Wikipedia talk:No original research/Archive 19

Archive 15

←

Archive 17

Archive 18

Archive 19

Archive 20

Archive 21

→

Archive 25

Posters

If there is a poster of a sports event (specifically, one like boxing or wrestling), is it OR to state who is on the poster? Someone keeps reverting my edits on one article saying it's OR to state who is on the poster (even thought the poster is right in the infobox). - unsigned

No. Nor can I legitimately say your last sentence ended with a period. That's forbidden OR: someone else must say it outside Wikipedia. I'm not even sure if I can say you had a "last sentence." Perhaps its OR to respond to you: I didn't read that you'd posted your question or read your question outside Wikipedia. - Minasbeede 19:11, 7 July 2007 (UTC)

It is original research but we allow original research with regard to images as a special case so long as no one claims different. For example, I can take a picture of a moth and put it on the moth page with the caption "black moth". Then when some one objects that it is in fact the rare dark red fuzzy butterfly, I must remove it from the moth page. Or move and relabel it if I reseach it and find I agree. WAS 4.250 19:31, 7 July 2007 (UTC)

Proposal: more slack for this guideline

I feel we should allow OR when something is bloody plain obvious. If an article were to say: in this video, "such and so" is said, wikipedia should be allowed to do so; even if no notable reliable source exists that says that the video says "such and so" provided that anyone can easily verify that the statement is true by simply watching the video.

In other words: things which are self-evident or almost trivial, should not have to pass the RS criterion. It is ludicrous to demand reliable sources for air is something we can breath or daylight originates from the Sun.

Rationale: the purpose of RS and OR is guaranteeing some level of reliablity. Slacking up as described does not compromise that. — Xiutwel (talk) 10:01, 9 July 2007 (UTC)

...air is something we can breath and daylight originates from the Sun. (Everyday Oxford Dictionary 1981 pp24,712) — This is not a case of OR — Newbyguesses - Talk 04:47, 11 July 2007 (UTC)

I'm also concerned about OR being trotted out in cases like this, usually to poke holes in pop-culture articles. But if someone is describing a video, the video is in itself a published source, so it's not OR. I do however think that, there should be somethinging in policy about "general knowledge" to take care of facts like "daylight originates from the Sun". Squidfryerchef 11:58, 9 July 2007 (UTC)

When people disagree, citing is our mechanism to resolve. If you say "the sky is blue" and no one contests it then no problem. If someone says is is black or grey more often than blue and sometims much of it is white then you need to cite a reliable published source. For some people it is general knowledge that there is a God. WAS 4.250 13:18, 9 July 2007 (UTC)

Citations need to be provided as per WP:V - "Editors should provide a reliable source for quotations and for any material that is challenged or is likely to be challenged" If something is general knowledge, nobody will challenge it and no source is needed. This isn't a question of OR. Tim Vickers 16:51, 9 July 2007 (UTC)

In the example you give, Xiutwel, the video is a primary source, and using primary sources is allowed so long as you describe and don't interpret. So you could watch the video and report "Sammy kissed Lisa when they were sitting in the ice-cream parlor." But you could not say: "And this was evidence that Sammy isn't gay, which lots of viewers previously supposed he was, because he seemed very close to Greg." If you want to say the latter, you need a secondary source i.e. a published article about the video that says those things. SlimVirgin ^{(talk)(contribs)} 10:38, 10 July 2007 (UTC)

Detailed example of synthesis

Having established that editors required an example based on real life experience, I constructed this example from a real dispute that I was involved in. It has been reverted as unsuitable writing for a policy page. It was at a level of detail required to make a discussion of the problems of NOR work, in my opinion. I may not participate further as I have other commitments but leave this as my catalyst to a proper discussion on how rules can end up with inappropriate results . Spenny 10:27, 10 July 2007 (UTC)

Spenny 10:27, 10 July 2007 (UTC)

Start of addition

Synthesis is a complicated concept and so it is worth working though a more detailed example. In the context of Wikipedia, careless use of the rules can lead to some unusual statements being claimed based on referenced sources. Often this might come about due to the editing process itself, where people seek to find citations that back up writing.

Synthesis recognises that although information apparently can be carefully sourced, if the actual detail of a statement is not properly referenced, flawed logic can result in unsupportable claims. Whilst many of these simple logical deductions are harmless and need not be challenged, synthesis is not allowed where it is challenged and seeks to advance a position.

The following example made deliberately bizarre to emphasise the problem closely follows a real life example, though the concepts of the real problem might not be quite as obvious at first sight. To understand the example, we need to do some scene setting so we can understand how the subtle problems of synthesis arise. It is also helpful to understand that our editor is a fervent campaigner against tinted windshields on safety grounds.

In Nirvana, there was growing concern about road safety issues related to driving but tinted windshields and sunglasses were seen as essential to coping with this land of perpetual sunshine. There were a sudden spate of deaths on pedestrian crossings and eventually a public enquiry was held into these. The so-called Blindfold Inquiry dealt in some detail with the various causes and noted that although tinted windshields and sunglasses had been raised as concerns, the actual cause of all the accidents was found to be the craze of young drivers proving that they could drive around blindfolded. There were no accidents that fell outside this cause and the Inquiry was accepted as a sound review of various research. The Inquiry came to no conclusion on tinting windshields or using sunglasses whilst driving, but there was a general public consensus that these were a problem, no viable alternatives were accepted. The Inquiry was available as a public document on the web running to 100 pages of detailed summary and assessment of various opinions. Responding to increasing public concern the President of Nirvana called for the banning of tinted windscreens and scientists also were concerned and wanted to see all devices that impaired visibility removed.

A respected news agency in Nirvana published the following release which was accepted by all as a fair high level summary of the issues.

Tinted windshields blamed for road safety disasters

The President of Nirvana called for the banning of tinted windscreens saying, "Tinted windshields are a disaster in the making, they must be banned." Scientists involved in the famous crossing disasters inquiry also demanded changes, "The Blindfold Inquiry shows that impared visibility can cause disasters." Scientists went on to say, "Sunglasses are a timebomb, they must be banned."

The Wikipedia article was written as follows:

Tinted windshields, also known as sunglasses_{(citation to above news release)}, are a harmful form of decoration for cars. Scientists have blamed sunglasses as the cause of deaths on pedestrian crossings._{(citation to above news release)}

Editors had originally complained that sunglasses were not the same as windshields and the Inquiry did not blame sunglasses, but the author had looked for references and found the press release from the respected news agency.

The two key problems with this synthesis is that the release never actually says that sunglasses and windshields are the same thing, it relies on an assumption that the title of the release, the President and the scientists are talking about the same thing. The writing of the release in this loose fashion does allow the casual reader to be confused. Once read in this way, this synthesis using the article definition is used in circular fashion to confirm the article from the release. The fact that the statements come from the same source is not a barrier to a false analysis of the source.

The second issue is that scientists never actually said that tinted windshields were the problem in the release, and although it is common sense that poor visibility is a related issue, the scientists did not tie the sunglasses and the blindfolding together in the release, there is no quote that actually ties the two statements together. There is nothing incorrect or misleading in the press release, the quotes are accurate, but the summary loses the context of the statements.

The Inquiry had wording in it that clearly defined the problem yet policy is used to dismiss the high quality inquiry statement. The argument is that the press-release must be more appropriate to Wikipedia as from the perspective of the press-release the Inquiry is a primary source and does not apparently have the validity of a secondary source which has tested it. It shows the danger of assessing reliable sources incorrectly. We can also be critical of seeking to suggest that casual comments of scientists are suggested as scientific consensus. Regardless of these policy pages, editing requires that critical faculties remain engaged at all times.

End of addition

I don't see how this is an example of SYNT. Can you say what the real example is? SlimVirgin ^{(talk)(contribs)} 10:33, 10 July 2007 (UTC)

I copied it to Wikipedia:No original research/example 1. WAS 4.250 14:41, 10 July 2007 (UTC)

Once again synthesis is painted black because of flaws in the A and/or B used in the synthesis. Minasbeede 18:05, 10 July 2007 (UTC)

Aside from the merits of the example, I think the word count on this is way too high. I think the explanation of synthesis is taking up a disproportionate amount of the policy page as it is. We need to keep the policy page short and to the point, so that people will read and understand it. Enchanter 19:20, 10 July 2007 (UTC)

Wikipedia:No original research/example 1 userfied at User:IanMSpencer/example ≈ jossi ≈ (talk) 01:40, 11 July 2007 (UTC)

An example to clarify

I am not suggesting adding anything to the policy but I want to make sure we all understand the same thing to be synthesis. So here is my example.

A 1914 encyclopedia article reads (Races of the World) that "[The Caucasian race] leads the other races in literature, commerce and all the arts of civilization." The encyclopedia is American and it is plainly designed to comply with the ideas of contemporary educators. (In its preparation we were able to secure the coöperation of specialists and educators whose standing will be recognized upon inspection of our list of editors and contributors. Advantage has been taken of suggestions which have come to us from many teachers during years of experience in the use of the former work, and it is believed that the present work will be found adequate and satisfactory.) It would be synthesis to put all this information together to say "White racial superiority was a commonly held belief among educators in America at the beginning of the 20th century".

The two points I want to make is that synthesis has just a good of chance coming up with the right answer as the the wrong answer. And secondly it can involve a single source. Synthesis is not disallowed because information found through synthesis is always wrong; it is disallowed because such information has a chance of being wrong. It much better practice (as well as policy) to find a source that directly addresses the issue rather than trying to make information that partially covers the issue fit by adding it together with other information. Synthesis goes hand and hand with lazy sourcing. If you take the time to find the best sources for the subject first, then you will have little to worry about regarding synthesis.--BirgitteSB 17:56, 12 July 2007 (UTC)

What is wrong with OR?

As long as the source is clear to the reader, what is wrong with OR in Wikipedia? I am ready to start slapping down some straw dogs, so please consider your response.01001 00:00, 11 July 2007 (UTC)

You can never be clear of who the source is, so have no way of assessing the claims. Outside Wikipedia I approve of OR, but I only publish mine in journals, so my peers can assess the evidence. Wikipedia cannot do that, so it doesn't allow OR. Tim Vickers 00:32, 11 July 2007 (UTC)

This is not a considered response. The question assumed that the source must be clear to the reader.01001 01:00, 11 July 2007 (UTC)

The easiest way to explain it is to consider the sort of cases it was invented for: you have a crank who has published all over Usenet and personal websites their crank theories about how the moon is made of green cheese, the clear scientific evidence for it, and the government conspiracies suppressing it. They can source all their statements to this self-published material. Because nobody cares about this crank, nobody has bothered refuting them, and so there's no sourcable material that can be used to even balance out the article - and they'll viciously defend it. Wikipedia editors being composed primarily of university students, we don't have the resources or expertise to review sources and judge their accuracy - nor should we leave it up to the reader to evaluate the source of such clearly factually ridiculous information. By relying on review by other organizations, we avoid this whole mess. Dcoetzee 01:33, 11 July 2007 (UTC)

So the problem is that a criteria is needed to determine that which belongs in Wikipedia and that which does not. For the sake of discussion, let's further assume that all entries into Wikipedia must be reasonable. Would this not handle the green cheese theory?01001 15:24, 12 July 2007 (UTC)

Then the policy against OR doesn't work against the crank statements (they're sourced, and bad sources don't instantly appear to be such, in general.) This issue deserves more attention but I get too wordy and too repetitive so I'll not try. (I read the essay suggested below. My take was that if there were flawed OR the collaborative community would eliminate it, in time, in theory. That is, in general, how the collaboration works, isn't it? Good stuff tends to be polished and augmented, bad stuff tends to be removed or corrected. I'm aware of the problem implicit in "tends to be" - that's partly why I used the phrase.) Minasbeede 03:34, 11 July 2007 (UTC)

10001: Read the essay Wikipedia:The_role_of_policies_in_collaborative_anarchy, that may help you with your question. ≈ jossi ≈ (talk) 01:37, 11 July 2007 (UTC)

01001, the problem with your OR campaign is that you consistently insist that your OR views are correct because you think they are obvious, even when they are manifestly contradicted by a great deal of published scientific research (e.g. here and here). We do not allow this sort of "OR" in wikipedia, it would be like letting the creationists insist that the articles on evolution by natural selection are wrong because they disagree with them. Encyclopedias are supposed to reflect the current state of the world's knowledge, not be a collection of uninformed opinions. What you are advocating is not even original research, it's the censorship of scholarly research in place of your POV. Pete.Hurd 16:27, 11 July 2007 (UTC)

conclusion / Proposal: more slack for this guideline

I^{— Xiutwel (talk)} am copying from a previous section, since the discussion had become rather lengthy with the excellent tractise on the dangers of synthesis inserted.

In the example you give, Xiutwel, the video is a primary source, and using primary sources is allowed so long as you describe and don't interpret. So you could watch the video and report "Sammy kissed Lisa when they were sitting in the ice-cream parlor." But you could not say: "And this was evidence that Sammy isn't gay, which lots of viewers previously supposed he was, because he seemed very close to Greg." If you want to say the latter, you need a secondary source i.e. a published article about the video that says those things. SlimVirgin ^{(talk)(contribs)} 10:38, 10 July 2007 (UTC)

question I want to check if this is a consensus viewpoint; I've also heard someone say it would be OR to watch a video when describing the video. — Xiutwel (talk) 13:18, 11 July 2007 (UTC)

You mean for example saying that the world trade center collapsed at free-fall speed? Or drawing the reader's attention to puffs of smoke, or supposed 'molten metal'? Tom Harrison ^Talk 13:43, 11 July 2007 (UTC)

The OR comes in (for one of your examples) when it is claimed that falling at free-fall speed implies a controlled demolition (and nothing else.) It is quite often the case that assertions of that type depend on a specific interpretation that excludes all other interpretations. The conclusion reached is the result of excluding all interpretations other than the one that leads to the conclusion, which is the desired conclusion of the editor. That process is intensely not NPOV. Minasbeede 14:18, 11 July 2007 (UTC)

Even leaving aside interpretation, I would be concerned about picking and choosing which bits to mention. The 'truth movement' is infamous for highlighting supposed anomalies to support their conspiracy theories, while ignoring other data. Tom Harrison ^Talk 14:57, 11 July 2007 (UTC)

You mean for example saying ... — No Tom, I do not mean that. I mean: "Bashar, according to Darryl Anka purportedly channeling Bashar, claims he actually represents an entire society of 230 million entities in constant telepathic connection with each other."^{[video of session]}
Or: "The lyre bird can mimic even chainsaw sounds^{[video of David Attenborough]}
I would agree with Minasbeede that drawing your own conclusions from a video does amount to OR, but literally citing a video is allowable, and to do so on wikipedia, no secondary sources are needed as RS — Xiutwel (talk) 08:01, 12 July 2007 (UTC)

User contributions Tom Harrison ^Talk 14:10, 12 July 2007 (UTC)

assertions, theories, opinions, and arguments

We do allow things in wikipedia that are not "assertions, theories, opinions, and arguments" therefore the sentence had to be changed to accurately reflect what was meant but was not actually said. WAS 4.250 21:03, 11 July 2007 (UTC)

"Unsigned articles may be less reliable"

I have a problem with this clause, since it is being used by persons to claim that an unsigned, collaboratively created Britannica-sourced article which reflects the consensus of experts is less "reliable" than the a signed article by a single self-styled expert out on the fringe of expert opinion. It directly conflicts with http://en.wikipedia.org/wiki/WP:Undue_weight#Undue_weight which says "If a viewpoint is in the majority, then it should be easy to substantiate it with reference to commonly accepted reference texts". Clearly, the Undue Weight policy says that a "commonly accepted reference text" can be used as a trump in a case of battling sources. But the policy on this page says that not only can they not be used as a trump against "secondary sources", they can actually be classed as "less reliable" than "secondary sources"! Just because some quack is willing to sign their name to some absurd claim, that does not mean that it thereby more "reliable" by virtue of that signature alone. Finally, you are just opening the door to further "original research" when you imply that an unsigned Britannica article can be improved upon by "secondary sources". A Britannica article normally represents the consensus of expert opinion, so should be deferred to be those otherwise inclined to go hunt down all the unrepresentative, fringy "secondary sources" that they can find. I'd call that "original research" in that they are trying to being a new, hitherto unappreciated "truth" to humanity.Bdell555 15:57, 12 July 2007 (UTC)

I agree. I've never liked that sentence, and in fact I think it's wrong. In Britannica in particular, the signed articles tend to be more POV and idiosyncratic. I'd be quite happy for us to get rid of that. SlimVirgin ^{(talk)(contribs)} 18:01, 12 July 2007 (UTC)

WP:NOR and free images

Over at Wikipedia talk:Non-free content, User:Mosquera posits that WP:NOR forbids the use of many free images because they have never been published anywhere. Please join in the discussion at Wikipedia talk:Non-free content/Archive 46#Fair Use Rationale: The Extended Mix. Thank you. howcheng {chat} 21:05, 12 July 2007 (UTC)

original research

Using search results (the number of results produced) to back claims of popularity, original research or not? There's an edit war stalemate regarding WP:V, WP:RS, and WP:OR at Stephanie (LazyTown)#Internet Phenomenon. — pd_THOR ^{| =/\=} | 12:29, 13 July 2007 (UTC)

It's kind of OR-ish, because it's hard to judge popularity from the number of Google hits. But it's the kind of thing that's often allowed so long as it's not causing any harm. The best thing would be to find a published source that says she's popular, and stick to what the source says. SlimVirgin ^{(talk)(contribs)} 02:22, 15 July 2007 (UTC)

Is this OR? Translation of ancient manuscript

An interesting discussion has come up at Talk:Rylands Library Papyrus P52#So what does it say?. This article is about the earliest NT manuscript, which is a credit card size piece of papyrus that contains roughly 25 words on both sides, with many of the words cut off. I looked at the Greek text found on a website, and I used Bible translation software to come up with a rough approximation of a translation of the fragment. But I used some artistic licensing on the lacuna. Another user took a known English translation of the whole verses in question, arranged the words to correspond to the Greek word order, and again, comparing it to the Greek, made a rough approximation of what words are included, and what words are partially cut off by bolding certain corresponding English words.

The issue is, we have been unable to find a source that gives a direct translation of just what is just on the fragment (we have sources that have reconstructed translations of the fragment and surrounding missing text, but not of just the fragmented text). One user suggested moving his work from the talk page to a website and citing that website, and that's obviously original research.

But now I think we've been thinking about this too hard, because isn't this just a simple matter of translation? Is it really original research if we translate text? Or even use a known English translation and note approximate word cut off points based on the original Greek? -Andrew c ^[talk] 17:48, 16 July 2007 (UTC)

Why would you want a translation of just the surviving words? I think there is a clear reason for the existing sources to have chosen to reconstruct the surrounding text and you should follow their lead. This fragment was not designed to only show 25 or so words and determining what was handwritten in an ancient version of a foreign language you do not read is something that is best gotten from reliable sources. A simple translation is when you translate something you clearly understand to a second language that you also understand. This is far from being a simple translation.--BirgitteSB 18:14, 16 July 2007 (UTC)

Well we do have a source that says what is there in Greek. We aren't trying to determine what was handwritten. The issue with saying "the fragment says..." and copying a "reconstruction" is that 80% of the "reconstructed" text isn't on the fragment itself. But I understand what you are saying regarding a simple translation. Thanks for your input._Andrew c ^[talk] 18:17, 16 July 2007 (UTC)

I guess one of my issues is regarding what you intend on doing with the information. If you want to make a mock-up of the fragment in English as an illustration, then I don't have a huge problem. However it can be near impossible to translate many words accurately out of context. (i.e. "fly" how would you translate this word without know if it is a noun or a verb from surrounding text) This difficulty is why I imagine all sources in translation reconstruct the whole document. It is just as much of a guess about the meaning of 20% without the context of the rest as it to guess what the 80% is. In fact the only way to translate that 20% with any hope of accuracy is to first reconstruct the whole document from other versions. When you just transcribe the Greek you do not have to decide on the meaning of any ambiguous words, when you translate you do. But as I said I don't have huge problem if this is for an illustration. On the other hand if want to use this translation and discuss it in the prose of the article. I do have a problem. Any discussion of this translation you have come up with will be original research with 100% certainty.--BirgitteSB 18:40, 16 July 2007 (UTC)

OR/Synthesis dispute

Is it correct to say that this section is full of OR/Synthesis of published material service to advance a position? Looks exactly like that to me... Dreadstar † 19:50, 14 July 2007 (UTC)

Can someone could check my logic? Thanks much! Dreadstar † 21:54, 14 July 2007 (UTC)

There is a dispute about OR/Synthesis in an article on the "What the Bleep Do We Know" talk page. Any assistance in clearing the matter up would be greatly appreciated. Dreadstar † 01:44, 15 July 2007 (UTC)

Yes, that section is OR. For the corrections to be included, we would need to find a source who had made those corrections in relation to the film i.e. a film review that said what that section is saying. SlimVirgin ^{(talk)(contribs)} 02:20, 15 July 2007 (UTC)

Excellent! My thoughts exactly! Thank you, SlimVirgin! Dreadstar † 02:23, 15 July 2007 (UTC)

I must say I find this decision extremely confusing. The example at Wikipedia:No_original_research#Synthesis_of_published_material_serving_to_advance_a_position seems clearly to be OR to me, and at least the first, second, and last items at What_the_Bleep_Do_We_Know!?#Factual_errors seem clearly not to be OR. If they are OR, then I don't understand how it's possible to write Wikipedia articles at all. What are the boundaries? For example,

• If the movie claimed logical proposition P (e.g. 2+2=5) and some reliable source claimed ¬P (e.g. 2+2≠5), but didn't mention the movie, would quoting it constitute a prohibited synthesis?
• What if the claim is merely logically inconsistent with P (e.g. 2+2=4)?
• The article on Albert Einstein says that his contributions included special relativity, "which reconciled mechanics with electromagnetism". That claim is unsourced. If it were sourced, could the source simply claim that special relativity reconciles mechanics with electromagnetism, or would it need to include Einstein's name?

