Wikipedia talk:Reliable sources

Transcript of a lecture hosted on a student organization's page

• Note: Moved to Wikipedia:Reliable sources/Noticeboard

Adding pages to list of Reliable Sources

Though I am not sure whether this is the right place to post my query/proposal, I decided to give it a try. While planning to put the external link from fxwords.com, I came to know about this page. I feel that the link can be considered amongst reliable sources and this applies to the entire site as well. I find the site blacklisted instead and the same cannot be posted even on this page. Please suggest.Diptanshu^Talk 15:23, 19 November 2010 (UTC)

Sources with user-generated content

Can the policy page include material on sources with user generated content? On the WP:V policy page, the section on Questionable sources and Self-published sources already mention sources with poor editorial control, and sources whose content is indeed user-generated, like open wikis, blogs, forums, etc. I asked on the WP:V talk page if those two sections could be combined into one, and was told I should take that discussion here. Nightscream (talk) 19:41, 7 November 2010 (UTC)

Small format question?

In the intro, The New York Times is used as an example of a reliable source (publisher). Should it be in italics if it is only being named and not given attribution for anything?

RfC on the relationship between the sourcing policies and guidelines

Should Wikipedia:Citing sources, Wikipedia: Identifying reliable sources, and Wikipedia:Identifying reliable sources (medicine) say (words to the effect of): "In the event of inconsistencies between this page and the policies, the policies take priority, and this guideline should be amended accordingly"? SlimVirgin ^{talk|contribs} 23:27, 11 November 2010 (UTC)

Comment

• Support. It's important for new editors to know that the sourcing guidelines should be consistent with the core content policies—NPOV, NOR, V, and WP:BLP—not the other way round. The content policies are strongly supported, stable, and heavily watched, whereas sourcing guidelines are not watched as much and are subject to more change.The proposed Wikipedia:Identifying reliable sources (natural sciences) failed to gain consensus precisely because people were worried about sourcing forks developing. New editors arriving at sourcing guidelines need to know that, if there's odd advice on them or they're unclear, they should be guided by the policies instead. SlimVirgin ^{talk|contribs} 23:29, 11 November 2010 (UTC)
• Oppose, there is no need for redundant and duplicate wording across multiple guideline pages, when all of those pages have a common template at the top of the page where general wording can be incorporated. SandyGeorgia (Talk) 23:46, 11 November 2010 (UTC)

• The template is non-specific and doesn't direct new editors to which content policies sourcing guidelines have to be consistent with. And apparently one of the editors who removed this from a guideline objects to it being in a template too. SlimVirgin ^{talk|contribs} 23:52, 11 November 2010 (UTC)

• Oppose in MEDRS. Oppose, I guess, in CITE. I don't care about RS. The disputed sentence currently appears in the lead of two guidelines (WP:RS and WP:CITE) and in one short essay (WP:Deletion of articles on living persons). Specifically, SlimVirgin added this line to RS on this time last year and to CITE in January of this year. It has been proposed, discussed, and rejected at other pages, including WP:POLICY—which says that all pages ought to line up with the community's actual consensus, and if that means changing a page with a "policy" template at the top rather than the one with "guideline" at the top, then so be it. See, for example:
  • Wikipedia:Policies and guidelines#Content: "When apparent discrepancies arise between pages, editors at all the affected pages should discuss how they can most accurately represent the community's current position, and correct all of the pages to reflect the community's view."
  • Wikipedia:Policies and guidelines#Conflicts_between_advice_pages: "If policy and guideline pages directly conflict, one or more pages need to be revised to resolve the conflict so that all of the conflicting pages accurately reflect the community's actual practices and best advice."
• SlimVirgin has repeatedly inserted the sentence into MEDRS over the objections of multiple editors. IMO it is unnecessary WP:Instruction creep, which particularly ought to be avoided in such a long and densely written guideline. WhatamIdoing (talk) 00:04, 12 November 2010 (UTC)
• Oppose per Sandy and WhatamIdoing. See DRY. Don't repeat bits of policy lest some bits go stale when the policy changes. This is an example of such. Nobody is arguing against "policy > guideline" but it doesn't need repeated in the third paragraph of the lead of every guideline SV edits. The solution to "policy ≠ guideline" is contentious but utterly irrelevant to the text of this guideline or MEDRS. Whatever the community decides in that regard should be documented in our procedural policies. Like WP:POLICY. Not here. Colin°^Talk 00:38, 12 November 2010 (UTC)

• Support. It has been implicit for years but many editors do not seem to understand the difference between the two and so and making it explicit with which policy the guideline give guidance for is a good idea. Far too much time is wasted on the talk page of articles when there is a discrepancy between policy and guidelines. Policy pages tend to be watched by lots of people, but guidelines frequently are not, it is quite easy for a small number of special interest editors to formulate a guideline that contradicts policy often without realising they are doing it. It is also difficult to keep guidelines up to date with changes in the nuances in policy. For all these reasons I think this proposal is common sense. -- PBS (talk) 00:40, 12 November 2010 (UTC)
  • The issue of how to resolve "policy ≠ guideline" should be discussed on policy pages. Discussing it here just complicates the issue with the question of whether it should be repeated in some or every guideline. Which brings me to the question: Which of our many guidelines should not have this text in the third paragraph of their lead? Colin°^Talk 00:45, 12 November 2010 (UTC)

• Support Many experienced editors seem to find this particular guideline/policy combination difficult to delineate, and mistakenly use the guideline to override the policy. Its a simple addition and might save a lot of discussion and contention. As well, for newer editors its a kind of "take note" statement that may alert them to the fact that policy/guidelines may intersect when with inexperience they may not have made the connection(olive (talk) 00:59, 12 November 2010 (UTC))

