Wikipedia talk:Reliable sources and undue weight

Prior discussion
Please feel free to add to this list.


 * Wikipedia talk:Reliable sources
 * Wikipedia talk:Reliable sources/rewrite
 * Wikipedia talk:Neutral point of view
 * Wikipedia talk:Attribution/Community discussion/Archive 1
 * Wikipedia talk:Attribution/Community discussion/Archive 1

Worst policy in this wiki
In my view, this is the worst idea we pursue here. Because we have defined only certain sorts of sources as "reliable", we end up with articles that are heavily slanted to the point of view of that type of source. Because many of our sources are, or derive from, the mainstream, Western, liberal media, in particular as reflected across the interwebnets, our articles strongly reflect the biases of those media. By saying we must not give "undue weight" to minority sources, we entrench the asymmetry of information on conflicts or disputes in which the majority is strongly supported by the MSM. Whether you consider this a problem is probably an outcome of whether you share those biases. To some extent, most contributors will, because they are Western and reasonably liberal. This sentence in particular is worrisome:

Still, we believe that the reliability of sources is usually the best way of judging due weight.

I think a problem is that we do not distinguish different types of article and different means of sourcing. If you are writing about a scientific subject, it seems clear that scientific papers are more "reliable" than whackjobologists' websites. Science is largely evidence based, after all, and the evidence should appear the same to anyone doing science anywhere, from any viewpoint. But transferring the cachet of scientists to other academics is problematic. Even more so is to transfer it to the media. Is the NYT more "reliable" because it is a bigger outfit than a small indy news-gatherer? Is it more reliable on local issues than someone who writes a website? We say yes. Basically, our policy on sourcing is heavily biased not just to "English-speaking" media but to money.

I note a particularly bad outcome from this idea. When we are writing about people, we can source things only to "reliable sources". Because we have defined that as "newspapers" so far as most people are concerned, our articles tend to focus on the "newsworthy". Well, being nice to your mum doesn't make the news. Killing her does. This leads to a tendency to sensationalise biographies, and robs them of the timeless quality that is a feature of good encyclopaedia articles.

I recognise this is pissing in the wind. No one is really all that interested in a "neutral" encyclopaedia. What we're aiming for is one that presents the Western, orthodox view in not too biased language. Doesn't make for such a snappy policy title though. But to my mind this idea is the tool of ensuring a lack of neutrality, and this area needs more thought.Grace Note 06:04, 26 March 2007 (UTC)


 * I'm confused. Are you saying:
 * WP:ATT is the worst policy in this wiki (the only policy that seems relevant, unless WP:V is on your mind)
 * WP:RS is the worst guideline in this wiki
 * Wikipedia:Undue weight (sources) is the worst proposal in this wiki
 * ? &mdash;  SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 07:25, 26 March 2007 (UTC)


 * I am saying that "undue weight" is the worst idea in this wiki, and I said it clearly enough. Do you really have nothing more to add than angels on a pinhead parsing of what I say? In any case, the policy I am discussing is the one this article begins with, which is the one that "undue weight". Perhaps you are not familiar enough with the policies to recognise it: this is the section in question. Grace Note 00:00, 27 March 2007 (UTC)


 * Why do you assume I wasn't asking an honest question? I genuinely had no idea what the exact scope of your critique was.  I'm familiar enough.  It simply wasn't clear to me if you were upset with the real separate policy's present formulation of it, WP:ATT/WP:ATTFAQ's handling (or failure to) of the issue, or this present page's attempts to reformulate it.  I don't think that's a question that warrants a response like yours, but thanks for at least making your point clearer. &mdash;  SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 00:05, 27 March 2007 (UTC)
 * Actually I did confuse myself a little! &mdash; SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 00:07, 27 March 2007 (UTC)