These are serious questions; I don't know the answers, and I don't understand the policy. -- BenRG 18:47, 17 July 2007 (UTC)

The correction section isn't OR at all. There isn't any synthesis. The film makes bogus claims. Each of those claims is refuted by science. All that is needed is a verifiable reference for the refutation. If the film claimed (for example), "the moon is made of green cheese", a reference to NASA studies about the moon's composition would not be original research. Kww 20:03, 17 July 2007 (UTC)

Best practice is to write wikipedia articles by taking claims made by different reliable sources about a subject and putting those claims in our own words on an article page with each claim attributed to the source that makes the claim. Our "original reseach rule says you are not allowed to come up with your own claim that is not in a reliable published source and try to prove it with cited claims that do not make the claim you are trying to support. "to advance a position" is the wording I remember in the policy. Don't think "Well its obvious that if this says that and the other source says this other thing then this important conclution must be true and I should claim it is true in this article." If its trivial and unsourced, don't mention it. If it nontrivial and there is no source anywhere that says it, then maybe you are mistaken. Lots of people don't see how something could not be true if so and so is true ... and are wrong. WAS 4.250 20:36, 17 July 2007 (UTC)

The relevant section is Synthesis of published material serving to advance a position. Some of the OR in the "Bleep" article is clearly synthesis because it takes a source unrelated to the movie, and joins it together with a statement from the movie - that is the very definition of synthesis. The resulting statement, that it is a a 'factual error,' whether it's right or wrong, is clearly a comment and conclusion being made by a Wikipedia Editor and not by a third party, reliable source. It is very clearly Original Resarch and a Synthesis of sources. Dreadstar † 21:41, 17 July 2007 (UTC)

But they aren't synthesizing. If they took the next step: "Since BLEEP can't get facts straight, none of it's conclusions can be trusted", they would be synthesising. They don't do that. They simply list claims that contradict scientific consensus, and cite factual references supporting the fact that the movie contradicts scientific consensus. Kww 21:49, 17 July 2007 (UTC)

That is definitely original research. Remember that OR is "any unpublished analysis or synthesis of published material that appears to advance a position". Taking inaccurate claims from the movie and comparing them to other published material (unrelated to the movie) to advance the position that the movie uses incorrect/inaccurate science is very clearly original research. It is certainly an "unpublished analysis" of the claims made in the movie. The movie has been widely reviewed and discussed. Many of the distinct claims and persons in the movie have been extensively written about. If you wish to advance the position the science in the movie is inaccurate, there are plenty of reliable sources available for such a point of view. Just be cautious to ensure that the article is balanced in proportion to the references. —Preceding unsigned comment added by Vassyana (talk • contribs) 19:39, 17 July 2007

Excellent point about 'unpublished analysis." Very significant observation! Dreadstar † 01:52, 18 July 2007 (UTC)

One problem is that there are no reliable sources with positive things to say about this movie. Anything that makes it appear that any portion of it possesses any credibility violates Undue weight, so there will always be an edit war with the true believers. Kww 02:28, 18 July 2007 (UTC)

Do you mean Undue weight? Or is there a violation of WikiProject user warnings (WP:UW) that I'm not getting? Dreadstar † 03:36, 18 July 2007 (UTC)

Serves me right. Whenever I'm nasty, I screw something up. Kww 03:40, 18 July 2007 (UTC)

lol! When it happens to me, I call it Instant Karma. It never fails! It's like Murphy's law... Dreadstar † 03:45, 18 July 2007 (UTC)

An RfC was opened for this issue. I invite everyone to comment on the issue in the RfC section of the 'Bleep' talkpage. Dreadstar † 09:45, 18 July 2007 (UTC)

Wikipedia:Requests for verification

Please see: Wikipedia:Requests for verification

A proposal designed as a process similar to {{prod}} to delete articles without sources if no sources are provided in 30 days.

It reads:

Some editors see this as necessary to improve Wikipedia as a whole and assert that this idea is supported by policy, and others see this as a negative thing for the project with the potential of loss of articles that could be easily sourced.

I would encourage your comments in that page's talk or Mailing list thread on this proposal WikiEN-l: Proposed "prod" for articles with no sources

Signed Jeepday (talk) 14:00, 18 July 2007 (UTC)

Or we could just use PROD and cut 25 days off the waiting period. -Amarkov moo! 18:23, 18 July 2007 (UTC)

Smile :) or you could use RFV and give editors 25 more days to fix it. Jeepday (talk) 01:24, 19 July 2007 (UTC)

"Tiny minority" view

Hi.

I saw this: "The inclusion of a view that is held only by a tiny minority may constitute original research." Does this mean that even if such a view has been published and documented in a reliable source, it still cannot be included simply because very few people believe it?! How exactly does few people believing something make it "original" research, anyway? mike4ty4 05:48, 6 July 2007 (UTC)

You raise a good point User:mike4ty4. However, the wording that you quote shouldn't cause a problem. If some information that is "published and documented in a reliable source" as you say is added to a WP article, it is unlikely to be challenged. The thing to look out for, especially if the information presented is likely to be contentious, is that, to conform to WP:NOR, no extra, or controversial, conclusion(s) should be drawn from the material from that source. And, that undue prominence (see WP:UNDUE) is not given to that source at the expense of other relevant sources included in the article. Undue prominence, for instance, could be mentioning a "minority view", while leaving unsaid the balancing or opposing or majority views.

All in all, I see no problem with that wording. Perhaps it needs a reference there to WP:UNDUE rather than WP:NOR, User:Newbyguesses - Talk 10:23, 6 July 2007 (UTC)

The problem is that it doesn't carry that implication. It seems implies that small numbers of adherents could make even something published in a third-party source "original research", even though that does not make sense. mike4ty4 09:05, 20 July 2007 (UTC)

Could someone with a good grasp of NOR, and a good deal of patience and time on their hand please provide a second opinion

The discussion refers to Allied_war_crimes_during_World_War_II#Treatment_of_POWs

And is currently and has been in the past conducted in

Previously: Talk:Allied_war_crimes_during_World_War_II#Historical_context.2F.22US_and_Australian.22

Currently:Talk:Allied_war_crimes_during_World_War_II#Prisoner_Execution_Sources

The paragraph itself has been the issue of quite a lot of recent discussions, some of which can be found in the talk archive page, but the discussion that we need a second opinion on now is at the end of:

Talk:Allied_war_crimes_during_World_War_II#Prisoner_Execution_Sources

If you have the patience and time, please read through the talk page and perhaps provide your comments because right now it feels as if we are talking past each other and walking in circles, we need some external imput. Thanks, --Stor stark7 ^Talk 22:55, 18 July 2007 (UTC)

I'll take a look shortly. Sorry to have missed this before. SlimVirgin ^{(talk)(contribs)} 23:33, 25 July 2007 (UTC)

Proposed Example of Synthesis - Restatement

The suggested example and supporting text from previous discussion is:

Jones says he obtained a degree in physics at university X between 1977 and 1980_{reliable source 1}, but according to Department of Education records, university X closed down in 1976_{reliable source 2}, raising doubts as to whether he ever attended_{no source}.

This analysis, suggesting that Jones has lied about his qualifications, is not permitted on Wikipedia, even though it is backed up by reliable sources, and the logic is sound. This is because coming to a novel conclusion requires a careful analysis of the facts which Wikipedia is not in a position to provide. It is possible, for example, that one of the sources was wrong (reliable sources are not necessarily infallible sources), or that they had been interpreted incorrectly (for example, the names of two universities could have become mixed up).

A researcher coming to the conclusion that he had lied about his qualifications would carry out a careful analysis of the facts, and would need to stand ready to respond if Jones challenged the allegations. Wikipedia is not in a position to verify this analysis, or stand by the allegations if challenged, as this is not part of its mission to summarise existing published knowledge. Therefore the allegations can only be published in Wikipedia if they can be attributed to a reliable published source. ''Italic text This is not intended to amend policy, simply to provide a more coherent example than the one in place at the moment. Spenny 08:26, 2 July 2007 (UTC)

For anyone monitoring the page who has not picked up on this due to the noise, it is suggested that this replace the current synthesis example. Assuming there are no objections over say the next 24 hours, I will assume there is a consensus that this is an improvement. Thanks Spenny 13:47, 3 July 2007 (UTC)

Although this is my own example, I don't necessarily accept that it is better than the one in the text right now, which is based on a real case of editor confusion. But let me add the missing part that was left out of my example: even if we exclude the final phrase "raising doubts as to whether he ever attended", it would still be (with or without the word 'but') 'synthesis by juxtaposition' (WP:JUXT?) and hence unacceptable - the point being that we need a reliable source making the connection for us between the school closing and Jones's attendance claim. So if this example is used, it must be used properly, by also demonstrating and explaining the JUXT part. Crum375 14:13, 3 July 2007 (UTC)

I agree (see above, if you dare :) ). However, we are trying to get across the basic principle and examples do not have to be comprehensive. (FWIW, the existing text does not cover this issue at the moment). I don't think there is much mileage in tuning the example (except by contrived avoidance). Perhaps you have a few words to add to the clarification? Perhaps a simple In this example, we can see that using the word but or even leaving these facts placed together could create an implied criticism so editorial care is required in these situations. Spenny 14:35, 3 July 2007 (UTC)

I don't think it's an issue of 'required care', we need to clarify that in this example, even without the word 'but' and without the final phrase, we may still not juxtapose those two independent sourced facts unless we have a reliable source making the connection for us. The problem (and possible advantage) with this example is that it highlights (and forces us to deal with) both basic OR by synthesis as well as synthesis by juxtaposition at the same time. We can't include it unless we are willing to address and clearly explain these issues, which of course slightly complicates it. OTOH, when properly explained, they do highlight the true (and typical) issues of synthesis and juxtaposition. Crum375 16:56, 3 July 2007 (UTC)

I think this is a terrible example of an analysis. The example appears to be one in which, for the purpose of the example, the two facts are to be accepted as facts. It's gratuitous and pursuing a position to claim that in this particular case maybe the facts aren't true, that a mistake was made in one or the other of them. The exact same arguments could be made for the exclusion of single statements (about single facts) - maybe they are untrue, making Wikipedia totally empty. The example dodges the real issue, which is that logic is excluded from Wikipedia unless the logic is included by quoting someone else who made the logical analysis. Note that in such a case the identical objection could be made: maybe the person being quoted didn't have the facts straight. In practice NOR is used by partisans to remove logical conclusions that expose weaknesses in the positions advanced by the partisans. The typical use of NOR isn't to improve Wikipedia, it is to foster particular beliefs by removing valid material (by any reasonable standard but not by Wikipedia's standards) from Wikipedia. I don't see a "basic principle" being communicated, I see ad hoc arguments used to justify a principle that really isn't compatible with the notion of an encyclopedia. Only in the specific case that A and B are combined to produce C is the rationale used that "well maybe either A or B is incorrect." A alone can be cited, B alone can be cited. Combining them does not make either of them less reliable.

Minasbeede 14:39, 3 July 2007 (UTC)

I sympathise with the sentiment (and this is a circle I have been through making most of the same points). The synthesis section is not about a new complicated principle, is is just a special case of NOR policy where people might not realise they have done OR as they have some related citations. In the long term, the whole section needs work and these issues need bringing in with some clarity. However, the main issue is advancing a position, blind logic is allowed, so I think that you have actually argued in its favour, at a pinch.

Another user identified the existing example as being obscure and it seems a simple first step improving to getting that section to be helpful (how naïve!). SYNTH is interesting as it highlights the slightly paradoxical position that Wiki gets into with its policies. However, it is something that we can give a steer on if we remember the basics. So yes, it is simplistic and gratuitous, but hopefully it makes the simple point to a simple audience, not one immersed in Wikilosophy.

Can I ask you a slightly different question? Does that mean you think the existing example should remain in preference to this one; or are you saying delete the SYNTH section in entirety; or ??? Spenny 17:57, 3 July 2007 (UTC)

The current (plagiarism-related) example is rather horrid in that it seems to hinge on a debate about the meaning of a word. I've already said the proposed example is bad. Perhaps either of the OR-tainted edits I made could be used in an example. In one I stated that Curveball (the informant) claimed to have been trained as a chemical engineer but that nothing in the intelligence he supplied appears to show a "glimmer" of engineering awareness. That was actually aimed at the apparent prohibition of original research in the press (to make the same connection and conclusion) and the apparently assumed similar prohibition within the CIA (which, of course, is entirely antithetical to the mission of the CIA.) That was particularly egregious OR since I basically assumed that an engineer's words would reflect the fact that he was an engineer without providing a citation.

The other was that the prohibition on retirees receiving any of their principal amount during their lifetimes, as was stated by President Bush on several occasions, was a means of prolonging the bubble that is inherent in the personal account scheme. Again, egregious. I cited no source for the cashing in of investments being the trigger for the bursting of a bubble when cashing in outweighs investing.

Both of these are worse than simple synthesis. But assuming I had sources for the uncited portions they'd still be forbidden since they both are synthesis.

So the simple thing to do would be to show, using either of these as an example, why they are incorrect in Wikipedia - other than by dogma. That is, illustrate why the dogma is useful in strengthening Wikipedia or preserving its useful characteristics.

But any simple example that shows the inherent flaw in synthesis would be fine. My objection to the proposed example is that it makes a special appeal to the possibility of error in citations that is not made for citations not part of a synthesis.

Minasbeede 20:29, 3 July 2007 (UTC)

Yes, but :) If you have an example where the citations are demonstrably correct, then the analysis is simple OR and a similar problem exists that the example is so simple that it does not work - there is no dilemma to debate. The contrived example is seeking to show at a simple level that the logical inference is correct, but cannot be justified without further investigation. Real world examples are likely to be too obscure, though perhaps superficially more justifiable, but they require too much back story to hold the point - "you had to be there". Logical inference that does not require further investigation (by a reasonable person - weasel words, perhaps) is allowed. I think everyone would agree that it is a difficult policy to work with and from some POVs is "broken", but it is a pragmatic way to deal with a conflict until the powers that be one day determine a different approach (Wikipedia is not a democracy either). What that does make me think though is that it perhaps is inappropriate to remove the existing example, and placing this example in as well as a starting point will help tell the story. Spenny 21:02, 3 July 2007 (UTC)

I can't think of any response that doesn't belong in the town pump area. If the example must of necessity hinge on some flaw that is not inherent in synthesis itself then it would seem (to me) that the example is flawed because of that dependence on that flaw.

As I understand the rule logical inference is not allowed. Logical inference may indeed favor a POV, just as 2 + 2 = 4 might favor a POV. That the facts support a particular POV doesn't make reliance on the facts incorrect, does it? It sure seems strange that in an encyclopedia it is forbidden to cite facts that support a POV. Am I missing something?

Minasbeede 23:59, 3 July 2007 (UTC)

I believe you are barking up the wrong tree. You seem to be criticizing the example, but in reality you are criticizing WP:NOR itself, since the example faithfully represents it. So forget the example and simply say that you disagree with NOR, which is a legitimate POV, and then try to get a consensus to overturn it. But attacking the example is like attacking the messenger. Crum375 16:59, 3 July 2007 (UTC)

I'm doing both. I am committing the sin of OR by noticing that the example cited is a bad example and is cooked to support the policy - but inconsistent and illogical. It's OK to cite any one thing but if you cite two things and link them logically then you have committed an offense against Wikipedia policy because one or the other of the things cited could be in error. That's supposed to justify the policy but it's wildly inconsistent. Let's see an example in which the policy is not justified by claiming that possibly one of the citations is in error. Such an example (I feel) illustrates the problem with the policy: it's illogical.

The example does not "faithfully" represent the policy. The example only works by making an assumption about the possible fallibility of one or the other sources cited. If a cited source can be fallible then that's a blanket problem for everything in Wikipedia, is it not? Take away the assumed possible fallibility and the example is an absurdity: you cannot logically link two things to reach a conclusion. That is the problem with WP:NOR, which is often used by partisans to remove material they do not wish to have published. I go along with such removal because, currently, that is the policy. I even removed some of my own editing myself because of the policy.

The nature of the world and of reality is not such that every possible valid conclusion has been published, particularly when the majority of sources themselves explicitly avoid reaching conclusions. It is valid for an encyclopedia to synthesize information, it is disastrous for an encyclopedia to avoid synthesis. An encyclopedia is not a huge compendium of "he said - she said" type material.

Thanks for your response.

Minasbeede

Wikipedia is not a traditional encyclopedia. The main policy is WP:NPOV, please read it. WP:V is also crucial, please read this policy taking the other two into account. ≈ jossi ≈ (talk) 18:13, 3 July 2007 (UTC)

Under the current policies Wikipedia isn't an encyclopedia at all. I go along with that. If the desire is to have a non-encyclopedia then that's that. You can't have a real encyclopedia if you exclude logic and thought.

The example is a bad example, the policy is a bad policy - if the desire is to have a real encyclopedia.

I don't think I need to read the dogma, thank you very much.

By the way, in this discussion there is a violent violation of WP:NPOV. Anything that doesn't bolster or bow to the dogma is opposed: the dogma is given favored treatment, even to the extent of cooking fallacious examples.

Thanks for the response.

Minasbeede 19:05, 3 July 2007 (UTC)

It is a much wider argument, but in a nutshell, the policies are designed not for logical correctness, but for a pragmatic approach, and these policy pages are, in my opinion, attempting to get across the spirit of the rules to allow the specific methodology of development to cope with a population of editors who are in general co-operative. Read around the policies and guidance such as "There are no rules" makes that abundantly clear. Attempting anything more is getting out of scope. I am quite content that the example is not perfect, and the explanation is not perfect. However, if we debate to perfection before implementing any change, then there will not be consensus. My aim when I accidentally arrived here was to make sense of a certain section did not express itself clearly. I think that is a useful contribution within the confines of current policy. Changing policy is for the Village pump in the first instance. Spenny 19:34, 3 July 2007 (UTC)

You asked if there were objections. I object to the example. The example depends totally on the possibility that one of the cited sources might be incorrect in some manner. That's a potential defect in all citations. To rely on the potential incorrectness of a source only in this case is neither logical nor "pragmatic."

You could go ahead and use the example but I see it as creating greater future problems: the actual issue has been skirted in that example. The assumed flaw that arises from synthesis doesn't come from the synthesis, it comes from citing something that is incorrect. Duh. The flaw is the citation of something that is incorrect, isn't it? The flaw would be just as grievous even if there were no synthesis - if the subject matter (and Wikipedia) actually matter so that the flaw matters.

If Jones says something that is at variance with the facts then while it might be excessive to say Jones has lied it would seem entirely fair to say that what Jones says is at variance with the facts - even if that has not been published elsewhere. If you want to get really weasel-wordy you can say that Jones has made a statement that appears to be in conflict with the apparent facts.

Thanks for your response.

Minasbeede 20:11, 3 July 2007 (UTC)

You are missing the point of the example, and the policy. We are not saying that necessarily one of the two sourced facts must be wrong. There are several logical possibilities. For example, Jones (or the newspaper that interviewed or quoted him) could have misspelled the name of the school, or the year got mangled somewhere. Or the Dept. of Education could have a typo or mistake in its records. Or the school that closed was actually re-opened a year later, and so on. The point is that there are lots of possibilities, and the bottom line is that we can't act as detectives, and conclude that Jones is lying, or even imply that Jones is lying, unless we have a reliable source that implies or says it for us, that we can quote. That's what NOR, SYNT, and JUXT are all about. Crum375 20:24, 3 July 2007 (UTC)

The justification for the policy given with that example seems to hinge on the possibility that one or the other of the citations could be in error. Fine, I can see that. But any citation could be in error. As I read the example the possibility of error is critical for edits that constitute synthesis but not really important for other edits. Obviously, that is not the Wikipedia policy - but the example makes a particularly weak case for the policy by specially relying on the possibility of error in the citations, apparently because they're part of a synthesis and not free-standing. I ask for an example in which there is no possibility of error in either citation but for which the synthesis attempted has not been published elsewhere. A trivial example would be 111111111 + 222222222 + 333333333 = 666666666. That's true. I could show it's true by elaborately showing the laws of addition and then showing how the laws lead to that sum - but that's forbidden. Unless I can show that addition leading to that sum to have been published Wikipedia forbids me to show it.

And if that sum had been published I'd still be forbidden to say, in Wikipedia, that 111111111.7 + 222222222 + 333333333 = 666666666.7. That's a synthesis.

My request is for an example of forbidden synthesis that does not have to look outside the synthesis for justification (either a dispute about the definition of a word or an error in one of the constituent citations or the like.)

I understand that the proper venue for the underlying issue is the town pump. The current example and the proposed example are bad. A better example is needed. I'll gladly drop the underlying issue here.

Thanks for your response.

Minasbeede 20:48, 3 July 2007 (UTC)

I don't think anyone would argue that simple logical deduction or arithmetic are disallowed, and it is in part dependent on context: is it contentious? Does it advance a position? in WikiJargon. Basically, every article is awash with these in its basic construction. An example of a problem would be one I saw recently where someone had published a table of average scores and cited mathematics as the basis. There was no argument that his source numbers were wrong, but he could not see that to go from simple sums to an average dragged in all sorts of assumptions, including the fact that his finger might have slipped on the keyboard and to verify it would require the same amount of effort. We assume that in most cases a source would be correct and be checked, (though we allow that mistakes can happen). Back to what I said early on: you have to be pragmatic. Spenny 21:15, 3 July 2007 (UTC)

Does anyone have any objections to the proposed example replacing the current example in the policy?

Crum375 makes the point that the example could specifically address the "juxtaposition" of sources to give an implied synthesis. I agree that there are situations in which this implied synthesis through juxtaposition of sources is inappropriate - and that the example is one of them. However, as the current policy does not specifically address this point I think we should set this aside as a separate issue to be addressed later. This is because proposed example is intended to explain existing policy, and not to change it or add to it.

Any comments? Enchanter 19:15, 10 July 2007 (UTC)

To be clear, does anyone object to the replacement of the existing example, or have any further comments or suggestions? Enchanter 23:22, 14 July 2007 (UTC)

Which is the proposed example exactly? SlimVirgin ^{(talk)(contribs)} 00:38, 15 July 2007 (UTC)

The one about qualifications, summarised at the top of this section, under: "Proposed Example of Synthesis - Restatement" Enchanter 10:53, 15 July 2007 (UTC)

I'm not sure I see the benefit of that example over the current one. The other concern is that it's perhaps too obvious an example. People tend to get the really obvious ones without help. It's the more subtle ones they have problems with, and that's what the example on the page is for. Perhaps others could chime in with their thoughts. SlimVirgin ^{(talk)(contribs)} 11:00, 15 July 2007 (UTC)

I have other concerns with it, which I discussed above. It is my own example which I have used previously, but it's missing an important part that relates to SYNT by juxtaposition, which is very important to understand, in my opinion. By leaving the JUXT part out, we'd be sending an implied message that JUXT may be acceptable, where it's simply not. OTOH, by including it, the example becomes more complicated, which may or may not be desirable. I also prefer real examples to synthetic ones. ;^) Crum375 12:14, 15 July 2007 (UTC)

To address these points in turn:

• The proposed example does not address "synthesis by juxtaposition". This is the case - but neither does the current example, so it is not a reason to prefer one over the other. The point of the proposed example is to provide a clear explanation of current policy - not to change it, add to it, or explain it in more detail.
• Real examples preferred to synthetic. The current example is not real (although it is loosely based on a real case), as I've noted in previous discussions. Please explain if you think I am mistaken here. Again, it is not a reason to prefer the old example over the new one.
• The example is too obvious. An example needs to be reasonably obvious to be understood. The current example is more than subtle - it is ambiguous and illogical, and does not provide enough information for readers to understand it. When readers who have failed to understand it have asked for it to be explained on the talk page, noone has, in my experience, been able to give a coherent and consistent explaination of it. And investigating in more detail does not help, as it reveals that the example is based on outright misrepresentation of sources, rather than a subtle synthesis. Again, if you disagree, please could you try to explain the current example in more detail, to help those who don't understand it.