• Support. How realistic is this scenario?
  1. A small groups of editors creates/modifies a narrow topical guideline so it matches their POV and helps them win edit wars.
  2. This group then tries to change the underlying policies to match their POV too.
  3. The group fails, because a much larger group of eyes is watching the policy, and realizes that these changes would have sweeping repercussions that would have unintended consequences on millions of articles, not just in the narrow topical area that concerns the guideline editors.
  4. The guideline editors insist that their guideline nevertheless overrules the underlying policy, and resists any attempt to have explicit statements in the guideline noting that policies take precedence.
• Is that too farfetched? I hope so. Jayjg ^(talk) 01:53, 12 November 2010 (UTC)
  • I'd say it's about as far-fetched as this one:
    1. Consensus actually changed on some point.
    2. A long discussion shows a majority of editors in favor of changing the policy page to reflect the current consensus, but they get stonewalled by a couple of wikilawyers who WP:OWN the policy page.
    3. People get disgusted and give up, leaving the policy page as an inaccurate description of the community's Real Policy™.
  • Do you think that's far-fetched? I believe it happens fairly often, at least for short periods of time. WhatamIdoing (talk) 03:57, 12 November 2010 (UTC)
    • It's actually quite far-fetched. Insisting that MEDRS doesn't need to comply with the core content policies (which is actually what's happening here) doesn't actually reflect any Wikipedia consensus or policy. Jayjg ^(talk) 05:03, 14 November 2010 (UTC)
• STOP. This proposal has no powers. Voting on it is pointless. The proposed text is in conflict with policy. WP:POLICY does not require guidelines to be fixed to become aligned with policy: the consensus there and indeed here is that common sense should prevail and editors fix whatever needs fixed. The delicious irony is that if SV wants the above text to appear on a guideline page, then it is a policy page (WP:POLICY) that must be fixed to allow that. But any discussion to allow that text would first have to admit that sometimes policy pages are wrong and guideline pages are right. Thereby negating its own argument in a puff of logic. Colin°^Talk 08:28, 12 November 2010 (UTC)

• Comment (semi-support): Anyone who tries to address policies rationally on wikipedia already follows this - guidelines are naturally subordinate extrapolations of policy. The only place you get anyone seriously objecting to it is MEDRS, and that's only because MEDRS is a problematic page to begin with, and the editors who use it don't want to give the other editors they argue with any leverage in policy. BATTLEGROUND mentality at its worst best. I don't really think it's needed as a statement, but I can't see that it would hurt. --Ludwigs2 02:12, 13 November 2010 (UTC)

• Oppose as proposed. Changing a few guidelines like that is not a robust proposal. The WP:POLICY policy is the right place to detail the "judicial review" process for solving conflicts between policies, guidelines, etc. Tijfo098 (talk) 14:45, 13 November 2010 (UTC)

• Support: What Philip said above in their opinion is very accurate. "Far too much time is wasted on the talk page of articles when there is a discrepancy between policy and guidelines" hits the nail on the head. I will add that it goes both ways however - I find myself explaining a policy in "plain English" and sending someone to the policy only to hear "The policy does not say that" so send them to a guideline that does say "that" only to get back "That is only a guideline, not a policy, so it doesn't matter." There has to be a way to stress they must go hand in hand. While this is about RS the wording would apply across the board if added to {{Subcat guideline}}. The other option is to simply go, guideline by guideline, and "conform" them. For example I feel Wikipedia:Notability (books) is set up very clear. Read the intro, right under this: These guidelines may be considered a specialized version of Wikipedia:Notability, applied to books, reflecting the core Wikipedia policies, including the following:; there are clearly laid out links to the core polices the guideline follows. The same with Wikipedia:Notability (films). However, for example, Wikipedia:Notability (music) is not laid out that way and discussion after discussion has been had about these guidelines in relation to policies, and more and more the guidlines seem to be winning over policy. Wikipedia:Notability (people) is not laid out as clear either, and that even explicitly states This notability guideline for biographies is not policy; The proposed wording may not be the only solution, but I think it helps. Soundvisions1 (talk) 18:13, 13 November 2010 (UTC)
• Comment Per SandyGeorgia's suggestion, resolving conflict between a guideline and policy should be explained in the template on all guidelines, but the explanation should be "If there is a conflict between a Guideline and a Policy, the conflict should immediately be brought to the attention of editors at both pages, and a centralized discussion should take place to resolve the conflict." per Blueboar, below. Anthony (talk) 17:54, 14 November 2010 (UTC)
  • Wikipedia:POLICY#Conflicts between advice pages seems to already include that. Tijfo098 (talk) 11:47, 15 November 2010 (UTC)
    • I'd like to see it at the top of all guidelines. Anthony (talk) 12:45, 15 November 2010 (UTC)
• Oppose as worded, though some kind of advice to readers along the lines of "if there's a contradiction, then most times rely on what the policy page says rather than the guideline page" might be useful. However we've no basis for asserting that the text of a policy page is always a better description of Wikipedia's practices than that of a guideline page (in practice we find that it usually is, but it isn't so as of right). --Kotniski (talk) 12:08, 15 November 2010 (UTC)

Threaded discussion

Change it - Structurally it makes sense to say "Policy is superior to guideline, so in any conflict the guideline must change to conform to the policy"... but that is too simplistic and does not reflect how things really work. It is proscriptive not descriptive. How things really work is that when a Policy and a Guideline conflict, we need to bring them back into sync... but how we do so is somewhat flexible. At least one of the pages will have to change what it says, and in most situations it will probably be the guideline that changes... but not always. There are times when the conflict highlights a flaw with the policy language that no one noticed before, and when this happens it may well be that the best solution is to change the wording of the policy so it matches the guideline. Yes, this is rare... but it is possible (And, indeed, it may be that the best resolution will be to change both the policy and the guideline). The policy on conflicts between Policy and Guidelines should be: "If there is a conflict between a Guideline and a Policy, the conflict should immediately be brought to the attention of editors , and a centralized discussion should take place to resolve the conflict." It is then a matter of the community reaching a consensus as to how best to resolve it. Blueboar (talk) 23:54, 11 November 2010 (UTC)

That is, in effect, what the community has already said (see quotes from WP:POLICY above), and what SV apparently doesn't want to apply to these pages. WhatamIdoing (talk) 00:05, 12 November 2010 (UTC)

Editors are expected to follow the spirit of both guidelines and polices, but since they are all descriptive and not prescriptive, none of them has any priority over any others. The choice of "policy" or "guidelines" is guided as much by tradition and bluster as by actual precedence. It's a distinction without a difference.

