Wikipedia talk:Verifiability/Archive 42

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 41 Archive 42 Archive 43

Burden of evidence - first sentence

Regarding the first sentence of the section Burden of evidence,

"The burden of evidence lies with the editor who adds or restores material."

Does "material" refer to any material or just unsourced material? --Bob K31416 (talk) 00:08, 23 October 2010 (UTC)

Well, it mean all material, but if the material is sourced, that usually satisfies the burden of evidence. Gigs (talk) 01:23, 23 October 2010 (UTC)
Thanks. The reason I asked is that if it means all material, then that seems too general and one might interpret it to mean that when anything is deleted, the burden of evidence is on the editor who wants to restore it, whether it was sourced or not. And what evidence is needed isn't entirely clear, since it could mean evidence that the material is worthwhile, sourced or not. Also, even if the deleted material wasn't sourced, that could have severe consequences for articles that aren't sourced well. If an editor knew about this part of policy, and was inclined to delete, a lot of Wikipedia could be blanked out for not being sourced and it could be a long time before anyone finds sources so that the material could be restored, if I'm understanding this sentence correctly. --Bob K31416 (talk) 03:55, 23 October 2010 (UTC)
It's difficult to AGF here, Bob. The edit I made changed no content. It was a structural edit. For example, adding a header to the material that dealt with NOR "Reliable sources and original research." Adding a header to the material that dealt with notability: "Reliable sources and notability." Adding a header "Other issues". Adding a header "When a reliable source is required." Compare for yourself. It is only just over 1,400 words and so easy to see the difference—before and after. I do this kind of copy edit periodically to remove repetition and odd phrasing.
But you revert that, yet restore a contentious and significant content change made by QuackGuru? It is hard to think of a good-faith explanation for that. SlimVirgin talk|contribs 04:22, 23 October 2010 (UTC)
If you're questioning my motives for starting this section, please note that I was considering this topic well before the current festivities started.[1] Regards, --Bob K31416 (talk) 04:58, 23 October 2010 (UTC)
Well, this is a sentence that I think people generally accept. If you add or restore something, no matter what it is, the burden of evidence for showing that it's not OR—that a source exists for it as described in the edit—lies with you, not with the person who challenges it. SlimVirgin talk|contribs 05:02, 23 October 2010 (UTC)
One of my concerns is that it seems to make it difficult to restore something that has been deleted for any reason or no reason at all. It doesn't really say what type of evidence is required. --Bob K31416 (talk) 05:08, 23 October 2010 (UTC)
A reliable source is required, and the policy goes on to explain what's meant by that. SlimVirgin talk|contribs 05:11, 23 October 2010 (UTC)
When I read it back on Oct. 8, trying to help the editors back then, the first sentence seemed more general and the part about a reliable source that followed it seemed to be only one application. I sincerely felt it was a problem. --Bob K31416 (talk) 05:24, 23 October 2010 (UTC)
This sentence has been in the policy for a long long time... and it enjoys a strong consensus of the community. It refers to unsourced material that is challenged or likely to be challenged. If the material is not sourced, then it is up to those who want to add the material (or who wish to keep) to find a source for it. If someone adds (or returns) the material without a source, they run the risk of having it be deleted for not complying with WP:BURDEN. Once the material is sourced, then WP:BURDEN is complied with.
However, there are other policies and guidlines besides WP:V, and other provisions of WP:V besides WP:BURDEN that must also be complied with. Any of these might cause the material to be removed. Complying with WP:BURDEN does not "protect" problematic material from being removed or deleted for other reasons. Blueboar (talk) 13:38, 23 October 2010 (UTC)

Perhaps you could give your opinion about the following scenario. Suppose some sourced material has been in an article for years and it is deleted. An editor restores it and it is deleted again. The editor who deletes it says there is a burden of evidence for restoring it and it can not be restored without first getting consensus. --Bob K31416 (talk) 16:12, 23 October 2010 (UTC)

I think what Bob is after is the difference between these two scenarios:
  1. X adds Q, Y deletes Q, X has the burden of justifying the addition before re-adding
  2. ? X deletes Q, Y restores Q, X who has the burden, the one who deletes or the one who restores?
On one hand, we'd like to assume that X has a justifiable reason deleting that material, but from an encyclopedic perspective we are more concerned about the veracity of material that's in the encyclopedia, so one could argue that the burden lies on Y (the one putting material back in) to show that the material belongs. This would (I assume) be to counter the "It's been there for a long time" type arguments (no offense blueboar) or other efforts to maintain a potentially less encyclopedic status quo. --Ludwigs2 17:02, 23 October 2010 (UTC)
in the "X deletes Q for being unsourced, Y wishes to restore Q" scenario ... the burden to provide a source falls on Y, ie the person who wishes to restore the (previously) unsourced material. Blueboar (talk) 17:42, 23 October 2010 (UTC)
I agree. I was just trying to clarify what I think Bob's question is. --Ludwigs2 18:44, 23 October 2010 (UTC)
I appreciate your attempt, but I still don't understand Bob's concern here. Jayjg (talk) 02:34, 24 October 2010 (UTC)
Bob, with anything you restore (even if it was deleted after being there for years), the burden of evidence is on you to show that it's appropriately sourced. It's not a question of consensus alone, but of supplying a reliable source. Consensus may be needed to determine whether the source is good enough, but that's a secondary issue. The point in the policy is that the burden of supplying that reference falls on the editor who adds or restore something. It doesn't fall on the editor who challenges material, though the challenge is expected to be reasonable. SlimVirgin talk|contribs 07:34, 24 October 2010 (UTC)

I am not sure if this is Bob's point, but apart from simply looking at whether the logic of the wording is acceptable to good editors taking their time on thinking it through "in theory", one way to test whether a better wording is at least worth considering is to ask whether that sentence is widely being used "in practice" in ways which either (a) go against the obvious intended spirit of the sentence or (b) go against other important WP priorities such as making sure edits do not make article worse WP:NORULES or less WP:NEUTRAL. I would also suggest that looking at how the rule by real editors on real articles. I would especially remind that typical editors, as opposed to editors who are likely to discuss policy wordings, are occasional editors, and typical articles are in pretty low quality conditions and perhaps do not get seen by many people. So we should not take our bearing from an over-perfectionist ideal of an article which is already in a good condition. So:-

  • Concerning (a) would I be right in saying that current wording has been carefully crafted so as to be advice to editors inserting material, rather than positive encouragement to delete? So that for example if someone proposed that the wording should be strengthened into a positive call for people to go out and start deleting this would gain no consensus? That is my understanding of the spirit of the wording.
  • Concerning (b) my observation is that the burden sentence is normally cited by people wanting to make a POV edit, without the bother of having to answer concerns about whether the edits they want will make the article worse (in violation of WP:NORULES or non-WP:NEUTRAL. It is in other words often used to "wikilawyer" and try to get around consensus. On most WP articles, where sourcing is a long term job, years away from being brought to a good level, it is apparently easy to construct a POV strategy in terms of deletions so that you can then use this strategy of citing the burden sentence.

I expect there will be people who think that one can or should ignore such concerns simply by either pointing to the long standing nature of the sentence, or by saying that all policies can be abused by wiki-lawyers, sort of. It is however true that some are more often abused than others, and that the aim of good policy writing should be to make the policies less prone to this. Just my two cents. I do not have a solution, but it occurs to me that WP almost needs two standards. Low quality articles need positive contributions, they need making every edit an improvement to be the main priority, and making every edit neutral to be a second priority. It is time wasting to argue about excessively about sourcing of material which no editors express any doubts about, when the basic test is still in flux. (Taking it too seriously would mean spending time getting sources for a sentence which is about to be deleted or changed anyway. Writing a draft somewhere, in order to give yourself time to get the text right and then get the sourcing right is not always a happy solution, and certainly not in the spirit of cooperative working.) This is often made difficult by policies written with high quality articles in mind. That's just a fact.--Andrew Lancaster (talk) 09:16, 24 October 2010 (UTC)

The sentence has been in the policy for years, so it's hard to remember what the discussions were when it was added, but it's intended as advice to people who are adding or restoring. It's not intended as a licence to remove, and in fact we do say people should not go around removing things without giving others a chance, except for BLPs. But the best way to protect material is to source it, and it's almost always faster to do that than to argue the point. SlimVirgin talk|contribs 09:59, 24 October 2010 (UTC)
While WP:BURDEN is not indented to be an encouragement to delete material... it is intended as a) giving editors permission to delete unsourced material when needed, b) a warning to editors that unsourced material might be deleted, and (as SV says) c) encouragement for editors to properly source their material when they add it. Blueboar (talk) 12:44, 24 October 2010 (UTC)
Yes SlimVirgin and Blueboar (you are both saying approximately the same thing I said in the relevant place) that is obviously the finesse intended by the current wording. One does not need to remember when it was written to see that, but only to look at it carefully. Anyone looking at it carefully, and presuming that all people will do the same, will likely find it very reasonable indeed. If WP policy had a legal system to be used by professional judges only it looks fine. Back in the real world, my question was whether this finesse is working or whether the wording write now encourages "normal editors" to use the burden sentence as a rule in the sense in which WP is supposed to have no rules - an excuse to make edits which other editors consider to make an article worse, without having to give justification. It appears Bob does not think this is the question he was pointing to but maybe my concern is still worth considering.--Andrew Lancaster (talk) 15:30, 24 October 2010 (UTC)
It's not a matter of what your intentions are. I raised this issue because I personally had a problem with the sentence in practice at the discussion that I linked to previously.[2] --Bob K31416 (talk) 14:09, 24 October 2010 (UTC)
looking at the discussion you link, the issue did not really relate to WP:BURDEN (as I gather that the material was cited), but had more to do with WP:UNDUE (a different reason to remove material, covered by a different policy). Blueboar (talk) 14:11, 24 October 2010 (UTC)
Suggest looking at the message before mine in that section. --Bob K31416 (talk) 14:26, 24 October 2010 (UTC)
Yes, I looked at the exchange. If I had been answering that question, my answer would have been... "WP:BURDEN doesn't really enter into it (as the material is apparently sourced, WP:BURDEN is complied with). However, removing the material might be justified by other provisions of WP:V and by other policies." Blueboar (talk) 15:16, 24 October 2010 (UTC)
If there is a citation, the deleting editor could claim that the restoring editor has the burden of evidence that the source supports the material and consensus is needed before it can be restored, otherwise, just the presence of any citation whether supporting or not, could be claimed to have satisfied the burden of evidence. --Bob K31416 (talk) 15:25, 24 October 2010 (UTC)
I see what you are talking about... yes, implicit in WP:BURDEN is the idea that there is a burden on the adding/returning/keeping editor to provide a reliable source... ie one that reliably supports the information.
Essentially the chain goes as follows:
Step 1 - Editor A adds unsourced statement
Step 2 - Editor B challenges statement and removes (Burden is on A (or C) to provide a source)
Step 3 - Material is returned with source (WP:BURDEN is apparently complied with)
Step 4 - Editor B (or D) challenges whether the source reliably supports the statement. Editors discuss and get other opinions.
Step 5a - If consensus is that source does not support info, source is removed... the material is now again unsourced (return to step 2). If source is found to support the statement, end of argument. Blueboar (talk) 16:03, 24 October 2010 (UTC)
Here's my original scenario adapted to the most recent discussion and in the format like yours above.
Step 1 - Editor B removes a sourced statement that has been in an article for years.
Step 2 - Editor A restores the material.
Step 3 - Editor B removes it again with the claim that editor A has not satisfied the burden of evidence for restoring it and will need to gain consensus before restoring it.
They could then have a discussion about whether or not the source satisfies the burden of evidence, with the result that they disagree. Editor B then reiterates that it can not be restored without satisfying the burden of evidence and that consensus is needed. So we have the situation where material that has been in the article for years has been deleted and can not be restored without consensus because of the burden of evidence. In the meantime, editor B has made similar deletions of 10 other items in the article. --Bob K31416 (talk) 16:43, 24 October 2010 (UTC)
There's an unwritten understanding in all our policies that they will be applied reasonably. If something is reliably sourced, and it has been there for years—and the only issue is that the removing editor feels the source is not good enough, but it does satisfy this policy—then the burden of evidence would probably shift to the removing editor. But we can't account here for every eventuality. SlimVirgin talk|contribs 17:05, 24 October 2010 (UTC)
Yes, but quite obviously some policy wordings will encourage reasonable behavior and some won't and this talk page is a place where such choices can be discussed I think. Your idea that a removing editor should be reasonable and engaging is indeed most likely a consensus understanding in the community, but it is currently missing from the wording of this page. Hence, I would say, this burden passage is one of the those most used by people looking for a policy page they can quote when they want to work against the basic spirit of Wikipedia.--Andrew Lancaster (talk) 18:43, 24 October 2010 (UTC)
I'd have no problem adding a sentence about how requests for references are expected to be reasonable, so long as it was worded very carefully. SlimVirgin talk|contribs 18:50, 24 October 2010 (UTC)
Current paragraph:

The burden of evidence lies with the editor who adds or restores material. Any material lacking a reliable source directly supporting it may be removed. How quickly this should happen depends on the material and the overall state of the article. Editors might object if you remove material without giving them time to provide references. It has always been good practice to make reasonable efforts to find sources yourself that support such material, and cite them. Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living persons or organizations, and do not move it to the talk page.

Proposal for discussion:

Within reason, the burden of evidence lies with any editor who adds or restores un-sourced material. Deletion of un-sourced materials is generally considered acceptable as long as it does not make an article worse or non-neutral. Editors deleting un-sourced material which they do not have any reason to suspect as unverifiable should be willing to discuss the subject patiently with editors of the article, if any, and to make reasonable efforts to see for themselves whether sources are available. Do not leave unsourced or poorly sourced material in an article if it might damage the reputation of living persons or organizations, and do not move it to the talk page.

Would be interested to hear what other think.--Andrew Lancaster (talk) 19:10, 24 October 2010 (UTC)
I would not say that B has a "burden of evidence" in your senario... rather he has a "burden of explanation"(ie B needs to explain why he removed sourced material - either on the talk page or in an edit summary). And if someone disagrees with his reasoning then there is an equal "burden of discussion" on everyone involved with the article to discuss the disagreement and reach a consensus. But those "burdens" are not what is being discussed by WP:BURDEN. That provision relates purely to where the burden to provide sources lies when it comes to unsourced information. Blueboar (talk) 17:23, 24 October 2010 (UTC)
You know, I think part of the concern here might be the (contentious and age-old) question of what should the page look like while discussion is ongoing, the fear (of course) being that once the article looks the way that editor Z likes, editor Z will loose all interest in discussing the matter and count on inertia to keep that preferred version in place. it might be nice to set out some kind of vague guideline on this that would both encourage editors to take a longer view (so that they weren't so worried about a 'bad' revision being in place 'right now') and place some reasonable limits on how long a revision can stand without proper discussion. But I don't think that's really part of wp:V, and should probably be discussed elsewhere. However, we might be able to salve some of those worries with a 'being verified' template: something like:
That might help everyone feel comfortable that about the state of the article while discussion is ongoing. If you think that will help, I'll whip it up in template space. --Ludwigs2 18:12, 24 October 2010 (UTC)
So will this be attached by default to all articles and then removed in those small number of cases which actually really have active editors that will notice it and respond? The working assumption here seems to be that tagging or deleting material is like starting a conversation. A conversation with who? I keep thinking that there is a problem developing on WP whereby the most active editors who consider questions like this are more and more exclusively involved in editing quite good article where a lot of work has already been done, perhaps having taken years, and they are no longer able to relate to the fact that WP is mainly made up of articles which remain in a poor state for very long periods and need help. Even articles about quite major subjects can remain like this for surprisingly long, and changing that needs more positive encouragement. I suspect that a big part of the problem is a subconscious fear of getting involved in a bad article. It is awkward to even become associated with a poor quality article because there are many more people willing to get in the way of such work than to do it.--Andrew Lancaster (talk) 18:52, 24 October 2010 (UTC)
I don't know about 'by default'. I see this more as a neutral way for editors to express their discontent with the state of a page during an active disagreement, without giving in to that very natural 'gotta fix it right now!' urge. The emotional side of this issue is difficult: Imagine a psych experiment where you took a mixed bunch of liberal and conservative students and forced them all to wear hats that say "Rush Limbaugh (or alternately Michael Moore) is absolutely right!" while they have a discussion about what their hats should say. You can imagine the consternation that would cause for one half of the group. But that is exactly what wikipedia does with every controversial discussion; it forces one set of proponents to swallow the fact that the point they disagree with is being broadcast to the world as truth, even while they are discussing removing it. I'm just suggesting this as a way to tone down the emotional impact to relieve some of that pressure. --Ludwigs2 19:21, 24 October 2010 (UTC)
I do appreciate your thoughts, but my practical point is possibly best expressed by saying that WP articles maybe need to be treated VERY differently depending what state they are in. MOST are in a state where they need more people willing to edit, not less. The last thing they need is an argument about page numbers or some other fine point of sourcing. Some will object that the current text is not about page numbers or fine points, but no, the current wording does not say it is not about page numbers. It is deliberately vague and therefore left very open to abuse. And the abuse is happening. Effectively the current "burden sentence" appears to be intended to make people think twice before adding material at all - or it can be interpreted that way, and the fact is that there are lots of people around to interpret it that way for potential editors, especially new editors, should they actually dare. People may one day look back and see that particular bit of text as the first step of WP towards becoming a more exclusive but less successful form of editing community.--Andrew Lancaster (talk) 19:41, 24 October 2010 (UTC)
Andrew, I could live with a lot of your proposed language... but I strongly object to the phrase " long as it does not make an article worse...". I foresee huge and endless arguments over whether removing some bit of unsourced information "makes the article worse" or "makes the article better". With the current language such arguments are avoided. If I think the article is better with the information in it... all I need to do is find a source. Blueboar (talk) 21:51, 24 October 2010 (UTC)
Well, I don't know if leaving something open to debate is always a bad thing, and nor will it always actually lead to more debate. Trying to close debate from a policy page is like trying to take the same side in a lot of different discussions at once including onces which have not yet taken place. To look at another policy WP:SYNTH is similarly covering a critical point that can be very difficult to define and very hard to agree about. It helps define the terms of any discussion, without having been written in a way so open to abuse by one side over another. So people cite it at each other and then they tend to have to find a way of collaborating anyway. And it is not the source of anywhere near as much abuse as the burden sentence? I'd also point out that the words you object to are already policy, in the sense that they are basically the same as WP:NORULES. As discussed above, I do accept that from a purely logical point of view, if WP were run by judges, the burden sentence has been written to allow the existence of that policy, but why does it only "leave space" for it while seeming to avoid mentioning it explicitly? Isn't that a bit odd in itself? It looks almost like a loophole left in a legal system in order for people who know the system well to use while not allowing others to see it.--Andrew Lancaster (talk) 07:00, 25 October 2010 (UTC)
Would it be acceptable to everyone to at least add to the first sentence of the current version of the policy, the word "unsourced" so that it would become,
"The burden of evidence lies with the editor who adds or restores unsourced material."
--Bob K31416 (talk) 06:22, 25 October 2010 (UTC)
That would seem uncontroversial to me compared to the current version. Another small adaptation would be to add a "Within reason...". I guess I am personally wondering whether the community should not consider being more ambitious though.--Andrew Lancaster (talk) 07:00, 25 October 2010 (UTC)
Introducing too much subjectivity, as in the "within reason" proposal, would generate a lot of useless conflict ("my removal was reasonable" "no it was not" "yes it was") and in the long term damage the reliability of Wikipedia, by encouraging editors to add unsourced information as long as they consider it within their own personal reason. The "no original research" policy was introduced precisely to prevent insertion of material that is only based on personal reasoning by Wikipedians.
As Slimvirgin mentioned above, all policies are expected to be applied with a sense of proportion (also per WP:IAR, as you say), but adding such an explicit qualification to a particular policy statement still amounts to weakening it considerably.
Regards, HaeB (talk) 08:18, 25 October 2010 (UTC)
Saying that policy wording do not need to be changed because all policies should be interpreted using a sense of proportion or common sense or whatever is also "subjectivity", surely? As I have argued above, trying to make a question which is inherently subjective fit in a Procrustean objective bed is what leads to useless debates. A policy which honestly treats a subjective subject as a subjective subject is much less likely to be abused for the simple reason that it will indeed lead to a dead lock and not help people "win" an argument which can not be won on other grounds. A policy wording which "kind of" "within reason" takes sides in a way which is "kind of" and "within reason" against other policies, language like we have now, is lawyer fodder. It is begging to be abused and it gets abused. Why may we not consider improvements?--Andrew Lancaster (talk) 08:29, 25 October 2010 (UTC)
From the practice I am seeing the explicit qualification is now becoming more necessary. If the environment allows WP:IAR to be called upon easily then it wouldn't be necessary but if WP:IAR is now to be used only as an extreme last resort then this policy needs to become more flexible. Lambanog (talk) 08:36, 25 October 2010 (UTC)
And do you think allowing other policies to appear to be in conflict with IAR will cause people to cite IAR more or less?--Andrew Lancaster (talk) 08:52, 25 October 2010 (UTC)

Poll - Proposal to add "unsourced"

Propose to add to the first sentence of the Burden of evidence section the word "unsourced" so that it would become,

"The burden of evidence lies with the editor who adds or restores unsourced material."

