Wikipedia talk:Reliable sources/rewrite

Aspects of reliability

I am a bit confused about these points:

Replicability of the material
Corroboration
Age and rate of change of the subject
Persistence of the source, particularly when an online source is used.
Confidentiality

Would you mind explaining? — Armedblowfish (talk|mail) 00:15, 6 November 2006 (UTC)[reply]

I've now amplified on each of my main points, they still need some refinement in terms of content and presentation but the structure is there. I'm really not sure what is meant by the remaining three, although my assessment is that they are probably already covered under the existing principles. The persistence issue needs a complementary section on convenience linking, since that's mainly what it affects. Personally I prefer dead tree sources rather than online because I don't trust online to still be there when I follow the link through, linking to online editions of dead tree is a useful compromise though.ALR 08:50, 9 November 2006 (UTC)[reply]

OK, I'm not sure about these. I think they're already dealt with and become specifics during discussion of examples:

The following may also be helpful, but are not as broadly agreed upon:

Declared legal liability (sometimes by implication)

Degree of establishment

Established history (track record)

ALR 08:38, 10 November 2006 (UTC)[reply]

I'm not really sure either, but I added them based on something Terryeo wrote. So I left a note on Terryeo's talk page. Armedblowfish (talk|mail) 00:39, 11 November 2006 (UTC)[reply]

Generally a good idea

In keeping with my usual preference for reliance on principles rather than hundreds of rules, I think this has potential. I'd like to see some more examples of sources which might be accepted and rejected, with reasoning, perhaps as a subpage. Guy 13:14, 6 November 2006 (UTC)[reply]

I was hoping we could limit this page to material we could agree upon enough to justify not demoting WP:RS to an essay. (Possibly, if we restrict it to material with a sufficiently high degree of consensus, we could promote WP:RS to a policy. But perhaps that is too much to hope for.) Somehow, specific application of WP:RS does not seem like something we are likely to achieve a high degree of consensus on, and even if we could, it would likely only be on extreme examples.

I, personally, think that whether or not a source is sufficiently reliable for inclusion has a lot to do with context. Is it reporting within it's area of expertise? How extraordinary is the assertation it is being used to support? How does it compare to other available sources? How necesary is the assertation for the completeness of the article (especially when it comes to WP:NPOV)? Is it being used to back up a negative assertation in a biography of a living person?

On the other hand, it might be a great essay topic (a "subpage" as you say).

— Armedblowfish (talk|mail) 19:26, 6 November 2006 (UTC)[reply]

I would agree that putting examples on a separate page is the best place for them. JYolkowski // talk 23:34, 9 November 2006 (UTC)[reply]

Use of sources

At first I changed this [1] some, because, at least in my opinion, links used soley for the purpose of providing futher information for readers are more Further reading/External links than sources. Then I thought it might go better in WP:CITE anyways, so I suggested it here. Armedblowfish (talk|mail) 18:58, 6 November 2006 (UTC)[reply]

I'd tend to disagree about not considering why sources are used in a discussion about reliability. The point makes clear why reliability is important, so potentially the section title needs changed. Whilst configuration control is an important issue, and to that end there should be minimal duplication, the reasons for reliability need contextualised without sending the reader elsewhere. It may be appropriate to include the section and provide a main article connection to CITE.

Also external links should still be held to a similar standard of reliability, otherwise it becomes too easy to avoid explicit referencing, because the source isn't reliable (or wrong), and inserting the inaccurate material in the article anyway under a different header. Externals are generally discouraged, that discouragement is enhanced through enforcement of reliability standards.

ALR 20:22, 6 November 2006 (UTC)[reply]

I believe External links / Further reading are beyond the scope of WP:RS... WP:EL discusses quality guidelines for external links.

As for why reliability is important, I just tried to write something about that, although you are free to delete it if you don't like it.