In the case at hand it's clear that guidance about sourcing a particular sort of article is likely to be more useful than general guidance about sourcing general articles. So considering all this I don't see why we need to add the text in question to the MEDRS page. — Carl (CBM · talk) 01:02, 12 November 2010 (UTC)

I agree... however, given the number of MEDRS related issues that are cropping up here and at other policy pages, we seem to have at least potential conflict between MEDRS and other policies... that needs to be addressed.Blueboar (talk) 01:23, 12 November 2010 (UTC)

I trust you these are popping up - I haven't been following too closely. If there is a real conflict, we should certainly address it. But the proposed text doesn't do that; it just says "if there is a conflict ...". Now I think the reasoning behind it is that some editor or editors thinks there is a conflict, but can't actually get agreement that there is. To the extent that the proposed text is meant to short-circuit the actual discussion on the conflict, the proposed text is misguided.

In the end, what all of our sourcing guidelines say is that we should use the best sources for each article, which will vary depending on the article topic and field. If the MEDRS page actually does describe what editors of medicine-related articles feel are the best sources, I don't see what the conflict is. Maybe this is just because I haven't followed talk pages closely enough, though. — Carl (CBM · talk) 01:31, 12 November 2010 (UTC)

If there's a problem with MEDRS, we shouldn't sweep it under the rug by saying "that guideline is just a guideline anyway". That strategy has not worked in the past. The way that WP:PROF is used to justify the inclusion of articles where no biographical sources exist directly conflicts with WP:V, and yet it continues to be used that way, and the administrators don't feel like they have the latitude to ignore the !votes that rely on it. I agree fully with blueboar, if there's a conflict between policy and guidelines, we need to resolve that conflict, not just try to assert absolute superiority of policy, because that will be ignored in practice. Gigs (talk) 02:23, 12 November 2010 (UTC)

That's a great and valid point, but the question is, what do we do in the meantime? Or any meantime for that matter where there is a discrepancy. Hashing out a policy debate can easily take weeks or months and while that is happening, it helps if there is a way to say, 'well, we're figuring this out, but until we do this is what takes precedence'. That is mostly only needed on the most controversial articles, but on those articles it really helps to focus debates on sources via the guidance of policy.

The MEDRS issue is not so much about whether or not medical sources should be used, but just how strong MEDRS own internal hierarchy between systematic reviews or case study reviews or literature reviews or randomized trials or mere case studies or recent studies, etc is. And, how that plays out on articles is one issue.

The other issue is where MEDRS applies. It's supposed to address only medical claims or advice, but on some alternative medicine articles like Chiropractic it has a fairly wide reach into sections that are not necessarily medical. This is part of the general NPOV v. SPOV/APOV (academic pov) debate that has always been lurking, and which is currently being discussed at WP:SCIRS. None of these are simple issues to resolve, which makes a clear delineation of policies over guidelines helpful. Ocaasi (talk) 02:39, 12 November 2010 (UTC)

"In the meantime", editors on specific articles should work out the right thing for those individual articles, following the spirit of the policies. Our policies are only descriptive of general best practice anyway, they aren't intended to cover every situation.

In general, and this is not intended to refer to you, I have seen too many editors say "policies take precedence" when what they mean is, "my personal interpretation of the policy takes precedence, so you need to do it my way". A "clear" delineation doesn't help if there is already disagreement, as seems to be the case, about whether there is even any conflict between the different policies and guidelines. — Carl (CBM · talk) 02:51, 12 November 2010 (UTC)

Well, the fundamental problem with MEDRS is that it overlaps way too much with existing policies and guidelines. The reliability of a source is always contextual. A medical claim being backed up by a source that is reliable and appropriate for that claim is already well documented in other policy and guidelines. I see absolutely no need for MEDRS to exist, and would support abandoning it entirely. As you said, editors need to work out the right thing for the individual articles, and indeed, the right source for individual claims. Gigs (talk) 02:56, 12 November 2010 (UTC)

Why on earth is MEDRS being discussed here. Please try to stay focussed. This proposal isn't about changing the natural order of guideline and policy but is about spamming guidelines pages with bits of policy and about some of that spam actually being in conflict with policy. Colin°^Talk 08:18, 12 November 2010 (UTC)

BTW, please consider being clear in the !votes above: Are you "supporting" or "opposing" the sentence everywhere, or just in particular pages? For example, I've got no objection to it being included on this page, but I do oppose spamming it into all ~30 content guidelines. I suspect that some of the "supporters" want it included in this page, but don't necessarily think it should be included in every single related page. WhatamIdoing (talk) 03:45, 12 November 2010 (UTC)

There was also SandyGeorgia's idea about putting it into the {content guideline} template. That makes as much sense, although it puts the statement on all of the content pages, obviously. Ocaasi (talk) 03:49, 12 November 2010 (UTC)

I suppose in that case, you could split the {subcat guideline|content} into two, creating a {subcat guideline|sourcing}. WhatamIdoing (talk) 03:59, 12 November 2010 (UTC)

I am quite disturbed that editors who have a clear agenda are hijacking this proposal because it appears to suit their own ends. There are two parts to this proposal. The first is to spam guideline pages with a reminder that policy takes precedence. That agrees with WP:POLICY but we don't need reminding of this in the third paragraph of the lead of every bloody guideline that SV edits. Those who think we need reminding couldn't be more wrong. If a few editors go astray, point them at WP:POLICY. SV seems to think that page was "abandoned"[1] but the last time I looked it was one of our key policies and guidelines, which "have wide acceptance among editors and are considered a standard for all users to follow." Perhaps some of the policy lurkers/editors here need reminding that many people edit on WP just fine and just because you've got into conflict with somebody over this rule doesn't mean we need it repeated everywhere. DRY suggests we don't pointlessly repeat bits of policy all over our guidelines, especially so with contentious bits. Which brings us to the second part to this proposal, which is that we must always assume the guideline is wrong and fix that. This is in conflict with WP:POLICY and indeed with common sense. On a Wiki like this where guideline and policy can be edited by anyone, neither page is perfect. I could understand such a rule if policy pages could only by edited by high priests but that isn't so. But anyway, any discussion to change that rule should be done on WT:POLICY and since it would be a significant procedural policy change, any such discussion should be advertised on community notice boards and village pump etc.