--Bob K31416 (talk) 01:58, 26 October 2010 (UTC)

Well... perhaps we should be even more specific... When unsourced material is challenged or likely to be challenged, the burden of providing a source falls on those who wish to keep the material in the article, not on those who challenge it. This gets rid of the potentially confusing word "evidence". Blueboar (talk) 02:00, 26 October 2010 (UTC)
LOL. Looks like the original proposal is doomed to defeat after only one response. --Bob K31416 (talk) 02:15, 26 October 2010 (UTC)
Sorry. :>} I don't object to your original proposal... just offering an alternative. (talk) 03:32, 26 October 2010 (UTC)

The burden is more than just providing "a source". It's the burden of providing evidence of verifiability. A source may or may not provide evidence of verifiability. For this reason, I oppose this change. Gigs (talk) 19:10, 27 October 2010 (UTC)

Copy edit

Here are the basic points: [3] -- 25 diffs in all.

  • The sentence: "This policy requires that anything challenged or likely to be challenged, including all quotations, be attributed to a reliable source in the form of an inline citation, and that the source directly supports the material in question."
  • Became: "This policy requires that all quotations and any material challenged or likely to be challenged be attributed to a reliable, published source using an inline citation, and that the source directly support the material in question."
  • I added a header, "Reliable sources and other principles".
  • Under that header I took the one sentence that dealt with notability, and I added a subhead, "Reliable sources and notability".
  • Under that header I took the three sentences that dealt with NPOV, and I placed them under the subhead, "Reliable sources and neutrality".
  • Under that header I took the sentences that dealt with OR, and I added a subhead, "Reliable sources and original research".
  • Under the NOR subhead I added: "The No original research policy (NOR) has three requirements relevant to the Verifiability policy:"
  • I also added under the NOR subhead an explanation for "attributable": "This means that a source must exist for it, whether or not it is cited in the article. Wikipedia must never be a first publisher."
  • I added a header "When a reliable source is required".
  • Under that header I added a subhead, "Anything challenged or likely to be challenged".
  • Under that subhead I added the two sentences (already there) about challenged or likely to be, etc.
  • Under the header I added another subhead, " Burden of evidence".
  • Under that subhead I added the burden of evidence material (already there).
  • Under the Reliable sources header (already there), I added a subhead, "What counts as a reliable source".
  • I added a header, "Sources that are usually not reliable".
  • Under that header, I added the questionable and self-published material (already there).
  • I added a header, "Accessibility".
  • Under the header I added subheads "Access to sources" and "Non-English sources" (the second was already there).
  • I added the header "Other issues".
  • Under that header I added the subheads that were already there about tagging a sentence, exceptional claims, reliable sources noticeboard, Wikipedia and sources that mirror it.
  • I removed See alsos already linked in the article.
  • I added a quote from Jimbo to a footnote.
  • I fixed a citation in a footnote.
  • I joined two paras in the lead.
  • I added to the lead "For how to write citations, see Citing sources. "

SlimVirgin talk|contribs 15:50, 24 October 2010 (UTC)

comments on SV's edits

SV, thank you for the break down... it helps. While I do wish you had broken your changes down into smaller chunks, to make them easier to digest, I have no objections to any of them. I don't see how any of them significantly change either the meaning or language of the policy in any way. Blueboar (talk) 16:12, 24 October 2010 (UTC)

I was very careful not to change meaning, and I'm pretty sure I succeeded. The purpose was simply to help people find things, e.g. the notability issue under a "Reliable sources and notability" subhead, and so on. The content was all there, but we were forcing people to hunt for it. SlimVirgin talk|contribs 16:14, 24 October 2010 (UTC)
SV's changes look good to me. I've added a link to the Misuse of primary sources section of the BLP policy to the new sentence saying that primary sources are welcome. --JN466 10:44, 27 October 2010 (UTC)
SV's edit look bad to me and it is hard to comprehend how it was a copy edit. There are problems with the changes. for example, this edit did change meaning. I don't see how the significantly changes inproved the page. There was no clear explanation how all the changes and moving around the text improved the page. QuackGuru (talk) 18:38, 27 October 2010 (UTC)
Could you give one example of the meaning having changed? SlimVirgin talk|contribs 18:43, 27 October 2010 (UTC)
I already did in my previous post. Do you think you need broader consensus for the mass changes. Do you still think the changes were a copy edit after I showed the diffs. QuackGuru (talk) 18:47, 27 October 2010 (UTC)
I can't see a change of meaning in that diff. Can you post here a new sentence or some words that you think might have changed meaning? SlimVirgin talk|contribs 18:51, 27 October 2010 (UTC)
You were not able to explain how your edits improved the page. Moving around the text to different places was confusing to me and did not flow as well as the previous version. You can't see a problem with the diff I provided is the problem. You did not explain very well how all those changes was an improvement. I don't see the purpose of the confusing changes.
Note: SV has not responed to previous questions. "Do you think you need broader consensus for the mass changes. Do you still think the changes were a copy edit after I showed the diffs."[4] QuackGuru (talk) 19:01, 27 October 2010 (UTC)
Okay, QG, I'm not psychic. :) You have to show me words that changed the meaning, by addition or omission. I don't see any, and I can't guess about what you've seen. SlimVirgin talk|contribs 19:51, 27 October 2010 (UTC)
According to your edit summary there were lots of little changes being made without clear benefit. That is your opinion all the changes did not have a benefit. QuackGuru (talk) 20:01, 27 October 2010 (UTC)
SlimVirgin never stated that her copyedits were "lots of little changes being made without clear benefit", so it's rather dishonest to claim or imply she did. And you have yet to show that any of her copyedits changed meaning, by addition or omission. QG, you can't re-write a fundamental content policy solely for the purpose of winning edit-wars on the chiropractry and similar articles. Wikipedia is not a forum for bashing "quackery". Yes, your attempts to significantly modify policy were reverted, because they did not have significant support; but that's no reason to say "if I can't edit the policy, then anyone who opposes my edits can't edit it either". Copyedits are not policy changes. Please move on to something more productive. Jayjg (talk) 04:22, 28 October 2010 (UTC)

requested edit to clarified wording


There are some acronyms in the headings in the Reliable sources noticeboard and WP:IRS section. It would be clearer to have a pipped link with some descriptive text linking to the policy page. Here are the applicable changes to the Reliable sources paragraph:


Reliable sources noticeboard and WP:IRS

To discuss the reliability of a specific source for a particular statement, consult the reliable sources noticeboard, which seeks to apply this policy to particular cases. For a guideline discussing the reliability of particular types of sources, see Wikipedia:Identifying reliable sources (WP:IRS). In the case of inconsistency between this policy and the WP:IRS guideline, or any other guideline, the policy has priority.


Reliable sources noticeboard and Identifying reliable sources

To discuss the reliability of a specific source for a particular statement, consult the reliable sources noticeboard, which seeks to apply this policy to particular cases. For a guideline discussing the reliability of particular types of sources, see identifying reliable sources. In the case of inconsistency between this policy and the Identifying reliable sources guideline, or any other guideline, the policy has priority.

Piped links for guidelines is confusing. A piped link looks like a wikilinked article. QuackGuru (talk) 00:39, 26 October 2010 (UTC)
I prefer it with "Wikipedia:" (or "WP:") in the front... when you see "Wikipeida:" in a link, you know it's pointing you to a policy/guideline page. (That was one of the first things I figured out when I first started editing Wikipedia). Blueboar (talk) 01:55, 26 October 2010 (UTC)
Or an essay, or a help page, or several other things. There's more to the WP namespace than policies and guidelines. See WT:POLICY#Shortcuts_masking_essay_status for a relevant recent discussion. WhatamIdoing (talk) 16:03, 26 October 2010 (UTC)
I guess none of the other links on this page use the WP prefix - they are all piped links. I thought it looked shoddy having the one piped link. meshach (talk) 17:21, 26 October 2010 (UTC)
I thought the other piped link to the noticeboard in the same section was confusing. QuackGuru (talk) 19:53, 27 October 2010 (UTC)

Scholarly and media sources redux

There is merit to both sides of the debate here:

  • In fields that are the subject of significant academic research -- whether it be hard science or the humanities -- we must make sure that we reflect the current status of that research. If all we cite is newspapers, we have done a poor job.
  • If there is significant discussion in the media around one of these topics, we must tell the reader about this discussion. If we don't, we have done a poor job.
  • Where media reports are based on scholarly studies, it is best to cite both the media report and the study, and where they disagree, go with the underlying study.

If we are agreed on the above principles, the task becomes a simple one: find a wording that encapsulates those principles. --JN466 22:11, 24 October 2010 (UTC)

  • I agree with number one, but it's a red herring, because no one has argued that we must only cite newspapers. But we must not reflect the current state of that research to the exclusion of other views, if other views exist in reliable sources. That is the essence of NPOV, and this policy can't say anything that contradicts that. There may be significant criticism of the current academic position in minority sources. A body of academics must never be allowed the only word on anything, and I'd be amazed if any experienced academic would disagree with that principle.
  • I agree with two, per NPOV—and by "significant," I would mean if the source was a high-quality one, not only if there had been multiple reports.
  • I agree with three in cases where the newspaper has made a simple error. Otherwise, where scholarly and high-quality media sources disagree—if it is clear that it is an active disagreement—cite both, with careful in-text attribution. SlimVirgin talk|contribs 22:21, 24 October 2010 (UTC)
It seems to me that the quickest way to show that Wikipedia has no ambition or ability to be a truly high quality encyclopedia is to treat media and scholarly sources equally when discussing scholarly subjects. NPOV is clearly upheld with the new proposals as any significant minority opinion in the scholarly community would almost always already be published in scholarly sources. That a NY Times health column should be equally as reliable on hard medicine as say, a Cochrane systematic review, if they are contradictory, is to my mind incomprehensible. Yobol (talk) 22:37, 24 October 2010 (UTC)
But you don't address the issue of why only scholarly minority opinion would count. This is not an academic journal. Scholarly point of view has been rejected many times by the community. I don't want to know only what drug company X tells me about the drugs it wants to sell me. I also want to know what the London Times says about that drug. I feel very certain that most of our readers and editors would agree with me about that. SlimVirgin talk|contribs 22:42, 24 October 2010 (UTC)
What would you say is the ratio of cases where there has been a significant cover up in the medical community that was not reported in a scholarly source to the number of times a media article got medical facts/nuance wrong in an article? Because that's really what we're balancing here; the unreliability of media to get scholarly subjects right (when compared to scholarly sources) vs. your perceived need for an outlet for those cover-ups to be exposed, right? Yobol (talk) 22:54, 24 October 2010 (UTC)
SV, why do you care what either of those sources say? You should care what the Clinical practice guidelines say. That is, you should pay attention to what the independent, experienced specialists say, not what the drug manufacturer tells you -- or what they tell a journalist. WhatamIdoing (talk) 23:47, 24 October 2010 (UTC)
  • Well, we are in agreement then, Slim. As for 1 being a red herring, see my post below. As for 3, the simple error was the scenario I was talking about -- the newspaper report being based on a study, but misreporting or overstating an aspect of the study's design, or findings. So let's get our thinking hats on and work out the wording. --JN466 02:54, 25 October 2010 (UTC)

Newspapers tend to be more biased in favour of drugs, whereas the academic community tends to be more critical and sceptical of the evidence or lack thereof; newspapers, especially now a days in the age of fast food journalism, just summarise press releases by the drug companies, hence the bias. So I think the newspaper viewpoint that you express is a red herring. For example newspapers may report an off-label, unapproved use of a drug, via some doctor working for the drug company. So are you saying slim that we should add in "new prescribing indications", based on what a newspaper says? How is that neutral? How is that sensible? We need to be careful on writing medical information. As I said in another section above, if a "non-academic/non-science" social issue arises in the the public domain such as "drug company found guilty of xyz", "concerns about biased/flawed research" etc, then a controversy section can be created and carefully sourced to newspapers.--Literaturegeek | T@1k? 23:29, 24 October 2010 (UTC)

As an aside, I find it hard to believe that "most of our readers" would share the belief that a journalist is equally reliable as a doctor with regards to medical information. But maybe that's just me. Yobol (talk) 00:18, 25 October 2010 (UTC)
Again, it depends on what is being said. Dr. Looney says in a TV advertisement that rubbing yourself with Looney's patented snake oil cream cures cancer... the New York times reports that researchers at Johns Hopkins studied the effects of snake oil on cancer cells and reached the conclusion that snake oil has no effect on cancer. Who are you going to call more reliable, Dr. Looney or the New York Times report? Blueboar (talk) 00:36, 25 October 2010 (UTC)
If anyone was trying to suggest we use Dr. Looney's commercial as a source here, you might have had a point. Yobol (talk) 00:44, 25 October 2010 (UTC)
OK, lets take a more realistic scenario... Dr. Plausible conducts a study in which he finds that snake oil cures cancer and publishes his findings in a reliable journal... however, the folks at Johns Hopkins say snake oil has no effect... the New York Times report is still just as (if not more) reliable, since they are reporting on the opinion of an equally (if not more) reliable source (Johns Hopkins). On the other hand, if the Weekly World News says Johns Hopkins is against snake oil, I will take it with a grain of salt and probably call Dr. Plausible more reliable. As I said... it all depends on what is being said and who says it, as well as where it is published.
I agree that in very very general terms, something published in an academic journal is likely to be more reliable than something published in a media report... however, there will be a lot of exceptions to that likelihood... enough exceptions that we can not make a fast and firm rule that says "academic sources are more reliable than media ones". We always need to look at the specifics. Blueboar (talk) 01:10, 25 October 2010 (UTC)
Actually, in the scenario you highlight, I'd bypass the media altogether and just use Johns Hopkins' study directly (that, of course, being the academic source). BTW, none of the new proposals say academic sources are "always" better. They all say academic sources are "usually" the most reliable, allowing for the exceptions that will inevitably arise. If, as I think almost everyone would agree, that academic sources are, in general, more reliable than media sources (with exceptions) about academic material, I don't understand why we would want to keep the current policies that treat them as equals. Yobol (talk) 01:32, 25 October 2010 (UTC)
I could support "academic sources are likely to be the most reliable sources on academic topics" .... "usually" is too strong a word. And that does not mean that media sources are unreliable. Blueboar (talk) 01:48, 25 October 2010 (UTC)
I don't have a problem with using "likely" instead of usually. But if our ambition truly is to write an encyclopedia, in the customary sense of the word, we should highlight the importance of scholarly sources. Academic sources are not just more reliable on average, they also do a better job at analysis. They are not just an "as-good" alternative to press sources, they are really indispensable for writing an encyclopedia. To respond to Slim, no one may have argued that we should only cite press sources, but this is what happens often enough. Examples:
Without accessing scholarly sources, we are missing a whole level of discourse that an encyclopedia should arguably contain. As long as we have a mindset that press sources (which are more easily accessible) are "just as good" as scholarly sources, this is unlikely to change. I have consciously chosen examples from the humanities and social sciences here. We have a similar lack of authoritative sources in engineering; see e.g. Transmission_(mechanics), Non-synchronous transmission, and many others. --JN466 02:24, 25 October 2010 (UTC)
By way of needfully complicating this: Not all scholarly sources are equal. A feature-length story in the Times might well be a better source than speculation in Medical Hypotheses or a letter to the editor in any academic journal.
And on that note, somebody might want to take a look at this unsourced change to the encyclopedia article, made on the heels of a dispute about whether academic psychologists violate WP:COI if they inconveniently notice that you're citing a letter to the editor by a non-expert. This change to MEDRS by the same new editor probably also requires some attention. WhatamIdoing (talk) 03:08, 25 October 2010 (UTC)
Sure. By the same token, a meticulously researched and fact-checked 60-page article in The New Yorker is far more valuable and authoritative than a conference paper by an obscure academic. Perhaps we need to indicate the scale of reliability for both scholarly and press sources. In press sources, this ranges from publications like The New Yorker to the yellow press; in scholarly sources, it ranges from peer-reviewed systematic reviews and authoritative standard works to conference papers and "peer-reviewed" articles in fringe journals. --JN466 03:56, 25 October 2010 (UTC)
QuackGuru posted some interesting sources further up on this page:
  • Communicating Medical News — Pitfalls of Health Care Journalism
    "Journalists sometimes feel the need to play carnival barkers, hyping a story to draw attention to it. This leads them to frame a story as new or different — depicting study results as counterintuitive or a break from the past — if they want it to be featured prominently or even accepted by an editor at all."

I think Blueboar's posts above are pretty much spot-on. The potential problem with this discussion is that it attempts to generalize too strongly about too many different types of case. And what is the problem being solved? What is the aim here? Are there lots of people arguing against using academic sources, or is the aim to get wording on a policy page which might eventually make it difficult to use anything but academic sources?--Andrew Lancaster (talk) 08:06, 25 October 2010 (UTC)

I don't think people are arguing against using academic sources... I think the subtext is an argument against using media sources. I object to that. Blueboar (talk) 13:13, 25 October 2010 (UTC)
No one's saying we shouldn't use media sources. I am saying we should use the highest quality source, regardless of which that turns out to be (media or scholarly), especially if there's a disagreement between the two. If it's on the social/political impact of a scholarly topic, the media source is usually going to be better. If it's on the hard facts and nuances of the scholarly topic, the scholarly source is usually going to be better. I see no reason why we shouldn't provide this guidance to our editors, rather than trying to imply a false equivalence to all these sources. Yobol (talk) 13:23, 25 October 2010 (UTC)
You are butting up against WP:NPOV here. When an academic source and a media source both say the same thing, it is fine to say we prefer the academic source (as it is a better source for that information). However, when an academic source and a media source say different things (and especially when they disagree) we must present both sides of the debate (giving each viewpoint due weight). The media source may be the best source we can (at the moment) find for a particular viewpoint. If we later find a better source for that viewpoint, great... but until we do, the media source is acceptable. Blueboar (talk) 13:53, 25 October 2010 (UTC)
A question I asked you earlier, which you did not comment on: "If, for example, a medical study comes out that says "Vitamin Q, in some selected patients, may in the future prove useful for treating Rare Disease X based on these preliminary studies" and Local Paper reporter reports "Vitamin Q cures Rare Disease X!", clearly the study is more reliable and the newspaper article should be ignored for the purposes of the wikipedia article on Vitamin Q, right?" If one source is clearly more reliable than another (about the same information), then shouldn't it take preference? Yobol (talk) 14:14, 25 October 2010 (UTC)
I don't think policy should be the place for this debate. If anywhere maybe at RS as a guideline. Even there I'd be a little hesitant. This is instruction creep. In my view good faith editors on talk pages should be able to resolve this issue and that would be the most natural place for it due to the case-by-case applicability. Lambanog (talk) 08:26, 25 October 2010 (UTC)
Jayen, just to respond to a point you made above (which I now can't find so apologies for slapping it here at the end), I have never anywhere argued against using scholarly sources. I know you know that, but I just want to stress it.
My argument is that they must never be given the only word on an issue. My argument is one of inclusion, not exclusion. Why do editors fear: "Drug Company X says our drug is really safe and wonderful, and the FDA and all clinical studies agree. Against that, the London Times says patients feel ill after taking it." What is wrong with letting the reader decide for themselves?
What I see in the "scholarly point of view" arguments is contempt for the reader. It is "we know best." But that is pre-Wikipedia, anti-Wikipedia. The whole ethos of Wikipedia is "we don't actually know best, so here's the information and you can make up your own mind." We must never lose sight of that, and I'm disturbed to see these discussions keep playing themselves out in various places, as though we have to reinvent the wheel every year. This issue was decided when the Wikimedia Foundation embraced neutral point of view as a founding principle. It means we don't get to exclude a point of view in the London Times, even if we really hate it. SlimVirgin talk|contribs 10:51, 25 October 2010 (UTC)
Slim, of course I know that you are not against using scholarly sources. I need only look at animal rights, an article whose main author you are, to see that. Look at the sources you used: Animal_rights#Notes. It is a well-referenced article that Wikipedia can be proud of. But ask yourself, could you have written an intelligent article like that just from newspapers? No. I remember checking through Abu Nidal when it was at WP:FAC. These are your sources: Abu_Nidal#Notes. You couldn't have cobbled that one together from news sources, either.
I am bound to mirror what you are saying here: I've never argued against using media sources in order to exclude a point of view. What I have said, and stand by, is that scholarly sources and media sources have different strengths. Writing on global warming or cancer using predominantly press sources would be a travesty and a joke. Where there is a substantial body of academic research, it's our job as an encyclopedia to reflect that body of research, using the best sources we can lay our hands on. In medicine, that may mean going to the study that the Times article saying people get ill is based on, and citing both the study and the press source together. That's giving the reader the accurate, high-quality information they need to make up their mind. --JN466 11:50, 25 October 2010 (UTC)
I agree with that 100 percent. We should reflect the academic research, and we should reflect the criticism of it from outside academia. Readers aren't stupid. If we link to Professor ImportantPerson of Kings College London, on the one hand, and Mr Neverheardofhim in some (high-quality) newspaper on the other, the reader can decide for herself who to pay attention to. And vice versa (because sometimes it's Professor Neverheardofhim). That's the essence of NPOV.
So, bottom line: the above is what this policy is geared toward. That is what the policy currently seeks to protect. My question: how do the proposed changes safeguard that principle of inclusion in a way the current policy doesn't? SlimVirgin talk|contribs 12:01, 25 October 2010 (UTC)
Slim, not really. We don't discuss both Jenny McCarthy (who has been quoted in many newspapers) and Dr. Francis Collins in the vaccine article, and let the "reader decide for herself" who to believe. NW (Talk) 12:21, 25 October 2010 (UTC)
The problem at the moment, NW, is that we also don't report research from universities as reported the BBC, even when a link to the study is included along with the secondary source. Certain biomedical articles have been handed over entirely to the scientists who are often being funded by the people who are selling whatever the product is. So we have a problem on both ends, and I hope we can at least agree on this point. On the one hand, we don't want to give serious space to supermodels with views on medical issues. On the other we don't want those articles to be controlled by companies who are selling the biomedical products the articles describe or refer to. SlimVirgin talk|contribs 12:45, 25 October 2010 (UTC)
I don't think anyone is arguing to use only media sources, or to only use scholarly ones (and I think we're getting too bogged down arguing against those straw men). The concern for improving the policy, from my perspective, is that currently it implies that media sources are as reliable as scholarly one, which in the case of academic material is almost certainly not true in general. If we are going to write a truly high quality encyclopedia we should be encouraging our editors to using the highest quality sources over lower quality ones (if they are in disagreement). I would hope that we're all here to write the highest quality encyclopedia we can, and implying, in policy, that a health column in a newspaper is as reliable as a review article in the New England Journal of Medicine flies in the face of that. Yobol (talk) 12:58, 25 October 2010 (UTC)


Wikipedians are normally highly attuned to conflict of interest, to who is benefiting. "Follow the money" is the old journalistic advice. But when it comes to science or medicine, that instinct is often lost. Scientists financed by companies that are making billions of dollars from a certain drug are regarded as the best sources of information about that drug, and may be the only sources currently allowed in some articles, where even to ask who has funded research is seen as inappropriate.

