Wikipedia talk:No original research/Archive 16

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 10 Archive 14 Archive 15 Archive 16 Archive 17 Archive 18 Archive 20


Other avenues

Does anyone know / can recommend similar sites like Wikipedia where OR is allowed (and yes I know all about uncyclopedia thanks). Magic Pickle 20:33, 4 January 2007 (UTC)

Yes. Wikiversity: v: —The preceding unsigned comment was added by Codex Sinaiticus (talkcontribs) 22:32, 23 January 2007 (UTC).


Isn't this policy redundant? What's covered here that's not covered by WP:V? Matt Yeager (Talk?) 05:44, 13 December 2006 (UTC)

Material can be sourced and yet still be OR, so this page explains what OR is. SlimVirgin (talk) 05:45, 13 December 2006 (UTC)

I have a question, If I found something in a computers registery that is a default value that I didn't change and could possible be put in a trivia section of an article. Does it count as original research until I can find a publication that mentions it? SleepyDan 22:39, 10 January 2007 (UTC)

As long as you only make descriptive claims, and don't use them to form unpublished arguments, including verifiable descriptions like that is not OR. But you have to ask whether trivia like that is notable enough to be included in the article. For example, I could list the first 20 bytes of a Mozilla Firefox binary, but there is no reason to put it in the article. CMummert · talk 19:21, 26 January 2007 (UTC)
Thanks, I don't know if it is notable or not I have to check other computers but I found that on at least my pc one of the DWord Names for crystal reports 11 is THX1138 like the movie.SleepyDan 18:51, 6 February 2007 (UTC)

Hypothetical Situation

I am the author of an article on Lipotes Vexillifer which has been published in a well-known scientific journal. Does the OC rule prevent me using material from that article in the Lipotes Vexillifer Wikipedia article even though no-one else would have such a restriction in using said material? Knight of Ashitaka 22:04, 24 December 2006 (UTC)

A Criticism of WP:NOR

I am not sure I agree with all of WP:NOR's tenets. In particular, I feel that primary sources, comprehensively and evenly summarized, ought to be able to stand for themselves. A review of a secondary source is a review of a review, so to speak, and even "competent scholars" are sometimes biased. — Rickyrab | Talk 06:47, 13 December 2006 (UTC)

I'm a bit puzzled too. For example I added, to a very short article about the Indian town of Palampur, the fact that paraglider pilots often land there. I know this because I have seen it happen -- in fact, I've done it myself. Adding that fact makes the article about Palampur a tiny bit more useful. Should I not have done it?

We are reporters of what other people say. Our own opinions, experiences, personal knowledge, mean nothing. That's what NOR is about. Wjhonson 16:46, 13 December 2006 (UTC)
Two reasons not to include personal experience in Wikipedia:
  • it's impossible to verify so Wikipedia would quickly fill up with urban myths if allowed
  • if it's notable someone will have written about it somewhere.
If you want to add that Palampur is used for landing paragliders, you can; you just have to include a source such as this one [1] Curtains99 17:23, 13 December 2006 (UTC)
If everything was sourced in articles there would be hundreds of footnotes for each entry. Magic Pickle 20:36, 4 January 2007 (UTC)
No, today's featured article has somewhere near 25 footnotes. The criteria is that "material that is challenged or is likely to be challenged" needs citation! So it is highly subjective. Many articles can indeed be very useful without citations, the important thing is that the burden of proof is on you to find a reliable source, and that unattributed material can be removed if somebody doubts its validity. As to the original posters idea that primary sources speak for themselves: they don't! :) --Merzul 21:56, 4 January 2007 (UTC)
Which would be awesome! Having many sources is a good thing:
  • As stated by WP:CW, Wikipedia is often not considered an acceptable source. By providing other (possibly acceptable) sources, we become more useful to those doing research.
  • Footnotes allow people to study a topic further. Because the footnotes are chosen by humans, they may make a Wikipedia article on a subject more useful than a Google search on that subject.
  • Having redundant sources mitigates the danger of linkrot.
--Clipdude 23:24, 7 January 2007 (UTC)

Ah yes, but do print encyclopaedias (in general) have hundreds of footnotes for each entry? No. Magic Pickle 20:40, 28 January 2007 (UTC)

Having footnotes for obvious facts causes clutter and makes it harder to find and use the worthwhile sources. But Wikipedia will usually need more sources for the same article than a print encyclopedia because we can't assume that the editor is unbiased or competent, so the extra sources compensate for the unknown qualifications of the editor. (Yes, I realize print encyclopedia authors may only be competent at general research, and not subject-matter experts.) --Gerry Ashton 22:55, 28 January 2007 (UTC)
If we could assume competency and integrity at general research, most articles could probably do without footnotes. Robert A.West (Talk) 23:49, 28 January 2007 (UTC)

Policy on Original Research

K, so i get what is original research but what does one do if one comes across it? like in Examination of Holocaust denial, which is up for deletion, most of the people agree that it is Original research but the high quality of that research could further benefit wikipedia. What should oyu do in this situation?/??Xlegiofalco 00:34, 16 December 2006 (UTC)

WP is an encyclopedia, which means it is a collection of summaries from other reputable and verifiable published sources. If WP had its own research, or improperly sourced information, it would cease being an encyclopedia, contrary to its charter. In the article you mention I see large stretches that have no inline citations. Whenever there is a contentious issue, sources must be cited inline to prove that each item is properly sourced to a casual reader. Crum375 02:02, 16 December 2006 (UTC).

What about the issue of translations? To extensively quote a published translation can be a violation of copyright, while a contributor may very well have the expertise to translate an important foreign work. Certainly in many cases it would not be difficult for readers to verify the reliability of a translation by comparing it to published translations. A great deal of Eastern philosophy is only accessible via translation and it would be very useful to quote extensive passages in translation. Not all translations require "original research," only technical expertise. I think this distinction of original research is not as clear cut as many believe as any good article requires a degree of synthesis and original thinking. Rthrelfall 05:29, 16 January 2007 (UTC)

Removed Paragraph

At one point the article contained the paragraph "Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia."

I gather that the first sentence was removed deliberately, but what about the rest? Was it removed by mistake? I can't find any discussion of it in the archives. User:Ben Standeven as 01:50, 17 December 2006 (UTC)

The only reason I can see not to use that paragraph is that the term "source-based research" may be misinterpreted to allow some kind of synthesis, or using sources in a selective fashion to advance a position. I guess it's better to stay away altogether from the word "research", as it has negative connotations in an encylopedia, which should be purely a tertiary source. It may be better to say that we "collect and summarize relevant and acceptable published sources", without calling it "research". Comments? Crum375 02:08, 17 December 2006 (UTC)
Can't just summarize because an article with just summaries would be unreadable. There has to also be some kind of synthesis of the summaries so they flow from one to the next, in one article, with joining phrases of some kind. Wjhonson 02:19, 17 December 2006 (UTC)
True, some minimal synthesis for "smooth flow" is always needed, but it must be a neutral presentation of the sources found. Crum375 02:49, 17 December 2006 (UTC)
Agreed. This is an important point, so let's make in the article? --Michael C. Price talk 07:39, 17 December 2006 (UTC)
Thinking about it some more, I think a "summary of existing sources" by itself implies some logical arrangement and smooth flow (and certainly does not exclude it), so the additional 'synthesis', which has a connotation of WP:OR, is not really needed. It may seem like splitting hairs, but I think we need to stay away from words that could imply or connote OR, like 'research' and/or 'synthesis', because once we open the door it could invite abuse. Crum375 15:31, 17 December 2006 (UTC)
I understand your caution but it is specifically "original research" that is banned, not "research" in general. Summarising existing sources is a form of research; trying to call it something else just makes it harder to see what is allowed and what isn't.--Michael C. Price talk 18:02, 17 December 2006 (UTC)
I guess it boils down to definitions and words. I can see it both ways. The problem that I highlighted is that perhaps you and I know exactly what kind of research and synthesis is acceptable, as just 'smoothing the flow of words'. But it's quite possible that if we actually use these words in the policy as permissible with some cautionary qualification, they will be abused or misinterpreted by some party during a future content or wording dispute. I think our goal in drafting a policy is to make it as clear-cut and as unabmiguous as possible, to minimize the potential haggling over meanings. It seems to me that if we just totally avoid the words 'synthesis' and 'research' as permissible, under any circumstances, and use 'collect and summarize acceptable sources', by proper anti-Murphy's Law engineering we eliminate this potential pitfall. Crum375 20:06, 17 December 2006 (UTC)
I came here to find out what OR means and I'm still not clear (witness my attempt at clarification that was immediately reverted). The term "original research" strongly implies that there are forms of research that are permitted. We need to be clear that this sort research is permitted and that sort is not; I don't think this policy document is clear-cut and it needs to be. Avoiding usage of the term "research" or "synthesis" is making the whole matter worse. It needs to be faced head-on. --Michael C. Price talk 22:56, 17 December 2006 (UTC)

(outdent) My understanding is that 'Original Research' is to be avoided, while summarizing and/or describing existing research, which was verifiably published in a reputable publication, is allowed. IOW, the distinction is that while WP editors can certainly present research, they cannot create it from scratch - they are only allowed to summarize and describe research that has already been published elsewhere. Crum375 23:01, 17 December 2006 (UTC)

Altered paragraph

Michael, please don't make edits that change the content policies. SlimVirgin (talk) 07:55, 17 December 2006 (UTC)

The above comment seems to be in relation to this clarification I made that SlimVirgin reverted. I altered "This entire paragraph" to "The first sentence of this paragraph", since the rest of the paragraph did not seem to be OR. It was a statement about the Chicago Manual of Style that is (I presume) verifiably true or false and can't possibly be OR:

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.

Surely it is less confusing to clarify exactly which part of the paragraph is OR? It was not a change of policy. --Michael C. Price talk 08:08, 17 December 2006 (UTC)

Michael, the entire paragraph is OR. That is why it's listed here, as an example. No source discussed the Chicago Manual of Style in relation to the dispute the article was about; therefore, to include it is OR. Please read the example. SlimVirgin (talk) 00:24, 18 December 2006 (UTC)

Restoring "source-based research" and "apple pie" paragraphs

This material has been around for a very long time, and presumably reflects a wide consensus:

Original research that creates primary sources is not allowed. However, research that consists of collecting and organizing information from existing primary and/or secondary sources is, of course, strongly encouraged. All articles on Wikipedia should be based on information collected from published primary and secondary sources. This is not "original research"; it is "source-based research", and it is fundamental to writing an encyclopedia.

In some cases, where an article (1) makes descriptive claims the accuracy of which is easily verifiable by any reasonable adult without specialist knowledge, and (2) makes no analytic, synthetic, interpretive, or evaluative claims, a Wikipedia article may be based entirely on primary sources (examples would include apple pie or current events), but these are exceptions.

Somehow they got lost this autumn. I feel they provide an important safety valve for WP:NOR, because without them there is no positive statement about what is allowed, only what is excluded. This can lead to editors invoking NOR for almost any summary they don't agree with.