It is deeply ironic that an attempt to change policy is being done on a guideline talk page. Colin°^Talk 08:47, 12 November 2010 (UTC)

POLICY is only "abandoned" in the sense that SlimVirgin herself stopped editing it a year ago, after most of her ideas were repeatedly rejected there, (e.g., her efforts to label guidelines like RS as being merely "advisory", and to declare that the page isn't a policy). The policy has almost 1200 people watching it, which is 165% the number of editors watching RS. WhatamIdoing (talk) 19:32, 12 November 2010 (UTC)

Oppose The relationship between policy and guidelines is well understood, and it is also well understood that the actual application of it requires interpretation. Within any type of violation of policy, some instances are more serious than others; within violation of guidelines, ditto. To take a flagrant example, the various sections of WP:NOT don't have the same degree of consensus. Some are continually challenged; some are universally acknowledged. To take another. COI is only a guideline, but drastic abuse of it is cause for a block. Really severe violation of WP:N are speedily deleted by the thousands, and it is only a guideline. This applies even to the strongest policies, BLP and copyvio. Gross copyright violations are speedied very quickly, and in a few cases, revision deleted. Small-scale violations are simply edited out of an article. Ambiguous or uncertain or challenged ones are discussed. Some violations of BLP are sufficiently harmful to warrant an immediate block and an oversight; most of them are dealt with less drastically. Our rules are already too complex, and their exact wording is too rigid. I recognize this as a very well intentioned effect to clarify the complexity, but it will just add to it. DGG ( talk ) 23:26, 12 November 2010 (UTC)

Comment Changes on policy pages have sometimes been opposed with the argument that policy pages have to suit all types of articles, have to be kept at the most general level, and that more specific advice should be given on guideline pages. That is a fair argument, but it falls down if any more specific advice provided on a guideline page is undermined by the statement that in the event of any discrepancies between guideline and policy, the policy alone, with its undifferentiated and general statements, will prevail. The sourcing standards in MEDRS for example seem quite sensible. Compared to WP:V, they privilege scientific sources over media sources, and give sourcing advice that would not apply to other topics, such as computer games or pop stars. That seems legitimate, and in the interest of the encyclopedia. I appreciate the idea proposed here, but I am also loath to pull out the rug from underneath guidelines such as MEDRS. It's a catch-22 to say that policies must remain general, that specific advice must be given in guidelines, and that whatever specific advice a guideline gives is superseded by the policy. --JN466 05:43, 13 November 2010 (UTC)

JN: it's really only a paradox, not a catch-22. guidelines should be an extension of policy: they should clarify what policy means for specific instances. This is (of course) a matter of interpretation, and the interpretations of policy are open to discussion and debate, but interpretations should never extend themselves to the point of rewriting policy - that should be done in policy directly. For (generic) example, if policy says "editors should do X when Y", guidelines can say "for this topic area, X means 'this' and Y means 'that'", but guidelines should not fundamentally alter X or Y - they should only exemplify and specify. for a concrete example, if we have a policy that say "Thou shalt not kill articles without due cause", a guideline might specify what 'due cause' is for a particular topic area, but should not ever suggest that some articles might be killed without due cause, or that some articles don't merit investigation of 'due cause'. That would violate the aforementioned policy.

There is never going to be a clean and neat statement of this, but if we want to rationalize policy then actual policy pages should be broad, universal statements of project intentions, and guidelines should expand, enhance, and enlarge on those intentions in specific instances.

As an aside, keep in mind that Wikipedia is young. The basic draft of the US Constitution took a couple of decades (from the inception of the idea prior to the revolutionary war to the codification of it in 1887), and really didn't reach its modern form for a 150 years after that. Even though Wikipedia is a much smaller project than the US federal union, it still involves a large body of people, and 10 years is hardly enough time to establish an accepted and consistent ruleset. One needs to take the long view in these matters. --Ludwigs2 07:07, 13 November 2010 (UTC)

Sure, it takes time. We tried to get WP:V in line with WP:MEDRS recently, by saying that where a body of scholarly literature exists, our articles should strive to describe the status of research in that body of literature, and that media sources are equally useful (and, where scholarly literature exists, complementary) sources for reflecting public discourse, societal reception, and current affairs (incl. current-affairs aspects of science) as well as BLPs. That seemed like a pretty universally acceptable principle. But it was blocked, and WP:V stayed with the general principle that all sources -- scholarly and media -- are equally reliable sources, with just a non-committal reminder that context determines what the best source is in any given circumstance. That sets up what are essentially POV forks in our policy/guideline system. Editors who want to privilege scholarly sources for science quote MEDRS, editors who want to cite media sources on science quote WP:V. We're not making progress that way. --JN466 08:51, 13 November 2010 (UTC)

@Ludwigs2, Just an aside, but not everyone in the world, not even every thoughtful person, would agree that the a single written constitution like the US one is the best approach. The British constitution seems to work fine in a more flexible way, arguing allowing both change and stability over a long period of time. The argument against the US approach is that it leads to what some people call "constitutional idolatry". In any case whatever one thinks about this subject, the British style of constitution is more like how WP really works, and is a better model here. PS I am not British and not American.--Andrew Lancaster (talk) 10:11, 13 November 2010 (UTC)

JN... you say: We tried to get WP:V in line with WP:MEDRS recently... I think that is what lies at the root of the issue here. There are those who feel that trying to get V in line with MEDRS is backwards... that it should go the other way... that the "correct" way to resolve the conflict is to try to get WP:MEDRS in line with WP:V (and also WP:NPOV). Personally, I think both approaches are flawed. What we need to have first is a well publicized central discussion that asks a) whether MEDRS conflicts with V and NPOV... b) looks into why MEDRS conflicts with V and NPOV (assuming it does), c) reaches a community consensus on how to resolve the conflict. Only then we can determine whether to change MEDRS or V (or both). Blueboar (talk) 16:13, 13 November 2010 (UTC)

