Wikipedia talk:Attribution

• Also see Wikipedia talk:Attribution/popculture

1 2 3 4

A critique

A good policy should have the following features:

• A name that makes clear what it is about (very important as most Wikipedians will never actually read the text of the policy);
• Brief and to the point (so that it can be read quickly - eg the policy box in WP:V);
• May be rule or principle based;
• Be expressed in positive terms - ie say what behaviour, content, etc is expected/wanted (note, in addition, it may or may not be accompanied by text explaining what is banned);
• Be accompanied by some brief explanation of the policy itself, but no more than 1,000-1,200 (no-one will read beyond that unless looking for loopholes, and a text that size will have loopholes;
• Be referenced to guideline pages that examine how certain aspects of the policy are interpreted in practice.

At present, this proposal is not yet meeting all of these criteria.

In particular, I feel the following needs to be addressed:

• The name needs to change so that it is clear what the policy is from the name alone. WP:Verifiability is clear. As is WP:No personal attacks. WP:Attribution is not. Maybe this proposal could eventually be renamed WP:Verifiability. Alternatively WP:Justify your edits, or WP:Always provide a source. No doubt others can come up with better names. But please bear in mind, the vast majority of Wikipedians will not ever read more than the policy title.
• The policy itself is not brief. It currently takes up 6 pages. Maybe expand the "policy in a nutshell" bit so that it covers everything and becomes the policy itself. The remainder of the text can then be a discussion about what the policy means in practice - pretty much like WP:V is ordered now.
• There is still too much emphasis on a prohibition, "no original research". The text would read more positively if that were the fourth point under "Key principles" and if a way could be found of removing "no original research" from the header. (This is a presentational point, I'm not suggesting not referring to "no original research" at all.)
• The page as a whole needs to come down in size. Mention only the key points here - few will read it all. By all means supplement it with a reference to other guideline or discussion pages for the few who do want to read more.

I recognise that no-one is saying this proposal is anywhere near complete. And my comments above are necessarily geared towards what I feel could, and should, be improved. I am certainly a supporter of reducing the number of WP policies, and hope this principle can be adopted further with other policy areas, jguk 18:10, 30 October 2006 (UTC)

The page needs to be copy-edited and tightened as much as possible before being presented, but comparing it to the policy box of V, which is 50 words long, isn't realistic. And "verifiability" isn't clear at all and has led to a lot of confusion, with people thinking they need to check whether things are true, rather than published. The NOR thing is crucial: no policy review that omits it has a chance of succeeding. NPOV and NOR are the backbone of Wikipedia. SlimVirgin ^(talk) 18:21, 30 October 2006 (UTC)

A short title that clearly communicates? Maybe Establish Credibility. After all that's the whole point. (I actually like "Says who?" best; but that's just me.) WAS 4.250 18:36, 30 October 2006 (UTC)

I was half thinking we could call it WP:NOR, if not WP:ATT. Because that's the point of it: don't add your own research, add other people's. SlimVirgin ^(talk) 18:42, 30 October 2006 (UTC)

I don't like WP:NOR as it doesn't, in the title, say what to do. After all, I could (theoretically) remove all the references to all the featured articles and they still would comply with the three words "no original research". Establish credibility does give a better impression as to what it's all about. I'd like to hear other suggestions too though, jguk 18:50, 30 October 2006 (UTC)

Using questionable or self-published sources

I would like to discuss a part of this section. One of the exceptions to using questionable or self-published sources is "Material from questionable or self-published sources may be used as primary source material in articles about that source's author". I understand the desire to allow non-disputed biographical information to come from the person who may know best. But, I think the statement offers too much freedom. I have seen several cases of this occur. In one article in particular it makes a statement about an event that occurs, suggesting that it was covered by a national newspaper and magazine, but then offers the self-published web site of the person the article is written about as the source, rather than the newspaper or magazine cited. Of course every single fact about a person should not need to be cited, but major events covered by recognized media are a better source. It is my belief that information from a persons self-published web-site, or from speaking with the person directly are basically no different that original research, and should be used extremely rarely, if at all. My concern is that this sentence gives free license for a biographical article to be written about someone and the large portion of the content will be from the person themselves with no independent verification. Atom 19:05, 30 October 2006 (UTC)

Something along the lines of this sentence make more sense to me: "Material from questionable or self-published sources may be used as primary source material in articles about that source's author only rarely, under exceptional circumstances, when other references for non-disputed events cannot be found elsewhere." Atom

Threshold

Instead of "…The threshold for inclusion in Wikipedia", might it not be better to say "…One threshold for inclusion in Wikipedia", since notability constitutes another, independent threshold? - Jmabel | Talk 06:20, 31 October 2006 (UTC)

(Other than that, I think this is shaping up very well.) - Jmabel | Talk 06:30, 31 October 2006 (UTC)

That kind-of implies to me that crossing either threshold is adequate. "A requirement" perhaps. (I think "necessary (but not sufficient) precondition" is probably a bit too techy, but would otherwise be ideal...) JulesH 11:21, 31 October 2006 (UTC)

How about "The minimum requirement?" That suggests the presence of other requirements. Moreover, meeting the minimum requirement does not sound very impressive, and so is unlikely to be used to defend material that has been attacked on grounds of irrelevancy, redundancy, unimportance, NPOV issues or poor style. On the other hand, "Does not meet the minimum requirement for inclusion," is damning. Robert A.West (Talk) 16:16, 1 November 2006 (UTC)

Wikipedia:Credibility

My proposal for a different name, and how this might affect the lead section. There is an existing page by this name, but I think it can be replaced or renamed. Robert A.West (Talk) 15:09, 1 November 2006 (UTC)

Wikipedia is an encyclopedia, not a publisher of original thought, nor are Wikipedia's articles written by acknowledged experts. To establish and maintain credibility, all facts, opinions, ideas, and arguments published in Wikipedia must be attributable to a reliable published source. The minimum requirement for inclusion in Wikipedia is whether material can be attributed to such a source. Wikipedia is not the place to publish your own opinions, experiences, or arguments.

Although everything in Wikipedia must be attributable to a reliable published source, in practice not all material is attributed. Editors should provide attribution for any material that is challenged or likely to be challenged; for such material, Wikipedia must answer the question: According to whom? The burden of evidence lies with the editor wishing to add or retain the material.

Wikipedia:Credibility is one of Wikipedia's two core content policies. The other is Wikipedia:Neutral point of view. Jointly, these policies determine the type and quality of material that is acceptable in articles. That is, content on Wikipedia must be both attributable and written from a neutral point of view. Because the policies are complementary, they should not be interpreted in isolation from one another.

For examples and explanations that illustrate key aspects of this policy, see Wikipedia:Credibility/FAQ.

Discussion

Submitted for your approval and/or merciless editing. I have no objections to edits to the above proposed text if people like it enough to seriously consider, but think it needs improvement before moving to the project page.

I think the term, "Credibility", embodies what we are after. We want to ensure that the information in Wikipedia is credible. No more, no less. In general, unattributed information is not credible, so we exclude it. Simple facts that can be looked up in obvious places are credible without explicit attribution, so we allow them. Original research is not credible because there haven't been enough expert eyeballs. Simple math and trivial syllogisms are readily evaluated for credibility, so we allow them. Many well-written but poorly-sourced articles in math are tolerated because there is a sizeable community of qualified editors to judge that the information is credible.

While I propose no exception to policy for the last case, the name change makes actual practice understandable: if the community credits the text, it will remain in practice, even without attribution. That can change as we get more eyeballs. The desirable degree of citation needed for any article is the degree required to establish credibility. Less is not acceptable, and more is wasted effort. Robert A.West (Talk) 15:29, 1 November 2006 (UTC)

I am not certain that "credibility" works, as it may imply a value judgement that is made on the basis of what our content provides, and that is better to be left to our readers to make. "Attribution" works better as it implies a goal that can be esily understood and applied by our editors'. ≈ jossi ≈ t • @ 15:34, 1 November 2006 (UTC)

But there is a value judgement going on here. Editors have to judge whether a source is reliable. That strikes to the heart of the credibility of Wikipedia articles. Any reader making a judgement of the article's credibility has to consider the sources, just like an editor does. Credibility is the right word to use here. A reader's judgement of whether they like the presentation, layout or readability of the article has nothing to do with credibility. Carcharoth 17:06, 1 November 2006 (UTC)

Although I am very sympathic to your aim, I am concerned that "credibility" will suffer the same difficulties that "truth" has. "Attribution" is a more objective policy goal though one more difficult to justify. And, giving little comfort to any of these concepts applied to WP, Wikipedia:Articles for deletion/Anthony M. Benis and more particularly NPA personality theory are worth considering by anyone concerned with policy in these areas. Thincat 15:53, 1 November 2006 (UTC)

I was fine with Verifiability, but there seems consensus that we should not re-use that name because of the misunderstandings it generated. Attribution has the defect (to my ear) of sounding bureaucratic, rather than a core principle.

No matter how we phrase the policy, Wikipedia will always remain vulnerable to complex vandalism and to clever and concerted efforts to subvert policy. In the cases cited, the articles not only appeared to satisfy requierments, but the NPA artcle was acclaimed a good article, which is supposed to mean additional scrutiny. Obviously, no one checked the article for NPOV issues, or they would have noticed a lack of criticism denoted a lack of references to the theory. Robert A.West (Talk) 16:11, 1 November 2006 (UTC)

I think "Credibility" is a great deal more susceptible to misinterpretation than is "Attribution." In particular, OR-pushers are likely to claim that their version is credible ("how can you say this isn't credible? can you refute my arguments? blah blah blah...") Of course, those who actually read the policy will understand that this is invalid, but few users will go that far. I think it's best to keep the name unambiguous. -- Visviva 08:55, 2 November 2006 (UTC)

We really need a verb in there somewhere. Establish credibility or Prove credibility. Maybe Reference every fact, or Demonstrate verifiability, jguk 09:46, 2 November 2006 (UTC)

Other proposal

How about: "Wikipedia:Validation", or in full: "Validation by qualitative external sources"?

Regarding the quality of the external sources:

• There is a threshold below which an external source's quality is too low to validate a contention/expression/... in Wikipedia (until now known as an "unreliable" source, but the new formulation is less negative). Wikipedia can set its own quality standards w.r.t. acceptability of source material for validation. The advantage of this less negative way of expressing the acceptability of sources for validation is that Wikipedians don't have to judge the world's knowledge by labeling it either as "reliable" or "unreliable". What do I know whether or not an amateur's blog is reliable or not? It *might* be reliable, in which case it would be an insult to call it unreliable. Or it might be totally unreliable, but needing original research to establish that unreliability. Setting forth quality standards for source material used for validation is less contentious. Bottom line qualities for external sources can include (as they do now): (1) being published; (2) being available to other Wikipedians so that these can check the validity; etc...
• Even if sources have been found that qualify just above the threshold set out in the quality standards, it is always possible (and should be encouraged) to replace them by sources of higher quality when such sources can be found. Thus far a lot of energy has gone into defining bottom values for the acceptability of sources (which is good, because that's where it starts, and there is still a lot of work to be done there), but Wikipedia's worth also increases when validation against just-acceptable-marginal-sources is replaced by validation against sources of a higher quality. This is also an improvement w.r.t. to the still more or less black-and-white reliable vs. unreliable approach. Once a source is deemed "reliable" Wikipedians easily become a bit lazy, and do no further effort to look whether more appropriate reliable sources are available. When focussing on quality of sources it is much easier and effective to say that we should strive for the highest quality of sources. --Francis Schonken 10:26, 2 November 2006 (UTC)

I like the name, although I'm not sure where you're going with the word "qualitative." "Validation" is more clearly normative than "attribution." -- Visviva 15:31, 2 November 2006 (UTC)

I'm still both hands for Verifiability, but IMHO, "attribution" is still lesser evil than "validation" or "credibility". Credibility is established by good publications, not by appealing to others' reputation. Validation is normally performed by fact-checking material; not "source-checking" ersatz. Attribution at least honestly describes what sourcing provides, because we won't try to redefine and bend terms. CP/M ^comm |Wikipedia Neutrality Project| 15:53, 2 November 2006 (UTC)

With a verb

Obviously, Credibility has gone over like a lead baloon, and I see the objections. Still, I'm not ecstatic about the title, so I'll make one final stab to try to find a better one. Someone suggested having a title with a verb, so I came up with some ideas.

• Wikipedia:Use reliable sources. Fairly short, and it says what we mean directly and in a way that is hard to misunderstand.
• Wikipedia:Not a publisher of original thought. Longer, and the verb is implied, but it states our core principle to report knowledge rather than creating it.
• Wikipedia:Assertions must have sources. Emphasizes that every assertion, not just every article, must come from somewhere.

FWIW, Robert A.West (Talk) 17:07, 2 November 2006 (UTC)

Wikipedia:Quote reliable sources?