For instance, if somebody is summarizing, I don't know, a cartoon or sitcom episode, and it has a very obvious Shakespeare reference, say Juliet on the balcony, then an editor ought to be able to make that connection and have it be left alone. But without those paragraphs, another editor could argue that the mention of Shakespeare was a "thesis" and therefore OR.

So, I'm reinserting those long-standing policies. Cool? Squidfryerchef 01:28, 19 December 2006 (UTC)

I for one agree. I'm all for the NOR policy but insisting on citations of scholarship when describing aspects of popular culture is problematic.--Father Goose 03:57, 19 December 2006 (UTC)
I agree with the restoration; we need to be positive about what is permissible and encouraged, as well as what is excluded, for the reasons you mention. I'm not sure what the first sentence means, though: how could you create a primary source in Wikipedia? I suggest we lose this (or clarify) along with the leading "However" from the next sentence. --Michael C. Price talk 12:06, 19 December 2006 (UTC)
I'd say that "creating" a primary source would be if you published your own study on Wikipedia. And we all agree that that is OR. I think that was used as an example to contrast with "source-based research". Squidfryerchef 14:33, 19 December 2006 (UTC)

"Synthesis of published material" questions

There's been a lot of discussion about the "synthesis" example and whether parts of it aren't really OR. I think that it's not so cut-and-dried, and if written a slightly different way, it would not be OR. How would you all feel about this?:

Mr. X, who wrote an article in a particular scholarly journal, was accused of plagiarism when it was discovered that he had copied a list of references from another journal article. However, accepted definitions of plagiarism vary; while the Detroit Style Guide ( I'm making these up ) says that any reference cited must actually be used in the text, the Cleveland Style Guide says that this is an acceptable way to create a working bibliography

It's my opinion that this is just good scholarship, not OR. It should be OK to bring in supplemental information like this, as long as it is only used to compare and contrast. For example, "plagiarism" is a very strong word, and it benefits the reader to have a clear definition of what Mr. X is being accused of. Squidfryerchef 01:38, 20 December 2006 (UTC)

I believe that this analysis and editorializing by WP is not encyclopedic - it sounds more like a journalistic column, and in any case is WP:OR. It is very similar to research that a district attorney would do to determine if a given individual is guilty of an infraction or a crime. The mandate of an encylopedia, WP included, is to present a summary of significant published research on a given subject, in a neutral presentation. In this case, if X is the subject who is being accused of plagiarism, then we can present published accounts of X's activities or that relate to him or his actions. If we cite a 'law book' Y as a generic source that does not mention X, and cite it alongside X, we are now performing original research, since we as WP are deciding that Y is applicable, and by merely presenting it alongside X we can depart from neutrality. If we can find a reliable published source that mentions X and Y together, we can of course cite that, but if we do the coupling of X and Y ourselves, we would be guilty of WP:OR. I think that's the crux of the WP:OR example, and what WP must not do. Crum375 02:05, 20 December 2006 (UTC)
Journalistic yes, but editorializing, no. The longstanding policy on what is excluded, is "It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or synthesis to a reputable source;" Even the title of the "synthesis" example says "Synthesis of published material serving to advance a position" This means that an impartial synthesis, perhaps explaining opposing viewpoints, is allowed. To clarify my example, I envisioned my two competing style guides as both being popular, authoritative documents with some differing points. It would be like contrasting an etymology from the OED and from the M-W dictionaries and having somebody tell you that was OR because the OED didn't mention X. Squidfryerchef 04:40, 20 December 2006 (UTC)
If source X said "Ivan committed plagiarism" but does not define plagiarism, I think it would be fine to mention a definition from source Y. But if instead source X said "the bibliography from Ivan's article is identical to an article by Henry" and then to list one or more definitions of plagiarism from other sources would encourage the reader to draw the conclusion that Ivan is suspected of plagiarism. --Gerry Ashton 05:15, 20 December 2006 (UTC)
Squid, your example is OR, because you, Squid, have decided to point out to the reader what plagiarism is, and you, Squid, have chosen which definitions to highlight. They may or may not be the definitions used by the accuser, and therefore may or may not be relevant to the story. If you know the definition used by the accuser and have a source, say so; if you don't know, don't guess or impose your own. Doing the latter is the essence of the "no new synthesis" provision of the NOR policy. SlimVirgin (talk) 09:51, 20 December 2006 (UTC)
But Squid also made the point that serving to advance a position implies that if multiple viewpoints are presented then this is not in violation of OR. (And if the viewpoints presented are not exhaustive it is easy enough for others to include more.) This seems to be the essence of encyclopedic: to be informative and balanced by presenting all sides. --Michael C. Price talk 09:59, 20 December 2006 (UTC)

The form of Squid's text is "SoAndSo was accused of plagiarism but his guilt is unclear since there are multiple definitions." That is clearly an opinion so it has to be sourced to a reliable source. But the only source given for this opinion is Squid, so it's OR without a doubt. The OR here is not the bringing of multiple definitions of plagiarism, but the application of them by Squid to judge the guilt of SoAndSo. --Zerotalk 14:42, 20 December 2006 (UTC)

Suppose it was rewritten:
Mr. X, who wrote an article in a particular scholarly journal, was accused of plagiarism when it was discovered that he had copied a list of references from another journal article. Of plagiarism the Detroit Style Guide says that any reference cited must actually be used in the text, whilst the Cleveland Style Guide says that this is an acceptable way to create a working bibliography
--Michael C. Price talk 14:49, 20 December 2006 (UTC)
Why would that make it any better? The question is: why does the editor want to insert a definition of plagiarism? If the def is important, the source will discuss it. If it isn't, why add it? People can look the word up for themselves if they don't know what it means. SlimVirgin (talk) 04:35, 21 December 2006 (UTC)
No, if the definition is important the source may not have discussed it, since it may be common knowledge with the original audience, not so for a general Wikipedian. A scientific journal would not discuss why the law of universal gravitation refutes Mr X, the flat-Earther, since they would correctly assume that is understood by their readers; Wikipedia has a more general audience. --Michael C. Price talk 10:22, 21 December 2006 (UTC)
But again, that's your opinion. It's your view that the definition is important; your view that it should be included; your decision which definition(s) to highlight. What we're meant to do instead is publish what reliable sources have published about the matter in hand — the allegation of plagiarism against X — and if they haven't discussed definitions of plagiarism within the context of that allegation, nor should we. SlimVirgin (talk) 16:44, 21 December 2006 (UTC)

The synthesis example seems to be taken from Talk:Alan_Dershowitz#The_debate_with_Norman_Finkelstein. But that version differs in one important way. In it, the reference to the Chicago Manual of Style is attributed to "James O. Freedman, the former president of Dartmouth College, the University of Iowa, and the American Academy of Arts and Sciences". In other words, the example isn't original research at all, because it is attributed to a source (James O. Freedman), and the source, not the Wikipedia editor, has drawn the conclusion. Ken Arromdee 15:43, 20 December 2006 (UTC)

Irrespective of what case the example may be based upon, the question is still open of whether the example is OR. However let's try another example:
Mr X claims to proven the world is flat, since otherwise people at the antipodes would fall off. However this claim misunderstands Newton's law of universal gravitation blah blah.
Is that OR because Newton did not address Mr X by name? (Assume that no one else has bothered to refute Mr X's silly claim.) --Michael C. Price talk 16:10, 20 December 2006 (UTC)
Yes, it is. SlimVirgin (talk) 16:45, 21 December 2006 (UTC)
Yes this is OR. To make it less so I would recommend the following

Mr X states "I have proven the world is flat, since otherwide people at the antipodes would fall off." (Mr X Speaks, Dutton & Co, 2002) Newton addressed this issue in his work De Principia Antipodiasticus when he said "... people at the antipodes will not fall off..." Wjhonson 17:18, 20 December 2006 (UTC)

Oh, I agree with Michael C. Price. A simple logical deduction should not be considered original research. That's one of the problems with the original research policy; it puts editors in a situation where saying "city A is west of city B" is original research when the source says "city B is east of city A". Deciding whether something meets a dictionary definition is straightforward enough that calling it original research stretches the policy beyond all usefulness.

To state that "according to the Chicago Manual of Style, Jones isn't a plagiarist" is wrong not because applying the definition to Jones is OR, but rather because the Chicago Manual of Style isn't an authority on Jones. It wouldn't matter if the CMoS gave a definition or directly said "Jones isn't a plagiarist". It's unacceptable either way, even though the second version clearly isn't original research. It's a reliable sources problem. (Of course, the version sourced to the former president of Dartmouth isn't either one.)

I wouldn't count the Newton example because Mr. X doesn't claim his theory follows Newton's Laws to begin with, so you'd have to say "Mr. X's theory would not follow Newton's Laws, because...." You can't decide for the reader that Newton's Laws must be accepted.

If you did word it that way, whether it's OR would depend on whether the conclusion that the theory violates Newton's Laws is a simple deduction. In the "Jones" example, people would argue that the Chicago Manual of Style's definition isn't relevant--not about whether Jones meets the definition. If whether Jones meets the definition can't reasonably be disputed, then it's a simple deduction like deducing that a city is east of a city in the west, and not OR. Likewise, if Mr. X's theory is so inconsistent with Newton's Laws that anyone could see it after a five second glance at Newton's Laws without knowing anything of them, it's not OR to call it inconsistent. If you could only determine that Mr. X's theory violates Newton's Laws by following an entire argument (even an argument in another work from Newton), then it would be OR and you'd have to use the "Newton addressed the issue in..." formulation.

Unfortunately, a lot of people disagree with me. However, you might want to check out the proposed Wikipedia:Attribution, which allows simple deductions. (It also uses the Dershowitz/Jones example in a FAQ, though.) Ken Arromdee 17:28, 20 December 2006 (UTC)

Some advice please

Please could someone (or somemany) take a look at the discussion on Talk:Command & Conquer 3: Tiberium Wars and the edits on the corresponding article. The article has been tagged as containing original research for as long as I've known so I went on a cull, removing anything that was obviously original research. Another user has now decided that my edits are too extreme and has reverted parts of it - and just added more original research tags. Can anyone provide some insight into what should be done? I am pretty sure that we aren't supposed to have sentences saying things like It is as of yet unknown if the race has also been officially designated as "the Scrin", or indeed if they effectively even are the Scrin. Thanks in advance! -Localzuk(talk) 17:52, 20 December 2006 (UTC)

Possible guidelines about when synthesis is and isn't OR?

It's been a very interesting discussion. Im wondering how these guidelines would be received:

- Just because the facts "A is a B" and "B is a C" are each asserted in reliable sources, does not by itself allow us to claim that "A is a C" ( this is a paraphrase of the "synthesis" paragraph ).

- Logical inferences that are made to advance a thesis are not allowed. For example, a literature term paper is OR just as a science experiment or a journalistic interview is OR. It is acceptable, however, to compare and contrast several points of view provided the other guidelines are met.

- Sources used in logical inferences are held to a higher standard than simply being reliable. They must also be an authority on the subject. This doesn't mean they have to be an authority on the main subject of the article, but to use a source for "A is a B", it must be an authoritative source on A or B.

- If there are several conflicting authoritative sources on the subject, the editor should not cherry-pick one that best supports an inference.