To be honest, Blueboar, despite the way I phrased things above, the proposal I made wasn't really motivated by MEDRS. I don't normally edit medicine articles. It was more to do with the whole climate change fiasco, and my longstanding impression that our articles sometimes tend to focus too much on press controversies – or on whether someone is or isn't gay, rather than what makes them notable :) – offering more journalistic spice than encyclopedic meat. Reconciling WP:V and MEDRS would have been a side-effect of that, because as it happens, some of those same ideas are enshrined in MEDRS. But I agree with what you say: we should look at the differences between WP:MEDRS and WP:V and think about how they match up with what an encyclopedia ought to be doing. We shouldn't be having POV islands. --JN466 19:01, 13 November 2010 (UTC)

Jeez... don't get me started on the climate change articles. My opinion there is that both sides in that debate are doing their best to keep any mention of opposing viewpoints out of the articles. Talk about POV islands! In that debate, press controversy is important to discuss to give our readers a complete understanding of the topic (something the "Pro" faction does not want to do)... but that press controversy needs to be discussed neutrally as being press controversy, and not as a counter to the science (which the "Anti" faction doesn't want to do). It is never a good idea to edit policy with a specific article or set of articles in mind... and that goes double for articles on controversial topics. Blueboar (talk) 21:08, 13 November 2010 (UTC)

You're spot-on in your assessment of the climate change situation. I still think though that the rationale you are describing, and which informed our proposal, describes the differing roles of press and scholarly sources in general. We are not a scholarly encyclopedia, and we cover current affairs, popular culture and media controversies more than a traditional encyclopedia would. That's what gives Wikipedia its unique flavour, and it is what makes us so popular. But we should do both jobs: present what scholarly research has been done, and give an overview of public discourse. In some topics, like differential calculus, there won't be any significant popular discourse. For other article subjects, like PopMatters, there will be no scholarly literature. In other topics, like Ernest Hemingway or climate change, there will be both scholarly and media discourse, and we should give an overview of both the scholarly and the current-affairs aspects of the topic. I agree that we should not edit policy with a particular scenario in mind, but I think what we have here are the foundations of a generic principle that holds true for any article we might possibly have. --JN466 03:34, 14 November 2010 (UTC)

JN: when I say it takes time, I mean it takes time. MEDRS is a particularly problematic case. I occasionally work on Alt Med articles (where they overlap with my general interests in metaphysics, science, and culture), and so it's clear to me that a lot of MEDRS is designed for AltMed thumping - in those cases it's used to undercut alternative medical perspectives by tautologically defining altmed sources as unreliable sources. Not that I object to MEDRS outright - there probably are some special sourcing issues for medical topics, and some pro-altmed articles need a thumping, now and then - but it will take time to convince editors that they have to give up what they see as a tactical advantage in order to conform better to policy. (and yeah, some editors will never give up that tactical advantage under any circumstances - separate problem...). I agree with BlueBoar that a centralized discussion is needed to rationalize all three pages, I just don't know where such a discussion would take place - a subpage of Pump (Policy), maybe? But if we're going to do that, we should open it up to all the verifiability and reliability issues on project (which are spread out over 3 or 4 policy pages that I know of and a few more guidelines). Basically it would be three issues: (1) what are the abstract (project-wide) principles of verifiability and reliability that we want to observe on wikipedia, (2) what is the relationship of wikipedia to science and the scientific perspective (which is the main area where reliability squabbles come into existence), and (3) what are the limits on the ways in which guidelines can extend policy. If we settle those questions, then we can come back and rewrite MEDRS with the new understandings that we develop.

@ Andrew (and our conversational sub-thread): I only used the US constitution as an example of how long it can take to establish consistent rules, not as a model. That being said, though, the wikipedia model as it actually stands is odd. it's loosely constitutionalized in policy, but mostly runs on a very informal, discursive, common-law type model. honestly, the closest equivalent to the current wikipedia system you'll find in the modern world is non-fundamentalist Islamic law as practiced in tribal regions: a loose set of written laws that are primarily implemented through experience-based interpretations (see fatwā), passed on through stories and examples, and debated by elders (i.e. sysops) when there's enough tumult to merit asking their opinion. The major flaw in our system, IMO, is that it lacks an effective memory: there's no firm and established core doctrine as in a constitutional system (except maybe the five pillars) that keeps the project on track, no condensed and rationalized set of past 'good' decisions (as there is in common law) that can be referred to, almost no systematic effort anywhere to ensure that what we do in this case is consistent with what we did in those cases or what we will do in that case. The administrative side of wikipedia has no real history (it just has a mess of more-or-less unconnected actions, if you see the distinction I'm making), and as such it tends to wander and wallow much more than it really needs to. But that's a much, much bigger discussion... --Ludwigs2 20:17, 13 November 2010 (UTC)

Actually, most of MEDRS was designed to deal with the case study and anecdotal evidence problem. You get a person (usually a patient with the disease) who finds some "peer-reviewed article"—oops, it turns out that it was a non-peer-reviewed letter to the editor, but they didn't notice that—that has a sample size of 1 or 2, and it's several decades old, and they use it to "prove" that the medical textbooks, all the practice guidelines, and dozens of randomized, controlled trials are wrong.

I do object to people using lousy sources in AltMed articles when high-quality sources are easily available, but I do that for any kind of article, and I presume that you do, too.

As an example, the multiple chemical sensitivity websites a couple of years ago were all excited about what they claimed the German government said about MCS's etiology. They said that the German government said that MCS was a purely physical disease, and several affected people tried to stick it into our article.

Problems: The letter they were citing was from the Austrian government, not the German one. Furthermore, it didn't say anything at all about the nature of the disease: It just said that for computer coding purposes, they copied the numbering scheme used in Germany.