Enchanter 20:12, 15 July 2007 (UTC)

I don't believe some of my own points have been addressed: if JUXT is not mentioned, it could be assumed to be acceptable by its omission, and if it is included, it makes the example more complex, which is OK, but we need to agree to it. Crum375 21:12, 15 July 2007 (UTC)

My concern about addressing "JUXT" is that JUXT is not in the current policy (at least, not explicitly), and the proposed example was not intended to change policy.

I agree with you on the basic principle, but I think that this point needs wider debate to make sure that any change we introduce has consensus and is well thought through. I therefore suggest that we discuss this proposal later, after having addressed the example. Enchanter 21:53, 15 July 2007 (UTC)

(outdent) I respectfully disagree, then. I believe that JUXT (not under a specific name), as a type of SYNT, is very much part of the current policy. It is a situation where by putting two sources together we are advancing a position that is not directly advanced by any specific source. It is a typical situation where people fail to understand NOR and SYNT, and thus is important to illustrate and explain. If we are going to invest an effort and replace the current example, it makes a lot of sense to use an example that addresses JUXT. Crum375 22:10, 15 July 2007 (UTC)

My main concern is to replace the current example, which has serious problems which have gone unadressed for too long. I agree with you that it might be better to have the example address "JUXT", but only once this has been debated widely so that we make sure that it has consensus and that the policy wording doesn't unintended consequences that you and me might not have thought of.

If you amend the example to address JUXT in a satisfactory way that gains consensus quickly, you would have my support. However, I see no reason to delay changing the example if obtaining consensus takes longer (which I suspect may be the case). Enchanter 22:30, 15 July 2007 (UTC)

This example was my own example, and it originally included the JUXT part, which was truncated at some point. I added it back in, as you can see higher on this thread. I see no reason to exclude it if we use the example, which is built for it, but I am still unsure about using a synthetic example in the first place, even it's my own. ;^) Crum375 22:39, 15 July 2007 (UTC)

I'm getting confused by the length of this thread - which piece of wording did you suggest adding back in? Could you resummarise how you think the example might look? Thanks, Enchanter 22:57, 15 July 2007 (UTC)

Take the current example, which is OK for basic SYNT. Then remove the words from "raising doubts..." to the end - it would then become an example for JUXT. We are still clearly implying that Jones is lying (with or without the word 'but'), and thereby advancing a position not stated directly by any source, so that is SYNT by JUXT. In BLP cases, at least, we would need to have a source using the two pieces together for us - we cannot be the first to bring them together, even without the final editorial part. Crum375 23:10, 15 July 2007 (UTC)

Well, I would much prefer explaining it without introducing the abbreviation "JUXT" to the policy page. How about:

Jones says he obtained a degree in physics at university X between 1977 and 1980_{reliable source 1}, but according to Department of Education records, university X closed down in 1976_{reliable source 2}, raising doubts as to whether he ever attended_{no source}.

This analysis, suggesting that Jones has lied about his qualifications, is not permitted on Wikipedia, even though it is backed up by reliable sources, and the logic is sound. This is because coming to a novel conclusion requires a careful analysis of the facts which Wikipedia is not in a position to provide. It is possible, for example, that one of the sources was wrong (reliable sources are not necessarily infallible sources), or that they had been interpreted incorrectly (for example, the names of two universities could have become mixed up).

A researcher coming to the conclusion that he had lied about his qualifications would carry out a careful analysis of the facts, and would need to stand ready to respond if Jones challenged the allegations. Wikipedia is not in a position to verify this analysis, or stand by the allegations if challenged, as this is not part of its mission to summarise existing published knowledge. Therefore the allegations can only be published in Wikipedia if they can be attributed to a reliable published source.

Where controversial claims are advanced, or in articles about living people or organisations, this policy should be applied strictly. In such cases, Wikipedia needs to be careful not to insinuate or imply new claims. For example, the above example would not be acceptable even without the final words "raising doubts as to whether he ever attended".

What do you think? Enchanter 23:01, 16 July 2007 (UTC)

I propose replacing the existing example with this revised wording. Does anyone have any objections, comments or proposed improvements? Enchanter 21:31, 20 July 2007 (UTC)

First, I think it's too long-winded. Secondly, I'd worry that it's too simple an example. Most people get it at this level. The one on the page is more subtle and is the kind of thing people have difficulty with. SlimVirgin ^{(talk)(contribs)} 23:18, 20 July 2007 (UTC)

I agree with SlimVirgin, the proposed version is too wordy; and although the current example is short, it is a very clear, concise and easily understood example of subtle original research and synthesis. I see no reason to replace it. Dreadstar † 00:09, 21 July 2007 (UTC)

Disregarding the final sentence about "juxtaposition" (which is not in the current example and I don't think is necessarily needed in the new one), the proposed example is of similar length to the old one (233 words compared to 224). I agree however that both are long and wordy in relation to what is there on the policy page, and would support suggestions on how to shorten either.

It's not that your suggestion is too long, just too wordy i.e. unnecessarily long. But my main concern is that it's too obvious an example. SlimVirgin ^{(talk)(contribs)} 19:14, 21 July 2007 (UTC)

Dreadstar, Slimvirgin, in response to your comment that you see no reason to replace the current example, please could you respond to these specific critisisms of it, which I raised earlier with no response, and explain if and why you disagree with them:

• The example claims that the Chicago Manual of Style defines plagiarism as "using a source's information, ideas, words, or structure without citing them". In fact, the Chicago Manual of Style contains no such definition, or anything that resembles it. Noone has contested this.

This is a real example from a real article, warts and all. And I recall that one of the participants posted a link to the Manual of Style on the talk page, showing that it did say that, at least at the time of the discussion. SlimVirgin ^{(talk)(contribs)} 19:14, 21 July 2007 (UTC)

Slimvirgin, I think you are frankly wrong about this. According to my understanding, the definition was taken verbatim from Harvard's "Writing with Sources" manual, and the relevant article included a citation both to the Harvard manual, and to a piece from "Smith" quoting it. You appear to be claiming that the editor who added the material had provided a citation to an earlier version of the Chicago manual, and that you have checked this at this at the time of the original discussion. I find this implausible, and would want to see good evidence. At the moment it looks to me like you have not understood the original case, and are misrepresenting what the case said. Enchanter 20:02, 21 July 2007 (UTC)

• The example is based on a real case on Wikipedia. However, it misrepresents the original case, at which different issues were at stake. The original case it was based on is very convoluted and difficult to understand - however, there was no misquoting of the Chicago manual, which was added later. (If anyone thinks I am mistaken here, please explain).

No, it very closely follows the original. SlimVirgin ^{(talk)(contribs)} 19:14, 21 July 2007 (UTC)

Again, I can see absolutely no evidence for this, and plenty of evidence that you have not understood the original case, as you are continuing to assert definitions from the sources that do not stand up to checking. Enchanter 20:02, 21 July 2007 (UTC)

• Quite apart from the way that the example misrepresents the source and the original case, it is unclear and does not follow logically. It frequently confuses new users, and I have asked for other users to explain it several times, with no satisfactory reply.

I'm still waiting for someone who supports the example to address this point. Enchanter 20:02, 21 July 2007 (UTC)

• The case it is based on is from a highly controversial dispute (vaguely related to Israel/Palestine). In essence, it is saying "here is an example of where an editor violated the policy". In this kind of situation, I think it is essential be careful not to misrepresent the editor who made the edits, not to misrepresent the sources used, and not to misrepresent the original dispute. Failure to do so risks damaging Wikipedia's efforts to be seen as impartial and objective. Honesty and objectivity are essential to Wikipedia, and bending the facts is not justified on the basis that this is "just an example".

I think the problems above are too fundamental to be fixed with modifications to the example, and in my view this complicated and controversial case is a bad case study to use. I would support exploring ways of improving the policy to explain it more clearly - perhaps including alternative examples. In the meantime, I again propose that the example is removed, as it was reverted without any of these issues being addressed. Again, if anyone thinks the example should stay, please give specific reasons. If I am right about the above, I think there is a compelling case for it to be removed - and if I am wrong, someone ought to be able to explain why.Enchanter 19:08, 21 July 2007 (UTC)

For the record, I'm strongly opposed to the synthesis policy, for reasons that have already been very well articulated by Minasbeede. I hadn't realized until now that we had this crazy policy, and I don't think there's a decent article on Wikipedia that doesn't violate it. The Jones example is potentially libelous and potentially POV; those are the real reasons to disallow it. We don't need the synthesis rule, and it makes no sense. -- BenRG 23:44, 20 July 2007 (UTC)

Most of the editors who argue against this example don't want the synthesis rule at all. The Jones example is not a problem only because it's possibly libellous. It's the structure of it that's not allowed, and that applies regardless of the subject matter. SlimVirgin ^{(talk)(contribs)} 19:14, 21 July 2007 (UTC)

CMOS discussion

Slimvirgin, following from the thread above, please could you provide a link to the talk page discussion, from which you recall that a participant had provided a link to an earlier version of the Chicago Manual of Style, which gives the definition of plagiarism in the example you wrote. From my understanding of the case and the source, I think it is very unlikely that any such version of the CMS, or any such talk page comment, actually exists, and I have strong reasons to believe that what you have written is inaccurate. I will gladly withdraw these comments if you can provide such a reference to the talk page and an earlier version of the manual with the definition. Enchanter 10:25, 22 July 2007 (UTC)

To check whether we are any closer to consensus yet, does anyone still support retaining the current "plagiarism" example in its current form? If so, please explain, with evidence, whether it reflects the sources and original case accurately, and explain the logic of the example for the benefit of those who think it is ambiguous and unclear. Enchanter 23:20, 23 July 2007 (UTC)

Yes, I still support it as explained several times above, unless someone can produce a better (preferably real) example showing one of the more subtle forms of it. SlimVirgin ^{(talk)(contribs)} 23:25, 23 July 2007 (UTC)

SlimVirgin, I asked you very specifically above for a link to the reference that you claim exists on a talk page, showing that the definition Chicago manual was accurately sourced to an earlier version of the Chicago manual of style. I do not believe that any such reference, or any such talk page discussion exists. Do you disagree with me, and continue to assert that the talk page discussion exists? Or do you accept that you may have been mistaken here? Enchanter 23:52, 23 July 2007 (UTC)

I'm not clear why you're keeping this going, because you seem to be the only one pushing for the change. It doesn't matter what was referenced, or even if it's completely wrong. What matters is what the real example said, and this is what it said. It's the structure that matters for our purposes. Do you see that? Also, as I said, my memory is that a source was produced either in the article or on talk, but I'm not going to go hunting for diffs, because it's irrelevant. With or without a source, it remains an example of SYNT. SlimVirgin ^{(talk)(contribs)} 00:35, 24 July 2007 (UTC)

Another editor very kindly e-mailed me the reference you've been requesting. It's Chicago Manual of Style, 2003, section 17.274): "References to the work of one author as quoted in that of another must cite both works." SlimVirgin ^{(talk)(contribs)} 01:12, 24 July 2007 (UTC)

Slimvirgin, to reiterate, what I said was: "The example claims that the Chicago Manual of Style defines plagiarism as "using a source's information, ideas, words, or structure without citing them". In fact, the Chicago Manual of Style contains no such definition, or anything that resembles it.". You then responded that "... one of the participants posted a link to the Manual of Style on the talk page, showing that it did say that, at least at the time of the discussion.".

You now supply a reference saying "References to the work of one author as quoted in that of another must cite both works.". I don't think that this proves, implies, or even remotely insinuates that that the Chicago manual of style defines plagiarism as "using a source's information, ideas, words, or structure without citing them". Do you agree?

Again, please could you provide a link backing up what your claimed earlier, or confirm if you now think that you were mistaken. Enchanter 17:26, 24 July 2007 (UTC)

Enchanter, as I tried to explain to you several times before, and SV also has, you are totally missing the point. It doesn't matter what the CMOS says. Whether it says A or B or C is immaterial in this case. The point is that an editor tried to use it as a reference where there was no source that connected the CMOS to the subject of the article. Making that connection, relating CMOS to Smith and Jones of the example, without basing it on a source that connects them for us, and thereby advancing a POV, is SYNT. That is all there is to it. You keep barking up the wrong tree and you may be missing the point of what SYNT is all about. Crum375 17:46, 24 July 2007 (UTC)

No, Crum375. As I have repeatedly said before, the editor did not use the CMOS as a reference for the sentence that "defines" plagiarism. If you disagree, please provide a link or a source. Enchanter 23:19, 24 July 2007 (UTC)

Enchanter, you've misunderstood the example, the structure, and I suspect the point of WP:SYNT. The example says: "If Jones's claim that he consulted the original sources is false, this would be contrary to the practice recommended in the Chicago Manual of Style, which requires citation of the source actually consulted. The Chicago Manual of Style does not call violating this rule "plagiarism." Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them."

There is nothing in there that is false that I know of, but the last sentence is uncited. Plagiarism "is defined," but without telling us who defines it. We guess they mean the CMOS, but we can't be certain. But no matter! It was a real example of editors cobbling together bits and pieces of material, some from the Chicago Manual of Style, some from their own ideas, and creating an argument of their own against the position of one of the people they were writing about. That's all anyone needs to see when they read that example — that the editors created an argument of their own. SlimVirgin ^{(talk)(contribs)} 18:07, 24 July 2007 (UTC)

Thank you for confirming that you now think that the sentence "defining" plagiarism was uncited. I assume that this means that no longer claim that there is a source linking it to the Chicago manual, as you had claimed repeatedly above. Please be careful to check your references more carefully in future, as mistakenly claiming that references exist which do not in fact exist is damaging to Wikipedia's efforts to be accurate, as well as a waste of other users time.

However, as I have said before, in the original example the definition was not uncited. It was cited to Harvard's "Writing with Sources" manual, and to a source from "Smith" which directly quoted the definition. It appears that you have missed this, which could well be the source of the misunderstanding. Please confirm if you agree that it was in fact cited.

You say that I have misunderstood the example and the structure. As I have said before, I think it is ambiguous and illogical, and not following the structure of the original case. According to policy page, synthesis means an argument of the structure "A and B, therefore C", where A and B are supported by reliable sources but C isn't. I have asked repeatedly for an explanation of other editors views of how the example is meant to illustrate the structure - in particular, what are points A and B? What are the reliable sources that support them? And what is point C? These straightforward questions have met either with no answer, or vague and contradictory answers which often make incorrect assumptions about what the sources say. The reason, in my view, is that the example is not coherent, not explained clearly, and is based on a case which is excessively complex and difficult to understand. If you disagree, please explain your understanding of the structure of the example. Enchanter 23:19, 24 July 2007 (UTC)

Enchanter, I daresay that Alan Dershowitz would disagree with you on your interpretation of what CMOS says on the subject of sources and plagiarism. However, and more to the point, Slim Virgin and Crum375 are absolutely correct: it doesn't matter what the CMOS says. An editor tried to use it as a reference where there was no source that connected the CMOS to the subject of the article.

Making that connection, CMOS to Smith/Jones, without basing it on a source that connects them for us, advances a POV, and is synthesis. It is from a real example of editors joining together bits and pieces of material, where there was no source connecting the CMOS to the subject of the article, thus creating an argument of their own against the position of one of the people they were writing about.

That’s the main point here; and the current example presents it in a very clear and concise manner. I have seen no proposed replacements that are superior to, or that even equal it. – Dreadstar † 18:53, 24 July 2007 (UTC)

I have never made any "interpretation" of what the CMOS says on the subject of sources and plagiarism. All I have claimed is that the definition of plagiarism cited in the example is not actually there in the CMOS (either in the current versions, or in earlier versions as was previously claimed by SlimVirgin). This is not a subtle question of interpretation; it is a simple matter of seeing whether the source says something or not. It appears that this is no longer in dispute and noone is claiming the the definition was in the CMOS.

You say it "doesn't matter what the CMOS says", because it is from a "real example" of what another editor wrote. However, as I have repeatedly said before, it does not reflect what the other editor wrote. The other editor used a different source (Harvard's "Writing with sources" manual), plus a source in which "Smith" directly quoted from the definition in the manual and implied that "Jones" had committed plagiarism.

If your argument is that it's perfectly OK to misrepresent the CMOS, and misrepresent the original editor and case while still maintaining that it is "real", then I disagree with you. I don't think it's OK to misquote and misrepresent sources and people for the sake of an example.

I also disagree that this example is "clear" - in the past, when other editors have explained what they think that the example means, the answers are vague and contradictory, and often make incorrect assumptions about what the sources said (as you did above). Enchanter 23:19, 24 July 2007 (UTC)

(outdent) Enchanter, this is my last post in this thread, as I think we are just going round in circles. The situation is very clear. It doesn't matter how accurate the CMOS reference is. It doesn't matter what the CMOS definition says about plagiarism, or what it doesn't say. You simply don't seem to get it. The only point that matters is that the editor used a source (CMOS) to try to advance a position, while the source did not refer to Smith and Jones at all. To connect CMOS to Smith and Jones, you need a source that does that for us, and there is no such source here. Nothing else matters. This is the essence of SYNT, and the fact that you still don't get it, shows that this example is excellent, as you are still hung up in the technicalities, not realizing the big picture - you cannot use the CMOS as a source for Smith and Jones, regardless of what it says or doesn't say, unless you have a source that connects CMOS to Smith and Jones. Thanks, Crum375 00:19, 25 July 2007 (UTC)

We are indeed going around in circles, because I have already answered this. You say that the only point that matters is that the editor used the CMOS to advance a position, and that a source was necessary to connect the CMOS definition to the Smith/Jones dispute. In fact, the source used for the definition of plagiarism was Harvard's "Writing with Sources" manual, and there was a source given, in the same paragraph, in which Smith directly quotes this definition of plagiarism, verbatim, from the Harvard manual. Your argument above makes sense. It just does not describe the actual case we are talking about.

(I recognise that I am perhaps oversimplifying here - the actual case is very complex, and there are a number of additional subtleties, such as that Smith does not directly accuse Jones of plagiarism on the source quoted, instead strongly insinuating it. However, the key point that the case is not being properly described stands). Enchanter 00:49, 25 July 2007 (UTC)

Actually, Enchanter, you did make an interpretation of what the CMOS says on the subject of sources and plagiarism. That’s what this is:

"References to the work of one author as quoted in that of another must cite both works.". I don't think that this proves, implies, or even remotely insinuates that that the Chicago manual of style defines plagiarism as "using a source's information, ideas, words, or structure without citing them".

But that’s totally irrelevant, so let me focus on what I see as the real core issue.

We are trying to discuss and distill a long, involved, complex discussion/dispute that extends back over a year in time and across several different articles and their discussion pages. So it’s easy to confuse the different elements in the original discussions around the CMOS-Smith/Jones issues. Let me see if I can identify the three stages of the issue’s history:

1. Original dispute outside Wikipedia between Smith/Jones. One side uses CMOS to back the definition of plagiarism. Whether it’s right or wrong, it is a verifiable event. [1], [2], [3]
2. Wikipedia Editors A and B’s dispute over the issue, and their combining of unrelated CMOS and Smith/Jones sources. (e.g. [4])
3. Editors X, Y and Z then used the above #1 and #2 to create the current example of OR Synthesis.

Just looking at the history, one can easily see that it doesn’t matter what CMOS says, it’s irrelevant. What matters is that various individuals at the beginning of the discussion thought it applied.

The real issue is not really even about what Smith and Jones believed; the relevant issue is how editors A&B applied two different sources to say that Smith was wrong in a Wikipedia article – thus expressing their own opinions by synthesis.

SV, do I have that right? You were there, taking part in all but the original Smith/Jones affair….so you are the true oracle of knowledge for all this...enquiring minds want to know!!...;) – Dreadstar †

I don't see where CMOS is being misrepresented; it looks fine to me. The statements are about how the two sides of the dispute viewed and used the CMOS description. – Dreadstar † 01:03, 25 July 2007 (UTC)

The above summary from Dreadstar is helpful. To add to this, and clarify how the current misquoting of the CMOS came about:

1. "Smith" made the following accusation against "Jones": 'It is left to readers to decide whether Dershowitz committed plagiarism as defined by Harvard University -- "passing off a source's information, ideas, or words as your own by omitting to cite them."'
2. This allegation from Smith was added to the article, citing both this source from Smith giving the above quotation, and to Harvard's "Writing with Sources" manual, from which the quotation that Smith used came.
3. The definition of plagiarism used was clearly cited to the Harvard manual, and clearly relevant because Smith directly quoted directly from it. Including it was not, in my view, original research (although the wording had been distorted, and the way it was written could certainly have been improved).
4. When the example on the policy page was written, the complex original case was misinterpreted. The definition, instead of being described as being cited from the Harvard manual, is misleadingly attributed to the Chicago manual. The Chicago manual does not contain this definition or anything resembling it. Neither Smith, Jones, nor the original article have ever claimed otherwise.

The net result is that Wikipedia has a definition on the policy page that Wikipedia says is sourced from the Chicago manual, but in fact comes from the Harvard manual. The source of this error is not the writers of the original case, but instead the writers of this policy page. To me, this is a straightforward error, which should be fixed, and I am surprised by the arguments saying that it doesn't matter. If we misrepresent a source, it makes Wikipedia look inaccurate. In this case, it is particularly ironic because we are misquoting a source which is specifically about maintaining high academic standards, and doing so on a policy page which is specifically about being true to sources. I don't accept that inaccurate quoting of sources doesn't matter because it is on a policy page rather than an article; being truthful and accurate is important all the time, not just when you are writing articles. I suspect that there is a tendency here to assume that if something has been on a policy page for a long time, it must be correct and accurate. I think this assumption should be challenged - there have been plenty of examples of inaccuracies persisting in policy pages for months without being spotted. Enchanter 17:30, 25 July 2007 (UTC)

Thanks Enchanter, I'm glad to know that my summary was helpful! I believe the conclusion you've come to may be incorrect; the policy is not misquoting or misrepresenting the content of CMOS, the policy is merely referring to the original disputing editors' opinion of what CMOS says.