— Armedblowfish (talk|mail) 23:50, 6 November 2006 (UTC)[reply]

I disagree that "External links" and "Further reading" are beyond the scope of WP:RS... Guidelines and policies often support each other, and WP:EL and WP:RS are not mutually exclusive. While we might not hold such sources up to the same standards as sources used to back specific statements of fact or opinion in the main text of the article, they should meet some level of reliability. We should be working with the WP:EL editors to address this in a consistant way. (And this reminds me ... we also need to discuss reliability for convinience links, a similar but distinct issue.) Blueboar 21:14, 7 November 2006 (UTC)[reply]

Part of the problem is that a source needs to be sufficiently reliable relative to the statement you are using it to support. An obvious example of this is a biologist, discussing either biology or nuclear physics. A biologist is more reliable when talking about biology than nuclear physics. Another consideration is how extraordinary the statement is. You would need a more far more reliable source to say that Ghandi was a serial rapist (and you'd still have to attribute the statement) than to say that Ghandi gave food to the poor.

When the link isn't being used to support a certain claim, you can't make any such judgements.

Another problem is that there is probably about as much controversy about WP:EL right now as there is on WP:RS.

Convenience links really don't have enough consensus to be discussed in a guideline one way or the other, but we do have an essay.

— Armedblowfish (talk|mail) 21:37, 7 November 2006 (UTC)[reply]

However the point does not mean that we should not consider any standards of reliability with respect to external links. As I recall, although I didn't look, I did caveat the statements, that I inserted and you removed without discussion, with the notice that the level of reliability needn;t be as explicitly high. Context is important, but ELs still have a context.ALR 21:42, 7 November 2006 (UTC)[reply]

Sorry, I thought this was the discussion about it. The fact that Blueboar agrees seem to indicate some consensus, although more participation would be helpful. I still think reliability / quality of further reading should be discussed in WP:EL, but perhaps I'm alone on that. Armedblowfish (talk|mail) 02:16, 8 November 2006 (UTC)[reply]

I think you're misunderstanding what consensus actually is, although that is common throughout Wikipedia. What we have here is a majority view, not a consensus. I'd agree that more participation here would be useful but since the RS page was protected, and not marked as such, three days ago there has been little activity. I'd also agree that the EL community probably need to consider reliability and once this is in shape they can use it, and potentially refer to it. mutually referential guidance is a Good Thing (tm)ALR 08:47, 9 November 2006 (UTC)[reply]

Well, at least we have consensus that external link reliability should be discussed... somewhere. Armedblowfish (talk|mail) 00:18, 11 November 2006 (UTC)[reply]

Armedblowfish, what that comment about participation aimed at me? OK, I'll try to participate more. :>) In the case of "further reading"... I think both WP:EL and WP:RS should discuss this ... hopefully in a way that support each other. Blueboar 19:25, 10 November 2006 (UTC)[reply]

Your participation is most welcome and appreciated, but at that moment I was referring to the participation of a relatively large number of people. I agree that quality guidelines for external links should address some sort of reliability, but I haven't thought much about the details, other than that it is somewhat different situation. I guess it should be flexible enough that editors needn't feel guilty about being bold and adding external links they aren't sure about. Maybe someone could bring it up over there when the discussion on WP:EL stablises, or write an essay, develop the idea, and see if many editors are convinced. Armedblowfish (talk|mail) 00:18, 11 November 2006 (UTC)[reply]

Shuffling the deckchairs

IT would be prudent to avoid shuffling the order that the various points are listed in until after there is actually some meat on the bones. Personally I'd say that explicit derivation and replicability are more important than source author bias, because bias can be mitigated through derivation and replicability.ALR 20:24, 6 November 2006 (UTC)[reply]

Okay. Come to think of it, declaration of sources / corroboration is sort of parallel to editorial oversight. It might be a sort of open editorial oversight, or a way of allowing the reader to do the oversight themselves. Armedblowfish (talk|mail) 00:10, 7 November 2006 (UTC)[reply]

I've re-sequenced the order of aspects of reliability, hoping to arrive at which is most important for reliability, which is second most important, etc. It is intuitive that some aspects of reliability are less critical than others. Terryeo 14:34, 11 November 2006 (UTC)[reply]

Convenience links

Should we really mention this highly controversial topic? (Or, at least it was controversial last I checked.) I, for example, think that convenience links should be distinguished from intermediate sources, where in the case of a convenience link an editor did verify the information based on looking at the original source, and in the case of an intermediate source, no editor looked at the original source. I don't know how many people agree with that, but there are a wide variety of other opinions.

