it overwhelms the article. If there is a point that really really needs a citation the the {{citation needed}} can be used to request it. -- PBS (talk) 10:43, 31 January 2010 (UTC)
I second that. I have been always for precise pinpointing what exactly needs to be referenced. In a large unreferenced article lost of {{fact}} would render the text unreadable. Therefore one prominent tag is good. In a small stub, it is easy and less intrusive to have 2-3 strategically placed {{fact}} tags. Mukadderat (talk) 02:19, 2 February 2010 (UTC)

Who cares whether it is "less intrusive"? It is supposed to be intrusive. All articles, stubs or not, should be sourced, and it is our duty to the readers and to everyone trying to cleanup articles that an article is completely unsourced. A stub tag has no meaning for most of our readers, but a big box indicating that what follows is unsourced is a good indicator that they should be wary. Being a stub is not the same as being unsourced, and many stubs are sourced. The two tags are not the same at all, and there is no reason why stub categories should not be taged as unsourced when they are completely unsourced. Removing unsourced tags from stub articles is a disservice to our readers and to Wikipedia as an encyclopedia, with the sole benefit that the article looks better. The lines stating that "unref" should not be placed on stubs should be removed ASAP. Fram (talk) 13:55, 9 February 2010 (UTC)

  1. ^ Jimmy Wales (2006-05-16). ""Zero information is preferred to misleading or false information"". WikiEN-l electronic mailing list archive. Retrieved 2006-06-11.