1. The first mention of CMOS in the Synth section is clearly an identification of material added to Wikipedia by one of the original disputing editors, and is a description of the actual OR.
2. The second mention of CMOS is in the explanation of that OR.

"This entire paragraph is original research, because it expresses the editor's opinion that, given the Chicago Manual of Style's definition of plagiarism, Jones did not commit it."

That does not state anything about the CMOS definion of plagiarism, it comments on the editior's opinion of what CMOS is stating. Perhaps a subtle distinction, but a significant one, nonetheless.

Per your analysis above; since neither Smith nor Jones mentioned the CMOS, and since it was indeed mentioned by another source, but not in relation to the subject of the article under dispute by Editors A&B; the resultant connection between the two by Editors A&B is at the very core of the synthesis. Not what CMOS actually says on the subject.

In this sentence, you describe exactly the OR comitted by Editors A&B, which is why their attempts to add it to the article were rebuffed.

"The Chicago manual does not contain this definition or anything resembling it. Neither Smith, Jones, nor the original article have ever claimed otherwise."

We're not misrepresenting the source in the policy, the original Editors A&B did that in their creation of OR. The policy merely describes that issue. – Dreadstar † 18:29, 25 July 2007 (UTC)

This explaination is clear and makes sense. However, it does not reflect the facts of the case.

You state that ": "We're not misrepresenting the source in the policy, the original Editors A&B did that in their creation of OR.". This is not correct, because the original "Editors A&B" correctly cited the definition of plagiarism to the correct sources (Smith, and the Harvard manual). They did not make the error or citing it to the Chicago manual. That was added later, in this policy page.

And even if the error had been committed by the editors of the original article - which it wasn't - it would still undermine the usefulness of this as an example. This is meant to be an example of someone arguing an unsourced point C from points A and B which are reliably sourced. It is not meant to be an example of someone writing material which is totally unsupported by sources. That's blatant OR, but not an example of synthesis. Enchanter 19:12, 25 July 2007 (UTC)

I am not going to search through potentially thousands of edits made over a period of almost two years on several different articles to find a single diff where Editors A&B attempted to add CMOS-related synth to Wikipeda. However, in the 'chicken or egg' category, it is clear that CMOS was mentioned prior to it's inclusion in the Policy (unless it was reverted and I didn't see it).

1. Talk page discussion about CMOS in relation to the plagiarism claim: 22:29, 5 April 2006
2. CMOS not mentioned in the Policy as of 14:55, 10 April 2006.

Even if you are correct in your historical assertions, IMHO, that still does not undermine the value or legitimacy of the current example. In any case, the example is fine and does not mislead or misinterpret anything, which I believe is the current consensus on the subject.– Dreadstar † 19:44, 25 July 2007 (UTC)

To address Enchanter's second point:

"This is meant to be an example of someone arguing an unsourced point C from points A and B which are reliably sourced. It is not meant to be an example of someone writing material which is totally unsupported by sources."

This above statement is incorrect, the Smith-Jones-Editor-A&B (SJEAB) example is exactly an example of someone arguing an unsourced point C from points A&B, which are reliably sourced.

Even if Jones or his supporters stated that he used the CMOS for his defense, the simple fact that Editors A&B then combined the two to say that "According to CMOS, Jones isn't a plagiarist". That is the synthesis.

It is not an example of someone writing material which is totally unsupported by sources, it is an example of editors synthesizing a new statement from sources.

Besides that, one would really have to examine the entire history of the dispute and know all the details around the event that this is all based on to even begin to see the issues you raise. Issues which are totally irrelevant to the fact that this is a good example that misrepresents absolutely nothing.

If you can write a better example, then please present it here. The ones already presented are inferior to the one currently in place. I guess the only other option is for you to escalate this issue up the chain, if that's the proper way to handle policy disputes. – Dreadstar † 21:46, 25 July 2007 (UTC)

You say that this is "exactly an example of someone arguing an unsourced point C from points A&B, which are reliably sourced."

As I have repeatedly asked before, please provide the reliable sources which you say back up points A and B. I started this thread asking for a source which asserted that the Chicago Manual of Style defines plagiarism as "using a source's information, ideas, words, or structure without citing them" (which I think was clearly from another source). You still haven't provided a source which remotely backs up this statement, beyond asserting that someone once said this on a talk page, without saying what talk page or where.

This is not a dispute about policy; it is a straightforward question about where you have got your information from. You say that something is reliably sourced; I ask what the sources are. Enchanter 00:15, 28 July 2007 (UTC)

Actually I have indeed provided a link to the talk page you mention. I'll provide it once again, for your convenience. I will provide two links to the page which contains the relevant discussions around the discussion of CMOS/plagiarism/OR/Synth: talk page link 1 and talk page link 2

I also provided external links to RS for the same issue: [5], [6], [7]

I must ask you to try and keep things separate in your questions, they seem to be mixing in elements from different venues, time-periods and intentions. As you and I have both pointed out, this is a long, complex issue we're tying to distill - and when generalities are used in requests for sources, it can be quite confusing to try and determine what you are asking for. From my understanding of what you have been asking for and stating, I believe that I've provided evidence for every single question you asked and for every single element of your case, save one - the exact diff that shows the OR being added to Wikipedia.

I'll also repeat this; the points you continue to hammer on are totally and completely irrelevant to the fact that the current version is a very good example that misrepresents absolutely nothing. I believe we're just continuing to argue a straw man at this point. – Dreadstar † 00:58, 28 July 2007 (UTC)

The specific question I asked was for a source backing up the statement that the Chicago Manual of Style defines plagiarism as "using a source's information, ideas, words, or structure without citing them".

You have provided several sources, without any explanation of how you think they back up the statement made:

• These sources [8], [9], [10], none of which assert that that definition comes from the Chicago manual.
• talk page link 1, which contains no mention of this definition of "using a source's information..."
• talk page link 2, which does mention this definition, but does not cite it to the Chicago Manual of Style. Instead, it cites it correctly to the Harvard Writing with Sources manual.

In other words, the only source that you have given that actually mentions the definition attributes it to the Harvard manual - which is what I have said all along.

Is it relevant that the example misquotes the Chicago manual? It is if, as at present, the example asserts that it is a real example of using points A and B which are reliably sourced from real published sources, to argue an unpublished conclusion. The Chicago manual of style is a real, well known, and reputable publication. If we are going to use it in an example, we should use it correctly and not imply that it contains definitions that are simply not there. Enchanter 01:33, 28 July 2007 (UTC)

I believe I've already addressed this issue too. The example does not misquote CMOS, it says that an editor misquoted it. Big difference. Show me the exact line that you believe shows that the Policy is misquoting CMOS, one that is not merely showing the editor's misinterpretation in the example, or the explanation of the example. – Dreadstar † 02:00, 28 July 2007 (UTC)

You say that it's OK to misquote the CMOS, because we not misquoting it ourselves but instead asserting that another editor had misquoted it. I think that this would be a weak argument even if it were true, because the synthesis section makes it clear from the outset that it is dealing with unpublished conclusions C supported by reliably sourced ideas A and B - so the reader is entitled to expect that the definition (point A) used to argue the conclusion that Jones was not guilty of plagiarism (point C) is supported by a reliable source.

However, more fundamentally, I don't think that your assertion that "the editor's misquoted it" is true. I have asked you for sources for this, and you have provided five. However, not one of these actually supported what you were saying. Enchanter 02:30, 28 July 2007 (UTC)

That's not 'more fundamentally', that's actually 'irrelevant'. Fundamentally, the issue is what the example says, whether or not it's understandable, and it's clarity. You refute your own point that the Policy is misquoting CMS. However, if you feel that strongly that it does...then by all means, propose a version that is better. – Dreadstar † 02:41, 28 July 2007 (UTC)

As for your other assertions about the sources I've provided, I disagree with you. I'm sorry they don't satisfy you, but I think they do indeed indicate that Jones and/or his supporters used the CMOS to help define what plagiarism is or isn't, and the editors on the talk page that I provided links to are clearly discussing CMOS in relation to plagiarism. I have done my due dilligence and presented clear and convincing evidence to back all my points. Editors here can read them for themselves and come to their own conclusions.

I must stress, however, that this discussion is a straw man argument and does nothing to show that the current example is in any way misleading, or that it isn't a good example. The issues you raise are irrelevant. I believe that's all I have to say on the subject, and I continue to oppose the current proposed versions, and I support keeping the current one until a better one is found and that meets consensus. – Dreadstar † 02:36, 28 July 2007 (UTC)

I agree entirely that the sources you provide show that editors had discused the CMOS in relation to plagiarsm - this was never in dispute and is obvious to anyone familiar with the history of the article. But that wasn't the question I had asked. The specific question I asked was for a source backing up the statement that the Chicago Manual of Style defines plagiarism as "using a source's information, ideas, words, or structure without citing them".

You claimed that the sources that you provide backed up the statement. I checked the sources, and can find nothing in them to support your position. You have provided no explanation of why you think I might be wrong. Do you still believe that the sources you provided back up that statement?

Enchanter 02:48, 28 July 2007 (UTC)

Just to circle around Synthesis section again

In a previous discussion on the example, I felt that one of the problems was that the policy page didn't really create a usable description of what the issue of synthesis was trying to address. I think if the words are right then the issue of needing a good example fades away. It seems that an example does not seem to address the problem, because if it is obvious, then it fails the subtlety test that people seem to like, and if it is subtle, then it is not obvious what the point it is making is. Let's see if we can get a better wording to explain the issue. I don't personally like the A, B C aproach, it is too algebraic and offputting. I suggest something like this:

When considering edits, it is important to understand that it is not just the introduction of facts that need to be considered, but the analysis of those facts too. In Wikipedia, the process of creating new ideas out of facts is called synthesis. An unsourced analysis of facts is just as much original research as the presentation of information without sources. A simple example would be to present a school's test results over a number of years and based on these make a judgment on whether the school was improving. The provision of citations for the test results still does not allow us to make any significant assessment of the school's performance as we cannot verify that any unusual issues have been properly accounted for.

While creating articles is about gathering information and there is a low level of analysis that cannot be avoided in that process, it is assumed that there should never be any significant new analysis. The test for what is reasonable is "Does it advance a position?". If the analysis required to present an idea is more than simple summary then an objector would be entitled to request the citation of the analysis, especially if it creates a contentious viewpoint. This applies however well grounded in fact the analysis may be. If no justification is forthcoming, then the analysis should be removed to ensure that content is verifiable.

I'm not wedded to this wording at all, but thought that some clarity on the main page would be useful. Comments? Spenny 23:24, 25 July 2007 (UTC)

The heart of this seems to be "advance a position." That seems to imply that an encyclopedia does not or cannot advance any position, that it is a dead repository of statements that it can be shown were made by someone else. Nowhere is it shown that the assumed (by Wikipedia) nature of an encyclopedia is correct. The notion seems to be that an encyclopedia can be created entirely out of things said by others about particular topics. Implicit in that assumption is the further assumption that if there is a controversial topic it is up to representatives of each side of the controversy to publish statements that can be cited to offset what has been said on the other side and that in all cases they do so. In the real world that doesn't happen. In the real world there are those who craft deception that passes muster in the press and media (today that is ridiculously easy.) While there are those who could perform an analysis of what the deceivers say there is no guarantee that they will (the deceivers also engage in tactics to suppress such analysis in the press and media.) Wikipedia can continue with its current attitude - no problem, free world, have at it. Continuing with the current attitude diminishes the quality of Wikipedia. No problem, free world, have at it. But diminished means diminished. Not that what I think matters but currently that seems to be what Wikipedia intends, and tit also appears that Wikipedia will have an essentially unending stream of apologists who affirm the policies even as they acknowledge that the policies defy truth and logic. No problem, etc.

I can get away, in talk, with saying that the proponents of privatization of Social Security have not provided the details of any plan. (The pertinent article itself acknowledges that fact, but in a manner that treats it as a mere detail.) What is impossible in Wikipedia is to point out that the proponents of privatization are engaged in a cynical scheme to manipulate public opinion. That's a fact of overriding importance - but Wikipedia wants to pretend that there are no such facts - unless identified somewhere else. If "identified somewhere else" you can bet that the "somewhere else" is associated with those who oppose such privatization. The objection then will be that the information violates NPOV, where NPOV means that lies must get the same coverage as truth. No problem, etc. That's what Wikipedia wants, and what Wikipedia wants is not to be an encyclopedia. Even in discussions in which one side is heavily engaged in manipulaiton and deception Wikipedia wants to have rules that treat both sides equally. (Rather like treating murderers on one side and police, prosecutors and courts on the other side equally.) At the very least it would seem some attention should be given to the question of whether forbidding all original research and all synthesis (that is, valid logic) is compatible with producing a good encyclopedia. (OK, Wikipedia doesn't want anything. I think my meaning is clear enough despite that unfortunately over-anthropomorphic phrasing.) --Minasbeede 02:40, 26 July 2007 (UTC)

I think my perspective is that the policy isn't very readable at times so people don't really know what the policy is. By cleaning it up, we can then see what it is. We then may not like what we see. So I see it as a two step process: make the policy clear, with that clarity it can be challenged.

I think some policy is obscure because it is uncomfortable. I think the whole citation of sources is a minefield, because we can have someone write a sound, neutral, unsourced summary which any reasonable being can see is a valid neutral summary, which can then be pulled apart by citation lawyers who confuse good writing with research. My particular issue is that the best articles are those written by those who understand the subject and simply can pull in the citations to justify those statements which they can recognise themselves to be controversial (or perhaps cannot, they just know their subject). By disallowing analysis in the first instance, it becomes difficult to evolve articles written coherently. I guess I am agreeing, NOR can undermine NPOV. Spenny 09:20, 26 July 2007 (UTC)

I'm comfortable with the two step process but what I've observed is that it is hard or impossible to give a good example for why synthesis should be forbidden (we may have to wait forever.) I think you are correct about articles written by someone who understands the subject and I think having such authors has been one of the strengths of the Encyclopedia Britannica. (Incidentally, not having such authors has been one of the most severe weaknesses of the US textbook publishing industry: the publishers often employ liberal arts graduates who do not understand science to write science textbooks. It's no wonder that the impression is created that science is "hard.") I also think that, in general, there's an arsenal of tools (yikes: mixed metaphor) that writers use and among those are logic and synthesis. Strip writing down to he said/she said and you weaken it, with the obvious result of weak articles.

There's also a need to use style and constructs appropriate to the material. Turning again to one of my favorite examples (the CIA/DIA white paper on the trailers) the style of that white paper itself is wrong - and revelatory. When I first read it I expected there to be a block diagram of the system: that's the starting point for understanding the function of a system (just as the block diagram is the starting point for constructing the system when it's being designed.) There was none. Instead, there was a lot of pseudo-scientific or pseudo-engineering mumbo-jumbo. Then look at the annex to the Duelfer report that deals with the trailers and prominent in that annex is a block diagram. Also prominent is an absence of mumbo-jumbo of any sort. Creationists operate in a similar fashion: they want non-scientists-level understanding and reasoning (mumbo-jumbo, or close to it) to trump science. Suppressing synthesis has the effect of promoting mumbo-jumbo - synthesis is a good tool to combat mumbo-jumbo. Forbidding synthesis makes it easier to get by with mumbo-jumbo.

Even with simple, straight citations there's a need for examination and consideration of what is cited to see if it is appropriate for inclusion in Wikipedia. Forbidding synthesis eliminates the need for similar examination and consideration for the material that is synthesized but there's no clear advantage to eliminating the need for examination and consideration in that one particular construct: it is still necessary - for every word, phase, sentence and paragraph - to subject new material (or changed material) to a process of examination and consideration. In that process new material is frequently removed. Why is that not possible (and ordinary) for material that arises from synthesis? Why throw out all synthesis, including all good synthesis, in order to avoid the task of examining and considering material that incorporates synthesis? --Minasbeede 15:04, 26 July 2007 (UTC)

The observations above are entirely correct: a traditional encyclopedia will be written from the perspective that the author knows the truth and presents it as an authority on the subject, opinion and all, without excessive referential justification of every single point. Depending on the subject, the absence of editorial opinion is a real weakness of wiki. It still seems to me that the inability to come up with a good example is suspiciously close to demonstrating that there exists no unflawed example. There simply exist circumstances in which drawing a conclusion is inevitable when a neutral presentation of facts is made. Sandpiper 23:05, 27 July 2007 (UTC)

I think it is actually slightly worse than that. Say I have a 100% perfect source that says "Sunburn is due to the sun's UV rays and is a naturally occurring thing." I have another source, say the BBC news website (considered "good enough" for our purposes), that says that "Scientists say, 'climate change is causing sunburn due to the failure of the ozone layer and something must be done.'" It is factually correct, perfectly reasonable in context, but it is not the whole story. I can write an article on climate change that says "Scientists blame sunburn on climate change" and actually it is a pretty reasonable statement (go with the flow, assume we are not wise in this subject and don't pick too many holes in the example, it works well enough for the point) and is reasonably well sourced, though we can see the flaw in the logic if we know our subject. If I try to refute this argument with the original source against a well-meaning editor who believes what the BBC says, I cannot as there is no link to climate change, even though it is an impeccable source, it does not appear relate to the subject at hand. Indeed, it may be such a basic premise that nobody has written something that says climate change is not the root cause of sunburn that I cannot disprove a sourced statement without doing the original research of that reasoning (and indeed there will be many comments saying increased sunburn is a problem of the ozone layer - it all fits together). If we have a no-loopholes policy on NOR then this is a genuine issue. One would hope editorial good sense would win out but policy is on the side of verifiability not truth - you can't always unverify a mistake.

For example, you can be in the position that an urban myth can gain credence and there is nothing you can do under strict NOR until someone spots the problem and publishes a refutation in a reliable source. So the principle introduced to stop fringe theories can be used to create and defend fringe theories. Spenny 23:58, 27 July 2007 (UTC)

Hypothetical question

Let's say you conducted an interview with the creator of a TV show or whatever, and then decided to personally put the information you gathered into a Wikipedia article prior to the interview being printed or whatever. Let's also say that the interview was not faked, and that it actually took place. This, in effect, could be consitered original reasearch and, as such, would prevent (possibly) valueable information from being added to an article. How would you get around something like that? --Wack'd ^{Talk to me! • Admire my handiwork!} 17:32, 27 July 2007 (UTC)

If you can audio/video proof of the interview to back up the stuff mentioned from the interview, then I don't see a problem. Corpx 17:39, 27 July 2007 (UTC)

Basically what Corpx said. The interview must be somehow published before it can be used; that is, the video or its transcript must be posted online or shown on television. Black Falcon ^(Talk) 18:17, 27 July 2007 (UTC)

Claims must be capable of being attributed to a reliable source. If someone challenges a claim it must be possible to provide a citation to a reliable published (accessable) source. WAS 4.250 18:43, 27 July 2007 (UTC)

Data-crunching OR?

Could do with some input over at Talk:Family First Party on the issue of presenting statistics/identifying patterns that are derived from sourced data but not explicitly stated at the cited source. --Calair 06:32, 28 July 2007 (UTC)

TV shows

Would it be OR to watch a TV show and write a plot summary about it? I would think so, because it is interpretation of a primary source. Corpx 03:22, 26 July 2007 (UTC)

The writing of summaries is recognised as acceptable practice on the assumption that the program is there to be viewed by others who can assess the validity of the summary. It is common practice in books, films and programs. The issue then comes as to whether you attempt some critical analysis. So to do a summary which states that Dr Who battled the Daleks on Earth in the 1960s will be fine. As soon as you start saying it was a weak episode or exciting or whatever, you are stepping into the world of research and you would need to find some reviewer to back you up.

More to the point in the world of WikiRules would be why this was being done. Is the series or episode notable? I pick a Dr Who example as, certainly from a UK perspective, it has a strong cultural and historical links and is notable in its own right, and there are certain classic episodes where they may be notable, either for the characters or the events. However, an episodic review of SpongeBob SquarePants might not be considered of such merit. Wikipedia is not a fan site and I think plenty of people would think that episode reviews crosses the boundary of notability. IMDB is probably the proper place on the web for that sort of thing. Spenny 09:04, 26 July 2007 (UTC)

As long as the plot summary is descriptive, rather than interpretive, there should be no problem with OR. That is, the summary may only contain information about what was shown; content about hidden themes, supposed similarities with other shows or real-life events (except those which can be visually confirmed), and so on must be attributed to reliable sources. Black Falcon ^(Talk) 17:58, 26 July 2007 (UTC)

What about the verifiability of the stuff mentioned from a TV show? Corpx 14:39, 26 July 2007 (UTC)

What do you mean? Are you referring to whether a TV show can be a realiable source for information about a topic other than the show itself? Black Falcon ^(Talk) 17:58, 26 July 2007 (UTC)

All I know about Lupe Vélez I first learned from Frasier. --Minasbeede 18:23, 26 July 2007 (UTC)

I mean, wouldnt watching a TV show and describing it be the same as a person witnessing an accident and writing about it on wikipedia? Corpx 18:30, 26 July 2007 (UTC)

Not exactly. A TV show is published and preserved, and therefore reproducible over space and time. A video recording of an accident could be used as a source if it was published (i.e., if it could be accessed/viewed by others). Black Falcon ^(Talk) 20:25, 26 July 2007 (UTC)

Same point I was going to make. A TV show is a discernable unit of fiction, reproducable and more importantly citable. When writing a summary of an episode it is generally assumed that uncontroversial claims about the episode (e.g. a summary of the events that take place within it) are being sourced from the episode itself, hence why an explicit citation is unnecessary. -- Y|yukichigai (_ramble argue ^check) 12:22, 30 July 2007 (UTC)

NOR and articles on TV and cartoon characters

I noticed many of the articles on TV and cartoon characters contains a lot of "facts" on their personality derived from their actions in various episodes. (The Simpsons, Family Guy, Daria etc...) Is this OR? I don't think there is any official word on the characters' behaviour, but certainly fans shouldn't just speculate this and that?--Kylohk 08:55, 30 July 2007 (UTC)

If it's based on their actions, then yes, it's OR since it draws conclusions. The exception is if the text in question is cited and sourced to some reliable website that makes that specific conclusion. Generally this is incredibly rare, and regardless if you don't see a citation it's almost certainly OR.