How is it that our instincts about money change when it comes to these issues? SlimVirgin talk|contribs 12:50, 25 October 2010 (UTC)

Raise it at the WP:MEDRS talk page. One of the sources QuackGuru posted above chastised media articles for not revealing which of the cited studies and experts were funded by the drug's manufacturers. That criticism also applies to us if we do not include this information. --JN466 12:55, 25 October 2010 (UTC)
And the impact of our bias towards massive publishing conglomerates and legitimated academic knowledge produced predominantly by the spawn of the bourgeoisie in pampered incestuous nests of limited "competitive" funding on our social sciences and humanities articles is? Wait two hundred years, the encyclopaedic correctness of our pharma articles will sort itself out. Fifelfoo (talk) 12:57, 25 October 2010 (UTC)
There is an interesting article on this: Angell, Marcia. "Industry-Sponsored Clinical Research. A Broken System", The Journal of the American Medical Association, September 3, 2008. She writes: "Over the past 2 decades, the pharmaceutical industry has gained unprecedented control over the evaluation of its own products. Drug companies now finance most clinical research on prescription drugs, and there is mounting evidence that they often skew the research they sponsor to make their drugs look better and safer."
And yet some of the proposals above would hand over control to them as prioritized sources in articles about their products. SlimVirgin talk|contribs 13:21, 25 October 2010 (UTC)
I think that Yobol is right, the argument here has been about straw men for far too long. I think it is reasonable to assert that newspapers are in general not as reliable as academic journals. In cases where newspapers are simply reporting about a study, the journal should be cited over the newspaper and the newspaper should be linked using the laysummary field of cite journal (This study, for example). But not always even that; newspapers are often pretty terrible.

I cannot think of any investigative journalism off the top of my head that has lead to major changes, but it would not surprise me to learn that it has led to recalls, questions about safety that are then investigated by people who have the ability to do valid research about the topic, etc. Slim, can you give an example where a preference for academic sources as proposed by these new guideline changes would exclude valuable information that should be in the article (that you alluded to before)? NW (Talk) 13:41, 25 October 2010 (UTC)

NW, I'll come back to your point, but could you first of all address mine? How do we handle conflict of interest in cases where the only studies we can cite have been financed by the companies who are selling the product the article is about? That is not an unusual situation but a common one. SlimVirgin talk|contribs 14:01, 25 October 2010 (UTC)
We should in-line attribute every claim, such as "According to a 2010 NEJM study financed by Big Pharma Company, Drug Alpha does blah blah blah. That same study claimed..." and so on. If we receive word later on that the conflict on interest has damaged the accuracy of the supposedly peer-reviewed paper in some way, then we should drop it as a source altogether. I would be surprised that such articles would even be notable, but that is another story, I suppose. NW (Talk) 14:09, 25 October 2010 (UTC)
I'm in full agreement with NW here. (Except that, for historical perspective, we should perhaps continue to mention the discredited study along with the research that discredited it.) Attribution is key, and it's something we do as a matter of course in many other areas. --JN466 14:20, 25 October 2010 (UTC)
I agree with User:Jayen466 - one of WP's most important functions is to debunk ideas that are now discredited, but still may found in obsolete books in the library, or in newspaper articles that were poorly researched. --Philcha (talk) 11:18, 26 October 2010 (UTC)
No that is clearly not WP's "function" because de-bunking involves original research and synthesis, which we try to keep out of. We can however report about de-bunkers if their work gets some notability and respect elsewhere first.--Andrew Lancaster (talk) 12:12, 26 October 2010 (UTC)
(ec) I don't know what you mean about such articles being notable. Most research into prescription drugs is paid for directly or indirectly by the pharmaceutical industry. If we prioritize scholary sources in those articles, and exclude the media, we are handing those articles over to the companies who make the drugs. Is it enough to rely on in-text attribution? "A 2008 study by Coca Cola revealed that it is the greatest drink in the world. A 2009 study, also by Coca Cola, suggested it was the most nourishing. A 2010 study, financed by a charity that is largely supported by Coca Cola, confirmed how marvelous it is."
We would not allow Coca Cola to take control of the articles about its products, with or without in-text attribution. We would not wait to see whether the conflict of interest had damaged accuracy. SlimVirgin talk|contribs 14:23, 25 October 2010 (UTC)
Examples like that are meaningless. For one, those kinds of claims would be prohibited per WP:REDFLAG. Secondly, what media source would be of use in cases like this? Actual examples would be a far better help. NW (Talk) 16:06, 25 October 2010 (UTC)
I am giving you examples. Almost all research into prescription drugs is financed by the manufacturers. If you exclude media sources from those articles, and only allow science sources, it will mean all or most of the sources will have been financed by the drug companies themselves.
We would not allow this in any other area, so I'm asking that you address it. How would you deal with it? SlimVirgin talk|contribs 16:35, 25 October 2010 (UTC)
It is a legitimate conundrum, but what is the alternative? Most pharmaceuticals are evaluated by the drug companies themselves in trials with controls that tend to give reliable results, but there are many documented cases of abuse that render those suspect. But what comes next? Homeopaths, naturopaths, and similar quacks that didn't even use a pretense of science to evaluate the drugs? It's a case of "lesser of two evils". Quackery must be resisted at all costs, and if it opens the door to some abuses, those at least tend to get corrected in time. Quackery remains quackery, and seems to be held tighter by adherents the harder you go at it.—Kww(talk) 16:42, 25 October 2010 (UTC)
The alternative is to do what this policy recommends, allow mainstream media sources as a balance if they have written about the drug in question. The proposals seek to ban that, or at least to throw obstacles in the way.
This is the whole point of NPOV, and the whole point of this policy: never to allow one individual or one group or one kind of source complete control over our flow of information. SlimVirgin talk|contribs 16:47, 25 October 2010 (UTC)
That's a strawman argument again though, isn't it. No one has argued that we should exclude media sources, and every recent proposal has clearly stated that quality mainstream media can be used for areas such as current affairs, including the socio-economic, political, and human impact of science. --JN466 16:58, 25 October 2010 (UTC)
They are already being excluded, Jayen, so that is the backdrop. Therefore any proposal must not make that situation worse, even inadvertently. Proposal 5 would make it worse (my bold): "In topics which are the subject of scholarly research, the most authoritative sources are academic works that have undergone scrutiny by a community of experts in that field. High-quality mainstream media are equally valuable sources for topics such as current affairs – including the public aspects of science – or biographies of living persons." It would be very easy for editors to argue that criticism of a drug is not a public aspect of science. SlimVirgin talk|contribs 17:15, 25 October 2010 (UTC)
I used your preferred wording: "current affairs, including the socio-economic, political, and human impact of science"; I am not sure it can be made much clearer, but I am willing to try. :) We can add "notable controversies", if you like. --JN466 17:17, 25 October 2010 (UTC)
My preferred wording is not part of Proposal 5. It was: "Non-academic sources, including high-quality mainstream media sources, may be used to report and interpret the socio-economic, political, and human impact of research in science, history, medicine, and other disciplines." But there is still a danger there in that media sources are being ring-fenced, and not in a way that clearly allows them to be used anywhere. And it still equates non-academic sources with media sources without defining them. We don't even define what we mean by academic source. Not defining these things is fine with the current policy, but where you start a policy of inclusion here, exclusion there, more precision is needed, something we've deliberately avoided because it will lead to fights. SlimVirgin talk|contribs 17:23, 25 October 2010 (UTC)
To be honest, I am not overly enamoured of the "most authoritative" wording for scholarly sources any more. I think what we got at in the discussion yesterday was more to the point: we should, as part of our work on an article, endeavour to present the current state of academic research if there is a significant body of such research. It's what QuackGuru has tried to incorporate into his version. It is far less exclusive: it just says, we have to endeavour to present the current status of that research. I think we could all agree that that is something we should be aiming for. That still leaves room to describe societal conflict surrounding that research. The sentence you got above in your most recent post is great, I think, or at least, its general thrust is. --JN466 17:36, 25 October 2010 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── I have no problem with that. But you'll find many of the editors who support these proposals want to exclude the media, not just make sure academic research is well-represented or prioritized. They want to hand over articles to those sources entirely. That's fine for some non-contentious issue in history, say. It's not fine for a contentious scientific issue in which private financial interests are at stake and financing the research. The mainstream media are the only watchdogs we have, for all their mistakes. Excluding them is what dictatorships always want to do, and the reason is always the same: they are irresponsible, not on-message, can't be trusted, make mistakes, they might say something inappropriate, they might mislead people, they're not under anyone's clear control, their poor stupid readers must be protected from them. These are all the arguments we're seeing on this page. SlimVirgin talk|contribs 17:45, 25 October 2010 (UTC)

It is probably preferable to discuss the proposals as they are presented rather than the perceived motivations of people who aren't here, and argue against something no one here has brought up.Yobol (talk) 18:03, 25 October 2010 (UTC)
Certain editors have been trying for years to exclude media sources from science articles, and are doing it de facto against policy, so to say no one has brought it up is somewhat bizarre. SlimVirgin talk|contribs 18:06, 25 October 2010 (UTC)
No one here has brought them up, as far as I can tell. These proposals explicitly lay out that media sources are allowed, so arguing against a phantom position really does nothing to further this discussion.Yobol (talk) 18:09, 25 October 2010 (UTC)
They are already allowed. The proposals seek to reduce where they are allowed. SlimVirgin talk|contribs 18:13, 25 October 2010 (UTC)
The proposals seek to use the most reliable sources. Dramatically comparing this to totalitarianism does not help the discussion. Yobol (talk) 18:20, 25 October 2010 (UTC)
I disagree. I think it's an excellent analogy. The point is that these proposals touch on a Foundation issue, the neutrality policy. You will argue that they don't, but as a matter of fact they do, and I have given an example above of the kind of situation they would leave us in in certain articles, forced to rely on sources who had a conflict of interest. So, because it's a fundamental issue, it can't be decided by a dozen accounts on one talk page, some of whom have very few edits. SlimVirgin talk|contribs 18:28, 25 October 2010 (UTC)
I guess we'll have to just agree to disagree then, though I would welcome community wide comments on this issue. BTW, was that last comment a reference to me? Yobol (talk) 18:47, 25 October 2010 (UTC)
(edit conflict) I am not sure that is so; my sense is more that editors want to have the state of scholarly research clear, and unfudged; i.e. without having what Jenny McCarthy (or Christopher Monckton, for that matter) said on a talk show set against a peer-reviewed systematic review, as though these two weighed equally heavy on the scales of knowledge ("the Academy of Sciences says X, but Mr M. says they're wrong"). I can certainly sympathise with that. My feeling is that editors are rightfully touchy about attempts to put pseudoscience on a level with science. This doesn't necessarily mean that non-scholarly viewpoints that have significant currency in the general population must remain unreported, nor that they are in every sense invalid. --JN466 18:33, 25 October 2010 (UTC)
I sympathize with that too, up to a point, but they take it too far. You have to remember that I've been watching this and the other sourcing pages for years, and I've seen numerous attempts to do this. I know you're doing it in good faith, but good faith butters no parsnips. What matters are the words, and the words suggested so far will encourage people to remove newspaper articles they dislike. I could come up with words that wouldn't do that, but several editors here would oppose them for that reason, so we're kind of stuck. SlimVirgin talk|contribs 18:39, 25 October 2010 (UTC)
I am not in a hurry to implement a proposal. We anticipated that this was going to take time to get right, and I think this debate is a useful step towards that. There is a number of things I like in proposal 8, and some I am not so sure of any more; I still hope there is a way we can all have our cake and eat it. --JN466 18:50, 25 October 2010 (UTC)
Slim, if you don't mind, why not post suggested wording for discussion, that could perhaps serve as some sort of a compromise?.--Literaturegeek | T@1k? 22:53, 25 October 2010 (UTC)
Yes, I'll try to do that, LG. SlimVirgin talk|contribs 20:25, 26 October 2010 (UTC)
About the "trying for years" comment: Perhaps these ongoing efforts mean that not only is it true that Consensus Can Change, but also that it has changed, and this policy is no longer an accurate description of the community's actual views. WhatamIdoing (talk) 17:15, 26 October 2010 (UTC)
I believe this is one of the strongest supported policies on Wikipedia. SlimVirgin talk|contribs 20:25, 26 October 2010 (UTC)
@SlimVirgin: is Wikipedia intended to be an an encyclopedia of scholarly knowledge, or a journal of current events?
Also, according to some websites, you have a professional (and respectable) background in journalism. I've long wondered how this might have influenced your longstanding position to deprecate the now-gutted WP:RS in favor of the newer WP:V, a position you are still arguing today. Is this not a case of, "where you stand is where you sit?" If you are a reporter covering Entomology, you are going to want newspaper coverage to count. (talk) 09:56, 26 October 2010 (UTC)
Obviously WP is not purely scholarly, nor is focused on reporting current events, and I think no one is arguing either of these extremes.--Andrew Lancaster (talk) 10:04, 26 October 2010 (UTC)
Actually, the double sized elephant in the living room is three gaping holes in wp:ver:
  • There is no source criteria for objectivity
  • There is no source criteria for expertise (with respect to the material that cited it)
  • There is no provision for increasing / decreasing the cite strength requirement based on the nature of the statement that needs the cite. For example, a stronger sourcing requirement for a controversial scientific topic statement than for a statement on Britney Spears' favorite color. North8000 (talk) 11:45, 26 October 2010 (UTC)
Expertise is addressed with respect to what we call 'self-published' sources (noting that 'self-published' is a term of the wiki art that depends on factors having nothing to do with the dictionary definition), so at least there's a little bit about that issue in the policy. WhatamIdoing (talk) 17:12, 26 October 2010 (UTC)

Threshold for inclusion

After a source has satisfied the criteria for being a reliable source, would the statement at the beginning of WP:V preclude any consideration of which source is more true?

"The threshold for inclusion in Wikipedia is verifiability, not truth—whether readers can check that material in Wikipedia has already been published by a reliable source, not whether editors think it is true."

--Bob K31416 (talk) 14:29, 25 October 2010 (UTC)

Yes, it would... if the source is considered reliable, then we can assume that it is accurately presenting a particular viewpoint on what the "Truth" is. Now... if there are differing opinions as to what the "Truth" is, we need to account for that as well... we should phrase what we write in the article as a statement of opinion... and we need to present other significant viewpoints as to what the "Truth" is... but we don't take sides and say "this is more true", or "this is true and that isn't". (see WP:NPOV) Blueboar (talk) 15:02, 25 October 2010 (UTC)
Short answer is yes, "truth" should not be how the case is discussed, because such discussion does not work on WP. Long answer is that this does not mean truth is not important to us somewhere in the background as people making an encyclopedia. In a sense Wikpedians aim at truth one-step removed, for practical reasons. If you think about it, "verifiable" comes from the Latin word for truth. What we are trying to identify, the verifiable, is truth defined in the limited way as what can be cross-checked and/or easily agreed upon by different people who do not know each other. I can not guarantee others agree with that formulation, but that is my observation of how people work on WP, and what the policies say. The concept of truth is contained indirectly within the concept of verification and is implied in descriptions on policy pages concerning article "quality". When people argue, for example on RSN, about the relative quality of sources they often give common sense explanations such as pointing to obvious mistakes or things which are wrong. I do not notice other editors saying that such common sense references to truth or error are totally irrelevant.--Andrew Lancaster (talk) 15:12, 25 October 2010 (UTC)
Good point... this is why I like to distinguish between the term "accuracy" and the term "truth". Even the most reliable source can contain a mistake or error when it comes to specific information ... for this reason, it is fine to question the accuracy of a reliable source in regards to a specific statement (as opposed to the "truth" of the statement). Blueboar (talk) 16:17, 25 October 2010 (UTC)
If there is a reliable source reporting on Jenny McCarthy's MMR–autism theory, we can say, by all means, here is this former glamour model who claims there is a link between MMR vaccine and autism, and went on Oprah Winfrey's show. But should we spend as many words on her assertions as we do on the statements of medical experts? Hell, no. --JN466 17:02, 25 October 2010 (UTC)
Here is a suitable source: a New York Times article. It describes the controversy well, and it's something that scholarly sources may not have bothered to address. It's an excellent source for covering an issue like this, which occupies a significant sliver of the public consciousness, in a disinterested manner. I see no problem in doing so, either under the present policy wording, or something along the lines of the proposed policy wordings. --JN466 17:13, 25 October 2010 (UTC)
@Blueboar, Interesting. Do you not think that the distinction between truth and accuracy which you make will work outside of Wikipedia or is this a case of using normal words in new ways, specially for Wikipedia?--Andrew Lancaster (talk) 18:16, 25 October 2010 (UTC)
Yes, I do think it works outside of Wikipedia. "True" is a subjective term... what I may believe is true may not match what someone else believes is true (take religious truths as the example... As a Christian, I believe the statement: "Jesus Christ is the Son of God" is true... A Jew or a Muslim would say the statement is not true. We have no way to demonstrate who is correct) "Accuracy", on the other hand, is objective... if a source misquotes someone, adds a column of numbers wrong, or makes some other factual error, we can demonstrate this inaccuracy. There is no subjective opinion involved. Wikipedia should strive for accuracy. Blueboar (talk) 21:36, 25 October 2010 (UTC)
Obviously religious belief is a special case because that is the one type of belief which people by definition don't come to by using rational argument and evidence alone, and therefore can also not easily come to agree upon. As I am sure you realize not everyone would even consider this the same thing as "truth" in any normal sense. The word "faith" is often used in such cases. In any case WP is clearly not going to start taking sides in those cases so I think we can put them aside in a discussion about editing WP. Normally I would think that the words truth, accuracy and objective are used like this:-
  • Accuracy: "Mr Smith was about 1 and a half metres tall" is less accurate than "Mr Smith was 1.65m tall". The statements are not different with respect to how true they are or how objective they are.
  • True: "Mr Smith was 1km and 65cm tall" is less true than "Mr Smith was taller than 1m and 65cm tall". The statements are not different with respect to how accurate they are or how objective they are.
  • Objective/Subjective: "Mr Smith was not a tall man" is written in a more subjective way than "Mr Smith was 1.65m". The statements are not different with respect to how accurate they are or how true they are.
Back to the subject, I think when people come to RSN and make a case against the merits of a journal based on the number of mistakes they found, what they are talking about is clearly "truth" even if they know to avoid using that word. It is also "original research" or "synthesis". In terms of policy it is perhaps borderline justified only by relatively vague policies like WP:UCS and postings like WP:CK. Nevertheless WP could not function without this tolerance and flexibility - or if you want to, we can call it "vagueness". Nearly every article and every edit relies on a bit of it. Sometimes having a fuzzy definition is going to create the least wikilawyering and encourage the best editing. Any vagueness which diverts talk pages away from wiki-lawyering and towards having to discuss what would be good for an article is a good thing in my opinion. "Burden of proof" is to begin with a lawyerish term, and not only that but a term which is specifically used in cases where laywers want to claim a right to avoid further discussion on a technicality. It is abused much more than WP:IAR in my opinion.--Andrew Lancaster (talk) 08:24, 26 October 2010 (UTC)
  • I think the elephant in the room here is the fact that much popular-media coverage of science and medicine is low-quality, misleading, and inaccurate. This is true even of generally reputable and high-quality news publications, and the situation goes downhill steeply once one gets to the second tier of popular media. Since this discussion seems long on extreme rhetoric about the impending totalitarian takeover of Wikipedia and short on actual concrete examples, I have one handy to illustrate my concern. Again, I have nothing against the judicious use of the popular press, but let's recognize the potential pitfalls, and maybe we could not accuse people who disagree with us of Hating FreedomTM? MastCell Talk 20:04, 25 October 2010 (UTC)
With all due respect I am not sure you are right about what the elephant in the room in, as there seems to be a bigger one some people are thinking even more about. To generalize, what I think others are saying is that when we deal in generalizations about types of source, all types of sources can be poor, especially when we look at one individual source at a time as opposed to a field as a whole, and that simple generalizations are sometimes dangerous.--Andrew Lancaster (talk) 20:11, 25 October 2010 (UTC)
Yes... that is the Woolly Mammoth in the room... because not all academic sources are of equal quality and reputation, and because not all media sources are of equal quality and reputation... it is a mistake to try to enshrine a comparison between the two types of sources into policy. You have to look at the specifics. And the simple fact is... Some media sources (those on the top end of the quality scale) are actually more reliable than some academic sources (those at the bottom end of the quality scale). Blueboar (talk) 21:43, 25 October 2010 (UTC)
Hence I'm wondering whether it is best to simply say that we should try to reflect, or summarise, the status of current scholarly research (if there is a significant body of such research in the topic area), using the best sources available. That way we might avoid making "totalitarian" statements like "scholarly sources are more reliable than press sources", which are bound to be wrong at least some of the time. --JN466 22:27, 25 October 2010 (UTC)
The question seems to be whether emphasizing the "scholarly" is always going to be appropriate. A field might be scholarly and it might not. Some fields are studied in several ways at once, both scholarly and non-scholarly. I think someone referred to investigative journalism sometimes breaking news about scientists for example. Of course in specific cases where a newspapers is giving a re-hash of an academic article, the academic article is generally going to be the better source, but can we generalize from this to all possible sourcing questions on WP?--Andrew Lancaster (talk) 08:28, 26 October 2010 (UTC)
  • Nice example, MastCell. One excellent reason why we should cite underlying studies rather than the press write-ups. --JN466 22:31, 25 October 2010 (UTC)