Wikipedia:Cite your sources? jguk 19:18, 2 November 2006 (UTC)

I was thinking of

• Wikipedia:Attribution
• Wikipedia:Source
• Wikipedia:No original research
• Wikipedia:No original thought (NOT) SlimVirgin ^(talk) 19:27, 2 November 2006 (UTC)

None of those make clear from their title what is required. The third one makes clear one thing that is not required, but gives no aid as to what should happen in its place (unoriginal research?!?), and so should be rejected on that basis. The fourth suggestion is inappropriate - original thought in layout and presentation are welcome, it is only in ideas and "facts" that it is not. On a more positive note, I do believe the second suggestion could be a way forward if modified to Wikipedia:Provide sources, for example, which would make things clearer. jguk 20:21, 2 November 2006 (UTC)

I like "WP:Cite your sources" - it gets to the point. The idea is to uphold WP:V, and thats the best way to do it. Everything else is an instruction on how. ---J.S (t|c) 17:36, 3 November 2006 (UTC)

I actually rather like "WP:Establish Credibility": It's what we want to do, and what they ought to do, and the policy itself contains everything about how to do it. But if there really is no support for that, I'd get behind "WP:Cite Your Sources". -- Thesocialistesq/M.Lesocialiste 22:45, 5 November 2006 (UTC)

Is it ready yet?

It's been over three and a half days since the last edit to the page. Are you on a truce or is it stabilising? Steve block Talk 20:48, 1 November 2006 (UTC)

Discussions like these could be endless. Someone needs to take the discussions by the scruff of their collective necks and forge towards a workable and acceptable document that can be thrown out to the wider community for final discussion, but can still be tweaked and modified in practice after it is accepted (if it is). Despite my tweaks and points made above, I think the document in its current form is acceptable (though I think the wording could incorporate 'validation' and 'credibility' to make clear what is being discussed, and the bit about how this all comes about because the reader has to rely on sources, rather than trusting us, the editors, is a crucial point that needs to be made). I always feel that if my arguments on a particular talk page are strong enough against a source or for a source, or against an unsourced item, then that will be enough to overcome any policy wonk pointing at this page and objecting (this is effectively my version of IAR: use common sense and treat things on a case-by-case basis). Going the other way, this page is suitable for me to refer people to if I think they haven't provided acceptable sources. Carcharoth 12:06, 2 November 2006 (UTC)

Having gotten tuckered out with the debates here, I did not follow it for a while. I see the pop culture exception has arrived by the backdoor: In rare cases, the best source for a particular article may be one that this policy would describe as "questionable". Editors are most likely to encounter these exceptions in the areas of popular culture and fiction, where professional sources offer only shallow coverage, or none, than for established subjects where professional sources dominate. In these cases, and where there is consensus that the sources in question are accurate, trustworthy, and reasonably free of bias, such a source may be used.

I'm still a "hell no" on this. Marskell 12:13, 2 November 2006 (UTC)

If I may, I'd like to interject an example from an area of popular culture that has a growing body of secondary literature, namely Tolkien studies. This is an example where fan speculation and internet forum speculation can be removed and replaced with references to reliable published sources. Ideally, this is what would happen in all such cases, but sometimes the "topic" never takes off in academic circles, and most resources remain self-published or website-based. But even in cases of popular culture without such resources, there often are general or specific studies done by obscure, or not so obscure, academics, but these are not always accessible. Carcharoth 13:30, 2 November 2006 (UTC)

I've been wanting to say something about this too, as I believe I also would be in the "hell no" camp. I missed the workshopping of this specific phrase, but having read over the talk page, I'm still uncertain as to what this clause specifically is intended to allow. My personal take on this issue has always been that this relates to the concept of notability. If AfD discussions, we point to discussion in reliable sources as part of our metric for determining whether we should have an article on something. Logically following that through, it seems like if Tolkien scholars have said commented on, say, Christian themes in Lord of the Rings, it stands to logic we might discuss that in certain LOTR-related articles. If that kind of critical commentary hasn't been done on Buffy the Vampire Slayer I don't think that's unfair to Buffy. And I'd be concerned if in the name of fairness we were allowing editors to go cite Buffy-websites to source interpretive statements about the show's theme, or worse, whether or not fans think two characters made a good couple. Point being: on the internet, analysis exists for everything, but in some cases I'd rather have no analysis than poor analysis from unreliable sources.

I'm not saying I don't support any relaxing of standards here, but honestly, I don't know specifically what kinds of situation the above pop culture clause was intended to be helpful in. I'd really appreciate it if someone could toss me a specific example of the sort of fact you might source to a questionable source that wouldn't exist in reliable sources but would make an article better? I'm not asking sarcastically, I'm genuinely curious. -- Bailey^(talk) 14:39, 2 November 2006 (UTC)

I think the idea is to encourage use of some sources, rather than no sources. I agree with you that no analysis is sometimes best, but sourced analysis is definitely better than unsourced analysis. Having said that, the typical sources used to support assertions about popular culture topics are sometimes really soft, poorly-researched, newspaper articles found online. These sometimes have standards little higher than found here, on blogs, or Usenet. An example is some of the Tolkien articles where I've seen numerous references made to online interviews and articles (due to the explosion of interest before and during the films) that, if you've read the primary and secondary sources, can be seen to merely be regurgitating and sometimes misrepresenting those primary and secondary sources. The expert knows these sources are unrelliable, and we should bypass them and go to better sources, but laypeople won't know that, and misinformation gradually spreads. I fear this is what would happen here if this 'popular culture' clause got through, so I too probably won't support it just yet. Carcharoth 15:04, 2 November 2006 (UTC)

This discussion might be illustrative. -- Visviva 15:29, 2 November 2006 (UTC)

I think these comments here have already demonstrated that, in the case cited above, there always will be no consensus that such sources can be used, jguk 19:20, 2 November 2006 (UTC)

I disagree. I don't think these comments demonstrate anything other than that these cases can't be settled on anything but a case by case basis. What is being argued for is that sources are evaluated in context. That's what academics do, why can't we? Steve block Talk 20:30, 2 November 2006 (UTC)

Then, I think we should say so more directly. Amateur researchers produce valuable material in fields other than popular culture: a lot of American Civil War and Napoleonic stuff has been written by serious hobbyists, students at the Army War College, etc. The historians I know who like to do military history complain that their field is a backwater -- no surprise that it is underserved. I tried to improve the phrasing. Robert A.West (Talk) 22:28, 2 November 2006 (UTC)

I'm still tempted to remove this. Where above is the consensus for a caveat here at all? The pop culture sub-page was set up precisely because there was no agreement. Marskell 00:47, 3 November 2006 (UTC)

Why I think there is consensus for my edit

Point by point.

1. Formal rules are not a substitute for good judgment and intellectual honesty. Formal rules cannot always determine whether material is reliable, notable, or relevant to a particular topic.
   • I believe that there is no serious dispute on this point, and the text has been there for a long time.
2. nor is there any infallable rule to distinguish reliable from unreliable sources.
   • The substance of several discussions convinces me that we nearly all agree on this point.
3. An eminent scholar in one field may be a worthless crank in another.
   • WP:RS#Beware false authority. I do not think this is one of the challenged sections.
4. Some amateur researchers produce material that is acclaimed for its depth and precision, especially in areas where professional sources offer only shallow coverage, or none.
   • This point, or its equivalent, has been made or conceded by enough people for me to suspect a lurking consensus, provided we find good phrasing and avoid the danger of being overly specific. If someone thinks he has found a reliable amateur source, this sentence allows him to argue the case. It doesn't allow him to argue that he has a waiver.
5. The concern should always be whether a source is accurate, trustworthy, and reasonably free of bias.
   • Does anyone actually disagree with this point?
6. If reasonable objections are raised about the use of a questionable source, it should not be used. Questionable sources must never be used to support biographical claims about living persons, or to support fringe theories or marginal positions in the areas of history, politics, current affairs, science, religion, and other academic disciplines.
   • On reflection, I think this does more harm than good, since it implies using questionable sources, rather than giving room for reasonable judgment about whether a source is reliable. I have removed it.

As an example of the sort of thing I am thinking about, Snopes is self-published, and there is little editorial oversight. Nevertheless, I regard it as a reliable resource in its area of expertise. Robert A.West (Talk) 01:26, 3 November 2006 (UTC)

This reduction is an improvement, but I would go further.

• That there is no infallible rule for reliable sources is obvious to the point that it doesn't need mentioning. I only see it as a subtle wedge on the page—"hey, no, I can use this cause WP:ATT says there's no infallible rules..."
• "Some amateur researchers produce material that is acclaimed..." Acclaimed according to whom? This is a repharsing of a devilish point. IMO, we should not allow amateurs to vouch for other amateurs and I don't see how else this will be used. Professionals can, of course, dig up amateurs and acclaim them in an essentially primary sense ("Gramma Moses had remarkably useful agricultural techniques scrawled in the notebooks under her bed"), and of course there's a place for that.
• "The concern should always be whether a source is accurate, trustworthy, and reasonably free of bias." Does anyone actually disagree with this point? Not me—if you can answer according to whom?

Erk, well, I'm still saying it should all go I suppose. Marskell 01:44, 3 November 2006 (UTC)

Oh, and: "An eminent scholar in one field may be a worthless crank in another." A little purple in the wording, but yes, this is fine. Marskell 01:52, 3 November 2006 (UTC)

About point 5: "The concern should always be whether a source is accurate, trustworthy, and reasonably free of bias." In a Key Principles section, we should not require that each source should be free of bias. Although the article as a whole should not be biased, some of the sources may be. --Gerry Ashton 02:06, 3 November 2006 (UTC)

(edit conflict) To answer "according to whom?", the reliability of a source is, and must be, determined by a consensus of Wikipedians. That's all we've got. There is no reliable source on what sources are reliable, and if there were, how would we determine that metasource was reliable? The entire rest of this page is guarding against the primary dangers: unsourced statements and original conclusions based on sourced statements. We can afford, and I think should, have a couple of sentences to guard against the other danger: bureaucratic rigidity.

To take a less obvious example, Henry Chapman Mercer wrote several works that are still valuable, yet he was technically an amateur antiquarian, and his works self-published. He was interested in a field (the history of tools) that escaped the professional scholarship of his time, and there was little professional followup, but much respect, for nearly a half-century after his death. Imagine we were writing the 1950 Wikipedia, using this policy. How should such a source have been handled and on what basis? Robert A.West (Talk) 02:11, 3 November 2006 (UTC)

Instead of "reasonably free of bias," what about "reasonably independent of the subject of the article?" That is the idea I was trying to get across: third-party sources. Robert A.West (Talk) 02:15, 3 November 2006 (UTC)

Biased sources can be used, so long as they are balanced by information from unbiased sources, or sources with different points of view. Also, "reasonably independent of the subject of the article" presumes the subject is an entity, such as a person or organization, from which the author of a source can be independent. The subject might be atoms, and a source author could hardly be independent from atoms, since she is made of them. --Gerry Ashton 03:04, 3 November 2006 (UTC)

"the reliability of a source is, and must be, determined by a consensus of Wikipedians. That's all we've got." Ack! *running for cover*. No, it's not all we've got. I realize that the infinite regress of amateur supporting amateur can be applied to any system of deducing of reliability (the dangers of "consensus science," for instance, are of concern to professional scientists) but if you're actually suggesting "let's throw our hands up and work it out on talk" I have been misreading your comments. We have a lot to go on—with lesser "rigidity," we have the same things that news sources, governments, and universities go on.

Your example is a good one, but (without a full look) is he the only one published in this regard? I mean, did no one reliable afterwards say "yes, this is the man for the history of tools"? Note, I feel fine in leaving room for the broadly amateur if the broadly professional recognizes them later as seminal. Marskell 02:29, 3 November 2006 (UTC)

I am not suggesting throwing up our hands or dropping standards. I oppose any phrasing that encourages the use of questionable sources. What I am saying is that we can lay down principles that distinguish a newspaper of record from a scandal rag, but ultimately the decision that the New York Times is the former and the National Enquirer the latter has to be based on a consensus. It can't be done mechanically. The same thing applies to who qualifies as a researcher.

As for Mercer, I picked him because for a time he was, "The man to go to in a field no real scholar cares about," but now they do. Maybe the right thing is to always wait until some accredited professional decides the topic is worth a look, but I don't think we should lock ourselves into that position. An editor arguing for the inclusion of amateur research material should bear the burden of convincing others that it is reliable, but that burden should not be infinite. Robert A.West (Talk) 18:44, 3 November 2006 (UTC)

But... what about criticisms of self-published sources?

I share Atom's concern. Shouldn't there be a provision to allow criticisms of the content of self-published sources? If I read the proposal right, an article about Fred Foobar would be allowed to say "According to Fred Foobar, he was born in Mars" or "Fred Foobar claims that he has lots of bridges to sell" (and indeed, Foobar actually said these things), and suddenly you need an expert to criticize Foobar's claims. Clearly this is absurd.

Indeed, this was an issue some time back with regard to the Neo-Tech (philosophy) article (now merged with Frank R. Wallace). Bi 11:15, 3 November 2006 (UTC)

The fact that we can include such examples doesn't mean that we must. If Foobar is notable enough for inclusion in an encyclopaedia, and if these claims are themselves notable enough to warrant inclusion, then you can be certain that another source will have already remarked on them.

On the other hand, if no one else has commented on the claims, then that indicates that they are not particularly noteworthy, and there is therefore no reason for us to include them in an article on Foobar. Jakew 11:30, 3 November 2006 (UTC)

The argument sounds nice in theory, but it doesn't wash. In the case of Neo-Tech, for example, many people have commented on the claims, but they're not authoritative "experts".