- There is an exemption for "general knowledge" inferences. Otherwise, ( here's a tip of the pen ) saying "city A is west of city B" would be original research when the source says "city B is east of city A". The only facts required to make that inference would be a lexicon that west is the opposite of east.

- I was told once that general knowledge, in a research paper, encompasses anything you might find in an encyclopedia. Obviously this definition wouldn't do for us. General knowledge for our purposes might be material included in a children's dictionary, encyclopedia, atlas, etc.

- It is permissible, but by no means necessary, to include a reference for a general-knowledge inference.

- The flip side of the "general knowledge" guideline is that if a source asserts that "A is a B", but uses an unconventional definition of what a B is, that is, one that contradicts the popular definition of what a B entails, then the editor is encouraged to compare and explain this usage of B.

- What constitutes "advancing a thesis" depends in part on whether opinion is involved, if there is a chain of inferences involved, or simply on the length of the synthesized material.

- Adding items to categories, "See also" lists of internal links, and deciding which sources to cite do not count as "advancing a thesis". Otherwise we could get into meta-meta situations where you need a source just to cite a source. This does not mean to enourage "listcruft"; what belongs in a list is subject to consensus, and some things do and don't belong. However, WP:NOR is not the right policy to settle such discussions. —The preceding unsigned comment was added by Squidfryerchef (talkcontribs) 04:01, 21 December 2006 (UTC).

Talk Pages

Every now and then I see someone respond to a talk page comment by saying the WP:NOR applies as much to talk pages as it does articles. However, that does not appear to be the case, and there is occasionally good reason to post original research on the talk page of an article for feedback as a means of later improving the article.

For example, let's say you uncover additional information for an article, or an error in the article, through your own original research. It would not be appropriate to simply edit the article using what you found. However, it would be appropriate to post your information on the article's talk page and ask if any other editors are aware of a published reference duplicating what you found. If a verifiable reference can be found by another editor for your original research, then it could be included in the actual article.

Thus WP:NOR does not fully apply to talk pages as it does actual article pages. Original research on a talk page can be ok provided it's done in the proper context of attempting to solicit additional verifiable cited information for the article itself. Dugwiki 17:16, 26 December 2006 (UTC)

NOR doesn't apply to talk pages, Dug, unless you're dealing with a living person, in which case WP:BLP kicks in and applies to articles and talk pages equally. Bear in mind, too, that talk pages are for discussing the article, not individual editors' theories. Although it's okay to outline some ideas in the hope that references can be found for them, it should be done briefly and sparingly. Wikipedia can't be used as a platform. Cheers, SlimVirgin (talk) 17:53, 26 December 2006 (UTC)
Right on all counts. The talk pages aren't supposed to be an open forum, they're intended to be discussion meant to improve the article. So original research on a talk page is ok provided it is done in that context. And you are correct that biographies are handled slightly differently than other subjects for various reasons, so as to avoid spreading possible false slander, etc. My post above is aimed mainly at the editors who simply blanket reply to any OR on talk pages that OR isn't allowed at all, even when the OR appears to be in proper context on the talk page. Dugwiki 18:08, 26 December 2006 (UTC)

Who is our model verifier?

There seems to be some disagreement about our model verifier and some confusion over the import of phrasing. The concept seems to owe something to the Common Law notion of a "reasonable man," sometimes now called a "reasonable person." This fictitious person possesses the knowledge and discretion that we commonly expect of a capable adult.

Similarly, Wikipedia policies such as Verifiability and NOR are based on the concept of a reasonable editor. This fictitious person possesses a reasonable level of knowledge, an accurate concept of logic, and is comparatively free of bias. It is therefore appropriate to term this a "reasonable adult," since Wikilawyers are typically less skilled than actual lawyers, and may not otherwise understand that a modicum of education is being assumed.

Despite what at least one editor has expressed, so phrasing things implies neither that all adults are reasonable nor that all children are not. Robert A.West (Talk) 19:25, 26 December 2006 (UTC)

Equations: Original research and verifiability

I think that there is a major gap in both the no origial research and verifiability policies when it comes to mathematical equations. There was a recent debate at full moon cycle and The village pump that brought up several good issues. Two questions that need to be addressed are: (1) Do all equations need to be "citable", that is, found in a "reputable" source? And (2) How much can you "change" an equation before it becomes original research, or is used out of context?

I would like to propose a few topics that should be considered for inclusion in both policies:

  1. All equations and constants should be verifiable; The equation or constant must be found in a reputatable source from which it can differ only by "trivial" modifications.
    1. Trival modifications include (1) a single mathematical operation, such as multiplication by a constant when converting between units, or converting radii to diameters, (2) a reorganization or regrouping of terms, (3) renaming of variables, and (perhaps) (4) the moving of terms from one side of the equation to the other.
  2. "Simple" modifications and "improvements" (unless already found in a reputable source) constitute original research.
    1. Simple operations would include (1) solving an equation for another variable (such as going from y=x^2 to x=\sqrt{y}), (2) inserting one equation into another, and (3) modifying or "updating" the numerical value of a constant employed in the original source.

While the above two rules might be considered to be too restrictive by some, please consider the following: If you are truly summarizing previous work, why would you have to make more than "trivial" changes to the equations and constants? If the constants do indeed need to be updated, then I would propose using the old (verifiable) ones, and mentioning in the text that these are could be out of date. Lunokhod 17:14, 27 December 2006 (UTC)

I fail to see how applying basic, elementary school mathematics(like the first two examples in part 2) to a particular form of an equation constitutes original research. Such manipulations do not consitute publishable results on their own; no scientist would take these results as being original unless a novel interpretation is then brought to bear. I would say that original research introduces new concepts or relationships which have not been previously considered or cannot be independently verified by a typical reader. Performing algebraic manipulations simply reorganizes known relationships and can be easily verified by the so-called "reasonable adult". Is there some particular reason(or example) why you have drawn the line on OR in this way? (BTW, I agree that 3rd item in part 2 is probably OR) -Joshua Davis 18:13, 27 December 2006 (UTC)

I agree. If a mathematical derivation is to be considered as original research then, assuming it is correct, it must be publishable in a reputable peer reviewed journal. E.g., some time ago I wrote a derivation of Planck's law that is self contained in the sense that all you need to understand it is what is by reading the section plus some things I refer to via internal wikilinks.
But this derivation is a completely different than any derivation you'll find in textbooks. In this narrow sense the derivation could be considered "original research". However, the purpose of wikipedia is to explain things in a way such that it can be understood by as many people as possible. To do that you need to adapt derivations you find in textbooks that are meant for university students who are familiar with more mathematical techniques and know more about physics than most wikipedia visitors.
The derivation is clearly not publishable in a reputable peer reviewed journal, so it would be wrong to delete it on the grounds that it is original research as there would be no way to get it back into wikipedia by getting it published first. The derivation is clearly not verifiable to most visitors of wikipedia, because they don't know enough maths. But then these people won't understand any derivation of Planck's law. All they could do to "verify" such a derivation is by comparing the equations one by one, symbol by symbol and see if they are literally the same without understanding one iota of what it all means. Clearly, this is not the kind of "verifiability" we need. A more reasonable definition of verifiability should be that the derivation should be verifiable to those people who know enough maths to understand the relevant topic. Count Iblis 23:30, 27 December 2006 (UTC)
The rationale for using "going from y=x^2 to x=\sqrt{y}" as being original research is the following: Let's presume that there is an equation that expresses the position of Jupiter's moon Io as a function of time, which has been published. One could invert this equation to give time as a function of the position of this moon, and use this as some form of "clock." Such a simple inversion has probably never been done before (but who knows) and this transformation, while simple, would represent a "novel" form of measuring time. I would argue that wikipedia is not the place to "publish" such "original research".
I understand that a novel derivation of Planck's law (as described above) might be more comprehensible than those published in most text books. Nevertheless, novel derivations that arrive at a familiar solution are original research. As I understand it, Einstein's derivations of special relativity is exactly what Count Iblis is describing: It is Einstein, and not Lorenz, that is generally attributed for describing this physical phenomenon.
In any case, the question remains as to where one "draws the line" between original research and novel re-descriptions of phenomena. I hope that this discussion will always keeps this important point under consideration. Lunokhod 01:29, 28 December 2006 (UTC)
I think the important point for novel research is a new interpretation, at least. Einstein re-interpreted known formulae and that lead to special relativity. Simply moving around symbols in well-prescribed rules for greater convenience of discussion does not constitute a new interpretation. I think an important aspect of new research is that it is disputable, at first. No one would dispute the result of a series of well-established manipulations applied to a well-established equation. Someone might dispute a new interpretation of the equation. As for your "clock example" above, it is a fact that one can use any periodic phenomena as a clock; no one would consider that research. It would be original research(and highly disputed) to pose that somehow the forward passage of time is caused by the Jupiter and its moon. This would be a new interpretation of the algebraically manipulated equation.
I would also like to point out that if one applies some algebraic relations to an equation, you don't really have a new equation, just a different form of the orginal. It's sort of like having rotated an object; the object isn't different, one is just looking at a different angle.
Thirdly, while not only disagreeing with this definition in principle, I think it is impractical. Having to comb through textbooks to get precisely the form of equation one is interested in to explain a certain point seems to be very inefficient, and I for one would never bother to do so.
As far as the necessity of "drawing a line" goes, I do not see this as a big problem that requires a more detailed rule/guideline. But this is my own experience; you seem to be suggesting that a problem came up elsewhere. -Joshua Davis 03:39, 28 December 2006 (UTC)
To clarify, I do not think that every equation should have a reference. It is only when there is a doubt as to the validity or originality that the above proposals would make sense. If someone required a reference for the "y=x^2 to x=\sqrt{y}", I could provide it, though I doubt that anyone would ask for it. Lunokhod 15:51, 28 December 2006 (UTC)

I generally support Lunokhod's efforts, but his proposal is too harsh. Lunokhod is trying to solve a problem seen in both full moon cycle and Halley's Comet, where amateurs used existing equations to create their own astronomical tables of object positions (epheremeses), which they subsequently added to Wikipedia. These calculations constitute original research; moreover, they would take me a long time to validate (and I am a professional astronomer).

However, restricting basic derivations (such as taking the square root of a simple equation) is too harsh, as virtually anyone with a high school education can validate such a calculation in less than one minute. Moreover, I would argue that some other simple calculations (such as taking averages or performing unit conversions) are needed to make Wikipedia more accessible for the general public.

As a compromise to Lunokhod's proposal, may I suggest the following:

  1. Simple calculations that can be performed by a person with a secondary school education in less than 2 minutes and that use only one or two referenced relations should be permitted. (For example, a calculation to determine the surface area of a sphere based on the diameter of the sphere should be allowed.)
  2. Complicated calculations that cannot be performed by a secondary school education in less than 2 minutes or that use multiple referenced equations should not be allowed. (For example, a derivation of the speed of light based on first priniciples should not be allowed unless taken from a published book.)