Actually, the double sized elephant in the living room is three gaping holes in wp:ver:

  • There is no source criteria for objectivity (with respect to the topic that cited it)
  • There is no source criteria for expertise (with respect to the material that cited it)
  • There is no provision for increasing / decreasing the cite strength requirement based on the nature of the statement that needs the cite. For example, a stronger sourcing requirement for a controversial scientific topic statement than for a statement on Britney Spears' favorite color. North8000 (talk) 11:45, 26 October 2010 (UTC)
We have always deliberately kept the policy away from details like that, because the more restrictions you add, the harder everything becomes for new editors, and the more loopholes we open up. With every word you add to policy, you risk adding several misinterpretations, or accurate interpretations.
The policy can't be algorithmic. We have to leave room for the editorial judgment that will change from context to context. SlimVirgin talk|contribs 12:16, 26 October 2010 (UTC)
I'm definitely 100% with you on the spirit of what you are saying. But keep in mind that my third point works both ways, especially if the policy is rewritten well to lay the framework for consenus-based interpretation vs. providing fodder for wiki-lawyering. For example, a statement by Britney Spears's sister would be a sufficiently good source regarding what Britney's favorite color is; vs. right now she would be amongst the ~90% of cites in WP that violate a (wiki-lawyer) thorough granular reading of wp:ver. North8000 (talk) 13:45, 26 October 2010 (UTC)
I agree with you. It's just finding the wording that's diffcult—wording that will cover most eventualities, and that won't lend itself to abuse. SlimVirgin talk|contribs 13:58, 26 October 2010 (UTC)
I think that it is doable. It's sort of my one year plan to promote and propose something. The framework is quite simple. To the current two source metrics (primary/secondary/tertiary and current RS standards) add two more: objectivity and expertise. All in the context of the statement that cited it. Define the strength of a cite/source as cumulative of all 4 of these metrics. The more questioned or controversial a passage is, the stronger the sourcinng required, and vica versa. And enforcement determination is be consensus application of this framework. North8000 (talk) 15:17, 26 October 2010 (UTC)
It sounds very worthwhile, but more like a 10-year plan to me. :) The objectivity part's particularly difficult, because the people who know most about issues are often deeply involved in them, at least in the sense of having strong views, and sometimes in the sense of having their research financed by interested parties. Requiring objectivity (i.e. disinterest) would cut out a lot of our sources. SlimVirgin talk|contribs 15:22, 26 October 2010 (UTC)
Yes, but keep in mind that the objectivity is just about the info in the cited statement. For example, in a "Republicans" article, for the statement "a common complaint about Republicans is xxxxxxxxx" the Democratic party might be considered an objective source of information for THAT info/statement even though they are not objective or disinterested on the topic of the article. North8000 (talk) 15:48, 26 October 2010 (UTC)
I see your point. I just worry that it would get very complicated trying to generalize. It touches on lots of philosophical issues about what knowledge is and how we recognize it, what objectivity is, whether it matters. SlimVirgin talk|contribs 16:04, 26 October 2010 (UTC)
Well, maybe it's a 1-2 year plan.  :-) And, IMHO, would solve a lot of problems. I guess that our WP opinions are formed by our WP experiences. I know that yours have included people being overly exclusionary about sources. Mine have been from trying to be a peacemaker is some "eternal warfare" articles and seeing how the policies lend themselves to misuse enabling that, and also having run into people like the other half of my split personality TheParasite [[5]]  :-) 17:41, 26 October 2010 (UTC)

Getting back to the original topic of this section, with the discussion following it in mind, here are a few points. Although there is the phrase "not whether editors think it is true", in practice a consensus of editors won't include material from reliable sources that they feel is not true. Perhaps MastCell's example is such an instance.

My personal feeling is that policy should give editors the guidance for determining what a reliable source is, but should not rank or stereotype groups of reliable sources such as news media, academic, etc. However, it may be useful to discuss elsewhere in an essay the various virtues and problems of different types of reliable sources, so that editors can use those ideas in selecting material from reliable sources and in their discussions on article talk pages. --Bob K31416 (talk) 15:00, 26 October 2010 (UTC)

Hmm... my intuition is leading me to think there is a problem of a different stripe entirely being discussed here (perhaps the stegosaurus standing behind the mastodon), namely: what is an encyclopedia supposed to do? I can think of a few different answers to that, which I've seen different people advocate at different points:
  1. an encyclopedia is supposed to present true information to the best of its ability. This leads editors to recommend that scholarly and scientific research be given a decided primacy, and that the encyclopedia take great pains to disabuse people of incorrect notions. Think of this as the pedological perspective, in which the goal is to impart correct knowledge.
  2. an encyclopedia is supposed to present all information without censorship. This leads editors to argue that all topics of knowledge (even false knowledge) have informational value and should be presented with their proper due. This is an informational perspective, in which the goal is to present the sum of verifiable information for people to digest on their own.
  3. an encyclopedia is supposed to present conventional information without bias. This leads editors to argue against including anything too new, too radical, or too 'opiniony' (narrowly conceived), and stick to things that are essentially uncontested 'facts'. Think of that as the phlegmatic perspective, in which the goal is to collect and catalog non-controversial truisms.
  4. (a perspective that's common, but not generally approved of) an encyclopedia is supposed to present interesting material. This leads editors to include material more on the basis of interest and salience than relevance or importance. This is a sort of divertissement perspective, in which the goal is to collect and present information that people are interested in, without any real regard to its significance in the world.
these are nowhere near mutually exclusive, obviously, and there are probably others I haven't thought of, but you can see that there are places where these conflict mightily in terms of what would be considered a reliable and verifiable source. this split is already present in the project in non-problematic ways - children's topics (pokemon pages, etc) are almost all written under #4, political pages are strongly defended as #2, math pages tend to be #3, hard science pages lean drastically to #1. it's where they come into conflict that we get all of our problems.
I'm not sure what to think about this yet, so... but there it is. --Ludwigs2 16:22, 26 October 2010 (UTC)
I sympathize, but part of the problem is that those who try to totally exclude all mainstream media sources often have some axe or agenda of their own. someone who is doggedly trying to add material which is sourced in mainstream newspapers and periodicals is someone who should be worked with; not someone who should be shown to the door. (I realize that many of the views expressed above on both sides of this issue would not necessarily disagree with my comment here.) --Steve, Sm8900 (talk) 20:03, 26 October 2010 (UTC)
Ludwigs2 has presented a very salient analysis here. In contentious topics, like pseudoscience, alternative medicine and so forth, it is tempting to have articles with distinct sections:
  1. "Here is the scholarly view." This is what science says about it.
  2. "Here is the public discourse taking place in reliable sources." This is more like the sociological approach – prominent views are simply described, in a disinterested and observational way.--JN466 02:40, 27 October 2010 (UTC)
Jayen I agree with that approach also. It would good if it happened more often. But I have seen people argue that some Wikipedia articles should belong to a discipline, and exclude things said about those subjects by people outside of it. It is something like an "OWN" problem, but not one based on individuals. An example that comes to mind is that linguists insisted that discussion of the Afroasiatic Urheimat should be excluded from the Proto Afroasiatic article because not a subject written about by linguists, even though this left the latter as a stub with one table, one introductory sentence, and 3 references.--Andrew Lancaster (talk) 12:48, 27 October 2010 (UTC)
I tried to weigh in on and keep up with this discussion early on, but other than it looks like it's going nowhere, I've lost the plot. No, we don't report (at least in medical articles except in specific circumstances) underlying studies and the laypress. The laypress almost always gets it wrong (or perhaps I should say I'm not aware of any case where they get it right, and that includes the BBC and the NYT), and underlying studies are primary sources-- we use secondary reviews, per WP:MEDRS, except in exceptional cases. On the other hand, we need to make allowances for high quality mainstream media in non-scientific topics, particularly politics, to the extent of due weight. I can't see this discussion going anywhere yet, so I hope if/when we get a semi-stable proposal, some kind soul will ping me. SandyGeorgia (Talk) 20:24, 26 October 2010 (UTC)

Per my last message above, I would remove ranking of reliable sources that occurs in the 3rd paragraph of the section What counts as a reliable source. This would involve rewriting the first sentence to remove the phrase "are usually the most reliable sources", and removing the second sentence. Specifically, I suggest replacing the first 2 sentences of the 3rd paragraph of What counts as a reliable source with the following sentence.
"Examples of reliable sources are academic and peer-reviewed publications."
so that the whole 3rd paragraph would become,
"Examples of reliable sources are academic and peer-reviewed publications. Material from reliable non-academic sources may also be used, particularly if it appears in respected mainstream publications. Other reliable sources include university-level textbooks, books published by respected publishing houses, magazines, journals, and mainstream newspapers. Electronic media may also be used, subject to the same criteria."
--Bob K31416 (talk) 21:24, 26 October 2010 (UTC)
If we do touch this paragraph, please let's get rid of the phrase "academic and peer-reviewed publications". Peer-reviewed publications are one type of academic publications. Suggest "Academic publications, such as peer-reviewed journals and books published by well-regarded academic presses ..."
I agree with the notion that we shouldn't make categorical statements about one type of source being more – or less – reliable than another, simply because of the category it belongs to. It's not always true, and it just sets up formalistic arguments.
What we could say is that Where a topic is subject to a significant amount of scholarly research, the article should aim to summarise the current status of that research. That's something readers would expect of an encyclopedia, and it can be stated without implying that one type of source is always better than another.
Another legitimate concern is that we should summarise, in a disinterested way, (non-scholarly) public discourse around a topic; press sources may be useful for that. That relates to approach 2 in Ludwigs2's classification above; I think it is fair to say that en:WP as a whole leans more towards this approach than towards 1. --JN466 02:19, 27 October 2010 (UTC)
Exactly right, Jayen. Blueboar (talk) 13:19, 27 October 2010 (UTC)
I've rewritten the third para in WP:V#What counts as a reliable source per the above. Please review -- does it read okay? --JN466 13:56, 27 October 2010 (UTC)
I like it... it nicely expresses the balance we try to achieve, and reinforces WP:NPOV. Blueboar (talk) 14:52, 27 October 2010 (UTC)
Make that two thumbs up; very nice. --Ludwigs2 15:52, 27 October 2010 (UTC)
It is now worse than the previous version. QuackGuru (talk) 16:14, 27 October 2010 (UTC)
How so? Blueboar (talk) 16:18, 27 October 2010 (UTC)
Maybe you can tell us how so after you read these comments that were on your talk page. I tried to explain to you there is a different with the reliability of the sources. Proposal 8 covers how to better word this. QuackGuru (talk) 18:51, 27 October 2010 (UTC)
Would anyone object to saying "peer-reviewed articles and books" rather than "journals and books"? I'm not aware of any journal that sends absolutely every piece through peer review. Editorials, letters, obituaries, etc. are not normally peer-reviewed pieces, even when they appear in a 'peer-reviewed' journal. WhatamIdoing (talk) 16:24, 27 October 2010 (UTC)
The word "article" occurs many times in that section, and always refers to the Wikipedia article; that is why I avoided it. I've made it "peer-reviewed publications and books by academic presses"; this is closer to the previous wording. Okay? --JN466 17:55, 27 October 2010 (UTC)

IMO, this discussion has become a bit chaotic and has resulted in a complete change in the 3rd paragraph so that it now has the 2nd and 3rd sentences that are digressions from the topic of the section "What counts as a reliable source".[6] When I made my suggestion for the 1st two sentences of the third paragraph, it was with the purpose of removing the ranking of reliable sources and I tried to minimize the effect on the rest of the paragraph. The resulting suggested paragraph would have satisfied the purpose of removing ranking and wouldn't introduce any digressions, and it would be at least as readable as what was there before. The present new paragraph does neither.

I would suggest making the more minimal change that I suggested above and then discuss here what more you would like to do, instead of the changes that have recently been made to the 3rd paragraph. --Bob K31416 (talk) 21:10, 27 October 2010 (UTC)

Hello!!! Everyone, there is no consensus for changing this. There is none above, there is none here. Propose significant changes to policy here, comparing the old and new, and then get significant consensus for it. Don't continually re-write policy on the fly. Jayen466, I'm talking to you in particular. This is what got the article protected before, two or three editors trying to ram through significant policy changes without consensus. Don't start it again. Jayjg (talk) 04:32, 28 October 2010 (UTC)

Although I agree with reverting to the older version, the reason for the page protection was edit warring, in which you were one of the participants with respect to promoting new changes that you supported. I recall that you have been desysopped[7] for edit warring and conduct unbecoming an administrator[8] in the past and I suggest that you not forget that lesson. --Bob K31416 (talk) 09:03, 28 October 2010 (UTC)
Two edits in a week is not "edit-warring", I didn't "promote new changes", and I've never been desysopped. Now, I suggest that you not forget that WP:NPA is policy, and stop focusing on other editors: no more comments about SlimVirgin, me, or any other editor. Comment on content, not on the contributor. Jayjg (talk) 12:27, 28 October 2010 (UTC)

Confusing "Self-published or questionable sources as sources on themselves" subsection

The subsection "Self-published or questionable sources as sources on themselves" is confusing and even actually ambiguous, because it utilises the double meaning of the word source (which can mean either a written text or other media, or the author of such texts or media).

It even actually appears to switch between the two meanings mid-sentence! In the first sentence, "Self-published and questionable sources may be used as sources of information about themselves, usually in articles about themselves or their activities", "source" would sensibly be interpreted as meaning publication initially (since we don't want to accept as sources unverifiable telephone conversations with people who happen to have been self-published authors at some point in their lives!) but then "articles about themselves or their activities" makes no sense if we continue to interpret the word "source" as "publication"! A publication does not have "activities" - it is a passive object - avant-garde experiments with robotics etc. aside!

Even the title of the subsection itself is confusing, for the same reason.

I think the meaning we want to convey - and here I am using the "text or media" definition of source throughout - is:

  • Self-published sources must be actually published, i.e. we cannot rely on e.g. telephone conversations with a (human) source.
  • Self-published sources can be used as sources about themselves, subject to the cautions listed in the policy as it is written now (those that are relevant, anyway).
  • Self-published sources can be used as sources about their author (individuals if the author is an individual; organisations or companies if the publication is clearly an official publication of an organisation), subject to the cautions listed in the policy as it is written now.

However, this assumes that a self-published source is only authored by one person - of course, this is not always the case. As the policy stands now, it seems to imply that we can use self-published sources as sources on any of their authors - is this really what we want to say?--greenrd (talk) 05:28, 28 October 2010 (UTC)

I have often thought the same thing, but have never been able to think of an easy way to fix it, and in actual practice it has rarely caused problems. But WP:BLPSPS covers some of the same ground, and seems to manage to do so without tying itself into a logical knot. Perhaps it can be done here too. --JN466 08:33, 28 October 2010 (UTC)

Google Books page links

WT:CITE#Linking to Google Books pages. RfC on whether WP:CITE should say Google Books page links are not required but are allowed in footnotes, and that editors should not go around removing them. All input welcome. SlimVirgin talk|contribs 20:37, 29 October 2010 (UTC)

Clarification needed on what a challenge is

I see some instances where this part of the verifiability policy is being invoked:

This policy requires that anything challenged or likely to be challenged, including all quotations, be attributed to a reliable source in the form of an inline citation, and that the source directly supports the material in question.