The other thing you need to keep in mind is whether or not some episode or other work within that realm of fiction has specifically mentioned within the work that the character has whatever specific aspect of their personality, e.g. the article on Adrian Monk can mention that he has numerous phobias because several characters across several episodes have specifically mentioned that fact. -- Y|yukichigai (_ramble argue ^check) 12:16, 30 July 2007 (UTC)

A lot of it looks to me like original research, but the people who make that judgement need to be the people who know the subject matter best, not people who know little about the subject. We get people deleting well known facts from bio articles and science articles who don't know the first thing about the subject and say they challenge the claim. Encyclopedia writing has to first be about actually knowing what you are writing about and not the mere blind application of rules. WAS 4.250 15:54, 30 July 2007 (UTC)

I sympathize, but the current policy specifically excludes truth and logic from primacy. On Wikipedia the rules overshadow all else. If that makes Wikipedia incompatible with "encyclopedia" then so be it: that's the choice that has been made. --Minasbeede 16:45, 30 July 2007 (UTC)

WP:IAR exists specifically to refute any claim that an improvement to wikipedia as an encyclopedia is allowed to be ruled out by some rule. 17:15, 30 July 2007 (UTC)

Super Duper, but not having the rule would be better. The issue isn't what is claimed about Wikipedia, the issue is that of what Wikipedia is - and actual practice defines that. To all appearances Wikipedia is a venue in which truth and logic are secondary considerations and that lower status for truth and logic is explicitly asserted. I'm all for forbidding/deleting feeble logic, "truth" known only to a few nutcases ("flat earth," etc.), conclusions based on special or strained definitions or redefinitions and specious reasoning. That isn't the effect nor the use of the rules. Still, vive IAR.--Minasbeede 22:02, 30 July 2007 (UTC)

Every year, Wikipedia is a better encyclopedia than it was the year before. If that ever stops being the case, anyone including for-profit enterprises can fork the content. Ain't freedom wonderful? WAS 4.250 22:56, 30 July 2007 (UTC)

What exactly deems one as an expert on a TV show? The expertise part should apply to real world subjects (science/math etc), not fictional topics. Corpx 17:01, 30 July 2007 (UTC)

No no no. The idea is not "expert". The idea is that you have to have enough familiarity with the what you wish to delete that you can be confident that you are being reasonable about challenging it. If you have never watched a show, it is unreasonable to believe you can tell what character descriptions need a printed source. WAS 4.250 17:15, 30 July 2007 (UTC)

I've seen these - the point is that it has to be straightforward synthesis, judgements that anyone would draw from viewing the source. One useful approach is to restate in your own words things that a character did, and allow the reader to draw character traits from that.

The other parts of this discussion are getting rapidly off-topic, but I'd assert that even someone not familiar with a topic can question and/or remove an unsourced and dubious assertion, like "he lives in the moon and his hair is made of spaghetti" in an article about a supposed real person. Dcoetzee 01:19, 3 August 2007 (UTC)

{{cite I looked at the bloody file}}

In the article The Microsoft Sound I tried to include the fact that the name and location of the start-up file on Windows 95 is C:\Windows\Media\The Microsoft Sound.wav. Another editor keeps deleting this because it is uncited. Here are the relevant extracts from the Talk page:

Somebody is now complaining because the fact that under Windows 95 the file is C:\Windows\Media\The Microsoft Sound.wav is uncited. Well, I run Windows 95, and that is where it is. I have looked for an appropriate cite template. I found {{cite book}} and {{cite web}}, but couldn't find {{cite I looked at the bloody file}}. The problem, here, I would suggest is too much reading of guidelines and not enough thinking about their purpose. HairyWombat 19:10, 4 August 2007 (UTC)

"{{cite I looked at the bloody file}}" goes against Wikipedia:No original research. SchuminWeb (Talk) 20:16, 4 August 2007 (UTC)

Use this to cite your claim Corpx 21:09, 4 August 2007 (UTC)

Software is published. One can look at the published software. But software is self altering, so you have to look at it in the published form and not the installed form. It could install itself differently on different computers. WAS 4.250 22:10, 4 August 2007 (UTC)

Is this the sillest citation request you have ever seen, or am I in the wrong here? Any guidance would be gratefully received. HairyWombat 20:58, 4 August 2007 (UTC)

It does seem a bit POINTy to ask for a citation for something like that, though I see Corpx found one for you. SlimVirgin ^{(talk)(contribs)} 00:26, 5 August 2007 (UTC)

Citing reliable published sources is used not only to establish credibility but also to establish relevance and notability. Perhaps the issue is not if it is true but whether or not is should be in the article even if true. WAS 4.250 22:07, 4 August 2007 (UTC)

On the flip side, a lot of highly notable information cannot easily be backed by credible sources. Most sources which are considered "credible" are oriented toward academic material, rather than popular culture and the like. --xDanielx^Talk 06:13, 5 August 2007 (UTC)

Yes, indeed. It's a big problem for people working in these areas. I'd be in favor of adding a pop-culture exception to this policy, but the last two times I tried, it got no consensus. SlimVirgin ^{(talk)(contribs)} 06:19, 5 August 2007 (UTC)

If you ever feel up for a third attempt, please let me know so I can offer my support. Just this morning I started an article for DaFont and within hours it had been tagged as a deletion candidate (explanation). --xDanielx^Talk 09:33, 5 August 2007 (UTC)

Me too, I'd support it. I'm not sure if it would apply here, but there is a recent Arbcom decision regarding pop-culture in paranormal articles: ArbCom Paranormal Pop. – Dreadstar † 10:31, 5 August 2007 (UTC)

When you try to pull off anything, sign me up. --Kizor 20:45, 6 August 2007 (UTC)

"Pop culture is an exception" did not get consensus. Perhaps something along the lines of "Claims with these characteristics (blah blah blah) are an exception" might get consensus. Can anyone here say exactly what those characteristics would be and exactly what requirement is therefore waived and why? Considerations: effect on fake wrestling bios; effect on self promotion in pop culture (this comic is known to be noteable; I've said so a thousand times in the blogs!); effect on history of pop culture. WAS 4.250 11:23, 5 August 2007 (UTC)

I would lke to see "Facts that are widely known and easily verifiable are an exception". An example used by Rich Farmbrough over on Talk:The Microsoft Sound is the fact that the Harry Potter series contains seven books, and is about a young wizard called Harry Potter. This would take care of the location of a particular file (which started this discussion). If you read between the lines of the guidelines, they almost say this already. All that is needed, I would suggest, is something more obvious to fight off the Wikilawyers. HairyWombat 20:16, 5 August 2007 (UTC)

I don't know that the "Harry Potter series contains seven books" and even if true it is only true for a certain definition of "series" and only true as of some time frame. So we need at least "As of 2007 the Harry Potter series of books sold by [so and so] contains seven books" and a link to [so and so]. I know some people who believe that evolution is widely known and easily verifiable and others who believe the opposite. There is a story of three men on a train who see a cow and one says "Look cows here are black"; responds a second "No, that cow is black"; responds the third "No, this side of that cow is black." WAS 4.250 23:41, 5 August 2007 (UTC)

Speaking of "pop culture", what are your opinions on Bigfoot in popular culture being OR? In the article's AfD, it is being argued that the article violates NOR. – Dreadstar † 18:03, 5 August 2007 (UTC)

Seems fine to me. I voted to keep. Xdanielx, if you have a specific idea of what kind of exception is needed for pop culture, feel free to draft something and we can take a look at it. SlimVirgin ^{(talk)(contribs)} 23:50, 5 August 2007 (UTC)

I agree with ya, SV..I voted keep too. – Dreadstar † 01:39, 6 August 2007 (UTC)

Misuse of Wikipedia Policies

Perhaps it would be instructive to consider ways in which those with an agenda (that agenda involving propagation of an untruth) structuring what they say so that it doesn't violate any Wikipedia policy but in a way that any rebuttal can be removed for violation of a Wikipedia policy. As an example, in the debate over the privatization of Social Security it appears that great care has been taken to separate any exposition of the possible details of such a program (it should be very well noted that the details are of overriding importance) from close association with President Bush. That way he can make broad, general, unsubstantiated claims and the other supporters can provide what they claim to be details about the proposed program without creating any association between Bush and the claimed details. If, subsequently, it turns out that the program got established on the basis of people believing those claimed details apologists can correctly point out that Bush (and by extension, all who pushed to get the plan enacted) never committed to those details nor promised that the program would be as described by any others (even though those others were acting in concert with the administration.) (Taking a broader view, politicians in general employ such tactics: they weren't invented by Bush.)

The net effect for Wikipedia would be that Wikipedia was misused as a part of a propaganda campaign. Wikipedia restricts itself to what is said by others. Analyzing what might be meant or what motives there might be for making deceptive statements is forbidden unless that analysis was first published somewhere else. (The kicker is that those doing the analysis would almost certainly be opponents of the proposal and the proponents of the proposal would object to the inclusion of their non-neutral statements - even though those non-neutral statements were truthful and accurate.) In the past that analysis of hidden meaning or motives was known as "reading between the lines." Reading between the lines probably is an individual obligation but when something that purports to be a neutral encyclopedia publishes material that adds support to a political deception a false appearance of truth and honesty is bolstered and fosters a disinclination to "read between the lines." --Minasbeede 17:01, 5 August 2007 (UTC)

It's a real danger indeed. It is the so called 'spin doctor' action. To not to talk about Global Warming issues: it's impossible to draw any statement that is no contestable.--Stefanomencarelli 19:05, 5 August 2007 (UTC)

Yes, spin doctoring. It isn't Wikipedia that is the target, it's the press and media, who more and more now operate in a (easily manipulated) he said-she said mode. The media, particularly, don't choose spokespeople for their analytical and honest approach, they choose them for shock value. It could be that somewhere someone has refuted one of the preposterous claims made in spin-doctoring mode - but chances are the press and media will pay zero attention to that. As a result, Wikipedia, restricting itself to published sources, is effectively forced into being an echo of untruth. --Minasbeede 21:28, 5 August 2007 (UTC)

You're right that it's a problem, Misasbeede. The way to deal with it is to dig deeper for alternative sources. In the example you gave above, I don't think it matters that it might be opponents who read between the lines: so long as the material is in a reliable publication, it can be used as a source. You may also find academics who analyse these kinds of proposals, or editorials in good newspapers. SlimVirgin ^{(talk)(contribs)} 21:48, 5 August 2007 (UTC)

My real hope was that the problem could be examined by others and the scope of the problem fleshed out. Thanks. --Minasbeede 01:45, 6 August 2007 (UTC)

Another example of 'spin doctors'. Don't hate me, but i must say it: in CNN and Fox site there are or were the list of deads in irak war.

CCN stated so: 'coalition losses of war in Irak', all the coalition losses and with photo and datas of their death.

Then i looked on Fox: the title of this section: 'Fallen heroes of war on terror'. Only US soldiers, general datas, but not photos (that, seen all toghever, gives you some displeasure for all those 20-years boys killed somewhere). This is a small example of what i say as spin doctors: one says impartially: coalition losses, another 'fallen (US) heroes'. Not the same manners to face this issue. An infid danger to who wants try to understand some situations.--Stefanomencarelli 12:06, 6 August 2007 (UTC)

Still on OR policy

Hi to all. I am a newbie of wiki.en and i have made in these days some contribution to new articles. For many reasons many of them were deleted by other wikipedians because this or that.

Lastly, there is an edit-war around an article, regarding an old japaneise aircrafts that i, seen the datas available on others sources, had rated better than a similar type with the same engine. So i have been accused to have said not a simple NNPOV, but a OR. I have made an 'original research' becuase, seen speed, range and some others parameteres, i made a perfectly reasonable sintesys claiming that that aircraft was overall better than the other. So i made an OR. Even an aeronautical engeneer, or so auto-claimed, have stated that my analisis was rational, but, still, violated OR.

Someone above have said a precise sentence: there is probably no good wiki-article, that not violate such policy. And i agree, it's madness.

I knew in wiki, that there is the proibition of 'Original researches' meant as: you cannot post datas and stufies made by your self. Obviousely, this was then traslated by some 'integralist' in the fullowing assumption: you cannot think on your own. So this is not a manner to make an encyclopedia, not even happened in the whole history of mankind. But is happening here, in wiki, the most big box of culture for the future generations. And who don't agree with this is a 'traitor' of the wiki-spirit.

I want to remark that this policy is not only madness, but also unpratical, because the original means were too stretched on all the possible ways, making a cure worse than sickness.

• So let's say, that nobody has ever stated that Pluto is smaller than Jupiter, right? I, or someone else, state this. So this is a 'forbidden analisis'?
• I claim that a tennis ball is smaller and lighter than Moon: so, because nobody has already written this (or no one that i can found everywhere) this is a policy violation?
• I claim that Ferrari Testarossa is faster and more costly than Ford T: is it a OR too?
• I claim that F-16 is a more modern and capable fighter than spitfire, this is also is a OR
• I claim that OJ Simson is not Micheal Jackson: also this cerebral activity is OR.

So you have the point. I agree to have a well referenced articles, atleast on important things. I not agree on the handpoles of Citation needed also because one says: the sky is blue.

But as data elaboration, there is no way that this works. It's an absurd rules, born by a reasonable concern, but grew as a 'patriot act' thinked to interdict the mind activity of the contibutors, as 'PRESUME GOOD FAITH' is reversed in 'PRESUME BAD FAITH, so ask references in every statement, even the less important'. Someone above has given the example of the Public security privatized. Sure, this is an hot argument, that needs to pay attenction IF Wikipedia wants to make 'culture', to let the people know. If there is consensous, a statement should had been written about even without sources:If someone don't agree, then Wiki is always 2 thing characterized:

• 1-it's freely modiphiable and
• 2-Disclaimer: Wiki cannot guarantee the absolute reliability of its information. Even with OR full operative, can eventually only worsened the quality of the articles.

Because this policy negates the valour of mind elaboration made by millions of persons, that has willing to help to create this project.

• Also, this is another issue: if one claimed one thing, maybe thanks to a reference found in a old article or book out of market, maybe misinterpreting it or deliberatly giving a different meaning, so for Wiki is ALL RIGHT.

If these are the terms of Wiki, if this is not an 'Encyclopedia' but a discharge of 'he/she said', so set up the overall failure of this project. Away from guarantee better informations, exculding logic, matematic and good sense, and above all, a dialectic between autors, so Wiki cannot simply cope with any of encyclopedias. It's not written by professionists, and if Plagued by such totally unlogic and destructive (See=TAliban) policies, that the informations findable here are of the same level of the 'good' food of Mc Donalds.To me, if i have clue of these policies on Wiki.en and the zelous application (even in ininfluent statements) made by some wikipedians, i simply didn't accepted to write a word in the NS0. Just as example.--Stefanomencarelli 11:29, 5 August 2007 (UTC)PS_Please change Wikipedia, the free encyclopedia: it's not free and not an encyclopedia (and no wikilove, also).

Your conclusions are are based on serious misconceptions of the nature of this policy. It does not preclude editors from making logical deductions (defined narrowly). Let's look at your five examples:

1. Maybe nobody has explicitly stated that Pluto is smaller than Jupiter. However, people have provided figures about the mass of the two. The mass of Jupiter is 1.898 6×10²⁷ and that of Pluto is (1.305±0.007)×10²². There is no original research involved in stating that the first figure is larger than the second (it is a simple mathematical inequality) and that, therefore, Pluto has a smaller mass than Jupiter. However, in order to arrive at the conclusion, you must first present the data (and sources).
2. The tennis ball and Moon example is not original research for the same reason as above (though I can't think of a reason that you'd ever need to note that bit of information).
3. The Ferrari and Ford example is not OR for the same reasons as above. Surely there is somewhere information about the top speed of each (usually provided by the manufacturer) and of their average or suggested retail price.
4. The F-16 and Spitfire example may be OR, depending on what you mean by "modern" and "capable". You can state that it has a higher maximum ceiling, a higher maximum speed, is more maneuverable, but the the terms "modern" and "capable" are themselves not unambiguous. Of course, if the terms have precise industry definitions, then it is not OR.
5. The OJ and MJ example is most definitely not OR. I am who I am and no one else. ;)

It might help if you pointed to specific dispute that gave rise to your disillusionment. Cheers, Black Falcon ^(Talk) 16:33, 5 August 2007 (UTC)

Very much. Thanks. PS. Did you know that B-2's cost (1993) is 5000 times higher than B-17s (in 1942)? It's a disturbing effect of the modern times tecnology.--Stefanomencarelli 18:57, 5 August 2007 (UTC)

• Black Falcon is certainly right that the policy does not preclude logical deductions, but the article could certainly use some improvement in clarity. It tends to be interpreted very broadly as a result of the ambiguity. I don't blame Stefanomencarelli for being confused, and I commend him for bringing the issue to discussion rather than interpreting it in a way that suits his interests. --xDanielx^Talk 09:34, 7 August 2007 (UTC)

Removal of No Original Research policy

I would like to state for the record my disgust with this policy of 'No original research', as it is comepletely ludicrous. All research at one point or another starts OFF being original research by some person or organization. To claim 'original research' is a bad thing is the height of hypocrisy. If nobody did any research to begin with, none of thise information would be available to disseminate on such a large scale. We would have no dictionaries, no encyclopedias, no atlases, no cenus counts.

Strictly speaking, if a person does not take the time to do research themselves, even if said research is simply the act of researching a SOURCE in order to make an addition to a page (which in itself is a form of original research because you have to FIND the source if it's not easily found), there is no research at all, and all pages have to have additions made based on new or rediscovered information by people who make these changes, often having to make a point of, you guessed it, doing original research to locate this information.

My opinion on the situation is this: This policy is hypocritical and must be done away with.

Almost forgot to sign this. Warwolf1 09:36, 6 August 2007 (UTC)

Fully agree. And let me say, that actually i am trathed contiously by an admin called Bzuk and some other: they rollbacked at first sight almost every stuff i post. They attacked me (that i am newbie, only her by 10 days,after all) with all the arsenal of reasons: before grammar herrors (too many..) then it's NNPOV, then it's needing citation every time i say one thing, then what i write is defined 'unuseful' (yes, they rated as 'unuseful' what i write, in a 10kb page, when there are 100kb all around). Fullowing some discussions about angel's sex, now i have rollbacked (for unknow reasons..) Reggiane Re.2001 fighter, and user Akradecki have treaten me in my talk with this words:

Secondly, you have been told repeatedly not to introduce original research. Per our definition, that includes "The term also applies to any unpublished analysis or synthesis of published material that appears to advance a position". So if you compare several things and come to your own conclusions, that is original research. If you cannot provide citations, do not introduce text. Consider this a final warning for that, as well. I will be watching your contributions, and if I find you adding unsourced material, I will revert it and will consider it vandalism, and I will take appropriate action. Your constructive edits are welcome here, your unconstructive attitude is not

So i am a sort of vandal, because i wrote that 'Ferrari is faster than Ford T'. Perhaps i have exaggerated abit, but i am beaten so many times, that i strongly suspect that not only Bzuk and others exaggerates a lot, but also that this OR policy is a weapon to strike everything with the more stupid motivation: really, if i write Pluto is smaller than Jupiter, i am beaten. I don't know how hell i can continue with this 'encyclopedia' that seems ruled in a totally crazy manner.--Stefanomencarelli 12:39, 6 August 2007 (UTC)

Standard responses: Reasonable synthesis of sources is permitted, such as "Pluto is smaller than Jupiter" or restating things in your own words. We don't claim that original research is bad (to the contrary it's a very good thing), just that Wikipedia is not an appropriate venue for initial publication - stick to professional journals for that. Dcoetzee 01:13, 7 August 2007 (UTC)

• I agree with Dcoetzee. The fundamentals of the policy in my opinion are reasonable, though not indisputable. Still, I think the article severely lacks clarity. No original research is too often interpreted as no original thought, resulting in a regression/reductio ad absurdum that many fail to see. I think a re-evaluation and cleanup would do it well, especially for the synthesis section, which is rather ambiguous and often interpreted very broadly. --xDanielx^Talk 09:29, 7 August 2007 (UTC)

Yes, no OR should be like: i cannot write that Shoa is never happened because i found new proofs (not sources). All the rest is more and more questionble as teh interpretation of this policy is strechted to the narrower levels. And with 2 millions editors, good luck!--Stefanomencarelli 09:49, 7 August 2007 (UTC)

No original research is intended to be a bit broad, and well it should be. Wikipedia is an encyclopedia, not a research journal. We should be reporting what the reliable secondary sources available to us say about the subject. This is a significant principle in Wikipedia. Subjects lacking reliable secondary sources are generally not considered notable and are subject to deletion. NPOV demands not that we formulate some theoretically neutral view, but rather that we present views in proportion to the available reliable sources and plainly report the claims & observations of others. We should not be drawing out our own conclusions or interpreting primary sources. There are plenty of wikis and Foundation projects where original research is not only welcome, but even encouraged. However, this is not one of them. Vassyana 10:46, 7 August 2007 (UTC)

Yes, it is indeed a significant principle in Wikipedia but that it is significant has no bearing on whether it is correct or wise. It is a huge stretch to claim that concluding, on one's own, that "Pluto is smaller than Jupiter" is in any sense "original research." But I agree: the current Wikipedia policy treats such simple logical conclusions just the same as it would treat a crackpot idea in physics. The only thing served by finding "available reliable sources" for such simple obvious conclusions as "Pluto is smaller than Jupiter" is a narrow-minded adherence to a rule. --Minasbeede 00:59, 8 August 2007 (UTC)

If the observations are "so obvious" such as "Pluto is smaller than Jupiter", it's a fairly sure bet someone has already published the observation. Try textbooks and general audience books on the subject. Textbooks tend to make a lot of "obvious" observations (Pluto being smaller than Jupiter, or the Earth is round instead of flat, are both great examples). General audience books tend to make many "plain-to-see" observations, because they are trying to present the topic to a very broad audience with minimal to no professional knowledge. Such observations and facts are far easier to reference than is often presented. Vassyana 10:46, 7 August 2007 (UTC)

Maybe "such observations and facts are far easier to reference than is often presented" but "often" isn't "always" - often is a weasel word used to duck the problem of there not being a source for a particular piece of knowledge that would be appropriate in a Wikipedia article because for the reader it enhances the understanding of the topic covered. Such observations and facts may in fact be totally missing in any published source yet be perfectly reasonable. While I don't know of a context within which one would want to say, in a Wikipedia article, that a basketball is bigger than a grapefruit a basketball is bigger than a grapefruit. (If you want to argue some super-large grapefruit then change that to "grape.") It borders on arrogant and insulting (and perhaps also aggressively ignorant) to send an editor on a chase to find a published source for that very evident snippet of knowledge. In any case, the apparent most-frequent use of NOR and the prohibition on synthesis is to remove material that some advocate of some cause would rather not see published in Wikipedia. It isn't a desire to improve Wikipedia that is the reason for wanting removal, it's a desire to perpetuate something that can be seen, with the application of only simple logic, to be untrue. The advocates aren't interested in truth nor in the quality of Wikipedia, they want to support an extreme position against a valid disproof of that position. It's the original justification for NOR turned on end: NOR is used to preserve the illogical and untrue. (It appears to me that there never will be an example provided for the prohibition on synthesis that exhibits a flaw in synthesis itself.) --Minasbeede 00:59, 8 August 2007 (UTC)

Yes, but unfortunately not every claim which is trivially true has been supported by a reliable source. If I were composing a list of the longest bridges in the world, I might have trouble finding a source to back up that "4,310 feet is longer than 3,917 feet." Probably a bad example, since there are other lists out there - but you get the idea. The WP:NOR article is severely vague, and that vagueness should result in a "do what's reasonable, keeping in mind some general principles" attitude, not "we may not think for ourselves." --xDanielx^Talk 17:45, 7 August 2007 (UTC)

WP:IAR is right up your alley. Do whatever makes wikipedia a better encyclopedia, understanding that sometimes you are wrong; and ignore the rules. The rules are for those who need rules, not for those who know better than the rules. So long as you back off when challenged, you can ignore the rules completely. WAS 4.250 02:02, 8 August 2007 (UTC)

Preference for primary sources for quotations?