I started a list of convenience link related discussions on Wikipedia talk:Convenience links.

Armedblowfish (talk|mail) 00:29, 11 November 2006 (UTC)[reply]

Given that this version of RS trys to avoid over-defining things, I think we should avoid going into the differences between "convenience links" and "intermediate sources" ... perhaps we should just call it "Linking to web-hosted copies of printed documents". Whatever we call it, the point is that providing a link to a page that hosts a copy of a document is OK as long as we can reliably assume that the copy is a true one. If the host site is reliable, then chances are that they are presenting a reliable copy. In this case there is no need for an editor to go digging for the original to double check it. However, as the reliability of the host site declines, so does the validity of assuming that any copy they host is reliable. At some point in this progression, an editor wishing to link to it should have to check it against the original.

I think we need to address this issue, but hopefully we can do it in a non-controvercial way (if we can do this, it may help end some of the controversy)... I attempted to do this by leaving it couched in the type of language that ALR has couched the rest of reliablility... noting that that there is a range of reliablility to such links, and that one should use the most reliable while avoiding the least reliable. Blueboar 01:13, 11 November 2006 (UTC)[reply]

I think that's really it, convenience links can be used but should be managed in a similar way to the rest of the issue. They may not properly reflect the source as cited. Catching up.ALR 22:08, 12 November 2006 (UTC)[reply]

Firstly, sorry for the long response time. My personal opinion is that the problem with the current version of WP:RS is the inclusion of material we have a low degree of consensus on, and is thus unsuitable to be guideline. I think the difference is important - it is the difference between citing the web-hosted copy, and merely linking to it for other reasons. While ALR did take a good, middle-ground approach, I think I would be happier if we talked about intermediate sources, not convenience links, and just linked to the convenience link essay. Armedblowfish (talk|mail) 15:54, 18 November 2006 (UTC)[reply]

Impact of revisions at WP:EL

(also posted at WP:RS) Please note that WP:EL has been revised... there is one section that may impact this guideline or the various rewrites being drafted:

Links normally to be avoided

Except for a link to a page that is the subject of the article or is an official page of the subject of the article, one should avoid:

2 Any site that misleads the reader by use of factually inaccurate material or unverifiable research. See Reliable sources.

This is the first time that I can think of where factually inaccurate material has been discussed in such a explicitly negative way in a guideline. OUR guideline certainly doesn't (nor the draft rewrites that might replace it). Given this, Perhaps the time has come for us to address the issue of sources that contain factually inaccurate material as well.

I know that we have to be careful not to contradict the priciples expressed in WP:NPOV... but should we state something similar to what is on WP:EL. I.E. any site that misleads the reader by use of factually inaccurate material or unverifiable research should not be considered reliable. Any thoughts? Blueboar 19:14, 17 November 2006 (UTC)[reply]

I think this is inconsistent with our verifiability, not truth policy. It is not Wikipedia's place to judge whether or not something a source (or link) said is "true" or not. And it's not the sources that need to be verified - the sources are the verification. Our only way of deciding how much faith to put in something a source (or link) said is reliability. If all of the reliable sources agree on something, we call it a fact. It might be wildly untrue, but at least we'll be wrong with the best. A very unreliable source might say something true. But we ignore them, because that's a minority point of view below our notice. Armedblowfish (talk|mail) 15:10, 18 November 2006 (UTC)[reply]