Besides, if the only claims that count as noteworthy are those that have been evaluated by recognized experts, then what's the whole point of allowing the use of self-published sources anyway? Bi 12:52, 3 November 2006 (UTC)

Now that is a good question. IMHO, using self-published sources (except, perhaps, for trivial, non-controversial claims) would tend to diminish our credibility. Jakew 12:55, 3 November 2006 (UTC)

• Have a read of the neutral point of view policy, that should answer the question. I would imagine the easiest solution would be to write that "although Fred Foobar claims on his website to be from Mars, most reports of his life point to him being born in Grimsby". Whether he has lots of bridges to sell would be removed as undue weight per the aforementioned policy. Hope that helps. Steve block Talk 13:56, 3 November 2006 (UTC)

As I said, the problem is if those "most reports" don't happen to be from authoritative sources. Again, there's the Neo-Tech (philosophy) example I mentioned. Bi 13:58, 3 November 2006 (UTC)

Move away from a rules-based and to a principle-based policy on sources

There has been much talk about trying to get a prescriptive rules-based approach to what sources are acceptable. Yet this approach will not succeed - any rules would need to be long, dealing with many exceptions, if they were to work. Really we need a straightforward principle on sources that is the policy - this should then be back up with the briefest of explanations on the same project page as contains the policy, and links to greater guidance (which would then be persuasive but not mandatory).

Might I suggest this policy:

A source for a claim is reputable if that source is capable of withstanding academic scrutiny

By way of brief explanation:

The for a claim bit is to imply that a source can be reputable for some claims and not others.

Academic scrutiny is meant to refer to a process where you ask yourself with due diligence whether that source is strong and believable enough to merit the claim it is being used to support.

If we move to a principle-based approach to sources, a lot of the text currently in dispute can safely be removed from this page (and relocated, if desired in another discussion page), jguk 12:58, 3 November 2006 (UTC)

• I like this up until the last two words. Steve block Talk 19:40, 3 November 2006 (UTC)

This bit makes me scream so much THAT MY HAT WOBBLES

In rare cases, the best source for a particular article may be one that this policy would describe as "questionable". Editors are most likely to encounter these exceptions in the areas of popular culture and fiction, where professional sources offer only shallow coverage, or none....

I have tried and tried to explain here that it is just not true that coverage is shallow for pop culture or fiction (at least not to the point where the shallowness of the coverage justifies using questionable sources). I've asked for examples, and so far only two have been produced that stand up (Spoo and that little mathematical gremlin thingie). I've demonstrated that plenty of sources exist for such fields as clockmaking and quilt making, and that there are articles on even the most obscure bands in music magazines.

The point the above extract of the policy (supported, possibly, by Phil Sandifer but opposed by many others carrying less weight) fails to take into account is that all forms of fiction, whether in published books or in produced films and television shows, can in the last resort (I mean, if no secondary sources exist, which in my opinion is unlikely) be used as sources for themselves. Someone gave the example of a particular planet of an episode of Star Trek, and I thought I'd showed how an article can be written on such a thing if it describes everything about the planet as shown in the episode (the reference is to the episode, with date and production credits). It is true that Wikipedia recommends writing from outside the fictional world, but it doesn't go so far as to ban writing from inside it, which at least precludes recourse to some amateur site that goes into arpeggios of nerdification on the ins and outs of said planet, virtuosic though such arpeggios might be. The real issue here is that the book, film, or TV show has gone through a process of editing, publishing, and producing which has involved much critical decision-making and collaboration by experts, whereas the website hasn't or can't be proved to have. So long as the Wikipedia article makes no original interpretation, it is acceptable.

Well, I've calmed down now (the above is about the nearest I get to a rant). We should just stop including the unproved—and, to me, dotty—assumption that popular culture (and fiction, of all things) is short of secondary sources in some way and stop trying to offer special dispensation to an area of Wikipedia that far from needing it is actually, to judge by the new article list, as healthy as a field of mushrooms. Doing so is like the Ghostbusters opening up drains, caskets, and floorboards and issuing the ghosts with licences saying, "OK, guys, here's New York. Off you go and swamp it while we get on with spraying ants".

qp10qp 13:05, 3 November 2006 (UTC)

This is precisely why we need a principle-based policy here - any rules-based approach will have people, probably justifiably, up in arms against some bit of it, jguk 13:08, 3 November 2006 (UTC)

To be honest, I don't think people digging their heels in helps reach a consensus. Since two examples have already been found where this is the case, and the sentence starts with the phrase In rare cases, I can't see how people are creating a problem. I'd offer more examples, but I'm terrified people will strip the information out based on the fact that it is sourced from a website rather than evaluating the information and it's source in context. If people seriously believe that allowing this wording in the article will mean that I can write up my grand theory of everything in a Wikipedia article based on my blog and the Wikipedia community will allow that article to stand, I would suggest they are taking an absurd position. As yet nobody has actually detailed why this is a bad idea. I keep hearing it opens up a loophole, but nobody has yet detailed in what way. How will this impact Wikipedia? Are people suddenly going to stop adding badly sourced material, or original research? No. What's the goal here, people? To build a page to stop people, or to build a page that details how we do stuff and like it done? If you want to build a page to stop people, then I'll shuffle off, because that isn't going to happen. Has WP:VANDAL stopped us getting vandalised? It's almost impossible here. People ask for examples, and then when they're provided they're disregarded as not counting. Let's try and reach a consensus here, not an argument. Steve block Talk 13:51, 3 November 2006 (UTC)

This comes back to the stability point. At some point, someone will read whatever you are referring to and, unlike the majority who pass by it because it has a referenced footnote (and therefore is 'probably OK'), will follow up the reference, conclude the website is unreliable, and remove the reference and information from the article. If you find something that will satisfy everyone, the snippet of information is far less likely to be removed, and stability of the article is increased. Carcharoth 13:57, 3 November 2006 (UTC)

No. The examples I refer to are in featured articles and are there because they aren't from unreliable places, and also there because they have a good consensus. At some point someone will remove them because they are following some poorly worded policy rather than evaluating the information in context. You will never find something to satisfy everyone, and again, if that is the goal let us please admit how absurd it is. Steve block Talk 16:38, 3 November 2006 (UTC)

It is my experience that many of these "authoritative" amateur websites would be picked to pieces if they were subject to a proper, professional review. There is a tendency for those who are not subject to professional review to write and claim stuff that they wouldn't if writing something that would be published and subject to professional editorial and peer review (I make such claims myself). There are exceptions, but all of these should be left for treatment on a case-by-case basis, on the talk page of the article concerned. All this policy needs, to cover the "popular culture exceptions" (if they exist), is to re-iterate that this policy does not over-ride valid case-by-case arguments on the talk page supporting the use of a particular source. Carcharoth 13:54, 3 November 2006 (UTC)

That is absolutely it. Some of us have said that, but others have replied that they can't support rules that invite you to ignore them. I see it Carcaroth's way.

I take Steve's point; but if the exceptions are rare, why state them? And if the exceptions are rare, why link them to the whole fields of popular culture and fiction? His argument that the exception clause will do little harm by staying in can be countered by saying it will do little harm by being taken out. I hope he doesn't think I am "digging in", because I don't even edit the article (which to me at the moment would be like taking a flying fuck at a rolling doughnut) but merely doggedly argue my case here (I also believe, if it's a question of consensus, that the consensus favours leaving the exception clause out, if anything).

We all recognise that occasional exceptions will earn their place at the editing level, and good luck to them. The trouble with leaving the case for exceptions in is that it renders the rules inelegant, compromised, conceptually imperfect, and, in my opinion, slightly unprofessional (my reason for interest in these matters is the dismay I felt at the muddled amateurism I encountered when first reading the central policies and guidelines: it's appalling that newcomers be given the impression that Wikipedia not only doesn't quite know what it's doing but that bad writing is the done thing).

Steve's point that Wikipedia is full of dodgy articles already is for me a reason to leave the exception clause out rather than to put it in; however, I agree with him that a principles-based policy is best (and we have the makings of one here, I believe). I also agree with User:jguk, somewhat of a voice in the wilderness around here, who sees the answer in greater concision and simplicity rather than in ever-decreasing twists of rococo scrollwork. qp10qp 15:52, 3 November 2006 (UTC)

I'd say that the exceptions aren't really all that rare. However, since deciding whether something is an exception involves human judgment, someone who opposes all exceptions can reflexively argue against each one and it'll be impossible to prove him wrong. Exceptions that are so obvious and so unambiguous that even the most diehard exception-hater can't say anything about them, are what are rare.

That's what's happening with the exceptions that "don't stand up". They do stand up, but it's too easy for someone who opposes them to declare, based on slanted judgment calls, that they don't.

Incidentally, the quilting and clockmaking examples were proposed by someone who stated them as a hypothetical, and wasn't actually trying to claim that quiltmaking and clockmaking require self-published sources.Ken Arromdee 16:33, 3 November 2006 (UTC)

Let's be fair here, I don't edit the article either, and argue my case here too. However, my mindset doesn't allow me to consider as viable a conceptually perfect ruleset, so perhaps we'd best leave that one aside. Good luck with that. Wikipedia is an encyclopedia built by amateurs. That it is amateurish should be worn as a badge of pride, not used to ridicule it. I note we still pat ourselves on the back when reports cite us as more reliable than professional encyclopedias, perhaps we shouldn't do so anymore? We used to assume good faith. It's a shame that we can no longer do that in our policies, but instead must close the door to reasoned evaluation. And I'd say the consensus is finely balanced, having followed the discussion from start to here. Taking the clause out does great harm, by stating things as not as they are. This stuff is important, because it is policy creep. If there are exceptions, be transparent. What reason do we have to hide such exceptions? I thought Wikipedia was a transparent place, not somewhere that had cupboards with beware of the tigers written on them, where important details might be found. Still, I've said my piece, Wikipedia is a shifting place and let's just see where the sands lie. Steve block Talk 16:38, 3 November 2006 (UTC)

Well, one way to achieve "greater concision and simplicity" is to simply rule out all self-published sources, period.

Or, rule out all self-published sources except when specially permitted by Jimbo Wales and his cabal, period. Or require every statement in an article which uses self-published sources to be marked as such. Each of these proposed rules are straightforward to apply. I don't see why someone'd object to handling material from reputable sources and material from self-published sources differently, unless that someone believes his source like a gospel, or has a POV to push. Bi 17:47, 3 November 2006 (UTC)

I hope no one minds if I ask, but... what was the other example besides Spoo? Spoo confuses me as an example. I can see that the Spoo article uses forum posts as a source, but they seem to be mostly posts written by the creator of the show. Wouldn't that be the same in principal as the exception for self-published material from a usually-reliable author (a well-known, professional journalist or researcher in a relevant field)? Published statements by the creator of Babylon 5 would usually be an acceptable source for Babylon-5 related information. I know this is just an example, but it doesn't seem like an example that demonstrates need for a special exception, or to allow forum posts in general. -- Bailey^(talk) 17:54, 3 November 2006 (UTC)

It was Buddhabrot. Once again, this is probably OK within present wording as an article about itself. I'm not quite sure where you stand on the matter, but my point in raising that and Spoo was that I'm all in favour of articles outside the rules if they are good enough and a group of editors agree to them; but they are likely to be very rare. I think, however that there would be no point in saying that in the policy; it would do more harm than good by creating mind-bending exception-think. Such articles, of acceptable quality, we've had a few, but then again, too few to mention. qp10qp 21:27, 3 November 2006 (UTC)

I believe I agree with that point. Buddhabrot may qualify as an exception article that's worth making the exception for. Spoo seems to be more or less within policy without an exception. If we're going to discuss making exceptions at all, though, doesn't it behoove us to come up with more than one example of the kind of thing we want to be excepted? Buddhabrot isn't even strictly a pop culture or fiction topic. We already have IAR to cover extrodinary and unusual cases; if there's a category that needs an exception that occurs frequently, we should be able name more than one example that fits that category. Is there a defender of this exception willing to step forward and name some more articles that this would cover? -- Bailey^(talk) 22:14, 3 November 2006 (UTC)

(As an aside, I think we'll need to define "relevant field" more clearly...) Bi 18:14, 3 November 2006 (UTC)

Yes, it's an amateur enterprise, Steve, and I love it so, but that doesn't mean it should be amateurish; there's a right way of attributing information, that's all (I'm into popular culture and believe it's possible to report it in a properly sourced way or to describe it from within its world). Unlike you, I do believe that a clean and simple ruleset is possible; and we've got tantalisingly close to one recently. I don't think we should hide or hunt down exceptions, just let them fight for survival on their pages.

The more I think about it, my opposition over this boils down to a tiny judgement call between the following two positions:

1. No exception: Rules should apply everywhere. Of course, non-adherence to the rules will occur, as to any rules. No reason to modify them on that account.
2. Exception: To make a set of rules in the knowledge that unmandated exceptions will occur weakens those rules. How could they be taken seriously?