I think there is a consensus that when an article quotes or (noncontroversially) paraphrases what somebody said, a primary source is preferable to a secondary source. Am I correct? For example, if you say, "In 1782, Benjamin Franklin said _____", you would want to cite, if possible, to the primary source (i.e., the newspaper, diary, or public record where the quote originally occurs), rather than a 2005 article in the The New Republic that itself cites the primary source. This makes it much easier for the reader to verify that the quotation is accurate, and avoids a level of indirection. COGDEN 19:06, 26 July 2007 (UTC)

I disagree with the above generalization. "In 1782, Benjamin Franklin said _____" is from An Address to the Public from the Pennsylvania Society for Promoting the Abolition of Slavery then I would usually agree. But if this quote is from private correspondence it would be better to attribute it to an biography or historical paper were it is published. I think the preference should be for the most relevant source which is not necessarily the primary source. Especially if the WP article follows with commentary on the quote the primary source will not be most relevant.--BirgitteSB 19:16, 26 July 2007 (UTC)

Quotes should be generally avoided in Wikipedia. Wikiquote would be the appropriate place for quotations. If a quote is really necessary (which is rare), it would be better to cite it to a secondary source, since such a source would also include claims and commentary about the quotation, which would be far more useful for building a solid article and avoid the temptations of providing commentary or presenting the text in a suggestive way. Vassyana 21:52, 26 July 2007 (UTC)

I'm only talking about verifiable primary sources that have been published or are accessible to the public for verification. This issue is important, because I deal primarily with religious history articles where direct quotes are usually the most non-controversial things in the article, and the secondary sources almost always color the quotes, and really strain hard to find subtle meanings in the text which not all scholars agree with. I've never been able to maintain a stable article in such cases without using the original quote or a close paraphrase, which the reader can verify and is not controversial. Then, the article can safely say what all the secondary sources preach about the quote, which is also not controversial, and everyone's happy.

At the very minimum, I think saying in this policy that secondary sources are preferred is an over-generalization, and does not reflect consensus. There's no such consensus, at least, in my area of the Wikipedia (controversial religious history). COGDEN 20:42, 27 July 2007 (UTC)

Direct quotes maybe non-controversial, but they are also not encyclopedic content. It is very hard to use direct quotes in an encyclopedic way without straying into original research. This is why the preference is that articles rely on secondary sources. I cannot say anything definitive without a real example in context of an article. However I do not see how you have shown any evidence the current policy as written is either an overgeneralization or without consensus. Note: I mean to say I find that assertion a very strong statement based on what you have shared with us here thus far--BirgitteSB 20:54, 27 July 2007 (UTC)

Let me show you an example from Early life of Joseph Smith, Jr., a featured article that appeared on the front page on Dec. 2005, which has the following paragraph, which despite being full of serious religious land-mines that are not apparent on the surface, has survived virtually unchanged since then, and everyone seems to be okay with it:

According to another story, in either 1819 (Tucker 1867, p. 19) or 1822 (Howe 1834, p. 240), while the older Smith males were digging a well for Clark Chase, a Palmyra neighbor, at a depth of more than twenty feet they reportedly found an unusual stone (Harris 1859, p. 163). This stone was described as either white and glassy, shaped like a child's foot (Tucker 1867, p. 19), or "chocolate-colored, somewhat egg-shaped" (Roberts 1930, 1:129). Fascinated, Smith reportedly took this stone and later began to see things inside it clairvoyantly (Tucker 1867, p. 20). Some scholars have concluded that these two accounts refer to two distinct stones found in 1819-1820 and 1822, and that these stories have in some cases been conflated (Quinn 1998). Other scholars believe that the two accounts refer to the same event in 1822 (Vogel 1994, p. 202). However, this has little support among his current followers.

In this situation (and there are countless similar examples), you have a number of issues that make primary reliance on secondary sources a problem: (1) the dates and descriptions of the stone(s) are critical facts. (2) Mormon scholars reject Tucker and Howe as unreliable, while secular scholars believe they are reliable. (3) Mormon scholars accept Quinn's conclusions and reject Vogel's. (4) Neither Quinn nor Vogel quote all of the above quotes verbatim. (5) The primary sources are in the public domain and on the internet, and therefore are more accessible than the secondary sources. Believe me, when this article was being written, lots of ways of expressing this and similarly-controversial paragraphs were tried, and they all failed to stick. Everyone who reviewed the article thought this was a good compromise, and the article was featured on the front page. COGDEN 21:19, 27 July 2007 (UTC)

The main problem with this method is the original research involved. You have several accounts of one or two incidents of a similar nature with conflicting details. An neutral encyclopedic treatment of this situation should compare the differing accounts and describe the two schools of thought about whether these accounts are about the same or different incidents (The above example has generally given it this treatment). But if you back up this treatment with primary sources you are straying into original research. In the very first sentence you explain that a story which may have happened at one time or another relates this incident and the sentance is backed up with 3 separate primary sources (I presume by the dates and my quick search of Tucker). Going with my presumption, I imagine that Tucker, Howe, and Harris do not ever refer to each others separate accounts. If that is correct, then it is original research to state that the two dates given by Tucker and Howe are simply different date about the same story. If these sources do not refer to each other then this statement attributing these dates to the same thing is not properly referenced. It needs to referenced to a source that speaks of both the accounts given in Tucker and Howe. The only thing connecting those two dates to the same story given these references is a Wikipedia editor's original research. Using other sources, however, I imagine it would be possible to compare and contrast different accounts given by primary sources without violating the policy on original research. If my assumptions about what these sources contain based on a cursory look are wrong please disregard this analysis.

You could use Tucker to describe Tucker's account and you can use Howe to describe Howe's account. However you cannot compare and contrast these account without using a references which already does this. I realize that relying on the research of Wikipedian's rather than having use the contentious work of scholars may make things easier in this topic, but I feel you and your fellow editors are going to have to find another way to handle the situation.--BirgitteSB 21:55, 27 July 2007 (UTC)

It seems to me that this is exactly a case of where it is proper to cite primary sources. As a general principle i would expect to be told where some information originates, not to be given a reference to a secondary source which tells me where the information can be found. That is indeed pointless indirection. This example does not use the primary sources to claim that there is contention, merely to show the things which are disputed. It goes on to discuss the nature of the dispute, and then secondary sources become useful. Ogden was not debating the merits of how well sourced the argument about conflicting accounts might be, but whether the primary sourced quotes were appropriately inluded. They were. If policy says this phrasing is unencyclopedic and ought to be struck out, then policy clearly needs revising. Sandpiper 22:45, 27 July 2007 (UTC)

But the information that Howe and Tucker give different dates for the same incident does not originate from either Howe or Tucker. Ogden was asking for more information on why this policy says Wikipedia articles should rely on reliable, published secondary sources. It is because writing a nuetral encyclopedic article with primary sources leads to original research. In the above example that is exactly what happens. The phrasing above is encyclopedic (which I thought I said), but it relies on original research. Anytime you have a single sentence with three different citations in three different locations within the sentence that is a red flag for original research. It likely means that none of the citations actually support the whole statement in it's entirety. And that is a problem--BirgitteSB 22:55, 27 July 2007 (UTC)

You don't need to cite a source for something so obvious as that 1820 and 1822 are different dates. Obvious statements don't need to be cited, nor necessarily do statements that are not controversial (like the fact that Tucker and Howe (and Harris and Roberts) were referring to the same story, which all the secondary sources take for granted as being obvious). There's also no requirement that everything cited in a given sentence or paragraph be discussed in a single source, if combining them is not controversial. COGDEN 01:29, 28 July 2007 (UTC)

I am sorry, I thought that the above was saying there was disagreement whether the two acounts were about a single event or seperate events. I thought it was contraversial and I agree that if it is not you do not need the level of citation that is causing the problem. I still would not cite the sentenance in three different locations but as uncontraversial material doesn't neccesarily need inline citations, I do not have problem in this case.--BirgitteSB 14:50, 28 July 2007 (UTC)

Another example from First Vision:

While in the vision, he said he saw one or more "personages," who are described differently in Smith's various accounts. In one account, Smith said he "saw the Lord." In diary entries, he said he saw a "visitation of Angels" or a "vision of angels" that included "a personage," and then "another personage" who testified that "Jesus Christ is the Son of God," as well as "many angels". In later accounts, Smith consistently said that he had seen two personages who appeared one after the other. These personages "exactly resembled each other in their features or likeness." The first personage had "light complexion, blue eyes, a piece of white cloth drawn over his shoulders, his right arm bare." One of the personages called Smith by name "and said, (pointing to the other), 'This is my beloved Son, hear him.'" Most Latter Day Saints believe that these personages were God the Father and Jesus.

Here, as before, all of the primary sources are online and in the public domain, and the precise nature of the vision is stongly debated as a matter of religion. This particular article is right now in the heat of a major debate, but none of the editors have a problem with this particular paragraph, which is entirely factual and contains nothing new, except that the issues have probably never been summarized so succinctly and neutrally before. COGDEN 21:51, 27 July 2007 (UTC)

This example is less problamatic as regards this policy. However you are lacking a reference altogether for the first sentance which is the only one with a comparison of the accounts. I don't think parts involving quotes are original research this time. But the paragraph as a whole is much less encyclopedic than the the first example. This is the other problem with direct quotes, they tend to fill space without giving much context about the importance of what is said. Why do we need to know the color of the eyes or the dress of this personage(s)? What is the point of recounting all this exact terminology (angel, personage, Lord} used by Smith seperately from the analysis of what his word choices mean in the larger picture? This is not an explanation of Smith's First Vision but simply a list of his accounts of it. Not origianl research, but not particulary encyclopedic either.--BirgitteSB 22:46, 27 July 2007 (UTC)

Once again, it's obvious and not-controversial that the vision is described differently. No need for a citation there. It's just a background introductory sentence. As to the significance of these elements of the vision, that's discussed later in the article. But for reasons of style, clarity, and readability, we needed a succinct account of the vision all in one place, rather than breaking up the vision account with commentary. COGDEN 01:29, 28 July 2007 (UTC)

Again I assumed it was contraversial, mainly because of your earlier comments here that you are using this method in areas of contraversy. I am not telling you your style is to seperate illustrative material from encyclopedic material is wrong, but showing you how primary sources can be problamatic if this was the main focus of the article. They are not a problem for illustrative material. Which is why they are not disallowed, however the large majority of the article should consist of nuetral encyclopedic material instead. Using primary sources to write nuetral encyclopedic material will tend to cross the line into original research. This is why the policy states, Wikipedia articles should rely on reliable, published secondary sources. You are fine using primary sources for illustrative quotes, but lots of illustrative quotes makes a poor WP article. Beware of using primarry sources for analysis or explanations in articles, and be certain the articles main focus is on explaining the topic and the contraversy rather simply listing what primary sources say about the topic. Does this all make sense to you?--BirgitteSB 15:05, 28 July 2007 (UTC)

We shouldn't confuse OR with NPOV. Actually, "neutral encyclopedic material" is not often found in religious-history-related secondary sources, which are typically controversial and opinionated. When all the secondary source does is interpret the primary sources, adding an opinionated viewpoint, you're getting farther away from neutrality. Moreover, unless there is a consensus, it's just as easy to create original research from secondary sources as from primary sources.

This policy page is supposed to reflect consensus, and I see no consensus to favor secondary sources as a general rule, especially a bolded general rule. I propose just deleting the bolded line, or perhaps replacing it with the following statement:

1. "Direct quotations or non-controversial paraphrases should be cited to accessible and reliable primary sources, if possible, so long as the context of the quotation is apparent from the source itself or from a secondary source. All other statements in an article should be cited to reliable, published secondary sources. Care should be taken to ensure that all statements are cited neutrally."

COGDEN 00:25, 3 August 2007 (UTC)

Any comments on this? COGDEN 02:37, 8 August 2007 (UTC)

I would not support this in any policy and it certainly doesn't belong in this one. My overall objection is because I find "primary source" and "secondary source" ambiguous enough to make this paragraph meaningless. I consider the use of terms in this policy a "crutch" to explain the issue with original reasearch in simplified manner, but it is not completely accurate. Expanding on this inaccuracy outside of the issue of original research would be a bad idea. The above proposal has absolutely nothing to do with original research however and definitely shouldn't be in this policy. You might get more suggestions in a new sections, this is a little buried.--BirgitteSB 13:32, 8 August 2007 (UTC)

I agree this is getting outside the scope of "original research", but so is any stated general policy preference for secondary sources over primary sources. So I'll withdraw the suggestion, but we need to remove the stated general policy preference for secondary sources (i.e., any suggestion that primary sources, as a rule, are disfavored), because clearly there is no consensus now to that effect. In at least one situation, several editors have agreed that primary sources are favored. If there's no consensus, it shouldn't be in the policy (at least yet). Therefore, I'm going to just make the edit in a way I think is non-controversial and start a new topic to discuss. COGDEN 19:12, 9 August 2007 (UTC)

WP:COI considerations apply where the claims are related to religious dogma. WAS 4.250 00:11, 28 July 2007 (UTC)

I don't think WP:COI is implicated just because of an editor's beliefs. That problem would arise, for example, if an editor were a church's spokesperson or a church official. There's no conflict problem with editing in areas where you have strong beliefs, as long as you follow WP:NPOV. COGDEN 01:29, 28 July 2007 (UTC)

Strong beliefs warp judgement. WAS 4.250 02:54, 28 July 2007 (UTC)

That may be true, but it's not a conflict of interest. If religious people were discouraged from writing about the Eucharist, Upanayanam, or Thetans, who would have sufficient interest and knowledge in these religious subjects to write the articles? COGDEN 00:25, 3 August 2007 (UTC)

A person's beliefs have nothing to do with COI, otherwise we'd have a situation where Wikipedians who are Roman Catholic couldn't edit Pope, or people who opposed the war in Iraq weren't allowed to edit 2003 invasion of Iraq. The definition of a COI is: "Where an editor must forego advancing the aims of Wikipedia in order to advance outside interests, he stands in a conflict of interest." Nothing to do with internal mental states. SlimVirgin ^{(talk)(contribs)} 00:32, 5 August 2007 (UTC)

I agree 100%. But sometimes COI is mistakenly use to refer to "pushing a POV". I know that I have sometimes said the one when I meant the other. The difference lies in proof that "pushing a POV" is caused by a COI; and when dealing with people who don't reveal their real identity "pushing a POV" is as close to COI as it gets. I had this thought lately; that a page where "pushing a POV" can be aired out might be useful. I would like a page where I can say that I think that slim and friends edits of Jew/Muslim/etc pages promote NPOV. I would like a page where everyone can say their opinion concerning the edits of Slim and friends edits of animal rights and agriculture issues. (I am in favor of her animal rights article edits and against her agriculture article edits.) WAS 4.250 02:09, 5 August 2007 (UTC)

Revised primary, secondary, and tertiary section

I've revised the section on primary, secondary, and tertiary sources to remove the stated "rule" that primary sources should not be used except in rare situations. There is no consensus for this proposition, and this policy article can only include rules for which there is a consensus. If someone wants to establish a consensus for this, then we can re-insert it, but for now, it should not be there.

Primary sources are widely used across Wikipedia, and several editors on this talk page have agreed that primary sources are even preferable to secondary sources (not just acceptable) in at least some situations. Moreover, most of the reasons people have proposed for favoring either primary or secondary sources in particular situations have nothing to do with original research. I've revised this section to reflect this, and to focus the section back on the subject of original research. As revised, the section makes no judgment about what type of source is preferable, so long as it is not used to support original research. Indeed, there are original research pitfalls in citing any source, primary or otherwise. Other than removing the non-consensus preference for secondary sources, I don't think the revision makes any significant changes, other than in terms of clarity and style. COGDEN 19:23, 9 August 2007 (UTC)

I am personally fence-sitting about this change and am interested in what other people think. I really don't think the old wording was so strongly against primary sources as you do, but if I think the old wording was less extreme than you did that only means I find this change even less significant than you do. Basically as long as OR is prohibited a lot of the usage of first-hand accounts is prohibited by default, so I do not see any real change in practice coming from this edit. This edit is basically re-focusing the issue on whether original research occurs rather than advice against the danger of original research when using primary sources. The only thing that I find is lost in this change, is that I think it is impossible to write an appropriate (i.e. following NPOV, NOR, NOT, and RS) Wikipedia article using only first-hand accounts. That old wording that articles should rely on secondary sources precluded that. However that point might fit in better on some guideline page on sourcing as it is really a common sense interpretation of the main policies rather than a main principle.

But if consensus supports this change, I wonder why have the section at all. Why take the space to define these terms (and to do so inaccurately by oversimplication), just to say they can all be used so long as there is no original research? What does this section add 'about original research not covered in the rest of the policy? --BirgitteSB 20:13, 9 August 2007 (UTC)

I agree that there is probably little practical change here, but I think it's important to keep the idea of a hierarchy of sources. In other words, to move beyond a primary source, you need a secondary source, to move beyond a secondary source in some non-obvious way (such as paraphrasing) you need a tertiary source, otherwise in each case you're doing original research.

I think it might theoretically be possible to write an article using only primary sources, but only when the secondary sources merely point out the obvious. (According to Wikipedia:Citing sources, you don't necessarily need to provide citations to propositions that are not likely to be challenged.) I can't really think of a good example, though. Any Wikipedia article, of course, has to be citable to at least one secondary source, otherwise the subject matter would be non-notable. COGDEN 21:37, 9 August 2007 (UTC)

(outdent) I've reverted the edit. The preference for secondary sources is long established in Wikipedia policy. Trying to alter this principle affects other policies and guidelines on a significant level. (And I would even argue it would be a fundamental shift in the focus and standards of Wikipedia.) For example. WP:V#Sources reads in the opening sentence: "Articles should rely on reliable, third-party published sources with a reputation for fact-checking and accuracy." (emphasis added) The preference for non-primary source is clearly stated. To address one specific aspect reverted: Reliable sources are "are authors or publications regarded as trustworthy or authoritative in relation to the subject at hand. Reliable publications are those with an established structure for fact-checking and editorial oversight." (emphasis added) This means that primary sources cannot be called "reliable sources" without contradicting the very guideline linked. In essence, the preference for secondary sources is simply a repetition of what is contained in the relevant policies and guidelines. Trying to make your sort of change will require a discussion about making significant changes to multiple policies and rules, or at the very least will have to be addressed at WT:V and WT:RS, as those are the principles that best cover this particular content rule. Vassyana 22:50, 9 August 2007 (UTC)

The WP:V#Sources quote regarding "reliable, third-party published sources" has nothing to do with whether the source is "primary" or "secondary". I don't think your understanding primary source is quite the same as other editors. A primary source can indeed be a reliable source. For example, (1) the Constitution of the United States, (2) the published Diary of Anne Frank, and (3) Darwin's Origin of Species are all very good primary sources that are extremely reliable sources of information about (1) U.S. constitutional law, (2) events in the life of Anne Frank, and (3) Darwin's theories. Journal articles, as well, contain primary information that are certainly reliable, and in some cases can be cited without original research (so long as no specialized expertise is required to interpret them).

Also, your reversion is improper because the secondary-source preference was added in late 2006, with essentially no discussion, after three years of having a neutral policy toward sources. No consensus was established for the addition in late 2006, and if you look at actual Wikipedia practice and comments on this talk page, I think it's clear that no consensus exists today. I'm re-reverting to protect the status quo as of late 2006. If we come to a consensus as to specific instances in which secondary sources or primary sources (or even tertiary sources) are preferable, then we can move forward. COGDEN 23:33, 9 August 2007 (UTC)

The examples you cite are not generally reliable sources for what you claim. Unless it is simply a quote or a very plain recounting of what the source says, it is probably original research in these cases. You cannot use the U.S. Constitution to argue constitutional law under policy. Since people who are professionally trained engage in vigorous debates about the meaning of various constitutional phrases, one can hardly cite the Constitution to back a claim. In that case, we should be citing reliable secondary sources which report on that discussion. The Constitution should only be quoted or mentioned as used in those sources, except potentially in rare exception, to avoid original research. Anne Frank's diary is widely read and discussed, but not a reliable source in and of itself. First person accounts are notoriously unreliable and there is a whole field of literature just addressing how accurate such accounts may be. In the case of famous accounts like Anne Frank, the available material is vast. If the first person account is accurate, we should rely on reliable secondary sources to report that it is indeed accurate and why. Origin of the Species is a much discussed and analyzed work. Much like the Constitution, there is incredible debate about the implications and intended meaning of various parts of Darwin's work, making it very difficult to present without original research. In all of these cases, reliable secondary sources are almost absolutely required, due to the interpretive issues involved. Beyond all that, it's just plain common sense that primary sources have the most potential for original research abuse, since they lack the analysis, commentary and so forth included in secondary sources. Vassyana 01:12, 10 August 2007 (UTC)

Often the recent sources, e.g. the Catholic Encyclopedia, introduce strong POV where the ancient sources have more varied, if not milder, POV. In this case, NPOV requires us to use the ancient sources (strictly speaking, many of these are themselves secondary sources). Moreover, an improper synthesis based on secondary sources is at least as bad as an improper synthesis based on primary sources. In this case, I think it would be better to describe appropriate and inappropriate types of inference, than to forbid otherwise-appropriate sources because some editors might misuse them. Jacob Haller 02:52, 10 August 2007 (UTC)

We are allowed to use some sense in evaluating sources. It's blatantly obvious that some sources are written with a very particular bias. However, there should be enough reliable secondary sources to both build an article and fulfill NPOV. If not, there's serious doubt as to whether that topic is even notable. In the case of Christian history, as implied here, there are a plethora of modern reliable sources that not only take into account the available texts, but also modern archaeological and anthropological knowledge. There are also numerous issues regarding the reliability of those ancient sources, including questions of authorship and accuracy of accounts. Those issues should not be addressed without employing secondary (or tertiary) sources. Even accepting those ancient texts at face value is a form of original research, since modern scholarship sharply investigates the authorship and accuracy of them. Vassyana 06:13, 10 August 2007 (UTC)

I'm reverting back to the most recent established consensus. For example, there was a discussion of recent changes in which only a single user objected. Changes like yours should be discussed, as those changes were above and as they were previously (see here for one of many examples). Looking over the history, it looks like a slow change towards disallowing primary sources. The changes was a result of a gradual process and much discussion. There was a point of change when some people noticed and/or objected. Much debate ensued, a consensus was reached and has been fairly stable since. The burden is on you to demonstrate that consensus has changed. Vassyana 01:12, 10 August 2007 (UTC)

I for one find the current text encourages wiki-lawyering to (1) remove material referenced to multiple primary sources and (2) insert material referenced to dubious tertiary sources. People regularly cite these pages when supporting opposite sides, and reaching opposite conclusions, in such edit wars. Therefore, this requires (1) enough clean-up that people can agree on what this means and (2) enough consideration of different perspectives, different concerns, etc. so that the rewrite doesn't run roughshod over appropriate practices. I suggest that primary, secondary, and sometimes tertiary sources are all suitable for general reference, given other indicators of reliability.