What are other people's thoughts? (Whatever people come up with, I do not want to see people publishing astronomical epheremeses in Wikipedia.) Dr. Submillimeter 12:21, 28 December 2006 (UTC)

When writing mathematics articles, I often need to rewrite formulas to make the notation consistent. I don't think Submillimeter's "two minutes" rule will be very helpful, and I'd prefer something which takes the intended audience into account. Anyway, I don't see why we need to specify this. In your case, where it takes a professional astronomer a long time (say, more than a couple of minutes) to validate the tables, that does seem to be original research. -- Jitse Niesen (talk) 13:53, 28 December 2006 (UTC)
As mentioned by Dr. Submillimeter, there are two distinct problems: novel derivations and the updating of constants that are used in equations. Perhaps I was too optimistic in trying to tackle both questions at the same time. Concerning the problem of "constants", we could just require that these be verifiable by a reputable source, just as everyhing else in Wikipedia should be. While this is implicit in the verifiability policy, I think that it needs to be explicit. Nevertheless, how does one write a policy statement saying that constants in equations need to be verifiable, but the equation itself does not?
While I would be happy with a two minute rule, as mentioned above, this does not address the competance of the intended audience. Besides, in general, I would not want to read a mathematical encylopedia article that required the reader to perform two minute calculations. Other possibilities might be an "operation count" between the original and quoted equation, or a limit on the types of operations one could perform (such as equation substitutions, re-organizations, variable renamings, etc.). Alternatively we could just require that the "spirit" of the original equation and/or derivation has not been altered, and that all constants within these be verifiable by a reputable source: This would at least address the problem of creating new ephemerides. Lunokhod 16:19, 28 December 2006 (UTC)
The policy-in-development Wikipedia:Attribution is on track to supercede this policy. On the subject of deductions from source material, the new policy simply says "It should be possible for any reader without specialist knowledge to understand the deductions. " This seems reasonable to me; specific restrictions like "number of operations" are unworkable in practice. CMummert 23:17, 28 December 2006 (UTC)
If you say it should be possible for any reader without specialist knowledge to verify the deductions, you open a can of worms: consider a statement that a musical piece is in some key (requires specialist knowledge about reading music), a statement about a foreign language source (requires specialist knowledge of a foreign language), or a logarithm (specialist knowledge since not everyone even knows what a logarithm is). Ken Arromdee 03:00, 29 December 2006 (UTC)
I have had no part in drafting the new policy; I just wanted to point out that it is coming onto the scene. You might consider raising your concerns on the talk page of WP:ATTR. CMummert 03:08, 29 December 2006 (UTC)
I agree with that. Getting specific will create problems. The WP:ATTR language seems just right to me. Crum375 23:22, 28 December 2006 (UTC)
I'm not certain that tabulated ephemerides, whether by professionals or otherwise, belong in Wikipedia at all. The volume of data those can represent is enormous, leaving aside the NOR issues. References and links to primary sources seem much more appropriate for Wikipedia content than tables of numbers. It seems like that sort of data, if the data are verifiable, are going to be simple copy/paste work and in danger of copyright violation, and if they are not verifiable, they'll violate NOR.
I'm interested in this discussion of equations because straightening out the discussions in source function, extinction coefficient, and opacity (optics) is, I expect, going to require me to blow the dust off my old graduate textbooks and cite page number and perhaps even equation number. And when Kurucz skips a few steps over an intermediate result -- which he does all too often -- then do I go looking for another textbook? I can't cite my old instructor's unpublished homework solutions... BSVulturis 23:30, 28 December 2006 (UTC)

Here is my second attempt at a policy statement concerning equations and constants. It is not as harsh as the original, and leaves sufficient vagueness to be adaptable to differing siuations. I believe that this should take care of the problem of creating new astronomical ephemerides, and could also be effective against users who attempt to insert "novel re-derivations" of a phenomenon. I have drawn the line at modifications that can be done on "the back of an envelope," but this could perhaps be made more precise by adding "(that is, containing one or two simple intermediate steps)". If you think this is suitable, feel free to change the text to your liking. Lunokhod 17:00, 30 December 2006 (UTC)

Mathematical equations and constants are subjected to Wikipedia's No Original Research and Verifiabilitiy policies. Cited forms do not necessarily need to be exact duplicates of those found in the original text: Trivial changes such as a regrouping or reordering of terms, renaming of variables, conversion of units, or other mathematical operations that working professionals could reasonalbly perform on the back of an envelope should be expected. What is not allowed are equations that differ from their original forms by several intermediate steps (irregardless of whether each step is simple) or numerical values of constants that are not immediately relatable by a trivial operation to the original cited value. "Updated" or "improved" values of constants that have not been published by a reputable source are strictly prohibited.

IMO, the foregoing has several stylistic problems. "Subjected to" should be "subject to." The "cited form" would be the form in the published material -- I am sure what is meant that the form expressed in Wikipedia need not be the form in the citation. "Irregardless" is a portmanteau word and a pet peeve of mine. I think the back-of-the-envelope quantification and reference to a working professional is wrong -- we should be thinking of what a professor would leave as an exercise for the student, not what he would leave unexpressed when speaking with other experts. Moreover, this could be clearer on the matter of inclusion of intermediate steps that the author left out. So, here is my attempt. Robert A.West (Talk) 17:35, 30 December 2006 (UTC)

Equations, constants and other formulae are part of the theories they express and must satisfy Wikipedia's No Original Research and Verifiability policies. This does not mean that the expressed forms need to be exact duplicates of those found in the original text. Trivial changes, such as a regrouping or reordering of terms, renaming of variables, conversion of units and so on, are acceptable, as is explicating intermediate steps omitted by the published material. In all cases, the equivalence to published material must be immediately obvious to any working professional and a reasonable "back of the envelope" exercise for the astute student.

Submitted for your approval or merciless commentary. Robert A.West (Talk) 17:36, 30 December 2006 (UTC)

I am happy with this. However, it might help to make even more explicit that constants in equations can not be modified. In the astronomical pages, we have been dealing with a curious situation where the equations were taken from a textbook, but the constants within them were "updated". This process requires several intermediate steps, which could take several minutes to hours for a professional to verify. It is for this reason that I think all constants must be immediately relatable to their published forms. Personally, I think that we should allow constants to differ by a conversion of units, or by the conversion of complementary quantities like diameters to radii, but not more. Lunokhod 18:34, 30 December 2006 (UTC)
Could you clarify? Suppose that we cite textbook A for the formula g=MG/r2, but it use old values for G and Me to arrive at a value for gn. Why not cite a newer source for the constants and redo the (trivial) calculation? Of course, the example is silly, but for the sake of argument, what would be wrong with it? Robert A.West (Talk) 18:51, 30 December 2006 (UTC)
I think that this example is acceptable, but close to the limit. It uses a single equation, requires only a few mathematical operations that could easily be performed on a calculator in 10 seconds, and it is immediately relatable to the original input constants (as long as the new values are given and referenced). The problem arises when the formula is not well known and/or when several intermediate steps are involved. For instance, let's calculate g for the Moon. We start with a published value of GM, however, a new topographic map was just published by the USGS about a month ago that we might want to use to update "R". This publication does not quote the mean radius of the Moon, and someone on wikipedia might try calculating this for themselves using the online data. However, there are several potential pitfalls associated with doing this: The raw data are not evenly distributed on the surface of the sphere, and the interpolated gridded data are presented in a "cylindrical" projection (the data are tabulated for every degree of longitude and latitude, even though this over-samples the data at the poles). You can see that this example is becoming problematical. Do we let the user average one or the other datasets (a simple operation), even though both methods of averaging have problems in terms of data sampling? Do we let the user explain what he did? Do we trust his value of the error bar for "R", if one is quoted at all? This calculation of "R" is original research, even though it is simple. The problem is how to describe the difference. It is for this reason I thought "immediately relatable to the published values" might be appropriate. Lunokhod 20:13, 30 December 2006 (UTC)
The proposals you are discussing here seem narrowly focused on astronomy, physics, and other situations where there are just a few "canonical" equations and each one is taken from a reliable source. The proposals may be appropriate in those settings, but are not appropriate for other settings. In mathematics articles, it is common to use equations instead of prose as part of explanations, and there is no expectation that every equation must be closely based on one in a printed source. So these proposals belong in a manual of style for particular types of articles, not here. The language that is already in WP:NOR and WP:ATT is intentionally vague in order to allow different sorts of articles to be written appropriately. CMummert 19:21, 30 December 2006 (UTC)
I agree. WP:ATT already makes it clear that some leeway is given to present information based on a published reliable source, e.g. calculate percentages from vote totals. Trying to nail it down any further will create problems down the road. Crum375 19:34, 30 December 2006 (UTC)
It appears to me that CMummert is describing the process of writing a secondary source, such as a review article or textbook. Wikipdedia is supposed to be a tertiary source, as is stated in the NOR policy, though the distinction between secondary and tertiary is not always clear. If you visit the Mathworld web site, which is similar to wikipedia for mathematics and physics topics, you'll find that the majority of equations there are referenced with page and/or equation numbers. Lunokhod 19:41, 30 December 2006 (UTC)
Mathworld is a very dubious model for us. It does not explain to the extent that we should, and yet it is notoriously inaccurate. Septentrionalis PMAnderson 00:19, 31 December 2006 (UTC)
This may be correct, but the errors that I have found and corrected in Mathworld were based on correct citations of an incorrect source. Stated otherwise, it is verifiable, and incorrect. Wikipedia has the same policy in this regard. Lunokhod 06:55, 31 December 2006 (UTC)

One of Lunokhod's grievances is with full moon cycle, which uses numerical values for the constants of motion of the moon, and which uses more precise values than cited in the source. Septentrionalis PMAnderson 00:17, 31 December 2006 (UTC)

Actually, every correct equation in a mathematics article could be related back to a published source rather trivially. The difficulty is more often to choose which textbook(s) to cite, and how to structure citations in the way that will be most useful to the reader. The derivations, even when original in detail, bear a relation to published texts that are obvious to anyone skilled in the art. Robert A.West (Talk) 00:44, 31 December 2006 (UTC)
I disagree that every equation can be "trivially" traced back to a published source. See this section of the citation guidelines adopted by the Math, Physics, and Chemistry projects for the consensus in those projects about what level of originality is appropriate in derivations. The Astronomy project is free to adopt more stringent guidelines for their articles, of course.
The situation with full moon cycle seems to be that it isn't possible to "easily" see how the derivations are performed - so why not just replace them with a version that can be cited, if it is a serious concern? CMummert 01:11, 31 December 2006 (UTC)

I agree with Joshua Davis' reply above (I haven't read all the other replies since I last visited this discussion, but I also agree with CMummert's points). Completely novel derivations of well established subjects like statistical mechanics are usually not considered to be "original research" by the theoretical physics community and are therefore unpublishable. Wikipedia should follow this standard and allow for such novel derivations. It is important to do this, because the standard derivations you find in textbooks are often not very suitable for wikipedia.

We should not create a gap that will prevent good explanations from appearing on wikipedia. If we say that a novel derivation is original research then the idea is that one should wait until that novel derivation is published and only then it can be included in wikipedia. Well, that's simply never going to happen in e.g. this case.