The problem I see is what constitutes a proper challenge is fuzzy. Theoretically every statement without an inline citation made in any article can be challenged on a whim. A vandal for example, instead of page or section blanking, could claim to "challenge" every statement without an inline citation and use this policy as cover for their erasure. Clearer guidance is called for. Lambanog (talk) 08:39, 20 October 2010 (UTC)

Yes, any of those could be construed as a challenge.
Editors are permitted to WP:IGNORE bad-faith challenges, and to use their best judgment in deciding what falls into that category. (However, it's worth remembering that sometimes it's actually faster to spam citations into an article than to discuss whether the challenge is valid.) WhatamIdoing (talk) 22:52, 20 October 2010 (UTC)
Lambanog, challenges are supposed to be reasonable, and usually are. It's implied of all our policies that they be applied reasonably, but it's not really something we can legislate for. SlimVirgin talk|contribs 09:35, 22 October 2010 (UTC)
When you analyze it, this would-be-very-important clause is worded so as to make it mean the exact opposite of what it was intended to say/ appears to say. In essence "must be cited if challenged" , but one can challenge something just for being uncited. A small change here would do a huge amount of good for Wikipedia. North8000 (talk) 10:34, 22 October 2010 (UTC)
The idea is that anything that is actually challenged must be sourced, and the unwritten assumption is that challenges will be reasonable. As an editor you're expected to anticipate such challenges, hence the "likely to be challenged" wording. What change to the wording would you suggest, N? SlimVirgin talk|contribs 10:54, 22 October 2010 (UTC)
Even allowing an unreasonable challenge to trigger a requirement for a citation would be a step better than what this structurally currently says, which is that NO challenge of the statement itself is required. So, I'd suggest adding something like:
A "challenge" consists of any challenge of the statement itself. Since requirement for a citation is the result of the challenge, it should not be given as the sole basis of a challenge. North8000 (talk) 11:17, 22 October 2010 (UTC)
We've long accepted that editors don't have to believe something is false. They may just want to know who has said something other than the Wikipedian. I'd be worried that adding language saying the content has to be challenged in some way (as false presumably) would weaken this idea. The point is that anything can be challenged for any reason, though hopefully not in a silly way. SlimVirgin talk|contribs 11:24, 22 October 2010 (UTC)
I believe that this is an important topic and that there is some merit to what I said. And of course, much merit to what you said. Unfortunately I'll be out of commission for 2 days and won't be able to discuss this further in a timely manner. Sincerely, North8000 (talk) 11:49, 22 October 2010 (UTC)
Not that I would want to argue for removal of that the statement which in spirit is amongst the best in all of wp:ver/wp:nor, but the status quo which you described makes that entire sentence meaningless except in spirit. North8000 (talk) 12:19, 22 October 2010 (UTC)
I confess I'm struggling to understand the exact point you're making here, but since your suggested wording emphasizes "the statement itself", perhaps the current word anything is the problem. I suggest we replace most of the sentence with the wording that appears in the main text, resulting in "This policy requires that all quotations and any material challenged or likely to be challenged must be attributed to a reliable, published source using an inline citation, and that the source directly supports the material in question." That additionally fixes the incorrect qualification of "quotations" currently implied in the lead. Would this help to address your concern? PL290 (talk) 12:43, 22 October 2010 (UTC)
Thanks for asking. Again, I'll be out of commission for 2 days and won't be able to discuss this further in a timely manner. But here's a example....the simpler more extreme nature of the hypothetical examples lets me get to the point quicker than with the pervasive more complex real world examples. Lets say I put an uncited statement in the Mexico City article that said: "during the 1990's, daylight occurred every day in Mexico City." Another editor, ThePrinceOfDarkness370 wants to Wiki-Lawyer anything about daylight out of the Mexico City article.
Right now he can just say "it's uncited - provide a citation or I'll delete it". I suggest a policy change that forces him to say: "I don't agree with the statement, and it's uncited. Either provide a citation or I'll delete it" While it sounds minor, if you look at the dynamics of the "eternal warfare" articles, I think that this would help the situation. While there is nothing to prevent PrinceOfDarkness370 from insincerely saying "I don't agree with the statement", in real life, to a great extent, this would reduce challenges to the sincere ones vs. the ones that are just warfare via. wiki-lawyering.
Again, thanks for asking. Unfortunately, other than maybe a sentence or two in the next 3 hours, I have to sign off for about 48 hours. Sincerely, North8000 (talk) 14:50, 22 October 2010 (UTC)
When you get back, can we please get a realistic example... I seriously doubt anyone would ever add the statement: "during the 1990's, daylight occurred every day in Mexico City" to an article. Without a realistic example, I really don't understand the complaint. Blueboar (talk) 15:23, 22 October 2010 (UTC)
Real world examples always have complexities which can allow derailment of the discussion. But maybe we can do both. In the Dorothy Molter article, In the first sentence in the "Evolution of the BWCAW" section I wrote "Dorothy's life, and her place in the public eye was significantly influenced by the evolution of the area where she lived into the Boundary Waters Canoe Area Wilderness". This statement is gorilla-in-the-living-room / pervasive in the hundreds of pages of material on her which I just read. (my decades of direct observation aside). None of the sources made this summary/statement directly. It is also one of the most informative statements in the article regarding gaining an understanding on the topic. Nobody is going to come along and say "I think that the statement is wrong", but it is quite possible that one of the misfits roaming Wikipedia (who may be unhappy with me for some other reason) would delete it for being uncited/uncitable, without even questioning the statement, and they would have the Wikipedia rulebook on their side. I suggest that the policy should require them to question the statement in order to delete it. I could get even more real than this, (where the such deletions actually occurred, with no challenge of the statement) but that would involve talking about the conduct of particular individuals which I don't want to do. North8000 (talk) 17:53, 22 October 2010 (UTC)
To be honest, the example you've brought is a very problematic sentence. It's quite vague, for one thing; what real information does "significantly influenced" impart? How was her life "significantly influenced"? What does "place in the public eye" mean? Did the "evolution of the area" affect both in the same way? The whole sentence seems unclear, and, frankly, doesn't seem to impart any specific information. Perhaps the reason it can't be sourced is that it doesn't actually add anything to our knowledge, and therefore no-one saw fit to make the statement. Jayjg (talk) 21:51, 22 October 2010 (UTC).
This is a useful generalization. The specifics that you say it lacks are the contents of 3/4 of the article, although it's a "can't see the forest for the trees" situation unless there is some overview type material like this. However, this is off of my main point. North8000 (talk) 20:01, 24 October 2010 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── One of the things I've thought of starting is a page where editors can briefly list examples with diffs of being asked to supply sources under this policy for material that's common knowledge, or just common sense to anyone familiar with the issue. It would be good to see some real examples of the problem and how pervasive it is. SlimVirgin talk|contribs 04:37, 23 October 2010 (UTC)

A "realistic" example has been requested. Here's a real example from 2007.
I think that it would be reasonable for editors to assume that this section applies only to challenges that are issued because the editor is concerned that the material violates a content policy (e.g., it is not verifiable). This means: If it's "cute" to fact-tag the number of digits on the human hand, or if your fact-tag is because your teacher won't let you cite Wikipedia, then nobody really has a BURDEN to supply the citation/do your homework for you. WhatamIdoing (talk) 18:06, 23 October 2010 (UTC)
Hmm. Well, there are sources for that claim: "pentadactyl Having five digits. The hands and feet of humans are pentadactyl, as are those of insectivores." George A. Feldhamer, Lee C. Drickamer, Stephen H. Vessey, Mammalogy: Adaptation, Diversity, Ecology, Johns Hopkins University Press, 2007, p. 559. Jayjg (talk) 02:54, 24 October 2010 (UTC)
Generally speaking, I find that's the case. If a flat-earther challenges that the Earth is spheroid, I can find dozens of sources to contradict that in very short order. Same if they challenge that it's the third planet from the Sun, or what have you. If material really is "obvious", finding a source for it ought to be trivial. And if it's so obvious that no one else has even bothered to say it (the "daylight in Mexico City" being a good example of this), we probably ought to question what we really should. As to the Dorothy Molter example, if none of the sources have said it, but you're asserting they imply it, you're engaged in original research and/or synthesis, neither of which are allowed. If it really is so obvious and so important, someone will have bothered to say so. Seraphimblade Talk to me 20:07, 24 October 2010 (UTC)
"If it really is so obvious and so important, someone will have bothered to say so." Are you sure? Maybe it is so obvious no one did. Or maybe someone did say so but it is so obvious everyone else didn't and the one who did say so is lost in the pile. For example would the statement "the black crust resulting from the char grilling of hamburgers is liable to contain carcinogens" be very controversial? Yet it is not as easy to find such a direct statement as I would have expected considering we are talking about hamburgers and a health issue. Imagine if we were writing about some exotic foreign dish that had the same properties but wasn't as well written about in the Western press and thus such details were not available for Google searches. Does that mean it is not true? Would a challenge based on that be constructive? Lambanog (talk) 07:32, 25 October 2010 (UTC)
The problem here is that people often oppose having to source statements by claiming that some statements are "so obvious that they're hard to actually source". However, when examples are brought, they all turn out to be either a) actually quite sourceable, or b) problematic. Yes, of course, "the black crust resulting from the char grilling of hamburgers is liable to contain carcinogens" needs to be sourced. In the immortal words of Joe Jackson, "everything gives you cancer". If it's notable that grilling specific products produces certain suspected carcinogens, then reliable sources will have said so, and no doubt in a nuanced way that will need to be reproduced, in accordance with WP:V and WP:NPOV. If reliable secondary sources have said nothing about carcinogens and grilling other food products, then Wikipedia should not do so, in accordance with WP:V and WP:NOR. Jayjg (talk) 05:24, 28 October 2010 (UTC)
Jayjg, that's great in theory but can get ridiculous in practice. You have written some really nice articles and have cited them heavily, but you offer little theoretical defence for yourself if I was to suggest your articles still lack citations because you haven't cited every fact and have a tendency to use one citation for sentences that contain multiple facts. Lambanog (talk) 13:05, 29 October 2010 (UTC)
Actually, it works perfectly well in practice, and the citations I use back up every single one of the "multiple facts" contained in the sentences they cite. Jayjg (talk) 01:00, 31 October 2010 (UTC)
Random persnicketiness applied to your FA article on Rudolf Vrba might find the source given (footnote 8) supporting the statement that "Canada" refers to an Aufräumungskommando as imprecise. That source says it refers to a store room. Should an editor take that as license to find more perceived inconsistencies to drive you up the wall? Lambanog (talk) 16:58, 31 October 2010 (UTC)
Well, I helped write that FA, I didn't write all of it, and the "Canada" bit in particular is one of the statements in that article that has been edited by a number of editors since, each with their own views on what exactly "Canada" was and meant. One cannot help it if a bit of wiki-rot sets in after several years. However, the point remains, it works perfectly well in practice, and the citations I use back up every single one of the "multiple facts" contained in the sentences they cite. Jayjg (talk) 19:09, 31 October 2010 (UTC)
Of course we can produce sources to support the statement in this example. Practically every anatomy textbook, including some ancient ones, contains information about the number of fingers on the human hand. But it's equally true that anyone smart enough to read the sentence or to add the tag already knows how many fingers are on a typical human hand. This is information that the vast majority of preschool-age children possess. There really is no conceivable justification for fact-tagging that sentence, and while an editor could produce a source, I think this editor was justified in removing the tag. WhatamIdoing (talk) 23:39, 24 October 2010 (UTC)

Wrong reason for citing quotations

The policy incorrectly states that the reason for citing the source of all quotations is to show there is no original research. The actual reason is to avoid plagiarism. If demonstrating no original research were the reason, there would be no need to cite quotations that were unlikely to be challenged, or that were supported by a citation to a source that supports the concepts of the quote without giving the actual words of the quote.

Incorrect example of demonstrating no original research: A shift register is a multibit memory with a capability of exchanging a single memory element's contents with its neighbors. <ref>Taub, H. & Schilling, D., Digital Integrated Electronics (New York: McGraw-Hill, 1977), p. 322</ref>

Correct example that both demonstrates NOR and avoids plagiarism: "[A] shift register [is] a multibit memory with a capability of exchanging a single memory element's contents with its neighbors." <ref>Katz, R. H., Contemporary Logic Design (Redwood City, CA: Benjamin/Cummings Publishing, 1994), p. 300 </ref>

By the way, the page from Taub and Schilling does indeed support the concept of the sentence, but uses different wording. Jc3s5h (talk) 15:41, 31 October 2010 (UTC)

The reason we cite sources is to demonstrate that some source actually said the information that wikipedia is presenting. That's NOR. plagiarism is a separate issue (Wikipedia doesn't want to steal text from people and present it as its own) which is handled by proper quoting and summarizing. Cites only help prevent plagiarism to the extent that a cite acknowledges that the text comes from a source, not from us. --Ludwigs2 17:45, 31 October 2010 (UTC)
"The reason we cite sources is to demonstrate that some source actually said the information that wikipedia is presenting" is wrong. The real reasons we cite sources are to demonstrate that soume source actually said the information, and to allow readers to obtain more information by reading the source, and to avoid plagiarism. If this policy were strictly about NOR, there would be no reason to even mention quotations. Jc3s5h (talk) 19:02, 31 October 2010 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── What it says is that all material must be attributable to a reliable source to show that it's not OR, i.e. a source must exist for it, whether cited or not. The OR issue doesn't carry over into the next sentence about attribution. SlimVirgin talk|contribs 21:05, 31 October 2010 (UTC)

The policy states

All material in Wikipedia articles must be attributable to a reliable published source to show that it is not original research, but in practice not everything need actually be attributed. This policy requires that anything challenged or likely to be challenged, including all quotations, be attributed to a reliable source in the form of an inline citation, and that the source directly supports the material in question.

I read the first sentence as saying all material must be attributable to show NOR, and the second sentence as explaining when material must actually be attributed. The second sentence expands on the first by explaining when material must actually be attributed. I read "including all quotations" as a (false) statement that all quotations either have already been challenged, or are likely to be challenged in the future, and that is why they must be cited. Jc3s5h (talk) 22:08, 31 October 2010 (UTC)
I've rewritten that sentence a few times, but for some reason it gets changed back, so I'll fix it again. SlimVirgin talk|contribs 17:11, 1 November 2010 (UTC)
Thanks, that's better. Jc3s5h (talk) 17:34, 1 November 2010 (UTC)

What counts as a reliable source

Following long discussions above, the idea arose to rephrase the third paragraph of WP:V#What counts as a reliable source as follows:

Current policy Proposal
Where available, academic and peer-reviewed publications are usually the most reliable sources, such as in history, medicine, and science. But they are not the only reliable sources in such areas. Material from reliable non-academic sources may also be used, particularly if it appears in respected mainstream publications. Other reliable sources include university-level textbooks, books published by respected publishing houses, magazines, journals, and mainstream newspapers. Electronic media may also be used, subject to the same criteria. Many topic areas are covered by a significant body of scholarly literature, such as peer-reviewed publications, and books by academic presses. Articles in these topic areas should aim to summarise the current status of knowledge and research reflected in that literature. Articles should present all majority- and significant-minority published positions, and may complement any available scholarly sources by drawing on non-academic sources, particularly high-quality mainstream publications, to reflect public discourse around such topics. High-quality mainstream publications include books from respected publishing houses, mainstream newspapers, magazines, journals, and quality electronic media; these are also preferred sources in topic areas that are not subject to academic research.

Perceived failings of the present wording include:

  1. The phrase "academic and peer-reviewed publications" in the present wording is awkward. Peer-reviewed publications are one type of academic source.
  2. The present wording states that "academic and peer-reviewed publications are usually the most reliable sources". Some editors feel that any wording that describes one category of sources as "usually most reliable" tends to encourage wikilawyering, especially with the objective to exclude media sources. Furthermore it's been pointed out that the quality of academic sources is variable; a meticulously researched and fact-checked article in The New Yorker for example will generally be of far higher value than an obscure conference paper by an unknown academic.
  3. Peer-reviewed academic publications, described as "usually the most reliable", are very often primary sources advocating new ideas that have not yet been tested established in the discipline.
  4. The present wording fudges the issue by implying that media sources are just as good as scholarly sources for scholarly topics, leading to media sources being pressed into the service for jobs that should really be left to scholarly sources.

Perceived benefits of the proposed wording include:

  1. Fixes the "academic and peer-reviewed publications" phrase.
  2. Avoids the "most reliable" wording, but states clearly that articles on topics that have an associated academic literature should summarise the status of published scholarly research.
  3. Makes clear that high-quality mainstream publications are reliable sources for giving an overview of the public discourse that takes place in the media – including public discourse around topics that are subject to scientific research. Some editors feel that being able to describe this public discourse is a vital NPOV concern. At the same time, this wording avoids giving the impression that media sources are good sources for describing the status of the actual scientific research – which has also often caused disputes, because the media sometimes report scholarly findings incorrectly, or dress up their reports in a way designed to maximise readers' attention.

Please discuss! --JN466 09:08, 28 October 2010 (UTC)


  • Oppose - The new version conflates source reliability with neutral point of view. By including every policy in every paragraph, we make it impossible to edit the policy, because any change will require every paragraph in every policy to be edited for any change. Jc3s5h (talk) 10:56, 28 October 2010 (UTC)
    I'm not sure I buy this argument, or if I do, then I'd start to think that Verifiability needs to be rewritten broadly. Verifiability (in its mindlessly simple, literalistic sense) is a no-brainer - If something has been published somewhere and we can cite it, it's verifiable. period. The minute we start talking about issues like reliability we've started entering into NPOV territory, since reliability seems to mean "how much trust and weight we should put on information coming from a particular source?", and that is a neutrality issue, not a verification issue. That means that roughly 3/4 of wp:V is cross-polination from wp:NPOV. do you see what I'm saying? --Ludwigs2 02:18, 29 October 2010 (UTC)
    Actually, Ludwigs, I think the bigger problem is that WP:RS isn't a core policy. Verifiability says if its published and you can cite it, ok. NOR says don't make it up, either. NPOV says represent it all fairly in proportion to its significance. But RS says what's significant. I don't see how to close that loop without bringing RS up to the same level of policy. Ocaasi (talk) 02:30, 29 October 2010 (UTC)
    Yeah, well... reliability isn't core policy (IMO) because it's such an intrinsically suspicious concept. I swear, on grumpy days I think of wp:RS as wikipedia's Animal Farm moment ("all sources are equal, but some sources are more equal than others.."). The idea behind RS isn't bad, mind you - even though we're not supposed to evaluate 'ideas' there's a certain logic to evaluating the sources that present 'ideas', because that way we can keep ideas in their proper prominence without actually evaluating them ourselves. But somehow the logic shifted: what used to be "Determine good sources and let them establish prominence of ideas" became "Define what sources are good so that we can use them to establish prominence of ideas", and RS started getting used as a kind of article siege engine. very strange... --Ludwigs2 06:48, 29 October 2010 (UTC)
  • Oppose - I don't think this achieves its aim of being less prone to abuse or misunderstanding: It is longer and more complicated, and this can cause more problems than it solves. Also, the implied aim of trying to remove all hints of preferences between types of sources, is probably going to far. --Andrew Lancaster (talk) 12:26, 28 October 2010 (UTC)
  • Support - however... Jc3s5h makes a good point, but, given the discussions above, it has become clear that this policy needs a clearer statement about maintaining a NPOV. Our policies do not exist in limbo. Each affects the other. And because of this, we already have a situation where a change to one policy will require a change in others (and yes, this does make it difficult to change any of them... that is a good thing... it ensures that any major changes really have been thought through thoroughly and really do have community consensus).
That said... I would suggest we keep JN's paragraph, but follow it up with something along the lines of the current text. What I really like about JN's paragraph is that it clearly explains how the context of what is being discussed at a particular point in an article impacts the reliability of sources. It explains how different types of sources will be best for different types of statements... an academic source is best for a statement as to what current academic thinking is, while a media source will be best for a statement about media coverage of the topic, etc. Once we say that, then we can talk about the best source within that type (ie finding the best academic sources for statements about academic thinking, and the best media sources for statements about media coverage.) It clarifies that academic sources are not in competition with media sources... because they are best used in different contexts and in support of different types of statements. Blueboar (talk) 12:35, 28 October 2010 (UTC)
Blueboar's remarks seem very reasonable to me. Relevance of context is important. Just not sure the current draft is better than the old version.--Andrew Lancaster (talk) 12:44, 28 October 2010 (UTC)
I suppose what I am trying to say is that JN's proposed paragraph may not be better as a replacement for the old paragraph... but it improves the policy as a whole, so should be retained. Blueboar (talk) 13:27, 28 October 2010 (UTC)
If the new paragraph is to be added (without deleting the present text) I think it should be added under the "Reliable sources and neutrality" heading, rather than "What counts as a reliable source". The section under discussion is about evaluating individual sources; evaluating the constellation of sources for an article is outside the scope of this section.
Also, I am concerned that if NPOV considerations are sprinkled throughout this policy, then editors who deletete material from articles will be apt to refer to this policy, when they really should be referring to the NPOV policy. Another editor will probably revert the deletion as unjustified, leading to needlessly complicated edit histories and talk pages. Jc3s5h (talk) 13:54, 28 October 2010 (UTC)
The thing is... While JN's paragraph may be influenced by WP:NPOV, what it is addressing is a WP:V issue... it is talking about what types of sources are best and most reliable within a given context. That is a WP:Verifiability issue. When an article discusses the public discourse on a topic it is best to use public (ie media) sources. When an article discusses the academic discourse on a topic, then academic sources are best. Whether an article should discuss the public (or academic) discourse on the topic and how you discuss it may be NPOV issues... but once you discuss it, properly sourcing that discussion becomes a WP:V issue. Blueboar (talk) 14:14, 28 October 2010 (UTC)
(as a side note... I personally think this is another example of why we need a unified policy that merges WP:NPOV and WP:V (and WP:NOR)... these policies are so interwoven that to discuss changes to one almost always requires a discussion of the others. But, since we don't have a unified policy... on with the debate). Blueboar (talk) 14:27, 28 October 2010 (UTC)
I agree that inter-policy mentions here are both incidental and also a good thing. I don't think I like the idea of a unified policy. As a bit of an anti-federalist, we help ensure against mass corruption by keeping the launch codes separate from the key separate from the briefcase. I think if we put together a 'unified' policy, it should just be for convenience and just be a transclusion of the three actual policies. I see no reason why new editors shouldn't be able to access THE policy page, where everything they need to know to be reasonably successful is actually on a single page. Ocaasi (talk) 01:27, 29 October 2010 (UTC)
I am sure that this whole effort is absolutely sincere and well intentioned and that all the users involved here are highly experienced. But, if this is a fundamental change, should we get a word from J?-Civilizededucationtalk 03:36, 29 October 2010 (UTC)

JN466, what do you mean "Peer-reviewed publications are one type of academic source." I've had one editor argue at great length that "peer-reviewed" and "academic" were independent qualities. That is, that some (or at least one) reliable sources were "peer-reviewed" without being "academic". Jayjg (talk) 05:32, 29 October 2010 (UTC)