• For quotations, primary sources are best.
• Where one source presents interpretive problems:

• Other works by the same author may resolve the problems
• Other works by his/her supporters and opponents will not resolve the problems
• Nonpartisan secondary sources may resolve the problems

• Where multiple sources present interpretive problems, including contradictions

• Nonpartisan secondary sources may resolve the problems
• If these are not available, it is better to present multiple sides than to rely on partisan sources or to create new syntheses.

• Where multiple interpretations, conflicts, and contradictions are present

• (not really sure)

This doesn't cover everything. One more thing: when various sources define movements, religions, etc., we shouldn't take these definitions too seriously, and should never take these definitions more seriously than how people consistently describe themselves. I'm bloody sick of the argument that Ben Tucker's ideas don't meet "the" "definition" of socialism. Jacob Haller 00:04, 10 August 2007 (UTC)

As WAS below mentions, it's how a source is used that makes it primary, secondary, or tertiary. From what I understand, and using your example, if you cite text from Ben Tucker to say in an article "Tucker describes himself as a socialist," then that's using Tucker as a primary source. That's OK. But if you cite Tucker to say "Tucker is a socialist," then you are using Tucker as a secondary source about himself. That's not OK if there are other sources that disagree with his self-description. You can use Tucker as a primary source but not as a secondary source. That is, you can use Tucker as a source about what he says but you can't use Tucker as a source to state in an article that what he says is true. That's my understanding of the way things are supposed to work around here. WAS 4.250 please correct me if I'm wrong. Southern Snake 05:05, 10 August 2007 (UTC)

No source is itself (qua source) primary, secondary, or tertiary. It is the use of a source that makes it primary, secondary, or tertiary with regard to that use. Without even basic understanding of the terminology we are using; WP:IAR is the only hope. WAS 4.250 03:32, 10 August 2007 (UTC)

Where in Wikipedia is this explained? Southern Snake 04:57, 10 August 2007 (UTC)

Is this what you're looking for? WP:NOR Primary, secondary and tertiary sources. There's also this degraded article: Wikipedia:Use of primary sources in Wikipedia. – Dreadstar † 05:05, 10 August 2007 (UTC)

No because that doesn't explain how usage makes something a primary or secondary source. And it doesn't explain what I just explained above. Southern Snake 05:06, 10 August 2007 (UTC)

Oh, sorry. I wasn't sure what part of this you were asking about. I missed your top post above. I'm also curious how use of a source makes it primary, secondary or tertiary. I thought that it was implicit within the source itself, according to how it was created. A transcript of testimony in a trial is considered a primary source, and it's a primary source no matter how it's used. I believe the same goes for secondary source's, and tertiary sources. The are what they are no matter how they're used. – Dreadstar † 05:14, 10 August 2007 (UTC)

I could be all wrong here so bear with me. It's my understanding that a person can't be used as a source for what he says as being true. He can only be used as a source as evidence for what he says. Is this explained in Wikipedia? For example, Tucker may say "I was the first socialist," but we can't use Tucker as a source to state in an article "Tucker was the first socialist." We can only use his writings as as source to show what he said. Is this sourcing issue discussed anywhere? Southern Snake 05:22, 10 August 2007 (UTC)

I guess what I'm claiming is that it's improper to use a primary source as if it's a secondary source. I haven't seen this explained on Wikipedia. Southern Snake 05:28, 10 August 2007 (UTC)

In WP:SELFPUB, it says "Material from self-published and questionable sources may be used as sources in articles about themselves, so long as...<list>". In the situation you describe, I would recommend attributing the statement to its source (Tucker) and cite where the statement was made, per WP:CITE Say where you got it. (e.g.) "In a BBC News interview, Tucker said his last name was Tucker".[1] Use of primary, secondary and tertiary sources is pretty well defined, so it would be improper to use a primary as though it were a secondary..it's inherently a primary and should be treated as such. – Dreadstar † 05:38, 10 August 2007 (UTC)

Exactly. This probably should be explained then in one of these policy articles. It looks like this is what the user Jacob Haller is having a problem with. Among other things, I think he wants to be able to say in an article "Tucker is a socialist," based on the fact that Tucker said "I am a socialist" when other sources say he's not a socialist. But, if he does that, he would be using a primary source as a secondary source. Someone needs to write up something in policy to explain this. I can see this issue coming up a lot in various forms, where a primary source is used a source for the truth of the source itself. Southern Snake 05:46, 10 August 2007 (UTC)

Thanks, Vassyana. I see you reverted the proposed change again today. I agree that there's not a consensus. I feel that any change in policy that upgrades the status of primary sources is going to degrade the quality of the encyclopedia. Here's the sort of thing that worries me: suppose I have personal information about some matter, information that hasn't been published, but that I would like to cite in Wikipedia. If any primary source is valid, then why couldn't I simply write an affidavit, file it at a courthouse, upload it to a commons, and then cite it in the article? In my mind, the current policy helps guard against this sort of thing. TimidGuy 15:59, 10 August 2007 (UTC)

We are allowed to use published primary sources. What protects us from your scenerio is "undue weight". So what that your affidavit says so and so. It is a reliable source that you claim so and so; but where is the reliable published source that says your saying so and so is relevant or encyclopedic? WAS 4.250 17:17, 10 August 2007 (UTC)

Here in Italy

Here in Italy we have a proverb: 'Made the law, decouvered the trick'. Is it so terrible, as overall rule, that the stuff published will be eventually challenghed with proper facts by millions other wikicontributors, to generate a concrete evolution and improvement of the articles? All the rules of the worlds will never assure the reliability and the truth over wikipedia datas. You cannot cite any source and do so a perfectly true claim. I can witness a politician contested by the crow (with even photos), and the day after a 'secondary source' as a newspaper will be 'available' to distort totally the facts happened, claiming the hot welcome he had. I can found a family dead under ruins, bombed by some fighter, and the day later see the claims on mass-media that they were 'all taliban warriors'. So go figure, the sources worths nothing if one support one or the other POV, there are all around even too examples to be made. If this screening of Wiki informations is impossible, then no reliable information could be guaranteed by wikipedia. Sorry.

More: a Personal Point ov View is not necessarly bad itself, seen that a judge, after all, express a sentence both 'personal' and 'neutral'. OTOH, there are the advocates or speakers of politicians that claiming things that they don't think but are directly the POV of they bosses. So a speaker, not sayng proper POV statements, 'should be' acceptable, and a judge not.

Perhaps i miss somethings, but let me repeat, the real needings for Wiki are: Be bold, Use goodsense, Ignore the rules and Search consensus plus the best: Search the truth. These should worths in Wiki, much more than 11112x10Small Textpolicies, often in contradiction one against the other, and good mainly to waste time and create a mass of burocrats. Let darwinsm do the work on the pages, and leave the 'creationism' party, that pretends to solve all with the 'table of law'. It dont' works as the free discussion (and as free discussion i mean talk with someone that don't treath handlyng the policies as hammers).--Stefanomencarelli 12:35, 10 August 2007 (UTC)

Kinds of claims first, kinds of sources second?

Perhaps we should define different kinds of claims, and then what kinds of sources can support each claim. As extreme cases: Jacob Haller 19:39, 10 August 2007 (UTC)

1. Albert believes this: Albert's own writings are the best way to support or debunk this claim. Even his own writings which may not meet WP:RS for other purposes are more reliable than others' writings which do meet WP:RS.
2. Albert is an brilliant physicist: Albert's own writings are useless here. Others' writings which assess his work are better here.
3. Movement A believes this: Different members of the movement may present different definitions. I think there are good reasons to favor internal definitions over external ones.
4. Movement A believes this; person B is a member of movement A; therefore person B believes this: Such reasoning involves improper synthesis and should be avoided. Direct citation of person B's own views is important.
5. Movement A believes this; person B does not believe this; therefore person B is not a member of movement A: Such reasoning invites improper synthesis and should be avoided. Direct citation of person B's non-membership in the movement is preferred, but general statements of person B's membership in group C and group C's non-membership in group A may be necessary if sources are scarce and disputed sources claim person B's membership in group A.
   1. Or such reasoning cannot support positive statements, without direct citation for the conclusion, but it can challenge positive statements.

Just some thoughts in response and general agreement:

1. Generally agreed. This is already covered in WP:RS and well-established. The exception to the general rule is when third party reliable sources make other claims. As a more common example, people often profess a particular faith due to social and legal situations, but often held a "hidden" faith due to circumstances. So, there might be a historical English figure that professed to be Anglican, even in many personal letters, but held to Catholic beliefs secretly to preserve his social standing and/or privileges.
2. Agreed.
3. Generally agreed. A number of groups have "sacred" and "secret" practices that are detailed by third party reliable sources. Also, analysis of those beliefs is best left to published reliable sources. For example, the Catholic Catechism is a very authoritative and reliable document regarding the particulars of the Roman Catholic dogma. However, it would be best left to outside reliable references to address how that dogma is put into pragmatic practice (or rather, how it is "really" used).
4. Agreed. For example, there are many practicing Catholics who are pro-choice, support the use of contraception and have other individual beliefs potentially at odds with church dogma.
5. Agreed, see above.
   1. I'm not sure entirely what you mean, but that sounds like it could potentially cross the line into original research. It would probably be acceptable to say something like: "Person X described themselves as a confirmed Roman Catholic. Reference Y disputes this stating that Person X did not attend Mass for most of their adult life." Just some thoughts. Be well! Vassyana 20:27, 10 August 2007 (UTC)

I'm not at all sure, but thought it might help to make positive assertions slightly harder and challenges/counterclaims slightly easier. Jacob Haller 02:15, 11 August 2007 (UTC)

Where are primary sources best?

Expand and/or criticize:

1. Determining what an individual believed or wrote.
   1. Other works by the same author may clarify the work.
   2. But works by their supporters and opponents often add new ideas or drop old ones.

Where are secondary sources best?

Expand and/or criticize:

1. Determining the importance of the person/event from the secondary source's POV.
2. Untangling the dates and order of events.
   1. Annals, diaries, and similar works can also untangle the dates and order of events, if they are trustworthy.
   2. (I have used unclear primary sources, in desperation).
3. Assessing the reliability of supposed primary sources.
4. Summarizing the development and/or use of a scientific theory.
5. Summarizing the assumptions and/or implications of scientific or mathematical/logical claims, but not political, theological, etc. ones.

Wikipedia is a secondary source on many articles; change the policy, delete the articles, or ignore?

I see a contradiction and I need to seek objective clarification. Specifically, many articles exist on wikipedia that are based on reliable primary sources. Many of these are at first blush benign, such as New Hampshire communities by household income. In this case, the only source is a reliable primary source: an official New Hampshire state government website that allows the extraction of data from a database.

Apparently this database is based on U.S. census data.

Effectively, the New Hampshire database is a restatement of U.S. census data. It contains a variety of fields.

The problem with the article New Hampshire communities by household income is that the original contributor decided which fields to extract, and what article to create. Using the same database, another contributor might create:

• New Hampshire communities ranked by number of households with children under the age of 5 or
• New York communities that have volunteer fire departments or
• New Hampshire communities supplied by Granite State Electric?

While ranking New Hampshire communities by household income seems logical and benign, such an article turns wikipedia into a secondary source, and not the tertiary wikipedia is supposed to be.

In deciding which of such articles to write, the original contributor is making an evaluation of the various fields, and selecting one.

Additionally, such lists are not encyclopedic, and are often substitutes for categories. E.g., why have an article titled "" when there is a category Category:Academics of the University of Cambridge?

Ramifications: Of course, many such articles exist. Deleting such articles will likely leave a sour taste in many wikipedians’ mouths.

To address this, I see two options:

1. Change the policy to explicitly allow, subject to consensus on an article by article basis, the ability to create such articles?
2. Start deleting the myriad of articles that turn Wikipedia into a secondary source.
3. Ignore the problem (which doesn't address it).

Looking forward to other's thoughts. (If changing this policy is out of the question, could you kindly direct where I should take this issue?) TableManners 06:02, 11 August 2007 (UTC)

See also section for above.

• New Hampshire Database
• New Hampshire Database sources

Closing and signing. TableManners 06:02, 11 August 2007 (UTC)

I think you are seeing a problem where none exists. Editorial judgement is a necessary part of creating this encyclopedia. We are always choosing to use this source and not that one; choosing to include this claim and not that one; choosing to juxtapose these claims but not those. WAS 4.250 19:29, 11 August 2007 (UTC)

Agree, agree, agree.See posts above of Minesbeene, just as example.--Stefanomencarelli 14:16, 12 August 2007 (UTC)

Wikipedia is not strictly a tertiary source. We're a mixed secondary/tertiary source based on both primary and secondary sources. There is always a certain bias in how articles are presented, including the choice of topics, and you have the right to question whether an article such as this one should exist via the talk page and/or appropriate deletion channels. Dcoetzee 23:43, 12 August 2007 (UTC)

Synthesis of published material serving to advance a position

Stevertigo's recent addition is a bit too abstract (sorry!) for me to follow. Are there some examples of articles that were abstracted upward as he describes? Tom Harrison ^Talk 19:33, 11 August 2007 (UTC)

Perhaps Evolution (disambiguation) and/or Evolution (term)? WAS 4.250 20:12, 11 August 2007 (UTC)

That addition was unnecessary and I am glad it was removed. This is a policy page, not a dissertation. Language has to be kept simple and to the point. ≈ jossi ≈ (talk) 16:56, 12 August 2007 (UTC)

I agree 100% with jossi on this. WAS 4.250 19:07, 12 August 2007 (UTC)

Meaning of Primary, Secondary & Tertiary Sources

"A newspaper article is a primary source if it reports events, but a secondary source if it analyses and comments on those events." Primary, Secondary & Tertiary Sources Anyone who wants to know the meaning of "Primary, Secondary & Tertiary Sources" should click that link and read. It is easiest to view sources as themselves primary or secondary but in fact at wikipedia it is important to realize that if we say "This source [so and so] said 'whatever'.[ref]" then we are using that source as a primary source for what it is saying. If a newspaper says the policeman said "Bob told me he killed Jay"; then we can use that report as a primary, secondary, or tertiary source depending on how we word it. If we say "Bob killed Jay"[ref] then we are using that newspaper report as a tertiary source. If we say "Newspaper [such and such] says the policeman said "Bob told me he killed Jay." then we are using that newspaper as a primary source. Which is better? Using it as a primary source. But some people here don't get that and insist that using primary sources is not as good as using secondary sources, when the key for wikipedia is not the source but how we use the source. WAS 4.250 14:09, 10 August 2007 (UTC)

I really disagree with your presentation of PSTS. The very quote you use seems to contradict the comments the follow. An article is a secondary source "if it analyses and comments on those events". In other words, it's based on the content, not the usage of the reference. The letters of Origen are all primary sources in religious studies, whereas a textbook covering his letters would be a secondary or tertiary source. In almost all instances, with some consideration for field, a work can be clearly called a primary or secondary/tertiary source. Vassyana 18:34, 10 August 2007 (UTC)

I think WAS 4.250 is basically correct in that if you quote the source, you are using it as a primary source for what it said, but if you state what it says as fact, then you are using it as a secondary source. If the fact is not really disputable, then using it as a secondary source is appropriate. But if the fact is in dispute, then the secondary source use may not be valid, and making an attributed quote is the better course. Dhaluza 02:02, 15 August 2007 (UTC)

Jelly bean, interpretation of SYNT, and request for opinions.

Hi. I've made some edits to the Jelly bean article, about the candy. In this article we have a table of colors flavors, which is meant for the basic store-brand jellybeans which usually have only 8 or 9 flavors. I wanted to clean up one column where we list "black" as "anise/licorice".

The background is that almost all candies popularly spoken of as "licorice" are actually flavored with either anise oil or a synthetic equivalent. At least in the U.S., only certain gourmet brands are actually made with licorice root or extracts thereof, and they generally advertise this on their packaging. So black jellybeans, which are popularly referred to as "licorice" flavored, are often actually "anise" flavored.

My change was instead of having the unwieldly "anise/licorice" in the table, was to put "licorice" with a footnote that says "Many "licorice" candies are actually flavored with anise oil", with a quote, citation, and a link to a journal article explaining the subject. I felt that better explains the issue to the reader, and clears up a popular misconception.

Anyway, another editor keeps reverting it claiming "OR". I tried to meet him halfway by discussing what he meant by OR ( and also hoping to convince him that the fact itself was true before trotting out the rulebook ), but that didn't work out, and he moved the discussion off his talk page mid-conversation.

This is only a guess, but the point of contention seems to center around an implied inference. For example, if I made an inference to join "Licorice candies are really anise" with "Jellybeans are a candy" to get "Licorice jellybeans are really anise", that would be an example of SYNT.

However, that's not what I did, and I'd like to counter with the following points:

• I only said "Many "licorice" candies are actually flavored with anise oil", backed up by a very reliable source.
• This footnote was in a section ( a table cell ) of the article about "licorice" flavoring. It's not a statement about jellybeans requiring a source on jellybeans, it's a statement about licorice flavoring.
• There is no "OR by juxtaposition" rule, despite what some editors seem to feel.
• "Jellybean" is a generic description and not a trademark. That means there are many different manufacturers and many different recipes, and it is OK to speak of them and how they are flavored, in generalities.
• Even if an inference were being made, it would not be a material inference because there is nothing about the size or shape of jellybeans that they would be flavored differently than any other candy.
• It's unreasonable to expect that there would be an article explaining how a particular _shape_ of candy is flavored. Squidfryerchef 23:27, 13 August 2007 (UTC)

Well, I feel pretty silly about this, but here's what I see in this Jelly Bean situation:

Policy: WP:NOR#Synthesis of published material serving to advance a position, which states:

Editors often make the mistake of thinking that if A is published by a reliable source, and B is published by a reliable source, then A and B can be joined together in an article to advance position C. However, this would be an example of a new synthesis of published material serving to advance a position, and as such it would constitute original research.[2] "A and B, therefore C" is acceptable only if a reliable source has published this argument in relation to the topic of the article."

Source A: Liquorice candies are flavored with Anise. (Source does not mention Jelly Beans)

Source B: Black Jelly Beans are liquorice flavored. (Source does not mention Anise)

Synthesis: Black Jelly Beans are flavored with Anise. (Original research/analysis/synthesis)

This is original research, because it expresses the editor's opinion that, given the fact that because "liquorice flavored candies" are flavored with anise, then 'liquorice flavored Jelly Beans" (which are candies) are flavored with anise.

The excuse that it's "just a link in the article, which doesn't draw a conclusion" is flawed and incorrect, because its mere presence in the article draws a conclusion, even if not explicitly stated.

Isn't there a single reliable source that actually says that some Jelly Beans are flavored with Anise? It seems simple - but I looked and it isn't at all simple to find a source that says so. I wonder why? Anyway, here are two moderately acceptable sources: [11] [12]. I think those might be usable, but not the one that's unrelated to the subject of the article...which is Jelly Beans, not "candies". Dreadstar † 23:37, 13 August 2007 (UTC)

Well that's actually my point. I don't think "source being related to the subject of the article" is to be taken so literally. I think the source only needs to be related to what is being footnoted. The other issue is that a jellybean is only a particular shape of hard candy, and wouldnt be flavored any differently than any other hard candy. If it were a certain _brand_ of jellybean than I would feel very differently and insist on a statement that that brand has a particular flavoring. ( CC from my talk page ) Squidfryerchef 23:42, 13 August 2007 (UTC)

You know, researching this really makes me wonder if Jelly Beans are flavored with anise at all...some of the sales sites say it, but why are there no reliable sources for such a thing? Mebbe it's not even true anymore..and they're all synthetically flavored. Maybe someone can find better sources for it. Dreadstar † 23:54, 13 August 2007 (UTC)

Well my take is that jellybeans are just a differently-shaped sweet, and would use the same flavorings as lollipops, etc, even sodas. If someone did the same thing about a brand-name confection, or even something like a liqueur that might not be a brand name, but had a generally accepted way of making it, then I'd be making the same edits and OR claims. Anyway it's more complicated than that, you'll probably find "licorice" candies made from anise, star anise, artificial flavorings, and true licorice. But the point is that in the U.S. "licorice" is so often used to refer to things flavored with anise that there is a lot of confusion, and the article ought to say something about that. P.S. that Canada.com article might not be usable for the "anise" issue but it does resolve a lot of the unsourced data in the article. Squidfryerchef 00:14, 14 August 2007 (UTC)

I fully understand your point, I just happen to disagree. It's entirely possible that all "candy-related" articles can have a source unrelated to that specific candy that just generalizes what all candies are flavored with. But I just don't think that's right. You cannot prove that Jelly Beans are flavored with anise, and these sales sites are not reliable sources. I dunno. It's just candy, but what applies to one Wikipeda article should apply to all. Dreadstar † 01:02, 14 August 2007 (UTC)

There actually is a page on the Wikipedia called Liquorice candy, which would be a fine place to put my citation to that chemistry journal, and I could change the list of flavors to link there instead of to the Licorice plant. But as far as policy goes I still say that there is no implied inference with the kind of explanatory note I was trying to add to Jelly bean. That doesn't mean all such footnotes are automatically pure gold; but having them does not create an inference by proximity. Squidfryerchef 03:17, 14 August 2007 (UTC)

Excellent! Liquorice candy looks like the perfect place for your citation! Dreadstar † 06:54, 14 August 2007 (UTC)

BBC Example

There is a current example of the problem which came up on the Administrator Incidents page discussed [here]. Someone spotted that a BBC employee apparently seemed to be editing the pages related to criticism. There appears to be a sensible response [here (there are a couple more corrections after)]. However, note that the argument is that they have checked the (primary) source which justifies the edit taking out the press bias of presumably deliberately misquoting the report. There are a few things here (does the remaining text in the article which notes that that the report was misrepresented in the press constitute OR - yes it appears to). It is also interesting as it brings in the question of reliable sources. We know that the Evening Standard does have vocal political campaigning, therefore for most purposes it is not a reliable source. On the other hand, we get into a problem as the article is discussing criticisms of the BBC and it is clear that these criticisms are being made. Wikipedia is being used to publish a political agenda, and potentially some interpretations of policy will encourage that, rather than resist it.