I can't really imagine anyone here thinking that such a source supports a claim that "Germany has classified MCS as a physical disease," even if you believe MCS to be an "alternative medicine" subject. WhatamIdoing (talk) 20:50, 13 November 2010 (UTC)

That makes sense. Like I said, I only run into MEDRS when I edit Alt Med articles; most definitely a selection bias on my part. It's still annoying though. I'm not really sure why the example you gave needs a separate guideline, however - wouldn't that sinply fall under wp:UNDUE? can't hurt to clarify it, of course, but... --Ludwigs2 01:36, 14 November 2010 (UTC)

See also MON 810. I've just corrected incorrect overgeneralizations on the [in]effectiveness of Bt cotton on the pink bollworm; The Hindu vs. Science (journal). See also: Schizophrenia: hoopla, disappointment and science journalism. Tijfo098 (talk) 18:58, 14 November 2010 (UTC)

Explaining UNDUE to a newbie is often complicated. "Proof" that MCS is a purely physical, non-psychological disease is hugely important to some of the affected people. They glance over the policy and conclude that you're saying that it is subjectively, in your personal opinion, not important enough to include—and what's your opinion, a non-expert, non-affected person, compared to theirs?

By contrast, if you point them at MEDRS and say that a blog's mistranslation of a letter to an activist isn't a strong enough source to overcome textbooks, peer-reviewed journal articles, and books published by academic presses, they might whinge about bias in the media and mainstream medicine, but they usually understand that their source really does fall on the wrong side of Wikipedia's rules. WhatamIdoing (talk) 18:06, 15 November 2010 (UTC)

You don't need a special guideline to say that. Our general guidelines on reliable sourcing are sufficient. Gigs (talk) 18:27, 15 November 2010 (UTC)

Well, you're welcome to try those conversation without referring to MEDRS. You are even welcome to resolve those disputes without reference to any policy at all, or only to IAR. What you do is your choice. Personally, I've found MEDRS to be the most efficient path to resolving sourcing disputes like this. WhatamIdoing (talk) 20:13, 15 November 2010 (UTC)

I'm sure that having a private version of the consensus guidelines that you and a few others can make say whatever you want to say does help you win arguments against fringe science people. That doesn't mean it's a good idea. Gigs (talk) 20:51, 15 November 2010 (UTC)

RfC on when apparently reliable sources are objectively wrong

Please see here for an RfC at WP:VERIFY on what to do when apparently reliable sources are objectively wrong? [2].

Thank you. --LevenBoy (talk) 15:59, 13 November 2010 (UTC)

If a source is wrong, it would then be un-reliable. GoodDay (talk) 16:23, 13 November 2010 (UTC)

Actually, no... "right" and "wrong" are not a factor in reliability. Sources often disagree, and when this happens it is not our job to determine which is "right" and which is "wrong". We note that there is disagreement, and who says what, giving them due weight... but we don't take sides in the disagreement by saying "this source is right" or "this source is wrong". See: WP:NPOV for more on this. Blueboar (talk) 17:35, 13 November 2010 (UTC)

How do we decide which to use? GoodDay (talk) 17:50, 13 November 2010 (UTC)

We use both, as appropriate. (or, if you mean which do we use as the article title, we go with whichever best fits the principles spelled out at WP:Article titles... commonality, recognition, brevity, etc.) Blueboar (talk) 18:35, 13 November 2010 (UTC)

Basically, if a consensus (or atleast a majority) is reached as to which source to use. GoodDay (talk) 18:39, 13 November 2010 (UTC)

Best keep discussion to one place I think. I have replied there. What you say works in simple cases of single incidents but it does not work where the question effects literally thousands of topics. In the vast majority of Wikipedia edits, editors are making unconscious judgments about what is right or wrong from grammatical, to stylistic/editorial, to factual according to their abilities. There is no need for us to shy away from such conscious discussion. Of course, they often get their grammar objectively wrong too but that is another issue ;-). --LevenBoy (talk) 02:28, 14 November 2010 (UTC)

I think that when a reliable source is proven wrong it should be made unreliable and taken out. For example when a correction is discovered in a news paper or magazine, that acknowledges what was printed was incorrect then the source, (atleast for the Wiki article it is used for should be declared unreliable. Another instance is where a book publishes a fact that is then countered by almost every single other source, then that book should be declared unreliable too. Elmmapleoakpine (talk) 01:07, 16 November 2010 (UTC)

OK, this is now about the more general question. Concerning the more general question then, please consider that there are sometimes more complex cases. I was involved in a debate with someone who found mathematical errors in the data table of a peer reviewed article, and then argued that the whole article should be erased from mention on Wikipedia, although the article was being cited by other mainstream sources and was apparently part of what the field knew. Context was that he started out from a position opposed to one of the conclusions of the article, and then read the article and found the problem in the data table. He cited WP:UCS. (And because he was deleting a source, not adding, and deleting a source gets benefit of the doubt these days, I found very little community interest in getting involved. Try citing WP:UCS as an argument for keeping a source!) Does your advice still apply in such a case or would you adjust your wording a bit?--Andrew Lancaster (talk) 11:05, 17 November 2010 (UTC)

Even the best (most reliable) sources can contain factual errors. If you discover that a source contains such an error, the error (or potential error) should be discussed at the talk page, and the editors should determine how to resolve the problem, by consensus. Note... the fact that a source contains an error does not necessarily make the entire source unreliable (that, of course, depends on the nature of the error).

The difficult situations are those where there disagreement over whether something a source says actually is an error or not. But the solution to the problem is the same... discuss the situation on the talk page and work towards a consensus.

To put it bluntly... the issue of what to do when a source potentially contains an error isn't something that can be "legislated" in policy, because each situation will be unique. Debates over potential errors can only be worked out on the talk pages. Blueboar (talk) 15:14, 17 November 2010 (UTC)

That sounds reasonable to me. I mentioned a complex example basically as a response to a simpler description which was perhaps suggesting that it would be possibly to write a policy on this.--Andrew Lancaster (talk) 15:26, 17 November 2010 (UTC)

Quotations

I just tried to edit this, but I'm not sure what it means:

Quotations should be cited to the original source if possible; when secondary sources are used, those that cite the original source should be preferred over those that don't.