So, the only reasonable standard for original research is as follows: An argument is original research if, assuming it is correct, an article based on it is publishable in a leading peer reviewed journal. I know that many contributors to the physics articles are professional physicists and many of them are referees for leading journals like Phys. Rev. (Lett.). So we have a good grasp of what is original research and what is not according to this standard.

Reasons to remove a novel derivation can be that it is actually incorrect or if it really is original research. We often say that edits by cranks that e.g. purport to show that special relativity is wrong is original research. This just means that if it were true then it would be publishable. The fact that it isn't true then doesn't matter and we can avoid having to discuss this issue with the crank editor, which saves us from a lot of trouble :) Count Iblis 01:32, 31 December 2006 (UTC) Count Iblis 01:32, 31 December 2006 (UTC)

Your reasoning is correct, so I want to rephrase it to avoid the idea of publishability, which opens a can of worms. The wikipedia definition of "original research" is unrelated to the meanings of those two words:
"Original research is a term used in Wikipedia to refer to material that has not been published by a reliable source."
This is because the entire point of the NOR policy is to be able to delete crank articles without proving they are "wrong":
Wikipedia's founder, Jimbo Wales, has described the origin of the original research policy as follows: "The phrase 'original research' originated primarily as a practical means to deal with physics cranks,...
So the policy overlaps in an essential way with WP:V, which is why other people are trying to replace both of them with WP:ATT (especially because many people erroneously believe that "original research" involves originality or research). The current phrasing in WP:ATT seems fine with me: everything must be verifiable in principle, but not everything needs to be explicitly sourced. Which genuinely original derivations are verifiable by means of their triviality, and which are not, must be left to a case-by-case analysis. The editors familiar with a particular article are in a better position to make this analysis than random editors drafting a policy. If the verifiablity is in any reasonable doubt, the material can be removed by any editor. CMummert 01:49, 31 December 2006 (UTC)
I'm in agreement with CMummert on the basic premise that these kinds of issues are best addressed on a case-by-case basis rather than formulating a policy which could backfire on exactly the sorts of edits that we'd like to see(e.g. "novel" derivations which more clearly explain topics to laypeople than their textbook counterparts). I do have to say that I'm not currently familiar with WP:ATT but I'll try to get to it soon. -Joshua Davis 07:15, 31 December 2006 (UTC)

Primary vs. Secondary sources

I would like some clarification for what constitutes primary and secondary sources. Sometimes a primary source is a great deal more informative than a secondary source; especially when the secondary source seems to be promoting an opinion. Specifically, would it be alright to use The Caddo Tribe of Oklahoma, et al., Plaintiffs, v. The United States of America, Defendants; Docket No. 226 to cite facts in an Oklahoma historical article in reference to how and why Caddos, Witchitas, etc., ended up on reservations in Oklahoma. Most secondary sources in this area are kind of opinionated, and quite a few have the facts wrong. DeepFork 16:40, 29 December 2006 (UTC)

It's difficult to tell with such documents whether they represent the final word on the matter, or whether later proceedings have overturned the conclusions, or whether historical research has uncovered new information. A recent secondary source would be better. However, if a later source refers to the decision, but misrepresents the information in the decision, that would be reason to exclude information from the later source from the article. --Gerry Ashton 19:08, 29 December 2006 (UTC)
In my opinion, primary sources may *always* be directly quoted. With quote marks, and a ref-type citation to the exact section quoted. This allows readers to check that you have indeed exactly quoted the primary source. Sometimes editors use selective quoting to push a pov and that should be avoided. Wjhonson 16:42, 30 December 2006 (UTC)
Of course, selective citation can be done with secondary sources as well. The distinction IMO is not whether the source is pushing an opinion (in a sense, nearly all sources are), but whether we have perspective on the biasses and limitations of the source, and whether an expert is needed to intepret the source for us. When citing a court case, we normally assume the judge unbiassed, but it may require a legal technician to disentangle ratio decidendi from obiter dictum. In other instances, sources were secondary when written, but should now be treated as primary because the sources on which they depended are no longer extant, and because the biasses of the author are no longer obvious. Rather than parsing a fine technical distinction, I think Wikipedia would be improved if editors concentrated on the intent. Which is the most authoritative source for this information? Which explains it most clearly? Do any of the sources need another source to critique or explain them? Robert A.West (Talk) 17:56, 30 December 2006 (UTC)
I would be very careful with "always". I would not quote the prosecution testimony in the Salem witch trials without a warning (readily sourceable from almost any secondary source). I would not quote Soviet production figures or the Protocols of the Elders of Zion, except in articles directly on the subject, with the consensus opinion of them made perfectly clear. Most other primary sources require less buffering, but this depends on how strong the consensus is that they are mistaken or lying. Secondary literature is often, not always, a good guide on where any individual pprimary source falls. Septentrionalis PMAnderson 00:13, 31 December 2006 (UTC)

Thanks to all. Your comments are helpful. My conclusion is to use primary sources unless a secondary source provides a primary source for it's statements. Secondary sources that just cite another secondary source (or no source at all) would not be conclusive enough to overturn a known primary source. There seems an endless loop of secondary sources repeating seemingly faulty facts from each other. Thanks again. DeepFork 18:50, 5 January 2007 (UTC)

NOR limiting the usefulness of WP

In my opinion this policy can and does limited the usefulness of Wikipedia. I understand how Wikipedia is strictly an encylopedia however how is it not benefical to everyone if you include information that has not previously been documentated or recorded. An example of this is a small village school. Living in the UK in a small rural community, our local school does not have a website and there are certainly no books avalible on the subject. Is it therefore wrong for me to post information regarding the school when Wikipedia seems like the ideal place to do it?

Previously, the history of the school has remained a mystery. The structure of Wikipedia is so that such an article would be perfect for people to contribute their knowledge of the school to and while the factuality of the information may be doubtful, i'm sure there is some kind of disclaimer or tag that can be put over this. If not then how about a new one. "There is no documented evidence to support this". The infomation would be in the best interst to everyone, isn't it better that there is some disputable information there than none at all? —Preceding unsigned comment added by [[User:{{{1}}}|{{{1}}}]] ([[User talk:{{{1}}}|talk]] • [[Special:Contributions/{{{1}}}|contribs]])

  • No. We're working on a quality encyclopedia, not an assortment of random data. >Radiant< 13:56, 3 January 2007 (UTC)
  • If you understand that WP is strictly an encyclopedia, you should also understand what an encyclopedia is, which is not a collection of first person reports or original research, but instead a summary of existing knowledge, as published by reliable sources elsewhere. When a user reads a WP article s/he would like to know that the information there has been published elsewhere and has been vetted for accuracy by objective persons. As you know it is easy for someone to just make up all kinds of false information; we want our readers to learn to trust us as a reasonably reliable source. Crum375 14:09, 3 January 2007 (UTC)
  • No, but dig deeper. While I do like the "intelligence gathering" possibilities of Wikipedia, and am not as strict about what belongs in an encyclopedia as some editors, that would be going too far. However, I'm sure there is at least some printed information available about the school that could be cited, such as the local newspaper, as well as primary sources such as property records, etc. Squidfryerchef 00:37, 6 January 2007 (UTC)

What is reliable?

I'm caught on this, as I can see a few others are (adding information about a school is not "random data" btw). This policy was brought to my attention when the use of an 'unreliable' website as a source was disputed in Talk:Inheritance. The website in question is [2], which I admit is a fan-site (or, I should say, hate-site) and not some published journal or recognized organization. Although I agree with some of the points given against posting it as a source, a quick glance at the site is all one needs (even for those who haven't read the books) to confirm that it is indeed a reliable one; it uses common sense and arguments made are backed with facts. Just because it isn't formally published doesn't make this any less true. Just because it was written by a couple of average joes doesn't mean that they're stupid and can't say anything that's correct or, in this case, 'reliable'.

At what point should published works win out over original research, so long as it contains common sense and undeniable truths? —The preceding unsigned comment was added by (talk) 01:38, 4 January 2007 (UTC).

I believe the test of a RS vs a non-RS is not so much apparent common-sense or intelligence of the authors, as the existing independent layers of objective validation as well as corporate responsibility. If it's a scientific issue, other criteria would also apply, like peer review, multiple citations by others etc. If you have a couple of apparent well meaning average joes posting info on their web site, they have no independent layers of review, and no corporate oversight and/or liability. Hence they would normally not be considered as RS by WP. A typical newpaper or magazine would normally have those mechanisms in place, hence it would generally be acceptable unless known to be extremist or one-sided. Crum375 01:46, 4 January 2007 (UTC)
Aha, this problem again, please understand, wikipedia is not about the truth -- WP:NOTRUTH -- reliability is really all we have! --Merzul 02:00, 4 January 2007 (UTC)
I took a quick glance at the site, and didn't notice any citations to other sources. As far as I'm concerned I won't even begin to consider a fact to be a fact unless I've obeserved it myself or it's been published by a reputable source (which two average joes are not). Even then, no fact is undeniable. Now, if the site gave citations to reliable sources, you could look up those sources and use them in a Wikipedia article, but apparently the site didn't. --Gerry Ashton 02:17, 4 January 2007 (UTC)
The page in question is Talk:Inheritance Trilogy? (Inheritance also lack sources) --Merzul 02:28, 4 January 2007 (UTC)

What is a Reputable Publisher?

Wouldn't deciding if something is a reputable source be original research? And as Wikipedia disallows original research (and anything not from a reputable source) this means Wikipedia would be disallowed containing any information at all. --Huffers 01:52, 12 January 2007 (UTC)

'Original research' is when you present a conclusion or statement that is not based on published verifiable and reliable sources, and is unacceptable on Wikipedia. The mere act of deciding what is or is not reliable, verifiable, or reputable publisher etc., and presenting it neutrally, when agreed upon by effective consensus of multiple independent WP editors, is a vital part of our encyclopedia building endeavor. Crum375 02:13, 12 January 2007 (UTC)
You mean, the only original research that is allowed on Wikipedia is deciding what is reputable? --Huffers 02:30, 12 January 2007 (UTC)
That is not research per se, it is deciding on and rating available sources. And even then, it is a collective consensus of all WP's editors, who must reach consensus among them, and even then, it is constantly subject to change with newer sources and newer editors. Crum375 02:34, 12 January 2007 (UTC)
Does this mean, for example, that the page on The Lancet can say that The Lancet is a reputable source for medical information without needing a reputable source for this? --Huffers 02:39, 12 January 2007 (UTC)
No. Any claim that a source is 'reputable' must be based on other independent sources. And then there must be a consensus about it by the editors. OTOH some common sense must be used too. You don't want every article that references the New York Times to have sources proving the New York Times is reputable, but having that information in the NYT article itself would make sense. Crum375 02:47, 12 January 2007 (UTC)
Then to claim in The Lancet article that The Lancet is a reputable source requires another reputable source, which in turn requires another (to show it is a reputable source), etc.
If these sources are allowed to form a circle, then any two independent sources should be considered reputable if they each say the other is.
If these sources aren't allowed to form a circle, then they all rely on a source that cannot be claimed to be reputable.
This sounds paradoxical to me ;-) --Huffers 03:03, 12 January 2007 (UTC)
I don't think so. There is a so-called 'web of trust'. I agree that if you had a simple case of sources A, B, C and A said trust B, B said C and C said A, that would not be acceptable. Also if the sources were related in some business relationship (say) they would not be able to vouch for each other. But if the sources are indepedent, and they do not form a clear closed 'circle', then they would be potentially acceptable. I would also add size and age - a large publishing house and a long established one, would also gain reputability 'points'. And no source is exclusive - we can have multiple sources with opposing or multiple views for example, as long as they are presented neutrally and with due weight. Bottom line, use common sense. Crum375 03:17, 12 January 2007 (UTC)

<reset indenting>
The reason the determination of reliable sources is not as paradoxical as it seems on first blush are two-fold:

  1. Judgement of reliability of a source is not a binary classification, but rather a (fuzzy) rating scale. So New York Times is generally considered more reliable than Tomonoga Times (circ. 10000), which is more reliable than
  2. The sources are connected to each other not in a chain but rather as a network. So a source that is cited by 100 other sources considered reliable is likely to be labelled reliable; while one quoted only once in a reliable source will perhaps not be rated too highly. One may think, that we are still stuck with the problem of deciding on the first reliable source; but some more thought will hopefully convince you that that is not really needed in order to get a roughly consistent assignment of reliability grades. An aside: google (and other search engines) use similar techniques (greatly simplifying here) while making page ranking decisions; and of course humans do it implicitly in daily life !