 * I don't assume you weren't asking an honest question. I applied Hanlon's Razor. I don't like the existing statement of policy and I like this restatement even less. Don't get me started on WP:RS, which is terrible. I'll give a for instance. I'm interested in creationism and intelligent design. A fantastic source in that area, written by knowledgeable, credentialled authors, is the talk.origins website. However, it's excluded by our policy. If one of those guys has an article in the NYT, we can cite him, but because he uses gasp a website, well, he's out. I have seen this view expressed. It's entirely wrong. Here's a dispute that is largely fought out online, not in the press, and we cannot hope to cover it adequately without reference to the many online resources: not just talk.origins but the Discovery Institute site and so on. The worship of newspapers is entirely misguided in my view when "editorial oversight" amounts to "write whatever you like, Judy" and when Wikipedians cite comment as though it was news because "newspapers are good sources". What worries me and what I think we need to think about here is that some -- and the proposer here is one -- want to abstract judgement of a source from "is the source contextually good?" to "is the source of a kind that we consider reliable?". The policies as they stand -- both NPOV and RS -- take this abstracted view. It has its place: I agree that news from a newspaper is likely to be more reliable than news from a blogger, if only because the quality of bloggers on the whole is lower than the quality of newspapers. But some bloggers are more reliable in context than any newspaper. I know that my kind of thinking leads us into a minefield, where there are no easy policy prescriptions, but it seems to me that rather than bickering over whether we should have one, two or three policy pages, we could more profitably be thinking about whether our policies do what they claim to in the first place, and we could traverse that minefield, or admit that we aren't interested. Grace Note 02:45, 27 March 2007 (UTC)
 * That I can agree with. Made a very similar and related argument the other day (and also raised the issue at both RS and ATT; I the threads with "goes too far" in the title.  I saw someone once attack the use of Jakob Nielsen's usability blog as "unreliable" when it is fact one of the most authoritative and current (and source-cited) sources on the topic in the world.  Boggles the mind really. &mdash;  SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 03:00, 27 March 2007 (UTC)


 * I am not sure what else would be better. Given the heavily slanted demographics of Wikipedians, it might be easier to find an Indian newspaper than an Indian Wikipedian interested in any given topic.  Or a Chinese newspaper if you want something AltaVista can translate.
 * This is only an essay, to be upgraded to a proposal when it's written better. If you can rewrite anything to make it less objectionable to you, please do so.  In any case, I think the idea, if not the current wording, will make this suggestion more useful in the search for neutrality than an RS written to be consistent with WP:V/WP:ATT.
 * — Armed Blowfish (talk|mail) 11:59, 26 March 2007 (UTC)


 * Not being sure what is better does not mean nothing can be better. My point was exactly that the "slanted demographics" create a systematic bias that runs a lot deeper than "you can only find English-speaking newspapers", including the notion that newspapers are good sources for encyclopaedias!
 * I do not want to rewrite the proposal because I think it, and the section of the NPOV policy that it expands on, should more or less be binned. I'm sorry, I do not mean disrespect to you, but I explained why above. I don't think this suggestion helps "the search for neutrality". As I discussed above, I think it makes neutrality impossible. Grace Note 00:00, 27 March 2007 (UTC)

I haven't looked deeply into the text here, but some of Grace Note's objections seems to make sense. We need to broaden the concept of what a reliable source, and it might have to be modulated depending on the topic. But in principle, I think the only workable way of determining how much weight should be given to something is by judging the source. The other option would be to make judgements based on the content of the material, and I think that's more problematic. Could you specify, what exactly you object to, and how else we should judge weight, if not by the nature of the source where it is stated? --Merzul 17:29, 26 March 2007 (UTC)


 * I completely agree that we should broaden the concept of what is a reliable source, and modify that depending on the topic. If you feel this proposal does not do that, please help improve it.  Currently, the explanation about judging reliability is in Reliable sources/Undue weight, and it originally came from Reliable sources/rewrite, a former attempt to overhaul WP:RS.  Note the expertise criteria in particular.  Thanks, Armed Blowfish (talk|mail) 17:38, 26 March 2007 (UTC)
 * Armedblowfish, are you following the discussion about renaming reliable sources to "acceptable sources", I think that's an interesting discussion. I added something along topicality criteria here, the examples are perhaps bad, they are more to illustrate the point. I have no idea what criteria popular culture editors apply. --Merzul 20:33, 26 March 2007 (UTC)
 * As I noted against your proposal (wherever it was you proposed it -- so many pages!), I think "allowable sources" is better. "Judging reliability" is an eye of the beholder thing. I think source X is reliable, you think it is partisan, she thinks it's the work of the Devil. That's kind of how it works. In my view, the concept of NPOV needs further refining, or at least restating in terms that make sense. Either we are creating a representation of the consensus Western view of the world or we are creating a truly neutral encyclopaedia. If the former, we should not pretend to be the latter; if the latter, we need a major rethink of what "acceptable sources" are and why. However, as I noted, because we are in fact the former pretending to be the latter, nothing will change. Grace Note 00:00, 27 March 2007 (UTC)
 * It is my hope that something like this would help people in such disputes reach compromises. Firstly, it gives people criteria with which to judge sources and support their arguments about how reliable it is.  Secondly it allows for middle levels of reliability, which can be included but must attributed and not given too much weight, which allows for compromise.  I agree that systemic bias is a problem.  However, I feel the systemic bias of sources, if we take the time to look for the wide variety out there, is less than the systemic bias of the demographics of Wikipedians.  —  Armed Blowfish (talk|mail) 02:51, 27 March 2007 (UTC)