Agreed that the two terms are over-lapping, but one is not a sub-set of the other. What the current text says is that the overlap of the two terms is generally pretty reliable. Peer review is generally a fairly good type of fact checking and academic sources are often reliable. That does not seem wrong to me as a generalization. I think this is another point where trying to go beyond generalization and cover all possibilities with a more complex text may not help to avoid the problems being mentioned. --Andrew Lancaster (talk) 06:01, 29 October 2010 (UTC)
Jayjg, could you give me an example of a source that is peer-reviewed, but not a scholarly source? --JN466 13:13, 29 October 2010 (UTC)
One editor I know has argued at great length that is exactly that. Jayjg (talk) 01:02, 31 October 2010 (UTC)
It's listed on university library pages [9][10][11][12][13][14]. It's cited in academic books [15][16] It's a new field of research, but I would have no hesitation in including it under the umbrella of scholarly sources. --JN466 16:33, 31 October 2010 (UTC)
Jayen466, being listed is not the same as being reliable, and being mentioned is not the same as being cited. Websites run by non-experts who work professionally in other fields don't really qualify as "academic". In any event, the point was that the individual in question insisted that the website was not academic, but was peer-reviewed. Jayjg (talk) 18:09, 31 October 2010 (UTC)
  • Jayjg is referring to something like a discussion he recently had with me.[17] For the record I never said that source was not "scholarly". I said it was not "academic" in the sense that there is no University programme for that field of scholarship (genetic genealogy). More to the point, I said that this fact is separable from being a reliable source using the same types of relevant facts you have cited, i.e. by showing what types of sources cite it as reliable, which includes strong academic ones. BTW there is also a positive discussion of it as a source for some types of information in a major genetics journal.[18]
  • Background. If I understand Jayjg correctly, which I may not, he wanted to remove all mention of the source because of the way it was once used on the Khazars article, as a secondary source about a theory concerning Ashkenazi ancestry being partly Eastern European. (Primary sources now source the same theory on Khazars.) When the subject then came up at RS/N he jumped in and tried to bulldoze people out of it calling himself un-involved. He latter claimed on his talk page that there had been a consensus of uninvolved editors taking his position, which was certainly not true in many ways. Summary of facts: [19]
  • Anyway, here is an interesting point which does connect to policy wordings, and to your response: at one point in recent discussions Jayjg against the relevance of citations in good sources by saying that neither WP:RS or WP:V "mention anything about being cited in reliable sources, they talk instead about being published in reliable sources".[20] When I tried to discuss this rather sophisticated point on RSN he jumped the thread, implying that the thread was not about what it seemed to be about, so the thread died. It is an interesting question whether this is what the policies intend though in my opinion. ?
  • OTOH the specific discussion Jayjg refers to above was from the last argument Jayjg was trying on his talk page before he gave up.[21] By that stage of the "discussion" I was only "allowed" to answer yes or no or he would delete from my posts, but he was apparently trying to build a truly remarkable argument that anything which says it has "peer review" for fact checking, and calls itself a "journal", but which is not the journal of an academic profession, is pretending to be something it is not, simply because of the use of these two terms which are normally academic. This appears to involve arguing that what appears to be a good type of fact checking can be assumed to be a sign of fraud. To make this argument requires the original research, so to speak, of claiming that academics who cite the journal and write about it have been tricked, whereas one Wikipedian has not been tricked, and can therefore over-rule them.
  • So.... Is the use of policy in that discussion we had consistent with how the policy is intended to be used? To me it certainly did not seem so.--Andrew Lancaster (talk) 17:32, 31 October 2010 (UTC)
Andrew, we're discussing WP:V, not your lengthy, self-serving, and fatally flawed comments and "summaries of facts"/revisionist histories, regarding your attempts to get a website on which you have been "published" accepted as a reliable source. I've said this before, stop talking about me, and focus on the subject of this page, which is the wording of WP:V. Jayjg (talk) 18:05, 31 October 2010 (UTC)
Not sure how you can justify that ad hominem aside in this context Jayjg. You have just gone out of your way to mention me and your recent discussions with me several times in a row, apparently because you think it has some sort of relevance? If it had no real relevance you should perhaps not have brought it up. If it was worth bringing up, then it is at the very least worth avoiding errors in your summary, and there would seem to be no harm in making the examples more clear and accurate?--Andrew Lancaster (talk) 18:22, 31 October 2010 (UTC)
Andrew, I did not mention your name, nor the source (until specifically asked), much less the dispute itself. I simply pointed out that at least one editor insisted a source could be peer-reviewed but not academic, which was both entirely accurate and relevant to the specific wording proposed for that section of WP:V. Focus specifically on WP:V and the wording of the section here, not other editors. Jayjg (talk) 19:16, 31 October 2010 (UTC)
If nothing else, it's a useful illustration of the sort of disputes editors get into over sources. But Jayjg, if your argument with Andrew was that JOGG is not reliable, then this makes your point above, that there might be peer-reviewed sources outside academia (which then would also be part of the "most reliable" category), moot. --JN466 18:41, 31 October 2010 (UTC)
Sorry, I don't understand what you mean, what would make what moot? I don't think there are real non-academic "peer-reviewed" sources, at least not in the sense people typically mean "peer-reviewed", but I've seen others arguing there are. Jayjg (talk) 19:16, 31 October 2010 (UTC)
We were discussing the wording "Academic and peer-reviewed publications" above. If you feel (as I do) that there are no real non-academic "peer-reviewed" sources, then the wording "academic and peer-reviewed publications" describes two sets of sources, the second one of which is wholly included in the first. It is logically comparable to saying, "American and Californian publications are ...", or "Mammals and guinea pigs are", which is logically untidy and somewhat tautological, as Californian publications are already included in American publications, just as "guinea pigs" are included in "mammals". It would seem more appropriate to describe peer-reviewed publications as an example of academic publiations, just like guinea pigs are an example of mammals, and Californian publications are an example of American publications. Does this make sense? --JN466 09:43, 1 November 2010 (UTC)
Well, "academic" and "peer-reviewed" are definitely not synonyms. Normally one would consider the latter to be a subset of the former - and, in general, the most-reliable type of the former. Jayjg (talk) 13:31, 2 November 2010 (UTC)
Would you agree with me then that the phrase "academic and peer-reviewed publications" is poorly worded, given that the latter are a subset of the former? --JN466 15:27, 4 November 2010 (UTC)
  • Support the Proposal: though I understand the objections. I'm just not happy with the current wording. For one, it's just plain wrong - academic and peer-reviewed publications are very often primary sources advocating brand-spanking new ideas, and while those ideas usually have a whole lot going for them, they're still not the kind of tried and true and true knowledge wikipedia is supposed to reflect. The proposal, at least, isn't quite so bluntly wrong. --Ludwigs2 06:30, 29 October 2010 (UTC)
    • I've added that point to the perceived failings of the present policy wording, above (please feel free to reword). --JN466 13:34, 29 October 2010 (UTC)
  • Oppose - The current wording is much more clear and straightforward. The proposed replacement seems at pains to avoid creating any actual criteria for selecting verifiable sources, and instead seems more interested in NPOV. The one exception is that it states that "High-quality mainstream publications" are the "preferred sources in topic areas that are not subject to academic research". Which subjects, prey tell, are not subject to academic research? Kaldari (talk) 21:20, 9 November 2010 (UTC)

Playing around with it a bit

I do not say this resolves it, but here is another approach:-

Current policy JN466 Proposal Another idea
Where available, academic and peer-reviewed publications are usually the most reliable sources, such as in history, medicine, and science. But they are not the only reliable sources in such areas. Material from reliable non-academic sources may also be used, particularly if it appears in respected mainstream publications. Other reliable sources include university-level textbooks, books published by respected publishing houses, magazines, journals, and mainstream newspapers. Electronic media may also be used, subject to the same criteria. Many topic areas are covered by a significant body of scholarly literature, such as peer-reviewed publications, and books by academic presses. Articles in these topic areas should aim to summarise the current status of knowledge and research reflected in that literature. Articles should present all majority- and significant-minority published positions, and may complement any available scholarly sources by drawing on non-academic sources, particularly high-quality mainstream publications, to reflect public discourse around such topics. High-quality mainstream publications include books from respected publishing houses, mainstream newspapers, magazines, journals, and quality electronic media; these are also preferred sources in topic areas that are not subject to academic research. Different topic areas are covered by different types of publications, some being more reliable than others in each field. For example in many areas peer-reviewed academic publications will be the best, and sources cited by those sources can also be considered potentially reliable, depending on context. On the other hand, even in those fields other sources can also be used if they can be shown to have a reputation for fact checking appropriate for the field. Wikipedia articles should attempt to present a WP:neutral summary of all majority- and significant-minority published positions, taking its main bearings from the best sources in that field.

Does this help in any way? To explain my bias, my main concern with JN466's approach is that it seems to attempt to give a listing of all possibilities, which I think can lead to more problems than it can solve. I have therefore tried to write more generally, but more simply. (Which has its own dangers of course.) Maybe even if my version is not considered good, it still helps find a better way forward.--Andrew Lancaster (talk) 10:03, 29 October 2010 (UTC)

I think the hand-wringing required to make it less explicit leaves it a bit of a marshmallow; I don't know quite what it's saying. The best idea, IMO, is the last sentence--that articles should aim to use the best sources available for a given subject, especially where high-quality, peer-reviewed scholarly sources exist, but that all sources with a reputation for accuracy should be considered and used where they fit the context. Ocaasi (talk) 11:02, 29 October 2010 (UTC)
Good point. It raises the question, doesn't it, about whether that is in fact the basic subject of the paragraph, and if not what?--Andrew Lancaster (talk) 11:08, 29 October 2010 (UTC)
I think you're getting at the key idea, though, which is that whatever the best sources are, we should aspire to use them--that this will often include peer-reviewed and academic sources, but it can also mean high quality journalism or other material with a reputation for accuracy. As for your general approach, I think rewriting the same ideas in a new way is very helpful to break up a logjam. Ocaasi (talk) 11:12, 29 October 2010 (UTC)
Yes, I was trying to strip it down to essentials, at least for this discussion. You could almost say that I am thinking that the first and last sentences, ignoring everything in between, is meant to be the "core" of what is intended. How far off is it?--Andrew Lancaster (talk) 12:47, 29 October 2010 (UTC)
One thing that is key to me, and which I wouldn't want to lose, is the notion that if there is a significant body of academic research on a topic, our article should describe the current status and findings of that research. The article may or may not do more than that – it may describe the public media discourse surrounding that topic, any controversies, its impact on politics, health policies, business, its cultural reception and so forth – but if there is a scholarly literature, we should aim to present what it says. --JN466 13:11, 29 October 2010 (UTC)
Sticking to minimalism for now, here is something based on my first and last sentences plus something based on the first sentence of the current policy which kinds of touches on your point (but which I thought you did not like):-

Different topic areas are covered by different types of publications, and within each topic area some of these sources are more reliable than others. Wikipedia articles often need to use several types of source because they should attempt to present a WP:neutral summary of all majority- and significant-minority published positions. The best sources are those which most helpful in judging which other sources can be considered reliable. For clearly defined academic topics, academic peer-reviewed publications are usually the most reliable sources. But they are not the only reliable sources in such areas.

Obviously I am "playing" with options like this partly to help people think about what the main purpose of this paragraph is actually meant to be. Basically is seems to be: "Some sources are better than others, and these should be used most of all, whatever they happen to be for any given topic." So why not just that?--Andrew Lancaster (talk) 13:54, 29 October 2010 (UTC)
There's something (IMO) that's getting twisted out of true here. Reliability is not meant to imply 'accuracy' or 'correctness', which are universals. Reliability implies 'conventionality' or 'orthodoxy', and those are very contextual. In most cases the two are more-or-less synonymous, but in some cases (particularly with new and fringe ideas), topics may be accurate but not orthodox, or the orthodox view on the topic may be entirely inaccurate. Examples:
  • New treatments that appear in a medical journal, which have strong research results but haven't yet spread through the discipline as a conventional practice.
  • Alternative medical practices which have long histories but are untested by modern medical science, so that the orthodox viewpoint is not the (assumedly more accurate, if it existed) western-medical viewpoint.
In my view we should really always be presenting the 'orthodox' or 'conventional' viewpoint of topics (with respect to the topic) and introducing 'accurate' or 'correct' sources merely to maintain proper weight and perspective, but either way I think we need to make the distinction between these two issues clear. I need to think for a bit about the best way to do that, though...--Ludwigs2 15:24, 29 October 2010 (UTC)
Right, but aren't the 'best' sources the very ones that typically espouse the orthodox viewpoint with most authority? Ocaasi (talk) 15:59, 29 October 2010 (UTC)
Well yes, but there is another problem with what Ludwigs2 says, I think, which is that mere orthodoxy is not the same as reliability. One can have a reputation for being reliable but not be orthodox. The sources who do not fit are the significant minorities. An author who gets respectful reviews, even though a field finds it hard to treat him or her as orthodox, need not be eliminated from WP.--Andrew Lancaster (talk) 20:27, 29 October 2010 (UTC)
lol - that's not a problem, that's part of my thesis. To take the example I occasionally get tangled up in, Traditional Chinese Medicine has something like 1500 years of practice and commentary, is a complex self-consistent theory that works well enough for what it is, and is largely unstudied by western methodologies (for an assortment of reasons). And yet, there are still editors who insist that TCM should only be discussed on wikipedia through the lens of what few western studies there are (because the western medical studies are assumed to be 'reliable'). That's bull. The most reliable source for describing TCM's practices and theories are in the large body of 'orthodox' TCM material, and western medical sources only have a place in the discussion where they refute particular claims or are required to keep advocates from presenting TCM as the dominant medical paradigm - they are not at all reliable for describing TCM, since they have not to any great extent examined TCM. Do you se what I'm saying? --Ludwigs2 23:18, 29 October 2010 (UTC)
Yes, I think so, and I doubt we disagree on the facts of such things, only how to describe it. But this is a wording question. Chinese medicine is not the type of case I had in mind. The type of case I have in mind might be for example the Multiregional origin of modern humans, which is a serious scientific theory that happens not to be the leading one at the moment. Chinese Medicine may or may not be considered unreliable as medicine generally by people who are expert in other types of medicine, but you are right that it should be in Wikipedia as an notable body of learning concerning which there can be publications which are reliable or unreliable about Chinese medicine itself. Western scientific fields, including medical science, aim as part of their own conscious aims, try not to aim at consistency with tradition or orthodoxy. For this reason, scientific fields accept theories as good science, "reliable", which are not orthodox. They do not have "heresies" (sorry, I can not think of another word) like traditional bodies of knowledge can. If an individual were to develop an alternative version of Chinese medicine, this would have to become notable on its own to be worth putting in WP, whereas if it was not notable, we'd probably have to treat as an unreliable or wrong type of Chinese Medicine, even if its results in terms of actual cures were approximately as successful. I hope this makes sense.--Andrew Lancaster (talk) 07:33, 30 October 2010 (UTC)
well we're talking about different types of cases then, and ought to accommodate both. Either that, or I'm just not understanding you... Face-smile.svg It's one thing when you have two theories within a particular discipline that are intentionally competing to be the dominant theorem, but it's another thing when you have incommensurate practices relating to the same general activity. two competing theories will naturally be subsumed together with one as dominant; incommensurate practices really can't be. put another way, multiregional and monoregional theories of human origins are part of the same discussion within the same discipline, and so naturally compare; things like TCM are simply disconnected from western scientific medicine - there's very little in the way of cross-communication, and so it makes no sense to discuss TCM in terms of a supposedly dominant medical paradigm: it needs to be described in its own terms and then contextualized with respect to western medicine. that's what I mean by using the orthodox viewpoint. Western medicine is arguably more scientific and reliable than TCM (satisfying the 'accuracy' sense of reliable), but western medicine cannot give an effective description of TCM because the models are just too different. --Ludwigs2 02:59, 31 October 2010 (UTC)
Sounds like we understand each other now. But not sure what this would mean in terms of this policy page.--Andrew Lancaster (talk) 07:16, 31 October 2010 (UTC)
A thought... I think part of the problem here is that the current wording is overly broad in its scope. It talks about the relative reliability of sources at a macro level... in terms of entire "topics". However, relative reliability often needs to be determined at the micro level... in terms of specific statements. While academic sources should be preferred for academic topics at the macro level (ie for sourcing the topic)... when we get to the micro level (sourcing a statement), a non-academic source may be the most reliable (that depends on what the specific statement is). Blueboar (talk) 16:12, 29 October 2010 (UTC)
Very good point. I was thinking about this when playing around with this. For example say you are editing on a scientific subject but there is a famous funny story about what Einstein said about it. Each bit of a source can be a different context for sourcing. But I left it because I thought "topic" can be "micro"? But then maybe the wording should make that clear?--Andrew Lancaster (talk) 20:27, 29 October 2010 (UTC)
That is why I wrote "Articles in these topic areas should aim to summarise the current status of knowledge and research reflected in that literature." This does not demand that every single source cited be an academic source, as long as the article correctly describes the status of academic research. --JN466 21:31, 30 October 2010 (UTC)

The "another idea" doesn't give particularly clear guidance at all; it's mostly "maybe this is reliable, maybe that is, depends on the circumstances". Certainly not an improvement on the current wording. Jayjg (talk) 18:14, 31 October 2010 (UTC)

  • I am aware that we are currently maintaining the present policy wording against a slight (11:8) majority of commentators on this page (including an arbitrator) who would rather replace it with Proposal 5. We may well come up with something better than that proposal eventually, but in the meantime I would ask some of those who have not given their verdict on that earlier proposal to do so now, so we have a clearer idea whether there is a consensus that proposal 5, while not perfect, is at least better than the present wording, and gets us a step closer to where we would like to be. --JN466 16:53, 31 October 2010 (UTC)
It appears that the "current version" has changed since it was first quoted for comparison with proposal 5. Regards, --Bob K31416 (talk) 18:24, 31 October 2010 (UTC)
Indeed, although the changes seem to affect word order only. --JN466 18:45, 31 October 2010 (UTC)

Just a general remark regarding these deliberations and the present state of policy. I felt that policy wasn't the place to rank or give guidance on which reliable sources are better, but rather to discuss that in a guideline or essay. In addition to reasons I expressed before, this is consistent with a general approach to policy that I think should be followed, which is keep it clear and simple. --Bob K31416 (talk) 18:37, 31 October 2010 (UTC)

That's a reasonable point—although the present wording, of course, also ranks the sources by saying "Where available, academic and peer-reviewed publications are usually the most reliable sources, such as in history, medicine, and science." And I could never figure out why we single out "history, medicine and science". --JN466 18:46, 31 October 2010 (UTC)
I understand that, and policy can be improved, for example to make it more clear and simple. However, I think that policy should be changed in small ways over time. This avoids problems of losing the good work over the years of many editors and avoids running around in circles and introducing new problems, some subtle and some not so subtle.
With this in mind, let me reiterate a suggestion for a simplifying change in the paragraph under discussion that I made previously which simply removed the ranking of reliable sources. Before continuing, please click here and read my previous remarks which included the simplifying proposed version which I will copy to here.
"Examples of reliable sources are academic and peer-reviewed publications. Material from reliable non-academic sources may also be used, particularly if it appears in respected mainstream publications. Other reliable sources include university-level textbooks, books published by respected publishing houses, magazines, journals, and mainstream newspapers. Electronic media may also be used, subject to the same criteria."
--Bob K31416 (talk) 19:10, 31 October 2010 (UTC)
That looks better than any of the other proposals I've seen so far. Jayjg (talk) 19:19, 31 October 2010 (UTC)
Looks fine to me too, Bob; except for
  • my longstanding point about "academic and peer-reviewed publications" (peer-reviewed sources are as a rule academic sources, and I've even seen editors interpret the present wording to mean that sources should be academic and peer-reviewed (because they wanted to exclude a book by an academic publisher) and
  • that university-level textbooks also seem to form part of "academic" rather than "non-academic" sources.
The rest can be sorted at the guideline level, as you said. --JN466 02:19, 1 November 2010 (UTC)
This particular proposal is only regarding removal of ranking of reliable sources in this section of policy. Let me state for the record that what I am proposing does not preclude any other changes that you would like to consider if this change is accepted and incorporated into the article. --Bob K31416 (talk) 03:30, 1 November 2010 (UTC)
I am all for simplification, which is what I think this is. I understand Jayen's concern but the simple "Example are ... and ..." is also rather neutral and not as easy as some wordings to abuse, I think.--Andrew Lancaster (talk) 07:15, 1 November 2010 (UTC)
Okay, Bob, I understand. Little steps ... --JN466 09:29, 1 November 2010 (UTC)

Current policy should stay. My personal experience is that peer-reviewed academic publications are the most reliable, all else being equal, naturally. Reason is that scientific and academic types are generally specialists, with specific background and training in the field. Reporters tend to be more generalists. One key is that newspapers, magazines, etc. get (e.g.) their science news from the scientific literature, but not vice-versa. Nucleophilic (talk) 02:27, 1 November 2010 (UTC)

Re "My personal experience is that peer-reviewed academic publications are the most reliable, all else being equal, naturally." - If that was put in the policy, i.e. the statement that "Peer-reviewed academic publications are the most reliable, all else being equal", how would someone apply that, since it wouldn't be clear what "all else" is? I feel that a good extended discussion of the merits and problems of various types of reliable sources should be in a guideline or essay and we should strive to make policy clear and simple. --Bob K31416 (talk) 03:50, 1 November 2010 (UTC)
Just expressing my completely-WP:OR opinion, not suggestng wording, which is fine as it is. Nucleophilic (talk) 04:11, 1 November 2010 (UTC)
The present wording of the policy is, "Where available, academic and peer-reviewed publications are usually the most reliable sources...". I'm not sure it is fine since the "usually" part of the statement isn't very definite and may not be very helpful in practice. From your first message, it seems that you perceive it to be fine because you have in your mind the condition "all else being equal" which seems to have a meaning for you which isn't clear to me. We should try to make policy simple and clear when we have a choice. I think the relative merits of different types of reliable sources should be given in a guideline or essay where the subject can be given the attention it deserves to more clearly and accurately portray the principles involved.
Going back to something you mentioned in your first message. "One key is that newspapers, magazines, etc. get (e.g.) their science news from the scientific literature, but not vice-versa." The news media also gets its science information from interviewing scientists in the field who interpret the information in scientific articles for reporters. In some cases, maybe those scientists interpret the science articles better than some Wikipedia editors? --Bob K31416 (talk) 04:39, 1 November 2010 (UTC)

Discussion above continued

These are merely practical real-world "rules of thumb" for making decisions in the face of incomplete information, nothing more. However flawed, they are all we can reasonably have. An old saw goes "The race is not always to the swift, nor the contest to the strong, but that's the way to bet."