Hope this possibly off-topic details are useful to you. Abecedare 03:22, 12 January 2007 (UTC)

Thanks. I'll have to think about this web stuff. --Huffers 03:39, 12 January 2007 (UTC)
You may find this link useful: PageRank (although keep in mind that this is just an analogy) Abecedare 04:05, 12 January 2007 (UTC)
And also as analogy, the Web of trust. Crum375 04:09, 12 January 2007 (UTC)


Can we add "religious scriptures" (texts such as Bible, Koran, Vedas etc) to the list of primary sources ? I know that this should be pretty obvious, but working on the Hinduism page I have seen editors debating on what the religion really says, based on quotes culled from the scriptures. It would be useful to point out the policy that specifically states that wikipedia articles cannot be argue a position (as opposed to illustrate a position) based on such quotes. Abecedare 15:31, 5 January 2007 (UTC)

I think that any recognized historical document is a primary source. Its published interpretation would be a secondary source. I think all religious scriptures or holy writings are therefore primary sources. In general, a primary source cannot be used to advance a position - it can only be used as reference or background material for secondary sources. Crum375 15:41, 5 January 2007 (UTC)
I agree that scriptures are covered by the mention of historical sources in WP:OR; but some editors working on religious articles would dispute that simple contention. So I am suggesting that we explicitly list "religious scriptures" as a primary source in the list, with the pro being that needless debates on article talk pages can be short-circuited and the only con that I can see would be the lengthening of the policy by two words. (I am assuming of course that no one here is disputing that scriptures are primary source. ) Abecedare 15:59, 5 January 2007 (UTC)
The main problem is that the "not truth" part of "verifiable, not truth" is so heavily underrated, and completely unappreciated! It is not so much the difference between primary vs secondary sources, the difference is how easily some idea can be attributed to a source. Secondary sources are very explicit, and you can cite a quotation from a secondary source saying "The meaning of Matthew x:yz is the following ... ". So it all comes down to attribution, NOT TRUTH! I wish the official policy was "No truth!" :) (please don't take that literally, but at least anybody who has been in a dispute over primary sources should know what I mean) --Merzul 19:10, 5 January 2007 (UTC)
I personally see no problem adding 'religious writings' to the list of examples of primary sources. Like any primary source, they would only be used as a reference or background for the relevant secondary source that discusses them and should not be used on their own to advance a position. Crum375 21:52, 5 January 2007 (UTC)
I think holy books are fine primary sources. However, there is one issue. I've lately been arguing against the stricter interpretations of OR, especially w.r.t. the "synthesis" paragraph, and I'm in favor of retaining the editor's leeway in comparing/contrasting sources. But I do think that for articles on religion, philosophy, or science, we must be careful not to cherry-pick verses to prove a point. Squidfryerchef 00:26, 6 January 2007 (UTC)
We should never prove a point with primary sources, whether 'cherry-picking' or not, as that would be 'advancing a position with a primary source', which we may not do per WP:NOR. It is fine to let a secondary source do it, and then we can supply the primary source as background and reference to the reader. Crum375 00:35, 6 January 2007 (UTC)

Possible clarification

Regarding this sentence:

Original research excludes editors' personal views, political opinions, and any personal analysis or synthesis of published material that appears to advance a position the editor may hold.

Should this be reworded slightly to say something like:

The no original research policy excludes editors' personal views, political opinions, and any personal analysis or synthesis of published material that appears to advance a position the editor may hold.

Or maybe:

The policy prohibiting original research excludes editors' personal views, political opinions, and any personal analysis or synthesis of published material that appears to advance a position the editor may hold.

It's hair-splitting, but I argue that those alternatives would be more precise than "Original research excludes . . ." Any thoughts, anybody? Yours, Famspear 03:23, 9 January 2007 (UTC)

Another option:
Original research includes editors' personal views, political opinions, and any personal analysis or synthesis of published material that appears to advance a position the editor may hold.
But I agree the current version needs fixing. Crum375 03:42, 9 January 2007 (UTC)
I tried to fix a similar gap in the introduction, same problem, the policy wording just doesn't adequately denote and describe what we actually mean by OR in practice. Two aspects not adequately referenced in the introduction: - "non-notable or little-supported personal opinions that have not been published or supported by a reliable (independent/third party) source", and "personal views of editors that lack notable external support". The intro needs to somehow make those notions clear as being included, since many references to OR mean those sort of things. Right now it doesn't. So I'd be happy to see a review of the wording of this policy too. FT2 (Talk | email) 01:14, 10 January 2007 (UTC)

Original Research vs. Collection of information

During a recent AfD the topic of Original Research came up, specifically the "synthesis of information" part of the policy. It was argued that citing a number of media references about a term and including them in a section titled "References in the Media" constituted synthesis, because the references made by the media were not noted by other media sources. Eventually the argument boiled down to whether or not the "mere collection of information" (my words) with no implied or directed conclusions drawn from this information constituted synthesis, thusly becoming Original Research. I have seen this same argument rasied a number of times since then.

The question I have is this: is the simple act of compiling information (say, background information on a person or place) from multiple sources without drawing any conclusions from that information a form of Original Research? My understanding of the policy is that provided there is no new conclusion created (implicitly or explicitly) by comparing or combining pieces of information it does not qualify. To paraphrase the example in the policy, if the text includes "A" and "B", so long as there is no "C" it is not Original Research. (After all, 99% of Wikipedia is a compilation of information from numerous sources)

Regardless of whether or not my understanding of the policy is correct, I believe that the article needs to be much more clear on this point. (e.g. a bolded sentence that states "the mere collection of information with no expressed or implied conclusions from that information (is/is not) synthesis." Thoughts? -- Y|yukichigai 20:20, 10 January 2007 (UTC)

The mere collection of information with no expressed or implied conclusions from that information is not synthesis. However, the mere collection of material with the implied purpose of establishing the notability of a neologism, is original research, and violates WP:NEO. --Merzul 21:05, 10 January 2007 (UTC)
What I'm trying to say is that once notability is established by a source discussing the term rather than using it, I can not imagine anybody objecting to a collection of interesting usages of the term. For example, the term 'new prog' can be cited as notable with this article because it says "The sound of 2006 will be called — in fact, already has been called — new prog", but a collection of reviews, where the term is used, should not be used for this purpose. Then once the notability is established with hopefully a few sources like the above, then these reviews can be used as evidence that individual bands have been labeled with this term. The collection of information without any implied conclusion other than what is contained in the original sources is wikipedia at its best. I hope this was clearer. --Merzul 21:33, 10 January 2007 (UTC)
Yeah, that's a bit more understandable. I should make it clearer myself that I've seen the "collection is synthesis" argument raised mostly in non-neologism articles since. Good to know, though. -- Y|yukichigai 21:44, 10 January 2007 (UTC)

An additional thought: perhaps the second use of the word "synthesis" should be removed from the text. (Provided my interpretation of the policy is correct) I think many people are getting confused by the fact that the definition of synthesis, combined with the way the policy is written, would seem to indicate that they can only use a combination of sources that somebody else has used before. Perhaps it should be changed to something like:

"It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or particular case to a reputable source;" (changed words in bold)

Further thoughts? -- Y|yukichigai 22:15, 10 January 2007 (UTC)

I'm not sure about the wording, but your idea seems spot on. For example, I don't like to see section headings like "Religion and intelligence" that cite research saying "most scientists don't believe in God." This isn't exactly synthesis, but it is using sourced material to imply something that is not attributable to the original sources (unless the study really did claim that religious people are less intelligent). One may think that this is covered by NPOV, but what usually happens is that for the sake of neutrality other studies are cited with contradictory results, and so the battleground is formed and the OR-wars for neutrality ensue. See the above discussion, where we have: the Detroit manual says this, but the Cleveland manual says that... Thus we have Neutral Original Research! Well, that's not quite NOR, or is it? --Merzul 22:41, 10 January 2007 (UTC) (sorry, don't mind me, I don't know what I'm arguing for... I'm going to sleep, because my answer didn't make sense to myself, maybe I didn't understand your response... hopefully somebody reasonable will answer you!) --Merzul 22:51, 10 January 2007 (UTC)
Yes, indeed, I did misunderstand the above. However, we both want the same clarification: do we accept Neutral Original Research? If we don't want it, then the example should be simplified to the following: "Mr. X, who wrote an article in a particular scholarly journal, was accused of plagiarism when it was discovered that he had copied a list of references from another journal article." And the neutral original research: "The Detroit Style Guide requires citation of the source actually consulted, but does not label violation of this rule plagiarism, whilst the Cleveland Style Guide does." If we do allow Neutral Original Research, then the wording of Yukichigai is truly spot on. --Merzul 01:32, 11 January 2007 (UTC)
As far as I can tell the existing wording permits "Neutral Original Research", because the first part of the sentence clearly indicates that the information must "build a particular case" to qualify as Original Research. However, I will point out that your example is not precisely neutral, as it would seem to propose (via vague implication) that whoever accused Mr. X of plagarism used the Detroit Style Guide over the Cleveland Style Guide. What would be more neutral would be a statement like "not all style guides define Mr. X's actions as plagarism," perhaps with the appropriate sections of the Cleveland and Detroit style guides cited or ref'd to back up that bit of info. -- Y|yukichigai 18:45, 11 January 2007 (UTC)

I'm the one who was brought up the "Detroit" and "Cleveland" examples in an effort to tone down the "Synthesis" example. I think there needs to be a counterexample, not "Neutral Original Research", but just good research. Because the "plagiarism" example is pretty mild OR, without a counterexample editors could infer that almost anything is OR.

This is OR: "Mr. X was accused of plagiarism...copied list of sources.... But because the Chicago Manual of Style does not consider this plagiarism, Mr. X did not commit plagiarism."