 * I do not agree at all, because I think the judgement of "reliability" entirely depends on those demographics. If you read this, you might have a better idea of where I'm coming from. I think that further steps down the road of codifying certain sources as more "reliable" than others are a bad idea. Grace Note 01:26, 28 March 2007 (UTC)


 * I don't think "reliability" has something to do with demographics, at least not in the physical sciences. In the case of "undue weight", I just encountered this issue very clearly. I expanded an entry under "viral resistance" for an HIV/AIDS drug, and I cited a paper. It was deleted soon after because of "undue weight". Well, wikipedia is indeed not a paper, and here lies the problem. A single paper or one finding does not constitute a truth because we need a number of papers from different corners (or institutions) of the world which arrive at the same conclusion (in medicine and science we also need clinical evidence in addition to in-vitro, in-vivo etc. findings). Only then, bit by bit, a truth is established. There are other examples. Once I encountered the statement that certain hop varieties contain morphine, I deleted or at least relativised it. (But I believe that there is one such paper claiming this finding). Of course, we need a number of papers that come to the conclusion that some hop varieties contain morphine. For now, the established truth is that hop does not contain this narcotic. At least that's how it works in the physical sciences. In the humanities, this might be different though. So, I definitely agree with the "reliable sources and undue weight" policy on wikipedia, at least for the sciences. The question is really what is a reliable or acceptable source and how many of them are needed to arrive at the "truth" of something. Whether we call it "reliable" or "acceptable" is more a question of semantics. Usually "reliable" means peer-reviewed within the context of academic principles. Other examples such as "cold fusion" illustrate this issue as well. Osterluzei (talk) 17:08, 5 April 2011 (UTC)

RfC on RfC :-)
I'd be interested in comments on an article RfC I responded to here, with further elaboration there. I think I got this spot-on (after the misleading nature of the RfC came out), as did User:David Lyons responding, at the first link, a little below me. If we didn't get this right, then either we are insane or RS is way more screwed up than I thought (I think both my and David Lyons's logic is correct, which would mean that RS's isn't), which has serious consequences for WP:N which depends heavily on the former. And of course for NOR/ATT in as much as they share RS's alleged sins. &mdash; SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 03:00, 27 March 2007 (UTC)

NPOV
Is there anything in here that isn't in the NPOV policy? SlimVirgin (talk) 20:54, 26 March 2007 (UTC)
 * Nothing that I can see. ≈ jossi ≈ (talk) 21:49, 26 March 2007 (UTC)
 * It seems to be explanatory, rather than inventing new policy. Grace Note 00:02, 27 March 2007 (UTC)


 * I thought it did.... —  Armed Blowfish (talk|mail) 02:14, 27 March 2007 (UTC)
 * Did what? Invent new policy? Grace Note 02:34, 27 March 2007 (UTC)


 * No, this is supposed to be an essay, and eventually a proposal. If successful, maybe a guideline, but never a policy.  : )  I thought it said things that weren't in the NPOV policy.  —  Armed Blowfish (talk|mail) 02:42, 27 March 2007 (UTC)
 * Would you like to pinpoint what exactly you think they were? Grace Note 02:45, 27 March 2007 (UTC)


 * I think I started synthesising things after the ToC.  This section  is from a a former attempt to overhaul WP:RS, although of couse it has been changed a bit since it was moved here.
 * I've been trying to stay consistent with the spirit of WP:NPOV, but I think there are new things there.
 * — Armed Blowfish (talk|mail) 02:57, 27 March 2007 (UTC)
 * Yes, okay. I've grasped that you think that it's new. But would you mind perhaps laying out in bullet form what you think is new? Because it's a strain to have to refer back and forth to the policy and your essay and try to work out what is different. Grace Note 03:17, 27 March 2007 (UTC)
 * Hmmm... I'll try to look at both of them side by side and figure it out more definitely. —  Armed Blowfish (talk|mail) 03:28, 27 March 2007 (UTC)


 * Good, because this will never fly as policy. FeloniousMonk 03:30, 28 March 2007 (UTC)

Requested move
Reliable sources and undue weight → Reliable sources/Undue weight — Present name will confuse editors that this is a new competing guideline proposal rather than an attempt at a redraft. Redrafts belong in /Subpages of what they are redrafting. Tried the move, but redir already exists and has been futzed with, so the move was blocked. &mdash; SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 23:53, 26 March 2007 (UTC)


 * You are confused about which policy is being addressed here. Grace Note 00:01, 27 March 2007 (UTC)


 * Oh, duh. I'll fix that. &mdash;  SMcCandlish &#91;talk&#93; &#91;contrib&#93; ツ 00:07, 27 March 2007 (UTC)