Looking for compromise, I'd be prepared to accept the principle of exceptions to the policy if particularisms such as the reference to popular culture and fiction could be removed. This would shut up my screaming that those areas do generate usable sources but would allow editors to use the principle in the clause to assist them in editing certain undescribable articles that I obviously haven't any idea about but which, thanks to the good-faith comments of others, I'm sure do exist. qp10qp 18:47, 3 November 2006 (UTC)

The problem from my eyes is that we already have the rule that exceptions exist, it's ignore all rules. I already asked earlier in this debate that we simply agree that Ignore all rules applied to this policy as much as it does to any other, excepting the neutral point of view which is a foundation principle, but that compromise didn't sit right with some. I don't know where the middle ground is, but it is a nice day when we can all agree that exceptions occur. I too have an interest in an area of popular culture, and I also think I'm pretty good at sourcing, so I know what I'm talking about too. Sometimes the best source you can get for an opinion is a website. It tires me out sometimes that 500 blogs may all criticise something on the same point, but people refuse to allow the criticism because of the weakness of the sources. Now I'm fully behind the idea of a simplified ruleset and have argued for such since I first got here, but I'm pragmatic enough to know a simplified ruleset doesn't equate to a conceptually perfect ruleset. I don't see that having exceptions weakens rules, because I don't see that we have rules on Wikipedia anyway. I always thought our policies were descriptive, and that declaring our contradictions keeps us honest, practical and encourages debate, the very method this enterprise is supposed to establish itself by. You're right it's about two positions, but to me they are

1. No exceptions: Rules should apply everywhere. Encourages a mindset of devotion to rules, stifles debate, discourages intelligence and establishes commandments. Creates a hostile environment where people become emboldened to punish for breach of rules.
2. Exceptions: Encourages debate, strengthens rules by allowing them to be applied reasonably and consensually, describes the practise openly and transparently and fosters an environment which encourages rational discussion and respect for all.

I don't mind how we establish it, I don't mind how we word it, but I think we have to be clear that exceptions happen, and are a good thing. Steve block Talk 19:35, 3 November 2006 (UTC)

Wikipedia is an encyclopedia, not an ideological experiment in free speech and free debate. Yes, I'm all for rational discussion, but there are some people who are simply interested in other things, like advertising their w4r3z. Letting these people have their way doesn't encourage rational discourse; far from it. We need a way to show these people the door. Bi 20:07, 3 November 2006 (UTC)

<edit conflict>We have one at WP:CSD. To address the point regarding what Wikipedia is, "Wikipedia's success to date is 100% a function of our open community. This community will continue to live and breathe and grow only so long as those of us who participate in it continue to Do The Right Thing. Doing The Right Thing takes many forms, but perhaps most central is the preservation of our shared vision for the NPOV and for a culture of thoughtful diplomatic honesty." Wikipedia is an ideological experiment in how free speech and free debate can create an encyclopedia. Steve block Talk 20:32, 3 November 2006 (UTC)

That's a blatant misreading of Jimbo's words. Note that the "most central" part of "Doing the Right Thing" is not free speech and free debate. It's "the preservation of our shared vision for the NPOV" and "a culture of thoughtful diplomatic honesty" (yes, it's "thoughtful", not "thoughtless").

And you can't deny that there are cases of bad behaviour which aren't covered WP:CSD: here's something from Wikipedia:Articles_for_deletion/Neo-Tech (philosophy) (3rd_nomination):

"Both ... articles were originally created ... by a now-banned user [User:RJII] whose express and only purpose on Wikipedia was a co-ordinated, sustained campaign to push a particular point-of-view across political and philosophical articles."

Is this the kind of behaviour you want to encourage? Bi 20:43, 3 November 2006 (UTC)

I'm at a loss as to how it's a blatant misreading to suggest that the open community must maintain a culture of thoughtful diplomatic honesty equates to allowing free speech and free debate. Perhaps you can explain which part it is of a culture communicating thoughtfully, diplomatically and honestly that disbars free speech and free debate. And I'm not sure why you point out to me that it's thoughtful, not thoughtless. I'd note Jimmy points out "Diplomacy consists of combining honesty and politeness. Both are objectively valuable moral principles. Be honest with me, but don't be mean to me." As to User:RJII, once again, you present the door he was shown in your argument, namely arb-com. The articles he created were created before this policy discussion started, before even Verifiability and Original research were policy. I'm not sure what point you are arguing here, but my understanding is that you believe that by not allowing an exception clause in this policy such articles will not be created. That won't be the case. If that's not your point, then I'm not sure what you are arguing. If we accept such articles will be created, then we both know that the only places to deal with extant articles are the deletion pages, and if it ain't a speedy it has to go through afd. RJII was very clever at working out how to meet policies in any event, I don't think his articles actually fall foul of this policy. What they do fall down on is the neutral point of view policy. As to the sort of behaviour I want to encourage, I have no idea where you get the idea I would encourage behaviour akin to RJ's from. His block log should show I blocked him over edit warring, I have written nothing here that would offer the slightest shred that I would condone such behaviour, and I can only suggest you are blatantly misreading me. Steve block Talk 22:14, 3 November 2006 (UTC)

As I pointed out, it's "thoughtful", not "thoughtless". User:RJII was clever: he tried to push a POV, but cloaked his attempts in seemingly NPOV language. And I never said you'd condone his behaviour; but your suggestion will effectively serve to encourage it. If you were "thoughtful", you'd realize this. Bi 05:16, 4 November 2006 (UTC)

Snopes

OK, let's take a concrete example of something this exception may be aimed at. Urban Legends Reference Pages (www.snopes.com) is a private site run and written by Barbara and David Mikkelson. To my knowledge, neither possesses an advanted degree in a relevant field. You could call them self-published, amateur or both. On the other hand, they are the place to go to check up on urban legends and web stories and scams. They have a reputation to protect, and try to protect it.

I would not call Snopes a questionable source. I would call it a reliable source that lies somewhat outside the traditional and that is useful because, it covers details and specifics, while professional scholars are properly interested in generalities and trends.

Why do I think the distinction important? Because I think that telling people that it is sometimes OK to use questionable sources is a mistake and unnecessary. All of the non-theoretical exceptions I have seen discussed are instances when a normally-questionable source is not questionable on a particular subject. People are normally pretty reliable about their own birthdate, the names of their siblings, and so on. The National Enquirer is a valid primary source for an article on the National Enquirer. A blog by a meteorologist hosted by Accuweather is not an official prediction, but there is no reason to call it "questionable."

So, I join my voice to those who object to a special exception for popular culture, and add my disquiet over the current phrasing. Robert A.West (Talk) 20:33, 3 November 2006 (UTC)

• Can you explain how we can use Snopes as a source in an article regarding anything other than Snopes itself, because my reading of the page is that we can't. Steve block Talk 20:44, 3 November 2006 (UTC)

This is completely wrong-headed. If Snopes is reliable, it won't be because it has a reputation to protect (after all, every popular web site can claim the same thing), or that it discusses details. If Snopes is reliable, it'll be because you can cross-check its claims, and they had better be cross-checked. Bi 21:08, 3 November 2006 (UTC)

Searching the New York Times website shows 14 hits for snopes.com. Since I don't subscribe, I could only read some of the articles, but the Times seems to regard snopes.com as a urban-legend busting website. If other reputable publishers also have a positive opinion of snopes.com, I think it should be considered a reliable source. --Gerry Ashton 21:30, 3 November 2006 (UTC)

I worked a bit on an article a while back where another editor wanted to use Snopes as a reference. I disagreed, but was tempted cave in since Snopes is quite a unique site. Then I realized that Snopes cites lots of perfectly reliable sources at the bottom of every article they write. I checked some of the books they used out of the library, and we all lived happily ever after. Seriously, I admit that Snopes is generally afforded a level of respect that most amateur research sites aren't, but I'd wager that's only because their sources are completely transparent, which means there's no practical reason to use Snopes itself as a source. -- Bailey^(talk) 21:41, 3 November 2006 (UTC)

Except for the (justifiable) preference for web citation where it's possible. Phil Sandifer 21:45, 3 November 2006 (UTC)

And the instances where they use unverifiable sources, such as phone calls to the relevant local police, or where they conclude from an absence of evidence that a thing does not exist, which would count as original research if done on Wikipedia. Robert A.West (Talk) 21:51, 3 November 2006 (UTC)

Here, Snopes is serving as a secondary source--one which conducts original research. One of the key issues regarding secondary source is that they frequently do things which may be difficult to verify independently--such as perform investigative activities like phoning the police--so we have to take their word for it. Which is why we have the notion of a "reliable source" in the first place--a source who we can trust on a matter which we cannot verify independently. When a newspaper breaks a big story, often times we (as an encyclopedia) are in no position to verify the story, thus we trust the newspaper's account. When a scientist makes a claim in a paper, most of us are in no position to repeat the research for ourselves, so we depend on the reputation of the scientist and/or the journal in question. Snopes is no different. It conducts original research, and we--as an encyclopedia--have to decide whether it is trustworthy or not on the research it conducts. (Snopes also acts as a tertiary source as well, of course). If we think it isn't, then we shouldn't cite it. Many consider snopes to be trustworthy. --EngineerScotty 22:38, 3 November 2006 (UTC)

Unfortuately, they've also got the habit of occasionally including false information just for kicks, like [1] for example. Don't get me wrong -- I like snopes -- but I'm much more confortable with just reading the books they use as sources, and working from there. Anything that's entirely based on their personal research probably shouldn't be cited, if only because it's difficult to ask our readers to blindly trust a relatively unknown source whose credibility is not obvious. -- Bailey^(talk) 03:46, 4 November 2006 (UTC)

So I guess we still need a rule-based approach to reliability after all? I think this is all too clearly shown by the kind of weird "reasons" advanced above by User:Robert A West for Snopes' reliability. Bi 05:16, 4 November 2006 (UTC)

The best way to cook a fish is as little as possible

Steve, you set up that contrast in a seriously POV way. It doesn't in the slightest follow that no exceptions means hostile thought-orcs on the rampage. The admins hardly bother to police attribution at all, and then only when refereeing disputes; they concentrate on vandalism, copyvios, advertising, and the like. If Wikipedia had a broken sink, they'd mend it; if a pigeon was trapped in the aircon, they'd free it. I agree that "rules" is probably the wrong word for policies, particularly as "ignore all rules" trumps the lot of them. For me, they're a set of principles, philosophies even, needing to be conceptually elegant, crystalline and memorable.

But here's the thing: "no exceptions" is not a policy point; it is an absence of a policy point—in other words, one less piece of policy. It cannot do the things you say—stifle debate, encourage commandments, discourage intelligence, create a hostile environment, etc.— because it doesn't exist.

It is a silence, an omission, a non-presence, a subtraction from the capillaries of the text, a Tao.

I shy from exceptions for reasons not far from those for which you seek them: to create an atmosphere free from pettifoggery, where articles can exist in a clean, open atmosphere, and where, if they're good enough, like Spoo, they'll stand. qp10qp 21:27, 3 November 2006 (UTC)

That's an excellent point, save for one key problem, which is that far too many people read pages - particularly policy pages - as having an implied "And no exceptions!" at the end. So long as such people are prevalent, and, more to the point, prevalent on pages for dispute resolution, an explicit space for exceptions is necessary. Pettifoggers do not need to be given an opening to cause trouble, and cannot possibly be closed out. On the other hand, well-meaning users often do need to have an opening to make positive changes (WP:BOLD, WP:BITE, WP:IAR), and can very easily be closed out. Phil Sandifer 21:34, 3 November 2006 (UTC)

• I'm not sure how to plead to the fact that I expressed my opinion in a point of view way. I quite clearly qualified it as being, "to me". That said, people already remove text from articles based on existing policies, and refuse to accept that exceptions occur. We already have wiki-lawyering. The policies on verifiability and no original research used to claim they were non-negotiable, I'm not sure they still do, but people used to attempt to enforce them across the board. I'm not arguing from a position of idealism but one of practicality. I've been involved in these debates before, I've seen people attempt to declare that policy is inviolate, and I'm not interested in seeing that again. Discussion has already been stifled through such methods. I'm quite happy for a set of principles to be established. I think the easiest expression is that "Wikipedia should delineate subjects from a neutral point of view and summarise sources accurately. Unsourced material may be removed. Sources should should support the material they are used in citation for, and should withstand the rigorous examination that the material sourced demands." I think we lose something when we attempt to expand beyond a short statement, because we end up in these debates over phrasing. No-one has still yet addressed the fact that using a weak source in a scientific article violates NPOV. Whatever people want to protect against is already protected. Steve block Talk 21:51, 3 November 2006 (UTC)

My own argument against an exception clause has been criticised for depending on "ignore all rules" to take care of exceptions, which it was claimed would make this proposed policy irresponsible. I can see the point. But the further I go back into the original thinking on Wikipedia, the more I appreciate its beauty and the reasons for its success against the odds. The founding principles and Walesian sayings are alert to the potential for hierarchies, frozen consensuses, and hostility to bold editing or newby editing to grow up and silt the freedoms. I believe a short, simple policy on this page would protect freedom more than institutionalised exceptions. I break policy rules every day, but I break them better by knowing the central principles. On article Talk pages, I've noticed that many editors have a garbled grasp of the rules, clinging to fragments of them only; perhaps that's because we've made policy pages so subclaused and iffed and butted that they end up making clever people act stupid. qp10qp 22:02, 3 November 2006 (UTC)

I agree entirely. Steve block Talk 22:18, 3 November 2006 (UTC)

The original research policy certainly does still claim to be non-negotiable. Actually, OR contradicts Verifiability right now; OR says the principles are not negotiable, and Verifiability says they are "negotiable only at the Foundation level". Not only do they contradict each other, but both contradict Wikipedia:Ignore all rules, which I've complained about, to no avail. Ken Arromdee 08:34, 4 November 2006 (UTC)

But, as you know, the present page is an attempt to combine No Original Research, Verifiability, and Reliable Sources, so the question is being addressed. I take no notice of the non-negotiable or don't edit-without-consensus pretensions of these policies, which aren't canonical (but Neutral Point of View is). Whatever we say, Ignore All Rules is always available as a joker. So I'm not completely out of sympathy with Steve's liberalism because the foundation principles and the Walesian statements not only list no rstrictions but seem determined to avoid them. The following are the non-negotiable m:Foundation issues)

1. NPOV as the guiding editorial principle
2. Ability of anyone to edit articles without registering
3. The "wiki process" as the final authority on content
4. Copyleft licensing of content; in practice, GFDL (working on changes via GFDL 2.0)
5. Jimbo Wales as ultimate authority on any matter (although some authority has been delegated to others; see Arbitration Committee and Board of Trustees)

Jimbo Wales states his principles this way {User:Jimbo Wales/Statement of principles):

1.	Wikipedia's success to date is entirely a function of our open community. This community will continue to live and breathe and grow only so long as those of us who participate in it continue to Do The Right Thing. Doing The Right Thing takes many forms, but perhaps most central is the preservation of our shared vision for the NPOV and for a culture of thoughtful, diplomatic honesty.
2.	Newcomers are always to be welcomed. There must be no cabal, there must be no elites, there must be no hierarchy or structure which gets in the way of this openness to newcomers. Any security measures to be implemented to protect the community against real vandals (and there are real vandals, who are already starting to affect us), should be implemented on the model of "strict scrutiny". "Strict scrutiny" means that any measures instituted for security must address a compelling community interest, and must be narrowly tailored to achieve that objective and no other. For example: rather than trust humans to correctly identify "regulars", we must use a simple, transparent, and open algorithm, so that people are automatically given full privileges once they have been around the community for a very short period of time. The process should be virtually invisible for newcomers, so that they do not have to do anything to start contributing to the community.
3.	"You can edit this page right now" is a core guiding check on everything that we do. We must respect this principle as sacred.