E.g., a reasonable rule of thumb is that material produced by defined experts and then vetted thru the academic publication mill is likely better than, e.g., a newspaper article by someone who never looked at the subject before and who probably got his information from (say ) a press release. This does not mean that "academic" sources are always better than popular literature sources, but merely "that's the way to bet". If I could be sure of stuff, I'd be a billionaire. Nucleophilic (talk) 13:56, 1 November 2010 (UTC)

It's the writers of that portion of the policy that had incomplete information. There's less reason for editors of an article to "bet" because they have more information. Consider the example you gave regarding "a newspaper article by someone who never looked at the subject before and who probably got his information from (say ) a press release." Suppose that press release came from the director-general of CERN? And regarding the other part of your example, "material produced by defined experts and then vetted thru the academic publication mill", suppose it was a study of the effects of a medical device on just a few patients. Would that information be a reliable characterization of the device?
So a general comment regarding a ranking of reliable sources doesn't seem useful since its usefulness would be obviated by the specifics of a particular case at an article. --Bob K31416 (talk) 14:29, 1 November 2010 (UTC)
Sure. Again, "The race is not always to the swift.... ". But what if that's all you have to go on ? The real world is full of circumstances where people use established rules of thumb in the absence of complete information. Physicians, who often make critical decisions in the face of incomplete information, call such rules of thumb "aphorisms". Things like "All bleeding stops".
Anyway, the case you describe, a press-release from Cern, easily qualifies as an academic source, assuming it gets correctly-passed on in the popular literature. If it differed from other WP:RS academic sources, WP:NPOV would require that all get reported. IMHO, the websites and official communications of major museums, research institutes, professional organizations, etc. also qualify as academic sources. So if you want to specify this, I have no complaints. Might even do it myself, if there are no objections.
BTW, over on talk:nobel prize controversies is a case where a New York Times reporter (!) gets the facts about a Nobel dispute wrong, arguably deliberately. A Nobel dispute allegedly-related to erectile dysfunction drugs is "sexier" than the real dispute, omitting a fourth discoverer. So does the popular literature get distorted. True, the academic literature also gets distorted, this being the real world. But less. Nucleophilic (talk) 15:35, 1 November 2010 (UTC)
Re "The real world is full of circumstances where people use established rules of thumb in the absence of complete information." - It's also full of circumstances where generalities are used in the presence of contrary information to make wrong decisions, e.g. not hiring a well qualified job applicant because the applicant has been stereotyped by gender or race. Please note that I am in favor of giving guidance to editors regarding evaluating reliable sources, just not with generalities that may be misused. As I mentioned, discussion of the merits and problems of various types of reliable sources would best be done with more and better explanation in a guideline or essay. --Bob K31416 (talk) 16:12, 1 November 2010 (UTC)
Sure. But those are irrational stereotypes, "used in the presence of contrary information". They are not useful and practical Rules of Thumb. But sometimes ethnicity is useful-- e.g., our physician example above would know "high" ( red-headed and freckled ) Keltic ancestry is associated with melanoma, while being of African ancestry tends to exclude it. Nucleophilic (talk) 16:32, 1 November 2010 (UTC)
All of which is why the policy should give broad guidance but still leave room for intelligent editors to apply specific judgments given the unique context of each article. Ocaasi (talk) 17:04, 1 November 2010 (UTC)
I think what you meant is that the policy should give broad worthwhile guidance, not just any broad guidance. So far, I don't think it's worthwhile and may be a distraction from the specifics of a particular article. --Bob K31416 (talk) 17:39, 1 November 2010 (UTC)
Again, these are guidelines and rules of thumb, nothing more. Like all such, they are adaptable to circumstances, within the general guidelines of WP:NPOV, etc., naturally. In fact, I am having difficulty understanding how our viewpoints differ much. Nucleophilic (talk) 17:56, 1 November 2010 (UTC)
Re "In fact, I am having difficulty understanding how our viewpoints differ much." - Maybe it would help if you summarized what you thought my viewpoint was. --Bob K31416 (talk) 18:34, 1 November 2010 (UTC)
You certainly know what you think better than I do. Perhaps you could summarize it for us all. Nucleophilic (talk) 22:00, 1 November 2010 (UTC)
From your remarks you may have been misunderstanding what my viewpoint is and your summarizing it would help identify the problem in communication. You certainly have the option of not complying, so thanks for the previous discussion and that will be the end of it for me. For now, I'll leave it to other editors to decide what to do regarding my suggested change of the first two sentences of the paragraph. Regards, --Bob K31416 (talk) 22:18, 1 November 2010 (UTC)
All the proposals I've seen so far seem to basically be replacing the established criteria for evaluating verifiable sources with a vague watered-down reiteration of the NPOV policy. Kaldari (talk) 21:28, 9 November 2010 (UTC)


RfC on the relationship between the sourcing policies and guidelines

Input would be appreciated at an RfC to ask whether the sourcing guidelines (such as CITE, IRS, MEDRS) should make clear that the core content polices take priority by saying something like: "In the event of inconsistencies between this page and the policies, the policies take priority, and this guideline should be amended accordingly." Please see the RfC here at IRS. Many thanks, SlimVirgin talk|contribs 23:44, 11 November 2010 (UTC)

Cannot be superseded

Nucleophilic has several times added the following, so I'm bring it here for discussion:

The principles upon which these policies are based cannot be superseded by other policies or guidelines.[1]

  1. ^ Wikimedia Concensus notes " Consensus should not trump NPOV (or any other official policy). However, a group of editors may be able to shut out certain facts and points of view through persistence, numbers, and organization. This group of editors should not agree to an article version that violates NPOV, but on occasion will do so anyway. This is generally agreed to be a bad thing and to be avoided."

  • Oppose. This was in the policy a long time ago, but was removed after consensus developed that it was inappropriate. It's also not clear which of the points in the policy are to be regarded as the "principles" that cannot be superseded. SlimVirgin talk|contribs 05:13, 2 November 2010 (UTC)
  • Oppose. The wording is apparently based on some wording added to a Meta page in 2005. He's been trying to add it to WP:NOT for several days, and more recently to WP:NPOV. To be quite frank, en-wiki policies are typically far more polished and well thought through than policies on other wikis, even meta. Jayjg (talk) 05:23, 2 November 2010 (UTC)
  • Oppose. It's just to say that what is written here is not always followed. It doesn't belong on the policy page. If there are enforcement issues that people are not addressing, noticeboards/RfC would be the place. Neither the U.S. constitution, nor any legal document, has a passage saying 'This law is abused and ignored periodically, which is bad'. Ocaasi (talk) 05:43, 2 November 2010 (UTC)
  • Query for more information Long time on wikipedia, but fairly new to the policy pages. I have apparently misjudged the situation. Could someone direct me to where this policy was rescinded here on wikipedia so I can catch up. IIRC, it was the policy last time I read it, some years ago. Admittedly a lifetime on WP. Nucleophilic (talk) 05:55, 2 November 2010 (UTC)
    • Nucleophilic, you've made just over 200 article edits to Wikipedia, and 430 edits in total. Since mid-September, you've mostly tried to re-write policies and guidelines. I'm not sure what you mean by "a lifetime on WP", but don't you think it would be prudent to get a little more experience, say, writing articles, or working in various other areas of Wikipedia, before trying to re-write all the policies and guidelines? Jayjg (talk) 06:00, 2 November 2010 (UTC)
      • Er, by "lifetime on wikipedia" I meant that the several years since I last read this specific policy statement here on en.wikipedia is a virtual lifetime on wikipedia in terms of change. This accounts for my ignorance of the change in policy, for which I apologize.
      • As I note and you reiterate, I am relatively new to playing an active role in the policy pages. True, I have paid some attention to policy in the past and reasonably-believe I know it in some detail. E.g., I knew about the above little bomb on wikimedia, when nobody else here seems to be aquainted with it. So your "little more experience" statement seems rather misplaced, if not slightly, and doubtless unintentially, uncivil. So as not to seem uncivil myself, I will repeat what many here have said- we all continue to thank you for your years of service to wikipedia.
      • Anyway, I would love to read up on the exact process by which this change in policy happened. Perhaps you can point me to the sources. As you know, I have been here off and on since 2006 under my present handle. Not many editors stay quite this long. Lurked and posted under ip number for a year or two before that. Nucleophilic (talk) 14:39, 2 November 2010 (UTC)
        • Maybe this is the previous statement?[22] "The principles upon which these policies are based are negotiable only at the Foundation level." --Bob K31416 (talk) 15:46, 2 November 2010 (UTC)
  • Oppose - No good reason for this given. Policies should say what is relevant and not filled with gobblydegook. Irrelevant pomposity. Dmcq (talk) 15:31, 2 November 2010 (UTC)
    • Stated-simply, the issue is whether a simple local concensus can override the three pillars of wikipeda. Hardly gobbledegook. The rest is a link to a quote from Wikimedia, the umbrella for the wikipedia empire. There seems to be general agreement that this also was the policy on Wikipeda. Unfortunately, nobody has provided information on exactly how and when it was rescinded. I.e., technically, it may still be policy here. This is a big enough change that it should have generated quite a sh!tstorm. BTW, we also thank SV for her years of service. Nucleophilic (talk) 16:09, 2 November 2010 (UTC)
Nucleophilic, what do you mean by "technically"? It appears to pre-suppose the existence of something above the community of Wikipedia, which can over-rule it's self-rule. I am supposing you are aware of WP:IAR and WP:POLICY? Those might be the places which explain what you are asking about.--Andrew Lancaster (talk) 16:21, 2 November 2010 (UTC)
Stated simply yes, the community could in principle agree anything except where the foundation specifically puts in its own policy. It hasn't on this subject. Please see WP:POLICY section on derivation. I'm pretty certain the foundation would step in if it went too far off the rails as far as they are concerned, but in essence wikipedia is self governing in its normal running. And it decided long ago that statement was unnecessary and removed it. Dmcq (talk) 16:26, 2 November 2010 (UTC)
See also [[23]] which states what the foundation expected from wikipedia. Dmcq (talk) 16:41, 2 November 2010 (UTC)
Re "Stated-simply, the issue is whether a simple local concensus can override the three pillars of wikipeda." - Yes according to WP:IAR. Policy isn't perfect. It's created by basically the same type of open editing environment as the hypothetical case you mentioned for an article. You might consider whether quasi-consensus can produce a flawed policy just like it could produce a flawed article, through alliances, gaming the system, etc. and simply the limited ability of us editors. --Bob K31416 (talk) 17:05, 2 November 2010 (UTC)
  • Oppose Given that the 3 pillars are interpreted/implemented by consensus, what operational meaning would this have? I CANn think of abuses that this would have. e.g. one person claiming that, per his interpretation , consensus violated this proposed new clause, and thus his opinion should trump the decision. North8000 (talk) 16:23, 2 November 2010 (UTC)
  • Oppose, having had my concerns answered. Nucleophilic (talk) 17:32, 2 November 2010 (UTC)
  • You're opposing yourself? :) Nice to see someone do that after further investigation. SlimVirgin talk|contribs 18:08, 2 November 2010 (UTC)
  • You mean editors here sometimes persist long after a point becomes moot ? I'm shocked, shocked !! Nucleophilic (talk) 19:15, 2 November 2010 (UTC)
  • Related note: this line existed in NPOV alone for a while, I went ahead and deleted it from there based on the discussion here, since the paragraphs involved are otherwise identical and should be consistent between all core policies. Gigs (talk) 02:48, 6 November 2010 (UTC)
Gigs, if I'm not misreading this, the discussion above was only about not to include Nucleophilic's addition, not the initial line itself. I'm not sure there is the same consensus to remove the line you did. I think it needs a separate discussion; for example, do we not want to give the core 3 supremacy, is there something which can supersede them?, etc. Ocaasi (talk) 04:13, 6 November 2010 (UTC)
I'm not sure what you mean, the line only existed in NPOV, not in the other two core policies, it makes no sense to include it in only one. And yes, there are things that supersede the three core policies, like copyright violations, libel, threats, in some instances, BLP, IAR, etc. The core policies can't be completely suspended by a local consensus, but if you'll notice the header box at the top, the word "normally" links to "use common sense". No policy or set of policies on here is absolute. Gigs (talk) 21:33, 6 November 2010 (UTC)
I suggest we remoe the line from NPOV (for consistency) and reinsert it on one of the pages that discusses the core principles as a unit. --Ludwigs2 23:17, 6 November 2010 (UTC)

The line has been restored to WP:NPOV and there is a discussion on the talk page there. Gigs (talk) 04:47, 7 November 2010 (UTC)

I think this discussion may need to be shifted over to NPOV and re-argued there... it seems we have people saying: "who cares what the consensus was over at WP:V. A consensus at WP:V is irrelevant to WP:NPOV, as this is a different page". (see: WT:NPOV#Sentence replaced re. non-negotiability of the principles upon which this policy is based) Blueboar (talk) 14:34, 7 November 2010 (UTC)

Above discussion continued

Agree with Ocassi. When I say the issue is moot, I mean that the poll was clearly against my change. So in the interest of WP:don't be a dick, I threw in the towel. My wimping out should not be the pretext for wholesale revamping of long-standing wikipedia policy. In fact, it is just the sort of "concensus" we see here that cries out for some rules that cannot be changed by a local concensus.

I am also concerned that some of the present editors ( who often, uh, act in concert ) have been severely chastized by Arbcom, even to the point of being desysoped and more. E.g., Wikipedia:Requests for arbitration/West Bank - Judea and Samaria. I hate for my unwise post be the excuse for some sort of coup while nobody was looking. Perhaps this ought to be the subject of more general debate. Nucleophilic (talk) 18:09, 6 November 2010 (UTC)

I'm not sure what your point is, but personal attacks like this are very inappropriate. If you want to accuse Jayjg of socking, WP:SPI is that way. Gigs (talk) 21:42, 6 November 2010 (UTC)
No, I am not accusing anyone of socking. The editors involved are clearly not socks, but merely regularly support each other. Far from being socks, SV even once put a block on Jayjg. In fact, I am surprised that you would make this accusation, which seems (well) straw. Rather, I am concerned that an important change in wikipedia policy may be made without wider input. I suggest it be referred to (say) "the pump".
Similarly, citing adverse Arbcom decisions is not a personal attack. If I really wanted to be provocative, I would have quoted Arbcom's rather inflammatory decision. In the interest of wp:civility, the only part I will quote directly is "Jayjg is also thanked for his years of service.".
If Arbcom did not want their decisions cited and taken into consideration, they would not openly publish them. In particular, some of the elements of this decision are relevant to the present issue. Finally, Wikipedia policy is that official "outing" is one of the consequences of and a major deterrent to bad behavior. If you disagree, I suggest taking this matter up with Arbcom itself. Nucleophilic (talk) 14:54, 7 November 2010 (UTC)
I'm not sure about the background, Nucleophilic, but I'm almost certain bringing the issue up here won't help improve this policy. Would you discuss it somewhere else, please? If you have a particular criticism, you know where the noticeboards are and how to start and RfC. Otherwise, this kind of speculation and insinuation seems highly unconstructive. Thanks, Ocaasi (talk) 15:28, 7 November 2010 (UTC)
Indeed... please discuss the policy not the editors. Given that this discussion is overlaping onto more than one of our core policies, I think calling for a centralized discussion and input from a wider sampling of the community (through RFC, pump discussions, etc.) is a valid request... but casting aspersions on other editors does not help. Blueboar (talk) 16:02, 7 November 2010 (UTC)
All true. I was merely responding to a ( somewhat straw ) insinuation that I was making accusations of "puppetry". In response, I cited evidence proving they are not puppets. We could go on and on like this ad infinitum. But it does seem a waste of time.
Again, my concern is that a major decision on wikipedia policy is being made by just a few editors. Similarly, the Arbcom decision bears on the issue here, independent of any personalities. Nucleophilic (talk) 16:49, 7 November 2010 (UTC)
Patience and AGF please. Nefarious schemes (should such exist) will come to nothing in the end, and (should such not exist) all major decisions are always made by just a few editors. You should take the long view, that short term edits will only produce long-term changes where the community as a whole is satisfied with them. that being said, what specific problem are you seeing that needs to be addressed? --Ludwigs2 18:13, 7 November 2010 (UTC)

I think the discussion of this clause has been skewed by the reaction to the provocative footnote. The point at issue is to establish that the principles of WP:NPOV, WP:RS and WP:V cannot be overridden by consensus when writing, for example, a guideline. SamuelTheGhost (talk) 12:16, 8 November 2010 (UTC)

Well, if you are going to try to define what is being discussed more clearly, which sounds like a reasonable proposal, presumably you mean a "local consensus"? OTOH, what is the problem being solved here? Do people really ever claim to have a right to over-ride those three polices? I guess that more often what people argue is that consensus defines what is in accord with them?--Andrew Lancaster (talk) 12:27, 8 November 2010 (UTC)
Yes, I mean a local consensus. Decisions about what gets written are often taken by quite a small number of editors, whose consensual opinion may well not reflect that of the wikipedia community as a whole. That is always a danger with "consensus". It is true that people usually try and present their views as based on the principles, and never directly challenge them, but in practice there are guidelines where the principles are insufficiently observed. I don't want to be specific here, as that would lead us into arguments which divert us from the main point. It's worth stressing the fundamentals on which we all agree. SamuelTheGhost (talk) 16:18, 8 November 2010 (UTC)
Maybe we should just restate which consensus we mean. A few editors can't override NPOV on an individual article; although, if the entire community decided to rewrite the NPOV policy, then, they could. Policy can change by massive amendment/agreement process, but that's not the kind of single-article scenario I think the policy is addressing. Ocaasi (talk) 16:50, 8 November 2010 (UTC)
I agree... we do need to be more specific. I think we are mixing up three separate issues: a) is this policy "written in stone" or can this policy change? (the answer to that is "it can, and does change") b) Can another policy/guideline contradict what is said here? (the answer is yes. In fact this can and frequently does occur... however conflicts between our policies shouldn't occur... so, when a conflict does occur we need to discuss the situation and work towards resolving it... bringing the two pages into sync. That may require changing the other policy or changing this policy... or both) c) Can the consensus at an article allow us to ignore this policy? (answer: essentially No, but... and this is extremely rare... an exception is made when there is reason enough to justify invoking WP:IAR). Blueboar (talk) 17:15, 8 November 2010 (UTC)
For a) Policy can change, but it requires consensus to do any more than minor tweaks; b) Guidelines which contradict this policy should be superseded by this policy--if the conflict is with NPOV or NOR, we need to resolve it; c) local article consensus should not override this policy except in very rare cases where it somehow improves the encyclopedia, avoids a copyright/blp issue, etc. Where do you want to put them? Do they belong here or at WP:Consensus? I think only B is relevant to this page, and that somewhere appropriate A and C should be mentioned at a central Policy page which addresses the big 3+ (maybe WP:Policy). Ocaasi (talk) 17:28, 8 November 2010 (UTC)
Yes, but in separating the three cases we seem to have merged others. We are using the word "policy" to cover both principles and guidelines, whereas as I see it the principles should be seen as effectively immutable, while guidelines are more fluid and less authoritative. If for example a guideline is in contradiction to WP:NPOV, it's the guideline which must be altered. SamuelTheGhost (talk) 17:41, 8 November 2010 (UTC)
No, it's there: 'Guidelines which contradict this policy should be superseded by this policy--if the conflict is with NPOV or NOR, we need to resolve it'. Capital-p Policy applies to NPOV, V, NOR, COPYRIGHT, BLP (I might have missed one). Guidelines is everything else. Indeed, Guidelines are somewhat more flexible than Policies, or at least less authoritative. The whole purpose of the 'superseded language' is to establish that (almost) nothing trumps the core policies, especially not guidelines, and not even a stable but local editorial consensus. Ocaasi (talk) 21:54, 8 November 2010 (UTC)
I agree that 99% of the time, if a guideline conflicts with a policy it is the guideline that will end up needing to change... however... there are rare occasions when it turns out the guideline is covering something we never thought of when writing the policy, and that may require altering the policy. Obviously such a situation would require a lot of discussion and consensus building... but because it can happen, it is a mistake to say that "Policy always supersedes guidelines". When any policy or guideline page conflicts with policy we need to avoid the knee-jerk reaction of "Policy trumps guidelines"... we always need to stop for a second and examine why the two pages conflict, and then think about what is the best way to resolve the conflict for the project as a whole. Blueboar (talk) 01:16, 9 November 2010 (UTC)
I think that even in the exceptional cases you mentioned, Policy has to change before the guideline can be considered official. I agree completely that guidelines can present improvements over policy, but if they contradict, I don't think the guideline can be considered 'binding' until the policy adopts its improvement. So... Policy still supersedes guidelines, however Policies can of course adapt to include improvements from guidelines, and when they do... well, they still supersede them. Basically, Policies are bullies and they do what they want until someone else comes along with a better idea and then the Policy steals it and beats them up. Ocaasi (talk) 01:30, 9 November 2010 (UTC)