SlimVirgin ^{talk|contribs} 19:50, 13 November 2010 (UTC)

Yes, the language is confusing. I think the passage is attempting to address what I will call the "Telephone Effect" (Ever play a game of "Telephone"? The further you get from the original, the greater you diverge from the original message). What I think the passage is trying to say is this: When quoting, it is better to take the text directly from the original, rather than taking your text from a secondary source that repeats the quote. (for example: if Winston Churchill wrote something you think worth quoting in an article, it is best to find a copy of Churchill's work and quote directly from that (and cite it), rather than relying on what some other source claims Churchill wrote). This is to prevent our inadvertently repeating a misquote or repeating a quote that has been taken out of context.

However, we also have to recognize that editors may not be able to obtain a copy of the original text, and in those cases we may have to settle for second best... which would be a source that itself cites the original (ie one close to the original on the "Telephone" chain... and since a source with citations is more likely to accurately quote the original than a source that does not include citations). Blueboar (talk) 20:50, 13 November 2010 (UTC)

Are direct quotes preferable to secondary citations?--Andrew Lancaster (talk) 21:17, 13 November 2010 (UTC)

Sometimes... sometimes not. Blueboar (talk) 21:26, 13 November 2010 (UTC)

I think this was meant to deal with a lateral bias problem. i.e., source X quotes source Y in order to assert claim Q, but we can't use that to say that source Y claimed Q, only that X claimed it. If we want to say source Y claimed Q, we need to quote from Y directly. That's just because X might be selectively quoting to produce a particular conclusion. --Ludwigs2 21:47, 13 November 2010 (UTC)

SV is probably right to raise the question then of whether this is clear.--Andrew Lancaster (talk) 07:32, 14 November 2010 (UTC)

I think Blueboar's reading matches mine. It relates to the usual best practices in source-based research. If you see that source A gives a direct quote from source B, and you want to include that quote, the usual best practice is to actually get a copy of source B, read it, and then cite it directly. In the cases when this isn't possible, it's certainly better if source A actually gives a full citation for source B, compared to some source C that just attributes the remarks to B without giving any indication where B might have said them. — Carl (CBM · talk) 02:09, 14 November 2010 (UTC)

In simple English. "Quote what the person said, not what someone says they said. Use the book that does so directly, not the book that does so indirectly and adds any interpretation or alternative context." --LevenBoy (talk) 02:35, 14 November 2010 (UTC)

No we can't actually do this, and we've been over it at length. We prefer reliable secondary sources to primary sources; the secondary sources, for example, may be quoting a different version or edition of the primary source. We can't use our own OR to try to "correct" a secondary source because we think we have a more accurate version of a quote. Jayjg ^(talk) 05:06, 14 November 2010 (UTC)

I'm not talking about "correcting" a secondary source -- we're talking about the situation where we are going to include the quote anyway. In the situation where we're going to include a quote from a person, it would be shoddy practice not to look up the actual place where the quote was made. I wouldn't let a paper go that I'm refereeing if it included a direct quote from person B that was cited to some other source by person A. If we are going to quote person B then we should be able to say exactly where person B said those words, and we should be able to cite person B directly. — Carl (CBM · talk) 13:06, 14 November 2010 (UTC)

We would definitely prefer reliable secondary sources for any interpretation or analysis of the quote (per WP:NOR). but... for the quote itself, I strongly feel we should prefer the primary source (to avoid misquotes and sources that take the quote out of context). To put this another way... when quoting, cite the original for the text itself... but cite a secondary source for any statements about that text and what it means. Blueboar (talk) 14:31, 14 November 2010 (UTC)

This would be normal practice outside Wikipedia and I dare say it sounds like common sense. I guess the question is whether the sentence will be understood correctly? It makes sense once someone explains it, but... I think my answer and also Jayjg's show how other WP issues come to some of our minds when we read it.--Andrew Lancaster (talk) 14:36, 14 November 2010 (UTC)

I completely agree that it needs to be re-written so it is clearer. Any suggestions? Blueboar (talk) 15:19, 14 November 2010 (UTC)

Add the word "direct" before "quotations"? — Carl (CBM · talk) 17:04, 14 November 2010 (UTC)

I agree with Blueboar's recent edit. If a secondary source quotes a primary source, we should cite the primary source for the quote, and the secondary source for any analysis and context. We must rely on secondary sources for additional context and analysis, but playing a game of hearsay telephone is not required when we have access to a primary source. Gigs (talk) 00:21, 15 November 2010 (UTC)

Actually, no... this is where it gets tricky... if we have not checked the text against the primary (ie we only look at the secondary source) then we should cite the secondary source and not the primary one (per SAYWHEREYOUGOTIT).

We are actually making three distinct points ... 1) we should (ideally) take the text that we are quoting from the original (or as close to the original as we can get)... 2) any interpretation or analysis must come from a secondary source... and 3) we should cite what ever it is that we used for each statement. Hopefully my edit made all this a bit clearer. Blueboar (talk) 00:54, 15 November 2010 (UTC)

Here We Go Again (Newspaper op-eds)

Just when I thought we managed to make a distinction between newspaper news and opinion, a recent revert called my attention to the following paragraph:

Some sources may be considered reliable for statements as to their author's opinion, but not for statements of fact without attribution. A prime example of this are Op-ed columns in mainstream newspapers. These are reliable sources, depending on context, but when using them, it is better to attribute the material in the text to the author.

This appears contradictory, and does not seem to fit in with the changes made to the earlier section on newspapers. The first sentence, to at least the casual reader, contradicts the second. Are they reliable for statements as to their author's opinion, or are they reliable sources, depending on context? The second appears to confer a lot more reliability than the first. Furthermore, my experience is that American op-eds often are meant to give a platform to a viewpoint that the newspaper considers newsworthy, while the content may be extremely POV and often factually questionable. I think the second sentence needs to be replaced. All American op-eds confer is notability, and then the content is as reliable as the author giving a public lecture, reported as such in a news source.