Wales keeps the "rules" as few as possible, and I can understand why. It means that the policies we are here working on are only policies because users say so. In fact they are little more than guidelines and can and will be ignored; their content will constantly change; they will contradict themselves from time to time, owing to human error.

Steve thinks it's restrictive not to have an exception here; I think it's restrictive to start "legislating" minutiae such as exception clauses, because to do so makes it seem as if the policy aspires to control things at the editing level. The truth is that what goes at the editing level is what good-faith editors agree on, or compromise over, whether it attends to attribution policy or not.

Ken, I happen to have read some proceedings of the case you were involved in around this issue. I felt that neither you or your opponent were going to get anywhere by referring each other to policies; you were doomed to thrash away like mastadons in the gurgling swamp until somebody won, uncomfortable and infuriating though that might be. Fortunately, such head-on clashes are rare.

qp10qp 14:38, 4 November 2006 (UTC)

What exactly are you talking about? I wasn't involved in any case around this issue. Ken Arromdee 19:30, 4 November 2006 (UTC)

Wikipedia:Mediation Cabal/Cases/2006-10-05 Bowling for Columbine. I thought Ignore All Rules and contradictions between policies were relevant to citation/sourcing issues there; forgive me if I misunderstood. I found reading that stuff very instructive on the practical implications of policies (but maybe that was just my take). qp10qp 22:25, 4 November 2006 (UTC)

If you'll look carefully, you can see that that happened some time after I complained about Ignore All Rules. Moreover, the other guy must have posted 10 times as much stuff as me, including long quotations of rules, which is what might make you think it was about rules.

I don't think either IAR or policy contradiction has anything to do with it at all. Ken Arromdee 01:49, 5 November 2006 (UTC)

Overfishing leaves nothing for supper

Well, it's gone again. I'm tired of this debate, it's circular and like I say, the heels tend to dig in. I'm not adding it back because I don't edit war, I discuss, and I've tried to make the case on this page for the exception in some form. I'm disappointed no-one has made a real case against the exception beyond the fact that it creates a loophole. I think the fact that anyone can edit Wikipedia is a bigger loophole that the one we're discussing here, but let's be honest, there's bigger fish to fry than this loophole. I'm not sure, however, that this page will gain a consensus until the matter is addressed through debate and compromise rather than removal and re-instatement. I've seen the people advocating some form of acknowledging exceptions exist bend over backwards. To me, there's been little movement on the other side. It used to be we could disregard people who were clearly not working towards a consensus. Steve block Talk 22:29, 3 November 2006 (UTC)

What kind of "case against" exceptions are you looking for? Your case for exceptions is based on fact-free generalities and painting critics as authoritarian thought-orcs. You were the one who claimed it's necessary to have exceptions, so the onus is on you to make a strong case for exceptions, not for us to make a strong case against. Bi 05:16, 4 November 2006 (UTC)

Just FYI, if you want examples of articles where the use of forums, fansites, and other non-professional self-published materials might be an problem, I can certainly offer them. I figured we all already realized there were a tons of cases where using these kinds of sources would be an issue. You could filibuster the Senate with that list. With apologies for possibly missing something, I never quite grasped what kinds of instances this proposed exception was supposed to except. We've talked about [spoo], but from what I can see that article stays relatively within the rules without an exception -- mostly it uses USENET posts by the show's creator, who is both a professional and an expert on the subject of Spoo. We've talked about quilting, but there's no lack of hard sources for a subject like quilting, and ultimately the example was withdrawn. Buddhabrot was the only example of an exception that seems like an exception to me, and to be perfectly honest, it's a kind of a cool effect, but I'm not sure it would sink the encyclopedian project if we didn't cover Buddhabrot. I'm not hositle to the idea of a hard-coded exception in theory, but no one seems willing to come forward with substantial proof that there are more logical exceptions of this kind than there are of any other kind. If there's no great trend in pop culture articles needing this kind of specific exception, then IAR should do it when it comes up. -- Bailey^(talk) 18:46, 4 November 2006 (UTC)

I currently have an article at WP:FAC that uses a fan forum and the creator's blog. Would this be an example? Firefly (TV series). The only time we used the fan forum was for this statement: "By June of 2003, actors Nathan Fillion and Adam Baldwin confirmed this on the official Firefly forum, as did Whedon in several interviews" --plange 04:47, 5 November 2006 (UTC)

Well, with the proposed pop culture clause, we're talking about exceptions other than the exception for "professional self-published material". The description of "professional self-published material" would already cover the examples you describe -- the show's creator would be assumed to have professional expertise on the article's subject, even if he's commenting a blog or forum. Since there's still a push to say more generally that pop-culture articles may sometimes use questionable sources, I'd like to see good examples that fall outside of the self-published-professional description. Regardless, good luck on your FA, Browncoat. -- Bailey^(talk) 05:56, 5 November 2006 (UTC)

Thanks! Just wanted to make sure I was ok :-) --plange 15:17, 5 November 2006 (UTC)

Or, you could argue that the exception "articles about themselves" applies: the official Firefly forum is citable in an article about Firefly. Robert A.West (Talk) 15:52, 5 November 2006 (UTC)

Yep, I guess though that it still shows it should be handled on a case by case basis, as I wouldn't cite the forum if it was posted by someone other than the cast or creator. --plange 23:43, 5 November 2006 (UTC)

Sources requiring expert interpretation

Many documents do not speak for themselves, at least to a modern audience of non-experts. The past is a foreign country: they do things differently there, and the documents of that realm are written in a language that differs from our own.

What exactly is the intention here? Are all historical primary sources off limits? Are Winston Churchill's memoirs unacceptable as source material for Winston Churchill or World War II? How about Bill Clinton's? More specifically, I have been working on several articles about Scandinavian medieval history. It is quite common to use medieval sagas, in translation, as references for articles in that area. Is this now to be held to be unacceptable, because "the past is a foreign country?"--Barend 14:17, 3 November 2006 (UTC)

That's not the intent, but the text you mention still needs to cleaned up to make it clear. Primary sources are not off-limits, just the use primary sources to support interpretive claims. It's fine to use The Illiad as a reference for the claim that Achilles and Patroclus fought together in the Trojan war, but not to say that they were one of the earliest models of the egalitarian homosexual couple in literature -- that would require a secondary source. -- Bailey^(talk) 17:15, 3 November 2006 (UTC)

I wouldn't worry too much. The question of "primary" sources has been greatly discussed here and elsewhere, and although policies make slightly deprecating remarks about primary sources now and then, you can't go wrong if you quote them from a published and edited source, in the form: Foo said, "Foo foo foo by so many to so foo" (dated and referenced to a published text, etc.). Not only is that on limits but it's excellent practice, in my opinion. If interpretation is needed—which it always is with the sagas, since they aren't impeccable sources for historical facts—it must be drawn from published commentaries and histories and not ventured by the editors (because that would be original thought or research).

On the other hand, I do see flaws in the wording of the prognostication you hold up to the light. The documents of that realm are written in a language that differs from our own: hmm, as you suggest, Churchill's language didn't differ much from ours; still less did Bill Clinton's, who, pace Hartley, did things not undifferently to a guy having an affair with his secretary nowadays (apart from phoning foreign heads of state simul satisfactus, of course). And what is the realm referred to here? ("the past", I presume—but need we get so flowery in policy advice?). The quote from The Go Between, resonant as it is, is inappropriate here—not least because to include an unattributed literary quotation as part of the text of a policy about the niceties of attribution could make us seem, at the very least, charmingly bonkers. qp10qp 17:29, 3 November 2006 (UTC)

The problem with "Primary" vs "Secondary" sources is that historians and scientists use the words differently. That is why I suggested "Sources requiring expert interpretation": . As for Churchill's language not being all that different, even that can be deceptively similar to our own. Consider simply the change in meaning of the word, "Gay." Consider the assumptions of race and class inherent in the language of 1940 that would be nearly impossible today. Much nonsense has been written because someone failed to understand that words change meaning. This is part and parcel of why we should rely on expert sources for interpretation even of recent history. I plead guilty to an attack of the purples, please improve the phrasing. Robert A.West (Talk) 19:03, 3 November 2006 (UTC)

I've rephrased to be more businesslike. Robert A.West (Talk) 19:26, 3 November 2006 (UTC)

I think this version is more effective. I'm still wondering if something if a broader statement can't be made somewhere about this, though. What would you think of something on the order of: "where relevant, widely-published documents may always be used as a source for purely descriptive claims about their own contents" ? That's a broad statement, but I think it may be accurate. -- Bailey^(talk) 19:46, 3 November 2006 (UTC)

In my opinion, it would be better. Let's avoid being too legalistic and leave easily comprehensible things short. It's not hard to tell descriptive claims from others. CP/M ^comm |Wikipedia Neutrality Project| 16:09, 4 November 2006 (UTC)

In my opinion (and not only mine) it's better to drop that often artificial and confusing classification of sources: it's perfectly possible to do without. Even "sources requiring expert interpretation" contains the same implicit misunderstanding about attribution. I'll make that a new topic below. Harald88 14:46, 5 November 2006 (UTC)

Can we save the debate?

I just the removed the exception. I may be reverted as I type this, but I'm thinking: can we, for the moment, agree to the page, while leaving aside the exception? We've already started a sub-workshop for the exception, but it's died. Does everyone agree the page is more or less OK otherwise? We can agree on that and move forward on the larger aim of replacing V and RS, while debating the contentious issue later? Marskell 22:07, 3 November 2006 (UTC)

I've been a bit absent from the debate due to being busy off-Wiki... the primary concern with that approach is that if we do so, consensus on the exemption will never be achieved--as those opposed to the exception will then have what they desire, and have no reason to further discuss anything. Further, I'd be concerned that were that to happen, and this policy were to go to the community as is and pass, is that it would then unleash a wholesale deletion spree against so-called "cruft" articles--many of which would no longer be kosher according to the (new) policy. Some may see that as a good thing, but there are many areas where the encyclopedia is currently both useful and correct, but which cannot be sourced under these provisions. Perhaps I'm failing to WP:AGF, but the proposal needs to have some sort of an exception for "unscholarly" topics in order to be acceptable.

To see why your proposal doesn't work, try this counter-proposal: "Can we, for the moment, agree to the page and include the exception? We've already started a sub-workshop for the exception, but it's died. Does everyone agree the page is more or less OK otherwise? We can agree on that and move forward on the larger aim of replacing V and RS, while debating the contentious issue later?" If WP:ATT were approved on these conditions, supporters of the exception would have no reason to reconsider the matter later--they win the debate automatically by virtue of stare decisis.

--EngineerScotty 22:26, 3 November 2006 (UTC)

• No, and I'm disappointed at the removal of the exception. Steve block Talk 22:33, 3 November 2006 (UTC)

"Those opposed to the exception will then have what they desire". Point taken.

"A wholesale deletion spree." No, this is a red herring (in this context). AfD has its "militias", which will operate independently of what gets determined here. There'll be no wholesale deletion (anymore than there is at present).

"Try this counter-proposal:..." This suggests accepting the unagreed upon, rather than ignoring it for the present. Now obviously my own feelings are clear if you read up the page, but I'm suggesting, fairly intuitively I think, that we proceed to innovate without an (IMO) radical move. We can still, broadly, remove V and RS, with this as the target, without having to agree on the point at issue. Marskell 22:37, 3 November 2006 (UTC)

• I think the issue should be resolved before we move on. How we resolve it, I don't know, but I think we should. What I don't get is that there is a consensus that exceptions exist. The sticking point seems to be in how to recognise that those exceptions exist. Perhaps the people who oppose the current language could suggest a way to break the deadlock. Steve block Talk 22:55, 3 November 2006 (UTC)
• Thinking cap on, how about this: We make three exceptions, and the third exception is Where community consensus dictates.? Steve block Talk 22:57, 3 November 2006 (UTC)

First Steve, I just have to say (repeat) that the "because it is, it should be" argument does not have a final basis here. This is the naturalistic fallacy applied to Wiki: because it is done, policy should allow it to be done. Yes, exceptions exist, but this does not mean we need to suggest them in policy.