Ocaasi states: "A few editors can't override NPOV on an individual article; although, if the entire community decided to rewrite the NPOV policy, then, they could. " Anybody question this statement ? The rest is just details. Nucleophilic (talk) 19:51, 12 November 2010 (UTC)

Sure they can. If what those editors truly believe to be an NPOV version of the article would, say, violate US laws (e.g., I believe that some court declared posting the code to crack DVD regions was illegal; similarly, perhaps they believe it impossible to write a neutral article about a specific picture of a naked child without posting the image itself), then they should post a (slightly) "non-neutral" but non-criminal article. One might also write an article that you thought was non-neutral if there were some specific, important reason—say, by not describing the kidnapping of that journalist a while ago. Or, for that matter, if the WP:OFFICE declares that you won't include something, then in the end, you won't include it. You might believe that a neutral article requires a link to a blacklisted malware website, but it's not going to happen, no matter how many editors at the article want it. The software just won't let you.
NB that I doubt this kind of exceptional circumstance will come up more than about once per every million article-years, but WP:Ignore all rules applies to all rules, not merely all rules except my personal favorites. WhatamIdoing (talk) 20:30, 12 November 2010 (UTC)
You don't even have to go to those extremes. If a group of editors strongly believe that it is in the best interest of the project write a non-neutral article, they can invoke WP:Ignore all rules and do so. Blueboar (talk) 23:17, 12 November 2010 (UTC)
But Blueboar I do not see this happening very often and what I see instead, to describe it in a simple way, is a deliberate tendency amongst people who influence how community norms are written and interpreted to try to make the community forget messy rules like IAR. Simple and clear rules written in a legal style and which always tend to give strength to one side in a dispute, without messy discussion, like Burden, have a certain bureaucratic attraction compared to messy but much more important rules like IAR. But does that make them good for WP itself? Consider the comments which came out during the recent discussion here on the burden rule. In my opinion, links like IAR, DUE and SYNTH have worked well for WP in a messy way, because they force people to argue in terms of the messy reality and give little opportunity for neat and clean wikilawyering. Wikilawyering is nearly always present during the worst problems on WP apart from things like vandalism. It happens because people think they see short cuts in the rules, which allow them to get what they want for the wrong reasons. So whenever policy norms are being written up, trying to avoid wikilayering should be one of the main aims, but to be honest I fear that, due to their administrative experiences, the wikipedians most likely to be writing policy pages are going to be people who tend to desire clear and simple rules, even if those these encourage wikilawyering, and are not good for WP overall. It sometimes seems to me that there are an increasing number of people who understand WP more like nucleophilic understands it (which BTW I agree to be wrong). Where are they getting their information from? I agree with you that any close reading of the policy pages should set him straight, but I hope people consider whether the policy pages are being deliberately written in a way which makes close reading overly-necessary. My apologies for a slight diversion here.--Andrew Lancaster (talk) 08:28, 13 November 2010 (UTC)
The last thing I want to do is encourage wikilawyering. Personally, I understand the guidelines as just that, general guidelines and rules of thumb to be followed except when there are articulable reasons not to do so. "As a general rule", " all else being equal " type stuff. Nucleophilic (talk) 15:46, 14 November 2010 (UTC)

When apparently reliable sources are objectively wrong

This policy states clearly, "verifiability, not truth" but what do we do when apparently reliable sources are objectively wrong? I am very deliberate in my choice of the words "objectively wrong" here.

Example, we have a problem on the Wikipedia with the term British Isles which is being discussed here [24]. A second RfC was opened here [25] relating to allegations of an ongoing campaign to remove the term but it was suggested this was actually a policy issue not a content issue. Our problem is simple but unique. The British Isles consist of 4 groups of islands; the United Kingdom, the Channel Islands, the Irish islands and the Isle of Man (actually a small group). The term is a translation of a 2,000 year old name given to the islands. Within the British Isles, there are two sovereign states; the UK and Ireland. Neither the UK (aka Britain) or Ireland include the Channel Islands and the Isle of Man which, while sharing a history and culture, retain significant independence. During that history, many other kingdoms and nations have existed and the borders between Britain and Ireland have been united and split apart. For the most part, the term "British Isles" is accepted as a non-political, geographic term on a par with Scandinavia. Despite minor disputes from some elements of Irish society, no alternative terminology has been widely accepted yet within academia*. A number of individuals on the Wikipedia are seeking to replace the term "British Isles" with alternatives, usually "Britain and Ireland". Legally and objectively, "Britain and Ireland" does not include the Channel Islands and the Isle of Man. We have a consensus on this. It is not a matter for opinion.

I think that the 38 archive pages consisting largely of discussion on this issue, without resolution say more than I can. North8000 (talk) 15:02, 14 November 2010 (UTC)

The problem is that for various reasons (historical, commercial and chauvinist etc), numerous references show that "Britain and Ireland" is used inaccurately as a shorthand for all of the territories within the British Isles. However, no references ever state that the Channel Islands and the Isle of Man are within either the UK or Ireland.

I do not wish to discuss the British Isles renaming dispute content issue here. I wish to raise the underlying policy issue. What do we do when apparently reliable sources are clearly and objectively wrong? Do we go along with the error on the basis that it is "common" and verifiable, or should we take a hardline on accuracy? Stylistically, it is impossible for us to use "Channel Islands, Ireland, the Isle of Man and UK", every time we normally use "British Isles". Due to the close proximity of the island, and the shared history and culture, many topics apply across the whole region.

(*Please note, for editors outside of the are, it is not possible to be nationalistic for the British Isle. The British Isles is not a nation such as Great Britain but a geographic term). --LevenBoy (talk) 15:46, 13 November 2010 (UTC)

I think geographical terms are a slightly special case on Wikipedia. Whenever they become a major issue we try to agree on a standard, the same way as we might agree on a style point, or concerning English or American spelling. For example I once used a source which referred to the region of Egypt, Sudan and Libya as northeast Africa, which is a usage that disagrees with the WP article by that name, and so I just avoided using direct quotes and wrote "Egypt, Sudan and Libya". Often such wording tricks avoid a problem.
Still, while making that decision about what our standard or standards will be (which seems to be the phase your discussion has been in for a long time?) we do need to decide what that decision should be based on. But is the problem really a case of needing to decide between what is true and what is verifiable? I think it is actually a case where the verifiable sources are still what you all refer to but (as is actually quite common) they simply do not all agree and do not all focus equally clearly upon this point. So it might be better to address this question as one of what the best and most authoritative sources are?--Andrew Lancaster (talk) 16:17, 13 November 2010 (UTC)
This isn't a WP:V issue (there are reliable sources to support both views on this)... it is a WP:NPOV issue. Neither side in this debate is "objectively wrong"... rather they are subjectively wrong (ie those who hold opposing views are of the opinion that the other side is wrong). That said... what you have here are sources tht disagree (some say British Isles, others try to be "PC" and say Britain and Ireland or some variation)... our job is to note that there is a disagreement between the sources and not to take sides in the debate. Blueboar (talk) 16:20, 13 November 2010 (UTC)
I concur that is not a WP:V issue, but a WP:NPOV one. I've commented out the RfC tag, because it's not asking for any modification to this policy. Tijfo098 (talk) 18:53, 13 November 2010 (UTC)
United Kingdom isn't an island. Perhaps you (LevenBoy) meant to say 'Great Britain'? GoodDay (talk) 16:21, 13 November 2010 (UTC)
I move we refuse to discuss this case on the grounds that the original poster claims that a general definition of British Isles can be objectively wrong. Since there is no entity that has, or even claims to have, the power to dictate the meaning of English terms in general, it is patently impossible for a general definition of an English language term to be objectively wrong. Jc3s5h (talk) 18:39, 16 November 2010 (UTC)
I am sorry, you are well off target on this one. I have never stated that.
I did not use the example of a "general term". I used the example of the specific political states of the UK, Ireland, Channel Island and Isle of Man are objectively defined by legal statute and nice clear lines drawn on a map. I wish to avoid discussing the example here. We are doing so elsewhere. Please just address instead the simple policy issue, what do we do when apparently reliable sources are objectively wrong? --LevenBoy (talk) 14:27, 18 November 2010 (UTC)
When reliable sources disagree, we note both (or all) positions in the article, with citations of reliable sources for each position. This happens all the time in science, history and other subjects. Experts in a field can disagree, often strenuously. It is not our place to decide which ones are "right". -- Donald Albury 11:21, 19 November 2010 (UTC)
And I would add that if a term is legally defined by some entity, that legal definition only binds the entity itself, and does not suppress any common or traditional use of the term. For an example of competing attempts to legally define a geographical term, see Persian Gulf naming dispute. -- Donald Albury 11:44, 19 November 2010 (UTC)

Discussion of potential RS error

I stumbled upon a similar situation while working on Marathon world record progression where the International Association of Athletics Federations (a reliable source) state that Henry Frederick Barrett set a world record in London on May 26, 1909, however, an actual newspaper clipping from The Times (another reliable source) shows it to have been set on May 8, 1909. Given that the IAAF is the official record-keeping organization, we have to go by what they say, but I included various footnotes explaining the discrepancy. This one is particularly troubling to me in that 1) the IAAF states Albert Raines broke the record in New York on May 8th, when it is likely that Barrett had already broken the record in London earlier that day, and 2) my explanation for the discrepancy borders on WP:OR. (I did report my "discovery" of this error to the IAAF, so we will see what they do.) Location (talk) 21:54, 13 November 2010 (UTC)

Looking at this talk section for its title and not its content, and after inserting subheads to distinguish this content (flipping a coin to choose them), I think Location makes a good V point. In this situation I think it would be simple enough to include both sources in article and make a brief non-OR note allowing readers to grasp the implications for Raines, if possible; dunno what wording would suffice, of course, and if we don't have evidence of the hourly timeframe of the May 8 events the implications themselves are not established. Basically "The Times states ... However, the IAAF records state". It's possible there is some unknown misunderstanding behind the IAAF's redating that would support their record listing, so we stick very closely to the sources as our solution.
In general on this very interesting question, a true "objectively wrong" source is usually glossed with a Homer-nods notation such as "sic" for typos, or, in the U.S. Code, "probably intending X" (codifiers are not permitted to correct the clerical errors of legislators, only to note them). It can often be corrected by replacement with another source. At risk of getting off-topic, it is a different case when the source is merely "technically wrong", meaning committing a solecism like sources using "British Isles" to mean two slightly different groupings of islands. (I'm glad I linked that: I just discovered that it's a solecism to use "solecism" outside of Cilicia!) I just had a source that referred to a region in the 12th century BC(E) as "Palestine", when at best one might say part of the region was then beginning to be known as "Peleshet" ("Philistia"), so I changed it to what might be the "proper" temporospatial term "Canaan". It is clear that the source was merely committing a shorthand reference to "the region later called Palestine". Some cases can be corrected silently, some need minitags, and some need the "semicolon" treatment to give two conflicting sources each their "fair" say. JJB 17:24, 14 November 2010 (UTC)

I watched that article closely for about a half year. Take a look at the 38 archives of debate on essentially the same topic. I even tried to facilitate a bit, and the regulars wisely advised me to run from that place, which I did. The core question of dispute is a matter of viewpoint/ advocacy rather than a matter of fact/ truth. (And, for those who try continue try to reverse engineer a Wikipedia mission from Wikipedia guidelines and policies, the foundation mission requires correctness, where objective truth exists, wp:ver is merely means to that end, not the objective) And the article involves folks who are similarly at odds in the real world. Current Wikipedia policies will, until they are refined, be used as a means to perpetuate that conflict rather than solve it. North8000 (talk) 23:54, 13 November 2010 (UTC)

Thank you Andrew, I agree that geographical terms are a slightly special case on Wikipedia and thank you Location for an alternative example and your thoughts, which I agree with, North8000. Within geographical terms the British Isles/Isle of Man/Channel Islands issue is a unique one for the English language. So far, attempts to work towards a MOS have failed and they are likely to continue to do so until this policy aspect is clarified. Therefore, I am returning this issue as a policy question because it is the policy question I am raising, not the British Isles issue. It is nothing to do with POV pushing and "sides".
In the case of the Britain and Ireland we have clearly and objective defined entities. There is no ambiguity. No room for subjective interpretation. Not even the smallest border dispute. Objectively, in this case according to the law, both the terms exclude the Islands of Man and Channel Island.
Yes, some individuals have and do use them erroneously to include a larger territory than they objectively describe. There is also consensus both that such use exists and is erroneous. The policy question is, therefore, where there is a dispute and likely other motivations to edit which aspects do we put more weight on? Common use, even if it is wrong, or the objectively legal definition?
Blueboar, I have read your comments here and at 'identifying reliable sources'. It was difficult to chose which page to post on as the issue falls between the two but I am sincerely raising the policy issue. In this case, it is unfeasible to use both or all because it would become far too unwieldy and it effects far too many topics. If you look closer at the discussion, you will realise why this is important. It is not a simple black and white edit war.
I think we need to develop the policy to specifically address such disputes. My proposal is that where there is such a dispute, the current, legally objective definition should take priority over common use. An encyclopedia should prioritise accuracy. Especially where it is easy verifiable. Which is why I posted here. --LevenBoy (talk) 01:25, 14 November 2010 (UTC)
(added later) I think that the 38 archives / 4-1/2 years full of an unresolved dispute on this say more than I can. I know I'm going to blow it trying to summarize in two sentences, but IMHO it's a term which has no legal/ country-type status, but which exists, but less now than before. One camp wants the term to be used and covered, the other camp wants to have use and coverage of the term minimized or for it to disappear, which could include wanting the article to disappear. And one core of fuel for the fire (aside from the usual Wikipedia warfare stuff) is that its a term with "British" in it's name and which includes all of Ireland. North8000 (talk) 15:23, 14 November 2010 (UTC)
We already have a policy for this... its called WP:Neutral point of view. More specifically, in that policy there is a section called WP:NPOV#Naming. Blueboar (talk) 14:20, 14 November 2010 (UTC)
Can I just point out the in isolation the two countries Britain and Ireland do not include the IoM or CI but when used as the term "Britain and Ireland" can and do so. It is also worth noting that from a purely geographic point of view the CI are NOT part of the BI's. It is only included by some unexplained tradition. Bjmullan (talk) 15:26, 14 November 2010 (UTC)
Wish you the best; you have a tough situation in that article. After 6 months of watching it, my advice would be to make it an article about the term "British Isles" rather than about all of the land & countries that that term encompasses. But as highly charged as this situation is there, a "middle of the road" proposal would probably be opposed by both sides. North8000 (talk) 15:40, 14 November 2010 (UTC)
Done already here :-) Bjmullan (talk) 16:05, 14 November 2010 (UTC)
(an aside) The Channel Islands are an interesting case... looking from a historical perspective, they are not part of the "British Isles" ... they are the last vestige of old Duchy of Normandy, and the fact that the King of England was (or at least claimed to be) also the Duke of Normandy. Indeed, there is a traditional "Loyal Toast" given on the islands which goes: "The Queen, our Duke!" (not Duchess). Blueboar (talk) 16:16, 14 November 2010 (UTC)
  • WP:PLACE covers this quite well. I fail to see how this RFC benefits the project - it's the 3rd RFC started within the last week by LevenBoy and LemonMonday on the British Isles Naming issue. It's starting look like forum-shopping. Wikipedia is not here to "right the wrongs of history" or of contemporary politics. We already have policies for assessing reliablity and the due weighting of sources, for geographic naming conventions, for naming wrt Ireland, and we have a space for discusion vis-a-vis the use of British Isles.
    Furthermore LevenBoy your above post frames this dispute as a campaign to remove the term 'British Isles' from WP - the mechanisms listed above are for usage not removal. Anyone atually involved in such a campaign should be reported per the topic's probation. Also, you are under civility parole yourself and need to stop casting generalized aspersions, and assumptions of bad faith, about others (whether named or not) - that's called poisoning the well and is a type of ad hominem remark. If there is a specific problem please report it with evidence (diffs) - I am (as every other admin is) very happy to deal with an evidenced issue which concentrates on diffs rather than speculation about motives. I understand your frustration LB but all involved in the Britsh Isles terminological dispute on WP need to work together towards agreement and consensus that is the only way forward--Cailil talk 16:31, 14 November 2010 (UTC)
BJM - I believe you are wrong in your assertion that taken together B and I do include the IoM and CI. There seems to be a drive on Wikipedia to ensure this to be the case but I see no evidence of it outside Wikipedia. I wonder if, for the sake of clarity, we should seek to replace a number of instances at Wikipedia where B & I is ambiguous, with the totally unambiguous British Isles (where our policy is to include the CI). LemonMonday Talk 16:55, 14 November 2010 (UTC)

I'd just like to take the discussion back to the policy issue I raised. Other aspects relating to the British Isles issue and history should be address elsewhere. Can I reiterate, this is not a simplistic "British versus Irish" thing as John J. Bulten has attempted to reframe.
I have read over the related guidelines and they do not address such issues. I am starting to think this particle example will have to be specifically decided elsewhere because consensus is not possible. I am just not clear where yet. Arbcom does not appear to address content decisions and it is a really Manual of Style issue.
Removing the British Isles equation from the equation, the situation we have is basically as Bjmullan presented it.
a) We have subject matters, entities with legally, specifically and objectively defined limits.
b) Those terms are then used erroneously in common use which have apparently acceptable references.
c) There is dispute over the erroneous use of the terms.
The question is, what do we do?
I suggest that where there is confusion or are disputes over the use of terms, policy should require us to observe the objectively accurate definition.
In short, that we should err to the side of accuracy in the first place rather than trends or popular opinion which are likely to change over time. --LevenBoy (talk) 22:58, 14 November 2010 (UTC)
LevenBoy I have noticed that on a number of occasions you have used the term "legal" with respect to the status of the term British Isles. Perhaps you could direct us all to the specific piece of international legislation that states this.... Bjmullan (talk) 23:15, 14 November 2010 (UTC)
LB's responded at talkpage of British Isles. -- GoodDay (talk) 01:39, 16 November 2010 (UTC)
It is messy to duplicate a response but the answer was, "I never have". I have though stated facts that each of the UK, Ireland, Channel Islands and Isle of Man most certainly are legally defined as independent, the only relationship between them all being by "international treaties". Fact. --LevenBoy (talk) 14:15, 18 November 2010 (UTC)


There is no standard term because there is no standard way in which the islands concerned are linked. Between 1603 and 1948, all could be grouped together politically, but from a geographical point of view, the Channel Islands are often better linked with France. It is therefore up to the editors of each article to reach a consensus of how to name the islands taking into account the subject matter concerned. Martinvl (talk) 13:19, 18 November 2010 (UTC)

Answering the titled question, it's very common sources that meet the WP:RS criteria to be objectively wrong. Aside from "nobody's perfect" the RS criteria overweights "editorial review" and primary/secondary/tertiary classification criteria at the expense of objectivity and expertise criteria. Under the current rules you just have to find more more and better sources and the resolve it by consensus or other WP mechanisms.
I think that the implied applicability to the conflict at the article is incorrect. I don't want to let that debate (which spans 38 archives) get started here too. But everything has to start with clarifying / agreeing on what the question is, i.e. What EXACTLY is the question with respect to article content? This stage is particularly relevant there. North8000 (talk) 13:54, 19 November 2010 (UTC)

Summary style

The Wikipedia:Manual of Style (summary style) guideline is often interpreted as providing an opt-out of WP:V for summary sections where the daughter article provides the sources. This is despite it repeating WP:V's nutshell. We have no policy, AFAIK, that allows articles to be sourced to other articles on Wikipedia, whether via normal links or via summary-style template links. Many of our FAs adopt summary style and I'm not aware of any that don't completely stand on their own legs wrt sourcing. Since guidelines should document best practice, it would be useful if anyone can enlighten me where our best articles follow the practice suggested by that guideline. Can I encourage folks here to join the discussion over at Wikipedia talk:Manual of Style (summary style)#citations. Colin°Talk 18:47, 22 November 2010 (UTC)

Even in summary style, sources are needed. I have expanded on this at the other discussion. Blueboar (talk) 20:23, 22 November 2010 (UTC)
I've added words to that effect to the MoS summary-style guideline. SlimVirgin talk|contribs 21:48, 24 November 2010 (UTC)