This, in my opinion, is not OR: "Mr. X was accused of plagiarism...copied list of sources.... However, it should be noted that academic authorities, such as the XYZ manual of style and the AQRS citation guide, are divided on whether this is considered plagiarism." This does not make any inferences, but some editors feel this is OR perhaps because of the proximity of sources. We can't get into a snake-eating-its-own-tail situation of only allowing combinations of sources that have been cited before.

What if it was: "Mr. X was accused of plagiarism in North Korea. However, it should be noted that bibliographic standards there are very different from those in the U.S. and Britain." Squidfryerchef 01:51, 11 January 2007 (UTC)

Well, on real articles one should not simply be a dick and demand sources for no reason, but on this page I have allowed myself to be one, and so I apologize, but the last example is original research, or rather original legal work. Are you Mr. X's defense attorney? Why should he be defended with a source that was not talking about him, it is the job of Mr. X's lawyers to engage in such research and point out these things, and then we can cite them. This is original research, if we don't have a source defending Mr. X on these grounds. -Merzul 18:51, 11 January 2007 (UTC)
The latter falls more into the category of "possibly tangential information" than Original Research. If anything though, that sort of info might very well be necessary to an article if Mr. X was accused of plagarism via a rule that was not present in the U.S. or U.K. In such an instance, not including that information could (unlikely) lead a reader to conclude that the N. Korean plagarism rule applied in the U.S./U.K. as well; in a sort of roundabout, non-NPOV way that lack of information would accomplish the same thing that the NOR policy sets out to prevent. -- Y|yukichigai 18:57, 11 January 2007 (UTC)
Legal definitions vary considerably from country to country. If one source's definition of plagiarism was different enough from what an English-speaking reader would assume plagiarism was, then we'd be almost obligated to explain the difference. Otherwise, we'd border on libelling Mr. X. PS. I think "possibly tangential information" is a great concept and maybe there should be a tag and a WP:TANGENT. Because I always see "unsourced" and "OR" tags added when what the editor really means is "justify or remove this tangent" Squidfryerchef 01:10, 12 January 2007 (UTC)

Foundation issues???

At the moment we have a sentence which says:

"The principles upon which these policies are based are non-negotiable on the English Wikipedia and cannot be superseded by other policies or guidelines, nor by editors' consensus."

I suggest that it is removed, because:

  • As far as I'm aware, this doesn't reflect an actual statement by the foundation. The document on meta that the statement links to does not mention original research at all, and in any case is not an official document issued by the foundation.
  • The real reason why no original research is one of our core policies is because there is an overwhelming consensus among Wikipedia contributors to support it. The sentence gives the impression the policy is there because of some edict by Jimbo or a committee, which is not the reality of the situation - like other Wikipedia policies, the policy was adopted primarily because there was a strong consensus in favour of it.

Enchanter 11:26, 13 January 2007 (UTC)

Now done Enchanter 12:44, 14 January 2007 (UTC)

Is Quoting from Wayback ( Considered OR?

Pardon me for intruding, but a question has come up lately on a WP page about an organization which has changed its Mission Statement. The current Mission Statement from that organization is cited in the article, and all editors seem satisfied with that. It is common on WP pages describing organizations to consider their history, which in this case requires some way of showing what their Mission Statement was for the first half of that organization's ten year history. This should, I would think, be straightforward due to the Wayback machine, in which this organization's previous Mission Statement has been found, and has been similarly quoted from with a solid citation to the Wayback machine (showing the path to the organization). There is no dispute that the citation is accurate, or that the citation points to the same organization.

Unfortunately there are a couple of editors for that page who are, I believe, misunderstanding the meaning of OR, when they state that this citation to the same organization's Mission Statement from the Wayback machine ( cannot be used since it is "WP:OR". Their argument is apparently that a citation to the page cannot be made, and that a reference to the organization's previous Mission Statement can be made ONLY if it is quoting some researcher or journalist who has written an article in a major publication that has pointed out that the Mission Statement has changed.

Frankly this seems plain ridiculous to me (and possibly an example of wikilawyering), but this well-cited section has been removed repeatedly in an edit war, and I am looking for opinions here from the editors of this OR page who are well acquainted with the OR guideline and can hopefully help me make a persuasive argument that OR is not meant to limit citations to ONLY published literature. Surely a citation from the Wayback machine is not only acceptable, but should be encouraged, since it is less subjective than a researcher's or journalist's statement, no matter how NPOV that researcher or journalist may be?

If the Wayback machine is as reliable a source as I believe it to be, would it be possible to add a reference to its use in the WP:OR page to make it clear that citing it directly is encouraged when it helps to illustrate a historical point? Or do you think it would be better to put a reference to the Wayback machine into the WP:RS page? Any ideas are welcome. Thank you, Jgui 04:33, 14 January 2007 (UTC)

I'm sure there are other good reasons why linking tothe wayback machine might be discouraged, but I don't think OR is the reason. OR is strictly defined as not quoting anybody, but for a wikipedian to come up with their own theories and write them here. So if they are quoting a defunct webpage only accessible through wayback or google cache or whatever, unless it's a "novel synthesis", that would more probably fall under WP:VER or something, but still at least it isn't exactly "original"... ፈቃደ (ውይይት) 04:42, 14 January 2007 (UTC)
The situation you describe is not and will never be Original Research. The Wayback Machine is a perfectly acceptable source, or rather is reliable only to the level of the content it has archived. Linking to a wayback'd version of the page to source information on a company's previous Mission Statement is the kind of use that The Wayback Machine was meant for. To reiterate, and so you'll have a little "sound bite" (or in this case a "text bite") to use if you want, Wayback'd webpages are perfectly reasonable sources, as reliable as the page(s) they archive, and using them in the way you have described is in no possible way Original Research. -- Y|yukichigai (ramble argue check) 05:11, 14 January 2007 (UTC)
Stop right there, technically this is a violation of WP:SYNT, and if we allow this, then all crackpot history theories are just as welcome. Essentially you are including a snapshot of a webpage to make the implied claim that the Mission Statement has changed, and this is a novel and substantial claim! You need to work with the editors that object to reach WP:CONSENSUS on this issue, because they have every right to be concerned about this activity. I don't buy this "it is only original research, when the implication is spelled out", wikipedia is read by human beings, if you first state the mission statement and then say "two years ago the webpage stated ...", the page is still making an unattributed claim that the company has changed the mission statement. Now, this may all be innocent, but if it is not, and if some editors object, then the burden of proof is on you to attribute this brilliant historical research to a secondary source. --Merzul 13:19, 14 January 2007 (UTC)
Some common sense is required regarding WP:SYNT. Merely pointing out that something has changed, and giving documention, is not OR. For example, if a company changed its name and mission statement over time (I'm thinking for example of Apple), a biography of the company would definitely be able to point to these because they neutrally document the company's change. What would not be allowed is speculation on "why" that company changed, or on the company's motivations, etc. That sort of speculation would require its own source.
Original research has nothing to do with research. It only means "material that cannot be sourced to a previous publication." In this case it appears that the material can be sourced, and therefore is not original research. CMummert · talk 13:37, 14 January 2007 (UTC)
I still don't understand why we can't wait for proper biographers or historians to do this job, I still don't understand why we need to dig in archives and primary sources to find new information and publish them on wikipedia. --Merzul 14:15, 14 January 2007 (UTC)
Technically you could interpret WP:SYNT to say that, but it is trumped by common sense and the spirit of WP:NOR. Citing a previous version of the webpage doesn't advance a position, it establishes factual information. If the use is to promote an agenda, position, opinion, or the like then that is Original Research, but a simple statement of fact is not a "position". Furthermore, even if you use the wayback link you are still (in this case) citing the primary source in both instances; no secondary sources must be cited to say "the webpage changed its mission statement" because the webpage itself has already established that and can be cited as such. -- Y|yukichigai (ramble argue check) 13:38, 14 January 2007 (UTC)
I'm saying that if there are editors who object, then they probably think this is promoting an agenda, maybe the agenda is to discredit the company. I don't know what is behind this, but if only judging the technical merits, this sounds like original analysis and synthesis of primary sources. That's why you need to work to reach consensus with the editors who object. --Merzul 14:15, 14 January 2007 (UTC)
Merzul, let us stipulate that the other editors agree that the Mission Statement has indeed changed and that the quote is not out of context or misrepresentative. When you say I still don't understand why we can't wait for proper biographers or historians to do this job, I still don't understand why we need to dig in archives and primary sources, are you saying that WP articles should be prohibited from using primary sources? That WP must wait until "proper biographers or historians" have taken the time and interest to interpret these primary sources before it can become suitable for a WP entry? WP currently has more than 1.5 million articles - do they all need to have "proper biographers or historians" interested in and writing about them before WP can write about them? And do you think, for example, that an article on the report of the Iraq Study Group should not be allowed to contain any citations from the report - but only citations from those writing about the report? I've found nothing in any WP guideline or the other editors who have commented to support this. Please explain your statement to me, or correct me if I've misunderstood you. Thank you, Jgui 03:45, 15 January 2007 (UTC)
Sorry, I had missed this direct question, but no, I don't think you have misunderstood me, I am indeed a citation hardliner and think that any contentious claim should have a proper attribution, and this does not only concern the written words on the page, but also implied ideas and conclusions, somebody has to be responsible for the implied claim that the mission statement has changed. However, I regret being so vocal in this debate, because I'm not so sure where the line between OR-synthesis and just fair synthesis actually lies. --Merzul 19:43, 23 January 2007 (UTC)


Yukichigai, please don't change the policy. Can you say here, please, what your edit means? SlimVirgin (talk) 06:13, 14 January 2007 (UTC)