To this point, a month on, no one has clearly articulated the clear and present danger the absence of an exception causes; quietly use IAR?—fine! But we don't need a get-out clause that amounts to "by the way, if you feel this policy is bullshit, use what you like." The ball has been punted around quite a bit, but your latest edit (while better than previous) amounts to more of the same. Marskell 23:20, 3 November 2006 (UTC)

Backing up a bit, and re-reading a second time, I see more merit in your change as a compromise (don't ignore my last!--I still mean it). I think I disagree with the headline point about community consensus, though. The community ups and decides to ignore The Times? Perhaps we can give more nuance to the headline. Marskell 23:31, 3 November 2006 (UTC)

• Edit away. But let's not forget that people still felt free to ignore WP:V and WP:NOR when they declared themselves non-negotiable. On that basis I can't buy the argument that the exception changes anything on that score. People who feel this policy is bullshit will feel that way whatever the policy says, and will have no trouble using IAR to justify their abuses no matter how well we hide it. Steve block Talk 23:34, 3 November 2006 (UTC)

We don't seem to disagree. You're asking for caveats despite the fact that much won't be changed. I'm saying, don't encourage what occurs, just because it occurs. I think if you read this page without thinking about the exception, it is a good page, and nothing you're saying contradicts that. And I must say: of course people will ignore this rule (or any rule). IAR is existential to the Wiki—it doesn't actually need a policy page, but will always occur spontaneously. The real irony I have found in this debate is people's attempts to delineate how IAR is/could/should be used... Marskell 23:53, 3 November 2006 (UTC)

(multiple edit conflict) I think that, paradoxically, IAR is itself the danger. We should never ignore the basic principles: attributable, descriptive research presented in as neutral a manner as possible. Those are the definition of quality for a quality Wikipedia article -- it is logically impossible to improve an article by ignoring them. If we don't say that loud and clear, some people won't get it. What we should ignore, on rare occasions, are the second- and third-order rules that usually make it easier to obey the principles. If we don't say that, some people will apply every jot and tittle of policy without comprehension. Robert A.West (Talk) 23:35, 3 November 2006 (UTC)

I made what I hope is a friendly edit, although I think (as I indicate below) that we need not be so specific. I inserted a no-exception-for-BLP sentence. I also rephrased and added "exceptional justifiction" in place of a repetition of "extremely rare". Robert A.West (Talk) 23:58, 3 November 2006 (UTC)

IAR is also a fundamental principle. And, actually, the others were and are neutrality and verifiability: material being not only correct, but verifiably correct. "Research", "attributable", "descriptive" - this is just one of the ways to set stricter limits. In fact, Wikipedia does and is supposed to contain a lot of things which are not connected to any research. If we say that it is impossible to improve articles while ignoring rules, most good contributors will ignore that; others will basically lose efficiency to negligible, zero or negative. All the policies need not just some exceptions, but general non-enforcing nature, serving as a justification for some otherwise questionable edits - sufficient, but not necessary.

A small example. Yes, 50% of material can be easily sourced. It will take 2-5 times more time than to create it at current 20-30% sourcing level. 90% can be, potentially, sourced. It will be at least 20-40 times more hard than at the current level. Now, imagine that someone can either write a 90% sourced 1-kB stub or a 25% sourced 30-kB article. These numbers are realistic, if not to say real. Now, how would he, actually, improve Wikipedia better - with that stub or a normal partially sourced article?

All rules need to be, at most, asymptotic ideals in regard to internal quality, but never laws. And all, actually, can and should be ignored when necessary, although held as general guidelines. I must admit that even neutrality, the least negotiable and the most accepted of the principles, has its limits: going beyond them destroys prose structure, readability and sense of an article. while it is more justified for some politicized topics, attempting to present some in-fiction information in a perfectly neutral way would be ridiculous. For that purpose, WP:NPOV already recognizes due weight and ignoring of non-notable minorities. But perfect, zero-tolerance weighing and equalizing can be not only extremely time-consuming, but pointless and harmful.

It is even more so for less important rules on how sources should be selected and cited, because excessive sourcing does not even improve correctness and verifiability level of the article. It can even harm the very idea it attempts to serve: checking 15 sources is easy, checking 300 sources is practically impossible, and misquotings, false sourcing, completely unreliable sources, fact cherry-picking and other deliberate or accidental errors find their way much easier into articles with large number of sources. CP/M ^comm |Wikipedia Neutrality Project| 21:49, 4 November 2006 (UTC)

IAR talks about making Wikipedia better, not worse. Adding original research or unreliable sources to an existing article takes it farther from the ideal, so makes it worse. Writing a stub today and improving it tomorrow is not IAR, it is the Wiki process (fourth pillar) at work. On the other hand, if 90% of an article "can be, potentially, sourced," then logically, the article must be 10% original research. That means 10% of the article is making Wikipedia worse.

Writing good, well-researched articles is hard work. It is also the right thing to do. IAR exists to help people cut through Wikilawyering and do what we are here for. Far too often, it is invoked to justify lazy, poorly-researched crap instead. That is why I call it, paradoxically, a danger. Robert A.West (Talk) 04:34, 5 November 2006 (UTC)

Not exactly. Not everything that can't be sourced (by most contributors) is original research. Most of the time these are things good but available sources just don't care to mention directly. FA, at best, may have up to 20% sourcing, 30% for summary style. Only some short lists approach 90% sourcing.

And, actually, if there was no IAR and everyone was preferring completely sourced stub to reasonably sourced article, we would not even be discussing this matter, and most likely would never hear about Wikipedia at all. Know the story of Nupedia? Wikipedia would reach the same level, at best. The rules would leave about the same number of people willing to do hard and pointless work. Most of our good articles (I'm not about that GA symbol) are written by people with certain knowledge of the subject. An average article has many hundreds of statements. Sourcing them all would make no sense. More importantly, it would destroy readability, as instead of writing prose and sourcing it editors would have to stick to barely coherent sourced pieces. Though, actually, the selection of editors would be much smaller, as the process is painstaking and pointless. Adherence to the general goal of writing encyclopedic articles, and not to the process, is what brought Wikipedia where it is today. CP/M ^comm |Wikipedia Neutrality Project| 07:25, 5 November 2006 (UTC)

With four achive pages and nowhere near consensus, this is a dead parrot --Audiovideo 14:42, 5 November 2006 (UTC)

Another thought

I'll risk trying patience and take another stab at the "good judgment" paragraph.

• Formal rules are not a substitute for good judgment and intellectual honesty. Formal rules cannot always determine whether material is reliable, notable, or relevant to a particular topic. Formal rules, applied without understanding, will exclude some reliable sources, include some unreliable ones or both. The key principles of this policy are non-negotiable and apply to all articles without exception, but the proper application of those principles may require careful thought.

No exceptions to the principles; the details may not cover all cases. If someone has a strong case for an unusual source's being reliable -- or for not trusting a facially-reliable source in a particular case -- that editor can quote this paragraph to be allowed to make his case. It will not make it for him. Robert A.West (Talk) 23:15, 3 November 2006 (UTC)

That wording works for me. ≈ jossi ≈ t • @ 00:32, 4 November 2006 (UTC)

I added a bit after that to require people using this paragraph to justify their case in detail. Bi 08:20, 4 November 2006 (UTC)

Clarification on use of self-published sources about themselves

I decided to be bold and add this bit of clarification:

• To clarify, the above conditions apply to the source material itself, not statements in Wikipedia articles about the material. Thus, if a questionable source makes a contentious claim, then even if an editor intends to rephrase the claim in neutral language, it is still not permissible.

This should address User:Atomaton's earlier concern -- as well as the problem I faced -- of people quoting gratuitously from self-published sources.

Although, the proposal still does sound a bit too woolly to me, and I'm not sure I'll fully support it even after this. Bi 08:20, 4 November 2006 (UTC)

How to cite (prose attributions)

This sentence -

"This is the attribution, within an article's text, of a claim to a specific source — "According to The New York Times, ..." — thus conveying that Wikipedia does not necessarily support the view."

- looks a bit improper to me. Wikipedia does not support views, after all. I suggest to reword it in some way. Ideas, objections? CP/M ^comm |Wikipedia Neutrality Project| 16:29, 4 November 2006 (UTC)

I rephrased to emphasize that prose attribution means either that the matter is disputed by other sources, or the editor believes that source bias should be clearly indicated in the text. Neither is a position of Wikipedia, which I agree has no positions beyond its own policies. Robert A.West (Talk) 21:53, 4 November 2006 (UTC)

Delete all talk about primary, secondary etc. sources?

Connecting to the above discussion about expert sources, see my analysis on Wikipedia_talk:Use_of_primary_sources_in_Wikipedia#Delete_all_talk_about_primary.2C_secondary.2C_tertiary.2C..._sources.

It demonstrates the general inutility of that classification. I take the liberty to also add part of a comment of Qp10qp as he said rougly the same in other words and added to it[2] [3] :

Everything is simpler without the classifications. [...] I agree with you that everything clears up if the distinctions are dropped. "Published" is the best criterion, in my opinion. A collection of letters, edited and published, is no different in reliability from what we call secondary sources. [..] And obviously we can't analyse primary sources ourselves, only quote the analysis of published historians, critics, commentators etc.

I also agree with that, but like to put attention to the fact that what matters for sourcing is not so much what the source is, but the use of that source -- the uses of a source may not go beyond the contents of that source, as is currently rather well explained in WP:NOR. Sources that do not interpret the contained data can of course not be used as reference for interpretation, while sources that only discuss interpretations of data are generally not reliable data sources either.

Note that the current sentence

For scientific subjects, Wikipedia should use peer-reviewed articles and scholarly books as sources, rather than raw data and unpublished works

is happily free of jargon already but IMO erroneous: it should either read

For scientific subjects, Wikipedia should use peer-reviewed articles and scholarly books as topics, rather than raw data, or

For scientific subjects, Wikipedia should use peer-reviewed articles and scholarly books as sources, rather than unavailable raw data and unpublished works (I'm not entirely sure what the intention is of that sentence, as the last is valid for any subject, not?).

All the time new (and not so new) editors are confused by this source-classification issue that shifts attention from the essentials of attribution to useless jargon that even misses the point. Therefore I propose to simply not mention those terms anymore except for a single mention so that people who remember that phrasing can find the relevant new passage.

Harald88 15:50, 5 November 2006 (UTC)

"Unavailable raw data?" That would make the passage more confusing, not less. Once one has internalized the policy, all these things are obvious: we explain them for the newbies and for those confused about specific issues. What is the meaning of "unavailable" in this context? Perforce, the data are available, at least to the editor that included them. Obviously, a distinction is being drawn, but what might it be? Does this mean that published raw data are fair game as the basis for an article?

The existing text is clear -- raw data are not to be used as the basis for an article because that way lies original research. It is not a new principle, it is a clear explication of a major instance of a basic principle.

As for "Source requiring interpretation," it says what we mean, and is therefore not jargon. Would you prefer "Some sources require intepretation?" The section militates against an error that is not infrequent. Let me give one real-life example out of many. There is unanimity that George Washington was effectively unopposed in 1789. One can cite the National Archives or any number of historians to the effect that the Constitution never would have been ratified if there was any doubt on who would be the first President. One day, an editor (of an article that made passing mention) looked up the raw vote, and noticed that there were electoral votes cast for other candidates, which there were, since the original Constitution made the runner-up the VP. Nevertheless, he added his own interpretation that the election was really contested, because the data made it "obviously" so. Sourced statements providing the consensus interpretation were "improved" with this editor's original research. The editor defended his use of the data because it was published in a secondary source, denied it was original research since he was "only stating what is in the data," and as a result, a passing mention grew to three paragraphs, including much original research, that gave the impression of a major historical controversy where there is none.

The whole point of the primary/secondary distinction is to make clear that this sort of nonsense is original research. Tax protesters muddle articles with OR constitutional analysis when their sources are attacked as fringe positions, for the good and sufficient reason that they have been rejected so often that they can get you fined for wasting the courts' time. ({{legally frivolous}} was designed just for those sources when we must include the fringe position under NPOV.) A court case interpreting the Constitution or the statutes is preferable to an argument about what is "obviously" in the text. I see much value and no harm in stating that principle, in more general form, in this policy, especially since it replaces RS as well as V. Robert A.West (Talk) 16:46, 5 November 2006 (UTC)

I think I'm persuaded by User:Robert A West's argument, especially since his position is backed up by real examples. Bi 17:34, 5 November 2006 (UTC)

Sorry for not providing examples myself, here's my reply:

Robert, thanks for the comments.

- About the "BTW": as that's really a separate subject to which you here gave useful input, I'll start it below.

- With jargon I meant "primary/secondary/tertiary sources". Sorry if that was not clear. I certainly find "Source requiring interpretation" a clear description in the context of "for the purpose of the citation".