Discuss! 10:41, 15 August 2007 (UTC)

Straw poll: Should No original research take a stand against primary sources and tertiary sources?

Let's see where we are here with a straw poll. The issue, summarized as neutrally as possible, is as follows:

From 2003 to late 2006, Wikipedia:No original research indicated that editors could use primary sources, secondary sources, and tertiary sources, so long as there was no original research. Around November [late summer] 2006, a statement was added to the policy indicating in bold that Wikipedia articles should rely on reliable, published secondary sources, and that use of primary sources should be "rare". [There was some opposition to the change at the time, but for whatever reason, the change remained]. No explicit statement was added against tertiary sources, but their disfavoring is implied by the above bolded statement.

Please indicate your opinions regarding the following questions:

Should Wikipedia:No original research oppose the general use of primary sources?

• Disagree. Primary source citations are the bread and butter of Wikipedia, and a prominent aspect of almost all featured articles. Any time there is a non-hearsay quotation, the citation is a primary source. Almost any citation to journal articles is a primary source citation. Citations from works of fiction, TV shows, movies, song lyrics, poetry, are all primary sources. If we discouraged citation of primary sources, very few Wikipedia articles about current events could be written. Moreover, the reasons given against citation of primary sources almost always have nothing to do with original research, and everything to do with NPOV. The rules for citing primary sources without original research should be the same as for secondary and tertiary sources. COGDEN 01:14, 14 August 2007 (UTC)
• Wikipedia articles should rely on reliable, published secondary sources<adding>primary sources should not be used for OR, synthesis via juxtaposition, etc.</adding> TableManners 01:53, 14 August 2007 (UTC)
• Strongly Disagree. Forbidding primary sources would make fact-checking impossible. I have been working on improving the coverage of 4th-century Christian history. Forbidding the use of older ecclesiastical histories (which reflect different sides of the controversies) would lock in the POV of later works like the Catholic Encyclopedia (which represent Nicaean trinitarianism) or that of modern revisionist works. Jacob Haller 02:21, 14 August 2007 (UTC)
• Disagree. We just need to be more clear when and why primary sources should be used. The use of primary sources alone almost guarantee WP:OR (and, not coincidentally, WP:NPOV) problems. However, primary sources are often a necessity as COGDEN and Jacob Haller point out above. Primary sources do not suffice, and may not be necessary, but there is certainly no reason to exclude them. --Ronz 02:32, 14 August 2007 (UTC)
  • I'm confused. Where in the policy does it state that primary sources are "forbidden" or "excluded" ? as per my reading WP:OR only says that such sources should be used "only with care". Abecedare 02:41, 14 August 2007 (UTC)
• Strongest Possible Disagree You need some secondary sources to establish notability, but ideally you have a mix of primary and secondary sources. Squidfryerchef 03:05, 14 August 2007 (UTC)
• No, as long as the editors don't analyze the primary sources. Rocket Socket 03:24, 14 August 2007 (UTC)
• Disagree Primary sources are invaluable for verification purposes. Secondary and tertiary help establish notability. MrMurph101 17:19, 14 August 2007 (UTC)
• Strongly Disagree If I want to use an occasional direct quotation, or excerpt from an interview with a luminary to illustrate a point, why not? This is getting ridiculous.--Filll 22:27, 14 August 2007 (UTC)
• Strongly Disagree Secondary sources used for analysis, but primary sources are preferred for data, and tertiary sources are also useful. It is not the type of source that leads to problems, it's the usage. WP should not assume that editors or readers are too stupid to use references properly, and make stupid rules based on this assumption. Dhaluza 01:30, 15 August 2007

(UTC)

• Strongly Disagree For current events the policy of relying on the secondary source material of entertainment networks masquarading as scholarly news outlets is asinine. Censoring available primary source material is endorsing a NPOV, Non-Informed-Point-Of-View that is. Primary source information of wikipedians for current event articles, as long as they log in and sign the post, should definitely be allowed. For instance I attended the Iowa Straw Poll last weekend and I can't post 95% of the information I have because it wasn't given out by some talking head in the spin room. I would propose that within 1 year of a current event wikipedians be allowed to add a primary source section to the bottom of the article, provided they sign their post and it maintains neutrality.GrEp 04:15, 16 August 2007 (UTC)

Should Wikipedia:No original research oppose the general use of tertiary sources?

• Disagree. Citations from tertiary sources such as treatises and textbooks are usually even less likely to be original research than secondary sources, particularly when it is a widely-used textbook or treatise cited by everyone in the field. COGDEN 01:14, 14 August 2007 (UTC)
• Wikipedia articles should rely on reliable, published secondary sources <adding>Tertiary are not as good as secondary sources. Primary sources are best for research, secondary are best for an encyclopedia, tertiary sources for a junior high school paper. Exceptions exist. </adding>TableManners 01:53, 14 August 2007 (UTC)
• Somewhat Agree. Tertiary sources, particularly dictionaries and sometimes also encyclopediae, reduce complex ideas and histories to simple definitions and summaries. People often draw syllogisms from these definitions, which is IS, and they are often wrong. Favorites involve "the definition of socialism," "the definition of capitalism," and "the definition of libertarianism" ... Jacob Haller 02:21, 14 August 2007 (UTC)
  • A tertiary source can't be "wrong." There is no wrong or right on Wikipedia. There is only what is sourced and what is not sourced. If you think a source is wrong, you just have to keep that opinion to yourself. Wikipedia editors aren't supposed to make judgement calls like that. Rocket Socket 03:19, 14 August 2007 (UTC)
• Disagree. I think the exclusion of tertiary sources is just a misinterpretation of the emphasis on secondary sources. We need secondary sources (or better, meaning tertiary). It's not a coincidence that this overlaps with WP:NPOV in the need for "the best and most reputable authoritative sources available." --Ronz 02:25, 14 August 2007 (UTC)
• Somewhat Disagree. There are many places, especially at the beginning of articles where there's a few general statements about the subject, or the definition or etymology of the subject title, and these opening statements can be backed up by citing to tertiary sources. Squidfryerchef 03:05, 14 August 2007 (UTC)
• No. As long as the tertiary source is from a reputable publisher there's no problem. Rocket Socket 03:24, 14 August 2007 (UTC)
• Strongly Disagree Tertiary and quaternary sources etc. should be fine. How do you even draw the line between secondary and tertiary sources, in any case? This is all pretty vague, and more nonsense from bureaucrats and others who want a weapon to use against editors who are trying to write good articles.--Filll 22:27, 14 August 2007 (UTC)

Discussion

• Primary sources are fine as long as they're only used to provide direct quotes or paraphrase sentences but not much more than that. Therefore articles should rely most on secondary and tertiary sources. The problem is when Wikipedia editors do their own analysis of primary sources. That's original research and not allowed. Rocket Socket 01:31, 14 August 2007 (UTC)
• Primary sources such as technical research papers, religious scriptures, raw data etc require a high level of understanding of the field and context to be interpreted appropriately, and this is not something should be done by lay, anonymous wikipedia editors. So articles should rely on secondary (and perhaps tertiary) sources, with primary sources treated with care. Aside: I consider review articles in academic publications that provide an overview of a topic and current consensus (rather than report the author's own research findings) to be secondary sources - do others agree with this interpretation ? Abecedare 01:42, 14 August 2007 (UTC)
  • The definitions used here for these terms are at odds with the way they are defined in the scientific community. An published article offering analysis of the authors own research would be considered a secondary source along with "review" articles. A primary source would be an interview with a scientist simply describing his work.--BirgitteSB 12:52, 14 August 2007 (UTC)
• I think we need to discuss appropriate uses and potential misuses of each type of source, or, as I'd suggested above, appropriate sources for each type of claim. Jacob Haller 02:21, 14 August 2007 (UTC)

P.S. In my experience, secondary and tertiary sources often present worse POV problems than the primary sources. Jacob Haller 02:28, 14 August 2007 (UTC)

It is not a problem if a source presents a POV. The problem is when an editor states his POV without a source for it. The policy against POV is for editors not sources. All sources are going to have a POV. Rocket Socket 03:16, 14 August 2007 (UTC)

There is no requirement that sources be neutral, only reliable. It is up to Wikipedia editors to produce a neutral article by balancing the information from sources from each viewpoint. We must use POV sources to do so. POV sources are a requirement.--BirgitteSB 12:55, 14 August 2007 (UTC)

• Consensus is not a vote. I agree with the first two commenting editors above. Vassyana 02:36, 14 August 2007 (UTC)
  • Comments. Review articles, as mentioned by Abecedare, are secondary sources. I would support including more encouraging language towards tertiary sources, but would still encourage caution due to potential oversimplification by the source. We can always provide clarification regarding those sources. I'm unconcerned with current events being lacking in Wikipedia. Wikinews is over that way if someone is interested in covering current events. Wikiversity, Wikiquote and Wikibooks are in those directions, if someone wants to work extensively with primary sources.
  • On the subject of early Christian history, there are a plethora of secondary & tertiary sources, and it's not our place to analyze and determine if modern sources are "revisionist". We should represent the views of modern scholarship and publications according to their prominence, not our own opinions. If someone wants to dispute the findings of modern scholarship, they are welcome to find a reputable publisher. Vassyana 02:36, 14 August 2007 (UTC)

• Comment about poll. This poll is inaccurate, to say the least, in its summary of the policy shift. There was an extensive discussion about the changes, which were implemented at the end of summer last year, not in mid-to-late fall as presented. (You can review the discussion here.) The problems with abuse of primary sources were recurring issues, and sporadically discussed on the policy talkpage over time. Primary sources are very difficult to use without engaging in original research, implicitly or explicitly, except when presenting them as used by reliable third party publishers. Vassyana 02:36, 14 August 2007 (UTC)

  I have to agree with Vassyana. Also, the phrasing of the questions are biased. BTW, I do not disagree with the use of 1 and 3 sources, but 2 sources are emphasized and more important. TableManners 03:07, 14 August 2007 (UTC)

  I have to agree. The questions are bit contrived and misleading. The policy does not forbid or exclude primary and tertiary sources. It doesn't even provide any injunction against using tertiary sources. On primary sources, it just says they should be used rarely and with great care to avoid original research. And of course, if people come across one of those rare situations (or just an exception to the rule), they're free to apply a little common sense. For example, it's common to plainly report census numbers for statistics on U.S. cities. It's just a plain reporting of facts and is purely "raw data", not presenting one interpretation, POV or another. Common sense tells us that's a fairly good primary source because it's use is very neutral and published by a source with a reputation for accuracy and fact-checking. (At least as far as the census number go, *chuckle*). Some primary sources are far more subjective, like personal memoirs and ancient historical accounts, and therefore need to be vetted and interpreted by modern third party reliable references. Vassyana 05:55, 14 August 2007 (UTC)

  If I may split hairs here, US Census Reports are not primary sources: they are syntheses of the Census returns -- those forms US citizens fill out every 10 years -- which makes it a secondary source. -- llywrch 18:22, 14 August 2007 (UTC)

Actually, they are primary sources. It's a reporting of raw data, collected and collated but still a primary source. Reporting the raw results of lab data, with little more than collection and collation is similarly a primary source. It may be original research, but it's not a secondary source. A report analyzing that data and drawing conclusions could very well be a secondary source however. Vassyana 22:44, 14 August 2007 (UTC)

• Comment, I agree with Vassyana and TableManners, the poll is biased and does not accurately describe the items under dispute. If a Wikipedia:straw poll is to be taken, then it needs to be worded in a way that accurately and neutrally describes the dispute. The current wording of the poll fails to do this. From what I understand, the basic dispute is whether or not to change or remove the wording of certain elements in the current policy, specifically:
  • "Wikipedia articles should rely on reliable, published secondary sources."
  • "Although most articles should rely predominantly on secondary sources, there are rare occasions when they may rely on primary sources."

It is my view that these two elements need to remain in the policy, and no changes be made to the current policy regarding the use of primary/seconday/tertiary sources. I vote to keep the current wording. Dreadstar † 05:35, 14 August 2007 (UTC)

• Comment,I too feel the poll isn't well stated. "Should Wikipedia:No original research oppose the general use of primary sources?" There's no sense in which primary sources are opposed. It's simply that secondary sources are preferred, because experts have have corroborated and interpreted primary sources. I've seen serious instances of abuse of primary sources. One problem is that the nature of primary sources and their utility varies widely from field to field. I believe the general principle should be to prefer secondary sources, and thereby avoid original research by nonexperts, while relying on common sense in those instances when use of a primary source is appropriate. TimidGuy 10:59, 14 August 2007 (UTC)

The key word here is general. The current version of the policy "allows" them in rare cases, but prohibits their general use. Keep in mind that this is a Wikipedia policy, not just a guideline. COGDEN 17:18, 14 August 2007 (UTC)

The comments made by other clearly indicate the poll is giving exactly the impression TimidGuy stated. Vassyana 22:44, 14 August 2007 (UTC)

• I also find this badly worded. I have not been able to keep up with all the discussion lately, but I do not think the statement being polled describes any version of this policy I have ever read. I find a poll on such a statement irrelevant. It is also inaccurate to pretend these terms are so solid. A "secondary source" can almost always be used a primary source some way or another. Also a single "source", however you define it, may have different uses as all three sort of sources. A modern introduction by a scholar, that could be used as a tertiary source. The main text as a secondary source for it's analysis, or primary source for the claims of the author. Appendices may also have copies of letters or other material that could be used as a primary source. These terms are all very ambiguous and a single book, or journal, or sometimes even a single article cannot be purely labeled with one of these terms. It all depends on how you use the source. It depends on what you are referencing, whether you are using a primary, secondary, or tertiary source. To all who are concerned about wikilawyers, if someone tells your source is no good because it is "primary". Then ask them "How exactly am I using original research in the article?" If they cannot tell you, then explain that the point of WP:NOR is to prevent original research not prevent primary sources.--BirgitteSB 12:48, 14 August 2007 (UTC)

• The problem with primary sources is the mile long plot summaries that users provide about movies/tv episodes, and especially soap opera characters' pages. Corpx 17:29, 14 August 2007 (UTC)

• That is not a problem with primary sources in general. That is a style issue with articles about subjects with plots. How long of plot summaries or whether they should exist at all in articles like Romeo and Juliet, The Apartment, or Frank Cooper Jr. is not relevant to a discussion on the usage of primary sources throughout all of Wikipedia. Please avoid strawman arguments.--BirgitteSB 17:49, 14 August 2007 (UTC)

• I'd tend to agree with you. The reason primary sources were discouraged is because of their widespread abuse. It is very difficult to use primary sources without engaging in original research. You're quite right that asking people exactly how they are being used incorrectly is the correct response. I do however somewhat disagree that the definitions are so flexible. Perhaps they are in a strict semantic sense, but really that's a semantics game. Most fields have fairly clear delineations between primary and secondary/tertiary sources. It's usually the distinction between secondary and tertiary that is unclear. Vassyana 22:44, 14 August 2007 (UTC)

I agree that many fields have solid definitions for these these terms. Unfortunately these definitions are contradictory with one another and Wikipedia encompasses all these fields.--BirgitteSB 13:54, 15 August 2007 (UTC)

• I believe Birgitte has the right idea: the problem is not with the material, but with how it is used. The only cases where primary sources should be excluded is if they are of a type that is not verifiable -- for example, unpublished family letters or personal interviews. If I use a source which no one else has reasonable access to -- for example records from a government or corporate archive, which often can only be used after a formal grant of access -- then no one else can verify that I quoted it correctly, let alone provided the material in harmony with NPOV. -- llywrch 18:22, 14 August 2007 (UTC)
• These questions are always subjective and the application of rules has to be flexible enough to not squelch good writing and discourage contributions of editors to WP. The idea that any one kind of source should be used to the exclusion of others is ludicrous. If I have an article about nursery rhymes, I cannot quote the public domain words of the nursery rhymes obtained from some public domain primary source? The music? I cannot quote from another encyclopedia that is summarizing the work of 5 scholars on nursery rhymes? Please people try to use your heads! This ridiculous nitpicking and rule-making will end up doing more harm than good to Wikipedia.--Filll 22:27, 14 August 2007 (UTC)

That is not at all what the policy says. This poll is extremely misleading and biased. Vassyana 22:44, 14 August 2007 (UTC) .

I agree, that is why I said the poll questions are irrelevant. The policy doesn't oppose any of these types of sources.--BirgitteSB 13:54, 15 August 2007 (UTC)

• Comment. We have to be careful that rules designed to limit cranks (Look, the Bible says Jesus rode a motorcycle right here!), should not infringe on the contributions of rational people. The fact is, there are many topics that by their nature require heavy use of primary sources—either because the text is the topic or just because the topic is obscure. Of course primary sources should be used with care, so that we don't originate new theories, but they do have an important place here.--Pharos 22:38, 14 August 2007 (UTC)
  • Just nothing that this position is well-within the current reading of the policy. It even acknowledges that there are cases where sections, and even articles, may need to be based on primary sources. It just encourages us to keep it to a minimum and do so with care. Vassyana 22:49, 14 August 2007 (UTC)
  • If the subject is so obscure that an article has to rely on primary sources then why is the article important enough to be in Wikipedia? If a subject is notable then it would naturally have enough secondary and tertiary sources that you wouldn't have to rely on primary sources. Rocket Socket 22:52, 14 August 2007 (UTC)

That's a very OR statement. See how useful OR actually is, see how it is nearly unavoidable? --Minasbeede 23:01, 14 August 2007 (UTC)

• Comment If we eliminate primary sources, we would not be able to quantify things--we would not be able to say how high a building is, or how long a bridge is, or how many people live in a city, unless someone else repeated it. And even if they did, we should always use the original primary source for data, because it is more reliable. Secondary sources are necessary for analysis, but raw data can also be presented, and left for the reader to analyze (often this is the best way to deal with competing POV issues). The problem is when a WP editor puts his/her analysis in an article--arbitrarily excluding sources won't help that. Stupid rules don't stop stupid people.

Also, a proposal along these lines would be unworkable. There is no clear distinction between what is a primary, secondary, and tertiary source. Scientists and historians use the terms differently, and there are other different perspectives as well. Even if an unambiguous definition were possible, references would not always fit neatly into these cubbyholes. For example, a scientific paper could include new experimental data, data analysis, and previous work in the field. So this one reference would be a primary, secondary and tertiary source all rolled into one. An encyclopedia may generally be regarded as a tertiary source, but it may have signed articles that are really secondary sources, and undigested raw data, which is still primary source.

There was an effort at WP:N recently to require only secondary sources for Notability that lead to an edit war with a lengthy edit protection, and a subsequent straw poll that showed no consensus for this version. The compromise wording that resulted strongly favors secondary sources, but without making them absolutely necessary. There was also an extended discussion of using primary source census data for creating articles on geographic locations a while back. The resulting consensus resulted in producing tens of thousands of articles on U.S. towns and cities based solely on the U.S. Census Bureau primary source data. This created a framework for expansion that was far better than having editors working at random. Dhaluza 01:30, 15 August 2007 (UTC)

This view and the consensus described at WP:N seems consistent with the current reading of this policy, which "strongly favors secondary sources". The current version also permits the use of primary sources on those occasions when they are necessary, provided they are used with care (particularly in avoiding original research). Vassyana 01:49, 15 August 2007 (UTC)

Discussion about sources

This is not a vote, but rather an invitation to discussion. I have endeavored to word the questions neutrally and generically, to invite broad comments about this aspect of policy. I believe these question cut to the core of what is being discussed and debated above, without leading a reader to draw conclusions about how the policy currently reads. Vassyana 23:05, 14 August 2007 (UTC)

• What distinctions, beyond the requirements of reliable sourcing, should be made between types of sources (primary vs. secondary, modern vs. historical, et al), if any?

Since we all can't agree on this issue, probably the best way forward is just to make no distinction between different types of sources. Ultimately, they all follow the same standard anyway, which is (1) that the accuracy of the citation can be verified by a person without specialized skill, and (2) that there are no novel synthetic or interpretive claims, either express or implied. This applies for primary, secondary, and tertiary sources.COGDEN 21:11, 15 August 2007 (UTC)

• If there is a consensus about distinctions between sources, which sorts should be encouraged or discouraged, if any?

All source citations should be encouraged. A good article uses both primary and secondary sources (and tertiary sources, if we are lucky enough to have good ones). The best policy is often to use primary sources to provide the unvarnished facts, i.e.:

• "The Second Amendment refers to a "right to bear arms" {see Second Amendment},

then back it up with secondary sources that add the spin, i.e.:

• "Some commentators believe this right provides an unrestricted personal right to own weapons {See We-heart-NRA (2005), manifesto.}"

and tertiary sources, i.e.:

• "According to the widely-cited Westmaster Law Treatise, 'the right to bear arms has traditionally extended only to the right of states to organized well-regulated militias' {see Westmaster 6:42}"

COGDEN 21:11, 15 August 2007 (UTC)

• What kinds of sources are more prone to original research abuse, if any?

Primary sources, but secondary and tertiary sources are more prone to NPOV abuse. The safest method, and least likely to lead to original research, is to use a combination of primary and secondary sources: primary souces for the what, and secondary sources for the how and why. COGDEN 21:11, 15 August 2007 (UTC)

• If there is consensus that certain types of materials are more prone to abuse, what kind of cautions should be given (or restrictions suggested), if any?

The standards of original research should be the same for any type of source. But we should caution readers that to "go beyond" what is said in a primary source, they need a secondary or tertiary source, and to "go beyond" a secondary source they need either another secondary source or a tertiary source. Allowance should be made, of course, for stating the obvious and the non-controversial in relation to the source material.COGDEN 21:11, 15 August 2007 (UTC)

Prioritizing secondary sources

Cogden, I reverted your removal of the sentence prioritizing secondary sources. This is a practice that has strong support. Can you say what your objection to it is? SlimVirgin ^{(talk)(contribs)} 21:18, 15 August 2007 (UTC)

Rather than removing it again, could you say here what you feel the problem is? SlimVirgin ^{(talk)(contribs)} 21:55, 15 August 2007 (UTC)