Just to clarify, I have never edited the article itself.Mzk1 (talk) 17:40, 14 November 2010 (UTC)

Almost everything is a reliable source, depending on context. I agree that statement isn't saying much, and yet manages to do it in a misleading way. Gigs (talk) 00:07, 15 November 2010 (UTC)

I've edited it. Gigs (talk) 00:16, 15 November 2010 (UTC)

"Demonstrate consensus"

A couple months ago, an edit was made that fundamentally changed the standards for scientific consensus: [3]

Prior to this, a claim of scientific consensus needed to be sourced to a reliable secondary source that actually declared that a scientific consensus existed. Now editors are free to "demonstrate" a scientific consensus by citing a bunch of disparate sources. This kind of meta-analysis is a violation of WP:NOR. We shouldn't claim that a scientific consensus exists unless someone else says it exists.

The edit was reverted, in part, but the "demonstrated" verbiage remains. I think we should revert to the left side of the above diff, and again require sourcing for claims of scientific consensus. Gigs (talk) 21:31, 15 November 2010 (UTC)

I agree, but that's question of proper citation and WP:OR. Strictly speaking calling someting scientific consensus without some source explicitly saying so is WP:OR by the author. In relatively undisputed cases such an approach might be acceptable, i.e. assumung the author is well versed in the subject and not POV pushing and essentially simply stated a common perception in the field (without having a literal external source at hand). But at least in disputed cases a source explicitly stating a scientific consensus is absolutely required. The current formulation could still be unserstood in a correct manner, i.e. one could read it as "requires reliable sourcing that demonstrates the consensus" = "a reliable sozrce needs to explicitly talk of a scientific consensus". However the original formulation was more explicit and less prone to potential misunderstandings. I'm fine with restoring the original version. --Kmhkmh (talk) 22:21, 15 November 2010 (UTC)

I disagree that "Now editors are free to "demonstrate" a scientific consensus by citing a bunch of disparate sources" because the "making blanket statements based on novel syntheses of disparate material" sentence specifically prevents that. This whole guideline paragraph reads like fallout from some great global warming dispute or whatever, rather than advice that is useful generally. Most facts are uncontroversial, or are claimed to be controversial only by a tiny minority. So, IMO it is usually best to avoid making explicit statements that "all or most scientists or scholars hold a certain view" and just state the view, citing the best source one can find. Those who believe they are not only right but in the majority tend not to say so: they just boldly make their case. It is those who know they are in the minority that make a song and dance about how many believe this and how many don't. Colin°^Talk 22:42, 15 November 2010 (UTC)

I agree that such statements are not often appropriate, which is why I think we should go back to the older, more explicit, wording. Do you think anything was wrong with the old wording? Why should we introduce a new concept of "demonstrating a consensus exists"? It sounds a lot like original research to me, even with the second sentence still in place. Gigs (talk) 23:16, 15 November 2010 (UTC)

I don't think this text is causing any problems (has anyone encountered a problem in a live article over this?), but it may not be optimally clear. We have two possible issues:

• You read a bunch of sources, you learn from the sources what the mainstream views are, and you write the article to correctly reflect those views: This is source-based research, which is explicitly endorsed by NOR, and all good editors do it. (POV pushers read all the sources, learn what the mainstream views are, and write the article to reflect their POV anyway.)
• You want to include a sentence that directly says "The consensus is ____": This is the (sole) point of this section. You normally need a source that directly says "The consensus is ____", or something remarkably close to that (e.g., some authority group says "After studying the issue, we have adopted the following guidelines"). It might be clearer to say "The statement that all or most scientists or scholars hold a certain view requires reliable sourcing that demonstrates the consensus directly says that all or most scientists or scholars hold that view. WhatamIdoing (talk) 20:54, 16 November 2010 (UTC)

• I've incorporated your suggestion, with a minor modification to add the word "Otherwise" in front of the next sentence. That satisfies my concern. Gigs (talk) 02:10, 17 November 2010 (UTC)

• Agree with Gigs. --JN466 23:01, 16 November 2010 (UTC)

I agree with the edit... but a word of caution is required... this is where IRS overlaps with both NPOV and NOR ... so we should double check that what we say here is in sync with what is stated at those policies. (I think it is... but we need to make sure... and if it isn't, then we need a centralized discussion to achieve a true consensus). Blueboar (talk) 15:29, 17 November 2010 (UTC)

Magazine articles?

I'm sure you all have discussed this, but I'm not finding magazine articles in the guidelines. They seem to be problematic, because most of them are "creative nonfiction", and they are not even intended to provide an NPOV (if they did, they'd be too lacking in drama to sell). Also, MastCell recently recommended to me a book that documents magazine articles (and news, but there's not much anyone can do about that) being commissioned or "planted" as a public relations tactic. Of course the whole subject of "public relations", or PR, is relevant to the reliable sourcing debate, but again, not much we can do about it - one person's despicable "obvious PR plant" is another person's admirable "reliable secondary source", and a shouting match isn't going to change that. So, let's just focus on the status of articles, that give the impression of facts, in reasonably respected magazines. Thanks, Postpostmod (talk) 14:28, 18 November 2010 (UTC)

Can you give some examples of the kind of articles you mean? There are, literally, thousands of magazines in English alone that do strive in many if not all articles to give an NPOV, and there is no problem in principle with using them as sources - I wonder if it is a specific subset of magazines you are thinking about? Barnabypage (talk) 14:43, 18 November 2010 (UTC)

Exceptional individuals wanted for challenging two-year assignment

You are:

Y an effective communicator with a sound grasp of policy;

Y able to see all aspects of a problem and find solutions;

Y courteous, disciplined and open-minded;

Y able to deal calmly with trolls, bigots and editors with issues;

Y able to make up your own mind under stress.

If you can answer "yes" to most of the above, you are probably arbitrator material. Learn more about standing in the upcoming election. But don't delay, nomination close very soon!

Tony (talk) 16:37, 19 November 2010 (UTC), for the election coordinators

• PS: YYou must also be able to prove your real name (with a copy of your passport) to "The Office" in case any litigation as a result of your actions arises.  Giacomo  17:54, 19 November 2010 (UTC)