Thus I was evidently not clear enough with stressing that it's not exactly the sources that require interpretation, but that instead Wikipedia does not allow interpretations of sources that advance positions --there must be consensus among editors that the source itself advances the position that the article claims that it does. It makes no difference if it's an unclear document with data or an opinion about such in antique English, when the meaning is contestable it can't be used for an expression-of-fact about its meaning. As Bailey and CP/M plus myself argued above under "Sources requiring expert interpretation", documents may always be used as a source for purely descriptive claims about their contents.

About your real-life example, it's very useful and I admit that I was too lazy to "sell" my arguments with an illustration of my own. Thus now we take yours:

The national archives and historians give a wealth of reference material that need to be cited of course. Also the election results can be stated, there can be nothing wrong with that either. And that's all. There is no need for claiming that some sources are somehow "better" or "worse" as that wholly misses the point of WP:NOR. Instead, no claims may be made that go beyond any of the sources - it's exactly the same for inferences that are presumably the opinions of historians ("secondary sources") but in fact go beyond them - in this case, claiming that results were contested without citing a source that made that claim.

These jargon-based rules are counterproductive and actually introduced their own nonsense; their introduction has not helped to avoid such incidents as I have encountered frequently, where referring to WP:NOR rarely made a difference. IMO it is more effectively replaced by practical instructions such as:

When in doubt, editors should ask themselves: who makes this claim, and in which publication? If requested, any claim must be verifiable by citation of a reliable source for that claim.

And of course, also such examples as "A court case interpreting the Constitution or the statutes is preferable to an argument about what is "obviously" in the text" may be clarifying. There is no need for introducing jargon.

BTW, did you actually read the proposal to make a whole guideline that tries to explain the meaning of that unnecessary jargon because there is so much confusion about it? Our purpose for this merge is to remove as much Wikipedia clutter as possible, and I'd say that that should include such causes for more clutter.

Harald88 18:28, 5 November 2006 (UTC)

The point, as already made in the proposal, is that "Material added to articles must be directly and explicitly supported by the cited sources. The analysis or synthesis of published material to produce unpublished interpretations or to advance a position counts as original research" and is not allowed. Introducing "primary" and "secondary" terminology does not make things more clear and in fact, as shown numerous times, simply leads to confusion. There is no reason for the distiction as all sources must be carefully used and one can use secondary sources to inappropriatly advance an original thesis even easier than a primary source. People do it with newspaper articles every day. WAS 4.250 17:48, 5 November 2006 (UTC)

So much talk about why this "primary"/"secondary" distinction has been shown "numerous times" to be so evil... but I've yet to see one single example where this distinction leads to genuine confusion which can't be cleared up with a quick explanation. Again, where are the real examples? Bi 17:52, 5 November 2006 (UTC)

Above I started with citing such a real-life example, where a proposed guideline that tries to explain it actually messes up due to it. Harald88 18:50, 5 November 2006 (UTC)

Read the talk pages of WP:NOR to begin with. WAS 4.250 17:59, 5 November 2006 (UTC)

I don't see any real-life examples there, just more of the same old abstract arguments unencumbered by facts.

Let me rephrase my question: Has there ever been any specific Wikipedia article whose editors were hampered by a genuine confusion over this "primary"/"secondary" distinction which couldn't be cleared up quickly, in the context of that article? You know, something like User:Robert A West's example of United States presidential election, 1789. Bi 18:07, 5 November 2006 (UTC)

The latest one, as WAS pointed out, you can find on the WP:NOR talk page: See [4]. You can find there such comments as "this isn't an OR issue, it's a primary source issue. Wikipedia is clear on primary source being preferable to secondary sources in the case of sci/med".

As I pointed out, the same instructions can be made much simpler and more accurately, so that editors can put essential concepts in their minds instead of cluttering their thinking with irrelevant stuff that they don't really understand. WP:NOR is in its essence very straightforward.

And on GPS we were confronted with an editor who kept interpreting overview articles in a way that advanced his own, and not the author's position ([5] plus other parts of Talk page). There is no need to make a special case against abuse of "primary sources". Harald88 18:52, 5 November 2006 (UTC)

Abuse of sources is a serious matter, and I, for one, regard it as a subtle, dangerous and pernicious form of vandalism, and I would be happy to see a stern warning on that front. This subsection tries to address something different that is often done by well-meaning but clueless editors. Robert A.West (Talk) 20:01, 5 November 2006 (UTC)

As you hopefully understood, I did not accuse anyone of bad intentions but original research which is abuse of sources in the context of providing references for the claims that the article makes. I do think that for clueless editors my above suggestion ("editors should ask themselves") will be more effective than a set of rules that is based on unnecessary definitions. Harald88 21:11, 5 November 2006 (UTC)

Then I misunderstood. I have seen many situations where editors have cited sources that said the exact opposite of what is being asserted – some transcend my ability to presume good faith. We seem to agree on the need to strike a balance between too much generality and too much specificity. Robert A.West (Talk) 21:36, 5 November 2006 (UTC)

The GPS and Depo Provera examples looks interesting... however,

• the GPS example, as West pointed out, is clearly not a case of confusion over definitions -- just a person trying to push his POV by acting abusive;
• the Depo Provera example is simply someone throwing out nonexistent policies to advance nonexistent positions.

Thus neither of these are exactly cases of newbies getting confused over the "primary"/"secondary" distinction. However, if User:Harald88 means that making such a distinction gives the wikikooks more leeway to do bad things to Wikipedia, then I may agree.

Anyway, as the proposed text stands, the "primary"/"secondary" distinction only kicks in for articles on history, not articles on science. Which makes a bit of sense: one must pass through peer review to get a scientific paper published, while any prominent person with an agenda can write up a historical document. Given User:Robert A West's Washington example above, I'd think that such a distinction is still useful in the case of historical topics. Well? Bi 11:06, 6 November 2006 (UTC)

I added some rantage below on the possibility of conflating this "primary"/"secondary" thing with the idea of reliability. Bi 21:00, 6 November 2006 (UTC)

improving sentence about scientific topics

I suppose from Robert's comments above that the intention of that sentence is as I first interpreted. Thus, do you agree with the correction as I proposed, by replacing the ambiguous "subjects" by "topics" and to remove the suggestion that for other subjects "unpublished works" may be used? And I happily include Robert's reformulation which helps to make even more clear what that sentence is about. I thus obtain:

'For scientific topics, Wikipedia should be based on topics in peer-reviewed articles and scholarly books rather than raw data.

What do you think, is that the clearest? Harald88 18:28, 5 November 2006 (UTC)

For scientific topics, use peer-reviewed articles and scholarly books as sources, rather than raw data and preprints.

This preserves the change subjects->topics, but speaks directly of choice of sources, rather than choice of topics for articles, which is not what I think Harald88 intended. It replaces 'unpublished works' with 'preprints', which is more specific, and I hope still makes clear that other preliminary documents (such as are sometimes available on departmental websites) should be approached with caution. Robert A.West (Talk) 19:11, 5 November 2006 (UTC)

Nice, I like it! Preprints are published but not peer reviewed, and therefore they are only reliable sources for scientists' opinions. Harald88 19:37, 5 November 2006 (UTC)

I made the change. I also changed the heading to "Some sources require expert interpretation" to emphasize that this is a caution and a special instance of a rule already specified, not a new rule nor an attempt to provide a rigid classification. Robert A.West (Talk) 21:42, 5 November 2006 (UTC)

Another facet of popular culture

Working through the DRV on Girly and the questions of its sourcing have led me to another solid example of a problem on this page, specifically the blog Websnark. Websnark is written by Eric Burns and Wednesday White. Neither of them are professional researchers in the field of webcomics because, well, there aren't any professional researchers limited solely to webcomics. The research community there is still based very heavily in the fan culture.

This is the case for a lot of things - Buffy the Vampire Slayer has an annual conference devoted to it that straddles the line between fan and academic, and comic studies (one of my fields) is deeply indebted to its roots in the fan communities (Something that's going on as a debate on the major English-language listserv for comics studies in the world).

I'm not sure that this is actually a special case - or at least, not one that isn't functionally identical to one we've already dealt with. The issue here is that things like webcomics or Buffy or superheroes are not just articles on the fictional texts, but articles about their role in culture - articles that are, in part, about their fandom. Buffy the Vampire Slayer isn't just about a seven-season television show, but about a phenomenon in popular culture - it is, in part, about the academic conference, and the fan response to the show.

This, I think, captures some of the heart of the popular culture exception. It's not that popular culture lacks good sources - yeah, there are books on Buffy, and articles, just like there are on webcomics and print comics. It's that the fan sources are fundamentally primary sources for the sort of article Wikipedia writes about these texts - articles that engage not only their fictional content but their real-world components, which includes the fandoms.

So it's not that comics fans are reliable sources about Batman. They're not. They're primary sources about Batman as a pop-culture phenomenon. And they can rightly and justifiably be reported that way. Likewise, it's not that Websnark is a vital source for interpretation of comics. It's not, and much as I love it, I'd never cite it in a paper on any of the webcomics it's covered. What it is is a vital source for describing what webcomics have done in popular culture.

Sorry if this is unclear. I'm somewhat drained, and probably not at full coherence, and it's a subtle point on top of that. Please let me know if there's something that doesn't quite follow, as I promise, it is a key distinction that really does disentangle the popular culture issues and the sort of intuitive sense we all seem to have that there's something going on with popular culture that we haven't quite accounted for here. Phil Sandifer 21:37, 5 November 2006 (UTC)

Actually, as I reread this, it may be that we need, instruction creep be damned, to go write a guideline about popular culture that addresses this. Because it is complicated, and it is a very weird aspect of "sources writing about themselves." And it also ties in with another chronic frustration of popular culture articles, the way they degenerate into textual minutae and fancruft. There's a lot of confusion over the fact that our article on a fictional character is a fundamentally different thing from the sort of article that ought appear in a fan-targeted encyclopedia on the topic. And the nature of that confusion, I think, obscures what a primary source for popular culture articles actually is. Phil Sandifer 21:40, 5 November 2006 (UTC)

A final, perhaps more succinct trio of phrasings here that might capture what I'm saying...

• Kirk/Spock fanfic isn't an unreliable secondary source on the gender politics of Star Trek, it's a reliable primary source about Captain Kirk as a cultural icon.
• A thread on a LiveJournal community for the TV show Supernatural where lots of people hate the most recent episode isn't an unreliable secondary source about the aesthetic quality of the second season, it's a reliable primary source about the relationship between the show and its fans.
• A Newsarama thread about how to resolve a continuity issue in a recent issue of Spider-Man isn't an unreliable secondary source about the Marvel Universe, it's a reliable primary source about the role of fan theories in comics fandom.

In all of those cases, it remains an open question whether the source is of sufficient notability to be included in the article, but that's a question of NPOV - whether the view is significant enough to report. But the answers to that question come from things like "Is the piece of fanfic one of the ones recognized as historically significant to the fandom," "Is the community a general one for Supernatural, or a slash community of people interested mostly in reading Dean and Sam as gay lovers," or "Was it a big thread on Newsarama with a large number of participants." Those aren't questions of reliability, and they're not the purview of this page. Phil Sandifer 22:01, 5 November 2006 (UTC)

I don't think this is unclear; you are very coherent, as usual. It seems to me that an article about Buffy fandom, or Star Trek conventions, or fan art, or Rocky Horror Show audiences, or comics subculture has to be a separate article from the one about the source art itself. In other words, you have one article about Buffy, referenced from secondary sources or from within the world of the episodes, and another about the fandom, which could be referenced through fansites because they would qualify as sources about themselves. The muddling up of the two is always uncomfortable and often difficult for someone who doesn't know the details to detect. This even affects historical subjects; for example, the articles on Lucrezia Borgia and Elizabeth Báthory are in my opinion contaminated at the moment by unsourced information that has its basis not in secondary sources but in popular culture; for which reason, I would like to make an article on Lucrezia Borgia in popular culture, one day, along the lines suggested by the recent list initiative of clever User:Durova. qp10qp 22:20, 5 November 2006 (UTC)

Exactly - and for smaller topics (Individual episodes of Buffy, and things with smaller fandoms), such splitting is probably unnecessary. Phil Sandifer 22:41, 5 November 2006 (UTC)

Phil's examples above are excellent, though I would note that in all three cases, the "primary source" use can only be used as factual references to support the respective assertions that: "Kirk is an icon who features in slash fanfiction"; "some of Supernatural's fans debate the show on LiveJournal"; and "continuity issues in the Spiderman comic are debated by some of the comic's fans on Newsarama". But Wikipedia can't say anything about the relevance or meaning of such assertions. That would need reliable secondary sources that respectively say (using different examples from Phil, and making them up entirely): "Professor X has stated that the existence of Kirk/Spock slash fanfiction reflects the suppressed homoerotic tension seen between Kirk and Spock in the original TV series"; "Commentator Z has speculated that the lambasting of the second series of Supernatural by fans in online fora may have contributed to the decision not to commission a third series"; "a study by researcher Y showed that differing background details in the Spiderman comic was the most common continuity error discussed by online fans of the comic". I think I've got most of the phrasing right there. Carcharoth 23:38, 5 November 2006 (UTC)

Quite true. I might argue, in the second and third cases, that one could do away with the descriptions of locality - "Many fans have invented theories to explain this inconsistency, ranging from X to Y," or "The second season of Supernatural met with a mixed reception among fans, most notably with episode Z, which was described by one fan as..." Still reporting factual details about the reception in fandom and in culture. Phil Sandifer 23:53, 5 November 2006 (UTC)

(edit conflict)"Professor X?" I find the image of Patrick Stewart lecturing to a group of young mutants on the cultural importance of Star Trek amusing.

Seriously, I see this as the camel's nose. To start, the iconic status of Kirk is, already, an interpretation. Granted it is pretty obvious, but saying even so much without being confident that a reliable source can be found is putting our toes over the OR line. Asserting the importance of K/S in that iconic status, or vice-versa puts at last one foot firmly on the other side of the line, and by the time the article actually says something interesting, we have left the line far behind and are doing blatant original research.

IMO, what little can be validly done on such topics from such sources is already covered under the "good judgment" rule. Robert A.West (Talk) 00:03, 6 November 2006 (UTC)

The word "iconic" may be working as a bit of a red herring here, in that (for obvious reasons) such descriptions suffer NPOV problems, among other things. Let me try again - Captain Kirk is an article about a cultural phenomenon - a cultural object. Without making any judgments of the importance of that phenomenon, one can observe that a significant part of that phenomenon is K/S, and one can further cite, say, a large archive of K/S stories as a primary source of evidence about that. Phil Sandifer 00:13, 6 November 2006 (UTC)

But, significant in what sense? Percentage of fans who read it? Who write it? Absolute numbers? If it is significant, that means cultural impact, but does slashfic really have any impact outside its own base? Those are all fascinating questions that (for all I know) may well have been studied, but they need careful study. Saying that K/S is "significant" without defining what that term means is, IMO, not to say much of interest. Robert A.West (Talk) 02:36, 6 November 2006 (UTC)

• I am assuming that the point is significant in relation to the article. The mere existence of some quantity of K/S fanfic would be valid as a minor point in a more general article, and could be validly sourced from a primary source. This much could be done under the proposal as it stands. Robert A.West (Talk) 02:43, 6 November 2006 (UTC)

I'm not sure I agree with the assessment that "the mere existence of some quantity of K/S fanfic would be valid as a minor point in a more general article" using the fiction itself as a primary source. This seems somewhat like original research to me. Generally, we use primary source documents to source statements about the contents of those documents, not about their general state of being. Saying that fanfiction has been written and then pointing at some fanfiction to demonstrate does not seem substantially different from stating that the sky is blue and then using the sky as your source.

The current text of OR actually only allows us to use primary sources that have been reliably published, and then only to to describe what is directly said by the primary source. These seem like perfectly reasonable restrictions to me. After all, we're not supposed to be using sources to prove that something is true; we're supposed to be proving that our claims have been made by reliable third-parties. This gives the reader some indication that our claims are likely to be true, but also that they're relevant and noteworthy in some way.

This is just my take on it, but for the most part, I'm of the mind that most primary-source documentation of fan reactions would qualify as "sources requiring expert interpretation" most of the time. Even statements like "many fans think" involve interpretation -- generalizing about the opinions of a massive demographic from a limited sampling of responses is something I'd prefer to have a reliable source do. -- Bailey^(talk) 03:59, 6 November 2006 (UTC)

That's a POV problem, then, not an original research one. You simply restate the phrasing to reflect the sourcing. It doesn't require expert knowledge to rephrase to reflect sourcing. If it did, we couldn't write anything period. It's not original research to show that fans don't like something, but it is a point of view issue if we use that in a manner which isn't relevant or conflates the issue above it's relative worth. Like I've said before in this debate, a lot of what people are trying to protect against on this page is already hammered out in the neutral point of view policy, which I might suggest we could perhaps look at once we get done here. And I think the goose is nearly cooked. Steve block Talk 18:24, 6 November 2006 (UTC)

Um... like "according to this survey which I conducted with anecdotal evidence combined with my gut using this wacko methodology I just slapped together a minute ago, most netizens don't like Star Trek"? Sounds like original research to me, and a very slipshod kind of research at that. I think I'll also prefer something more rigorous. Bi 18:57, 6 November 2006 (UTC)

Um... seriously, if you want me to pull that apart, I will. First up, conflict of interest, second up, neutral point of view, third consensus. Whatever you want that is more rigorous exists already. Hey, here's one for you, why don't I just make up articles on everything and anything I want? Steve block Talk 21:00, 6 November 2006 (UTC)

My point was, this "most netizens don't like Star Trek" kind of thing is original research. In fact, it's pretty much right up there with legally frivolous interpretations of the US Constitution. Bi 21:12, 6 November 2006 (UTC)

Regardless of whether it falls under NPOV or NOR, we're still only allowed to use primary sources to describe the literal contents of that source -- otherwise, third-party 'expert interpretation' is required. A thread on a LiveJournal community can be a primary source about the relationship between a tv show and those specific fans, but not about the relationship between the show and fans in general. Why? Because those LiveJournal users aren't saying anything about any other fans, and they also aren't qualified to say anything about any other fans. Even if you've got a really huge thread, with many, many posts, without generalization, you're almost always going to be talking about a statistically insignificant percentage of the any television viewing audience. This would virtually always violate NPOV. Stating the professed opinion of 2 or 12 or even 200 fans is always going to cause issues of undue weight if it's not possible to demonstate in the article why the opinion of a tiny fraction of non-expert fans is relevant. Yes, this partially an NPOV issue, but the question becomes, why allow a special exception for these kinds of cases with (regards to the reliability of forums, fansites, et cetera) if these kind of exceptions will never reasonably pass NPOV? When will this be helpful? -- Bailey^(talk) 23:31, 6 November 2006 (UTC)

It confuses things to claim a source as primary or secondary when it is how you use the source that makes it one or the other. WAS 4.250 09:02, 6 November 2006 (UTC)

It is not needed to distinguish between popular culture and other things when it comes to sources because all sources need to be carefully evaluated concerning what they are used for. Don't quote 1911 Britanica about racial differences; do quote slashdot about a security problem with mozilla; don't quote some blog about who is gay; do quote the writer of a piece of fiction about aspects of that fictional universe. WAS 4.250 09:02, 6 November 2006 (UTC)

The jihad against "primary"/"secondary" aside, I'm also not very sympathetic to those people who want a Wikipedia article for each of their favourite TV show characters. There's a reason why Memory Alpha was created, people. Bi 11:18, 6 November 2006 (UTC)

I'm inclined to agree, which is the flip side of this. If a character, episode, etc exists below the level at which there is any observable cultural impact (whether among a visible and verifiable fan culture or in the general public) of it specifically (as opposed to of the overall show), it does not deserve an article. Our articles about fiction are about fiction in the real world. Fictional biography is a small part of that, and not the most important part. Phil Sandifer 14:56, 6 November 2006 (UTC)

Conflate "primary"/"secondary" with reliability?

Given the ways that people have tried to muddy the waters over "primary" and "secondary" sources, I'm starting to think that it may be better to simply define everything in terms of levels of reliability. As far as Wikipedia goes, I can discern 3 levels of reliability:

• S-reliable: Claims in the source can be paraphrased directly as if they were fact. E.g. "Abraham Lincoln was an American politician..."
• Q-reliable: Claims in the source can be quoted, but cannot be paraphrased as fact. E.g. "Vorilhon claims to have encountered a flying saucer..."
• Unreliable: Claims in the source cannot be used at all. E.g. tiny minority claims.

Thoughts? Bi 14:04, 6 November 2006 (UTC)

(Make that "Some claims in the source can be quoted...") Bi 17:22, 6 November 2006 (UTC)

I don't think it's a question of reliability of information but of reliability of publishing process. If Michael Crichton is reported in reliable newspapers saying that global warming is a myth, we can use his referenced point of view in an article confidently, unreliable though a pop author might be on the subject, and the editorial community will doubtless oppose it with more authoritative voices from other sources. "Published" is in my opinion the threshold criterion. Of course there will be arguments about quality of publication and indeed about what publication means, but the basic concept is easily graspable: for material to be used on Wikipedia it must already have been reliably published. The content is a matter for editors to grapple with at the editing level of articles, not for policy. qp10qp 14:48, 6 November 2006 (UTC)

Actually, the reliability of the publishing process is what I'm getting at, as well as the topic of the Wikipedia article. Thus whether a particular source s is (to use my notation above) S-reliable, Q-reliable, or unreliable will depend only on

• the process by which the source was published
• the topic of the source
• the Wikipedia article which the source will be used for

And nothing else. Also, if a source states that p is true, then the article should only do one of the following:

• If s is S-reliable: "p is true", or "according to s, p is true".
• If s is Q-reliable: "According to s, p is true."
• If s is unreliable: Just keep quiet.

For example, Pittsburgh Tribune-Review says, "According to Crichton, global warming is a myth". This leaves only 3 possibilities:

• If Pittsburgh Tribune-Review is S-reliable for the topic at hand, one can simply say "Crichton said that global warming was a myth."
• If Pittsburgh Tribune-Review is Q-reliable for the topic at hand, one has to say "According to Pittsburgh Tribune-Review, Crichton said that global warming was a myth."
• If Pittsburgh Tribune-Review is unreliable for the topic at hand, keep quiet.

This means, one won't be allowed to "lift out" Crichton's claim from inside the newspaper quote and just write "Global warming is a myth." Which will be right, since the Pittsburgh Tribune-Review never once claimed directly that "global warming is a myth". Bi 17:20, 6 November 2006 (UTC)

Just as reality is more correctly understood when entities are not multiplied beyond necessity; so too, communication is more effective when terminology is not introduced unless needed. The concept that things like reliability and expertise and accessable and credibility have a range and are not black/white is/isn't binary adjectives/evaluations and further that they apply to the use made of a source and not the source standing independent from a specific use is not better expressed in an invented terminology. WAS 4.250 19:22, 6 November 2006 (UTC)

Well, the reality is that all this fluffy-puffy Zen talk about "range" and all that does nothing except to confuse the hell out of honest contributors and give wikikooks more weasel room. And as far as I can see, whatever kind of "range" stuff you have, ultimately any statement in any source can only be used in the 3 ways I stated above. Bi 19:59, 6 November 2006 (UTC)

(after edit conflict) The place of judgment is deciding where a source is S-reliable, Q-reliable, or unreliable on a given topic. For example, one of the standard histories of a certain American State was written in the early forties, and has a tendency to go wandering into the author's unfavorable opinion of the New Deal. It is S-reliable on the last royal governor; I doubt it is even Q-reliable on FDR, because there are much better sources on that line of argument. Septentrionalis 20:13, 6 November 2006 (UTC)

Interesting... I must check it out. Which source is it, by the way? Bi 20:44, 6 November 2006 (UTC)

While I find the S/Q/U distinction intriguing from a theoretical viewpoint, and would love to argue its finer points, the mode of thought necessary to grasp and interpret it will be foreign to most editors. Explaining things in terms that will require explanations is not useful. Robert A.West (Talk) 20:27, 6 November 2006 (UTC)

Hmm... but surely you agree it's useful to distinguish between sources whose claims can be stated as fact, and sources whose claims can't be stated as fact? Or something.

As things stand, there are quite a lot of things in the existing Wikipedia policies which need some explanation anyway. Of course the policies tend to use phrases like "reliability", "primary source", "neutrality", etc. which are comprised of common words, but these phrases have specific idiosyncratic meanings in the context of Wikipedia. And it seems that the proposed WP:ATT won't be that different. Also, I see a danger in using common words to denote more specific meanings -- the specific meanings can easily get confused with their more common meanings in everyday language. (But maybe this worry is unfounded...)

Well... I don't know. Maybe call the different levels something else? Perhaps something like reliable, quotable, and unusable.

(OK, OK... this whole thing is just a random idea I got. Perhaps I should just stop now...) Bi 20:44, 6 November 2006 (UTC)

(edit conflict)

• R-reliable: Because of remoteness of time or language, direct quotation from the source without expert interpretation is apt to mislead. e.g. "Ruth uncovered Boaz's feet and lay down." (Ruth 3:7) The translation is literally accurate, but those words have figurative meaning as well. A reliable source should be used to explain. Robert A.West (Talk) 21:01, 6 November 2006 (UTC)

Now that's a problem. :( Bi 21:05, 6 November 2006 (UTC)

Not to be obtuse or anything, but what exactly is being implied here with Ruth and Boaz? I can guess, and if I'm guessing right, this is a very good example! I too find this theoretical approach very interesting, and I hope someof it can make it into the FAQ, or "more detailed explanation" bits that can hopefully be accessed "behind" this policy, for those who are interested, or want the detailed explanation. Carcharoth 23:43, 6 November 2006 (UTC)

The word "foot" is often euphemistic in both ancient and modern Hebrew. In modern Hebrew, this phrasing can be slang for fellatio, and some commentators have assumed that it has always meant that. Others have pointed out that a foot is sometimes just a foot, and lying down at a man's feet so that he notices you when he wakes up is quite credible as a combined act of submission and demonstration of chaste virtue. Absent commentary, a modern English speaker would miss the innuendo, and a modern Hebrew speaker might not be aware how different Archaic Biblical Hebrew actually is from the modern language, just as some speakers of modern Greek overestimate their ability to understand Attic. Robert A.West (Talk) 00:08, 7 November 2006 (UTC)

I strongly suspect that is the sort of thing that people had in mind when they first started making the primary/secondary distinction. The source need not be millenia old to have such a problem. IIRC, one letter of F. Scott Fitzgerald reads, "He is one of my gay friends." This meant, more or less, "party animal," and many readers would understand just that. Many have never seen the word used in this sense and would be misled. Robert A.West (Talk) 22:09, 6 November 2006 (UTC)