Scroll down; there's a whole section on what my edit means called "Original Research vs. Collection of Information". -- Y|yukichigai (ramble argue check) 06:14, 14 January 2007 (UTC)
Please just explain what you mean. This is a longstanding part of the policy and you can't change it without extensive consultation, in addition to which it's not clear what you mean anyway. SlimVirgin (talk) 06:17, 14 January 2007 (UTC)
The edit seeks to clarify the meaning behind the use of the word "synthesis". As it is, the current wording, combined with the dictionary definition of synthesis, can be interpreted to mean that we are only able to use combinations of sources which have already been used before. i.e. we can only have an article which uses information from sources A, B, and C if some other source has also used the specific combination of sources A, B, and C. This is obviously not what NOR intends to prevent, but (as I stated in the above section) I am occasionally presented with the argument that because an article or section compiles information from multiple sources - despite the fact that the information promotes no conclusion, opinion, or particular case, implicit or otherwise - it constitues Original Research. Invariably that bullet point which I was editing is cited to support this article, often with the accompanying argument, "collecting information is synthesis." (which, from a dictionary standpoint, is technically correct) The edit I was making seeks to prevent that misinterpretation from occuring. -- Y|yukichigai (ramble argue check) 06:26, 14 January 2007 (UTC)
That is what it means, if the material serves to advance a position. The policy says: "Original research includes any ... analysis or synthesis of published material that appears to advance a position the editor may hold. That is, any facts, opinions, interpretations, definitions, and arguments published by Wikipedia must already have been published by a reliable publication in relation to the topic of the article." That's what it says and that's what it means. I'm still not clear what your edit means. SlimVirgin (talk) 06:30, 14 January 2007 (UTC)
I am aware of what the policy means; the problem is that the "synthesis" example is worded in a way such that it is often misconstrued or misinterpeted to mean that the mere collection of information constitutes Original Research, regardless of whether or not it advances a position. The edit I am trying to make is intended to be a simple way to make it clearer that an "original collection" of information is not by itself Original Research unless it advances a position. If you can think of a better way to accomplish that I would be glad to hear it. -- Y|yukichigai (ramble argue check) 06:42, 14 January 2007 (UTC)
Here is your edit: "[Material is not allowed if it] introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or particular case to a reputable source ..." Can you say what the insertion of "or particular case" means? SlimVirgin (talk) 06:32, 14 January 2007 (UTC)
In this case "particular case" means "the stated or implied conclusion established by the synthesis of information." It is that second "synthesis" that appears to cause most of the "synthesis is original research" misinterpretations I run across, and were it replaced with a different word or phrase it would make the policy clearer to more people; the term "particular case" was just the best idea I could come up with, and it was acceptable to the very small consensus of people who bothered to get involved with the discussion. -- Y|yukichigai (ramble argue check) 06:42, 14 January 2007 (UTC)
Hi again, because this is one of the three core content policies, if only a very small number of people take part in the discussion, it means there's no consensus to change the policy, even if you all agree. I don't see how the second "synthesis" causes a problem. Can you give me a concrete example of what you feel the current version would not allow that your version would, and why that would be a good thing? SlimVirgin (talk) 07:24, 14 January 2007 (UTC)
I think you're missing the point: this doesn't change policy, not at all; it simply changes the understandability (to make up a word) of the policy. I don't know how much clearer I can be. -- Y|yukichigai (ramble argue check) 07:30, 14 January 2007 (UTC)
It makes the sentence unclear, whereas it is clear as it is. That's a change. I still don't know what your sentence means exactly. The synthesis issue is explained elsewhere on the page too: an unpublished synthesis of published material is not allowed if it serves to advance a position. SlimVirgin (talk) 07:33, 14 January 2007 (UTC)
I meant it doesn't change the effect of the policy, or at least that is my intent, and while you and most other editors seem perfectly capable of understanding what that sentence means still many others do not, or even worse think they do when they don't. Specifically, they seem to think that the sentence means that any collection of information from multiple sources constitutes Original Research. If you feel that my edit did nothing to clarify the sentence... well then it obviously isn't the right change to make. :P Regardless, what I'm attempting to do still needs to be done: make it clearer that for a collection of information to count as Original Research it must establish or promote a specific position or conclusion that cannot be referenced to a reputable source. -- Y|yukichigai (ramble argue check) 07:42, 14 January 2007 (UTC)

On a side issue, these are my observations on the policy wording:

  1. In the intro, it is unclear what purpose the words "that appears to advance a position or" serve? Presumably the issue that makes something OR is whether the combination is a novel narrative or interpretation; if it's not a "novel narrative or historical interpretation" but "appears to advance a position", then it might well be breach of NPOV but it is probably not OR. I'm not sure what these words add in practice. (And deleting these words would probably fix User:Yukichigais concern too).
  2. Again, under "What is excluded?", "and any unpublished analysis or synthesis of published material that appears to advance a position" and "It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or synthesis to a reputable source;" In both cases the fact it favors an editors position is not, by itself, germane to OR, that would be breach of NPOV only. These might better be reworded: "and any unpublished analysis or synthesis of published material that appears to be a novel narrative or interpretation" and "It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a novel narrative or interpretation, without attributing that analysis or synthesis to a reputable source;"
  3. Why is the {{main}} for section "Reliable sources" not the policy on reliable sources itself?

I think these wordings might be the substantive source of the problems noted by Yukichigai. The issue on OR is the idea's novelty, it's not relevant to novelty whether the idea is also favored or otherwise by some editor. FT2 (Talk | email) 08:07, 14 January 2007 (UTC)

I can't say for sure whether or not these are contributing factors to the misinterpretation that seems to happen, but I do feel that the sections of text FT2 has identified here could use some clarification or rewording. There is, after all, nothing wrong with making the policy a little easier for the plebes to read. :P -- Y|yukichigai (ramble argue check) 10:03, 14 January 2007 (UTC)
I completely agree with FT2, advancing a position has nothing to do with OR. The crucial concept is novelty, and this whole issue of advancing a position only serves to confuse. --Merzul 13:28, 14 January 2007 (UTC)
I agree. I find the whole section on synthesis confusing, especially the example about plagiarism, which to me is hard to follow and not an obvious clear-cut case of OR. I would prefer that the section were rewritten using a new example. Enchanter 19:44, 14 January 2007 (UTC)
With the above comments, it seems worth testing consensus on whether to replace references to OR "advancing a position" by references to OR "advancing a novel narrative or interpretation" in the policy wording. Dipping a toe in the water, would people view discussion of this specific question favorably? FT2 (Talk | email) 00:27, 15 January 2007 (UTC)

It seems this is a minority view

It seems my stringent interpretation is a minority view. Can we discuss one final example, and I will then accommodate my view of synthesis to what you say. Here is the example...

The New York Times reports that Mr. X has been accused of Plagiarism in North Korea. However, according to the Korean Academic Citation Manual, the concept of plagiarism is "different in North Korea from that of the Americans".

So, assuming there are no secondary sources who defend Mr. X using the Korean Academic Citation Manual, do you prefer a Wikipedia that adds the extra sentence of analysis or one that only follows what has been published in secondary sources? Let's take a poll, I go first:

  • I think the second sentence is original research and violates WP:SYNT. Basically my understanding is to be neutral with respect to reliable secondary sources, not to be neutral with respect to ultimate reality, which I consider a dangerous path to follow. --Merzul 14:52, 14 January 2007 (UTC)
  • I think that WP:SYNT definitely needs to be clarified much, much more, since it could be conceivably argued that ANY given Wikipedia article is a "novel synthesis" when taken as a whole, except for the ones that are exact copies of the 1911 Britannica, et al. Indeed, it could be argued that Wikipedia as a whole is a "novel synthesis". Obviously some degree of synthesis is going to occur here, as different editors add different viewpoints together, we need to be much clearer in drawing the line of what constitutes an unacceptable novel synthesis. ፈቃደ (ውይይት) 14:59, 14 January 2007 (UTC)
You example is a perfect illustration of how this is a gray area. It's easy to see that assuming the topic is something like "Plagiarism of Mr. X", these might both be relevant facts. So if relevant fact A from source aa is already mentioned on the page, does this mean relevant fact B from source bb cannot be mentioned, because there is as yet no other source that mentions both relevant fact A and relevant fact B? Can they both be mentioned, but not in the same paragraph? Can they both be mentioned as long as they are added by different editors and not the same editor? The crucial test seems to be more concerned with whether or not there is a novel conclusion C that is being drawn from putting A and B together. The only thing joining your two relevant facts in your example is the word "however". So if we took out the word "however", would it still by SYNT? Or just stating two facts relevant to the case from two different sources? ፈቃደ (ውይይት) 15:54, 14 January 2007 (UTC)
These are excellent questions! Many of them would have to be solved on a case to case basis, but I still have a very hard time justifying where I personally draw those lines. For example, I think List of songs containing covert references to real musicians is not WP:SYNT, but the above example is, even if the word "however" is removed. There is no logical way I could explain why I think so, I guess we are again back to whether it is arguing a point of view. But maybe here my difference is that I have this strong belief, which seems, I freely confess, absurd in the highest degree, that meaning depends on context ;) --Merzul 17:57, 14 January 2007 (UTC)
I too feel that the synthesis explanation is hard to follow and difficult to enforce, unless one is ready to throw the baby (usefulness to reader) with the bathwater (OR). To expand on the example cited above, (1) is it still synthesis without the word "however", since in that case arguably no conclusion is being drawn ? (2) what about if we simply wikilinked "plagiarism" (wikilinking is an editorial choice) and that article began "Plagiarism is defined differently in different countries, for example according to the Korean Academic Citation Manual ..." ?
IMO the only difference between (1) and (2) is that in the second case the information is hidden from the reader, and therfore a worse choice given the fundamental goals of wikipedia. Abecedare 20:54, 14 January 2007 (UTC)
Per an above discussion of the same example, I think in this case it comes down to context. Were Mr. X cited for plagarism under a rule that only existed in Korea we'd be duty-bound to mention that the Korean rule does not exist in the U.S./U.K. on the english Wikipedia, because otherwise readers might inferr that the rule existed in the U.S./U.K. locale and we'd be bordering on libeling Mr. X. This starts to get into an odd "damned if you do, damned if you don't" situation where you have to chose between being compliant with WP:OR and being compliant with WP:NPOV if we go with the proposed interpretation of the policy; it would seem highly illogical, though, to maintain two policies which can conflict with each other in such a case, which is why I am suggesting that Original Research only applies when there is an implied or explicitly stated conclusion. -- Y|yukichigai (ramble argue check) 22:15, 14 January 2007 (UTC)
The issue is more complicated, because I see no contradiction with NPOV: "representing fairly and without bias all significant views that have been published by a reliable source." (italics added) In this example, no reliable sources is making this defense of Mr. X! Is it then my duty to research primary sources that I personally find applicable to the case? I'm not sure. But I am sure of this, if I represent the NY Times neutrally, then any case of libel should be taken up with them, not me. --Merzul 22:41, 14 January 2007 (UTC)

Example again, without "however"

Essentially to reset indentation :). The example that tries to get to the core of the issue:

The New York Times reports that Mr. X has been accused of Plagiarism in North Korea. According to the Korean Academic Citation Manual, the concept of plagiarism is "different in North Korea from that of the Americans".

As I said above, I think this is WP:SYNT, and here I hope to see people telling me that I'm right or wrong. Well I hope people telling me I'm right, but I would also be glad if people told me I'm wrong, and then we could include this to the policy as a positive example of not being original research. Opinions? --Merzul 22:49, 14 January 2007 (UTC)

I don't think it is SYNT. Again, assuming the topic is "Plagiarism of Mr. X" then these are both relevant facts, and legitimate as a "collection of information". Although juxtaposing them in the same paragraph like that may make some implications, it would certainly be fair to mention laws in Korea somewhere in the discussion, because it's relevant. I don't think the argument about "not defending Mr. X this way if noone else ever has" works because we aren't obliged to attack Mr. X any more than we are to defend him, this is only supplying directly relevant info, and common sense should prevail, or else almost any article here could be SYNT on grounds that no other source has exactly the same collection of information as we have. ፈቃደ (ውይይት) 23:06, 14 January 2007 (UTC)
THis is just A and B. In order to be an example of SYNT, there would have to be a third sentence C saying some explicit conclusion like "Therefore, he would be innocent under international law." ፈቃደ (ውይይት) 23:12, 14 January 2007 (UTC)
I don't believe it is OR, but I would delete the second sentence because it is irrelevant. It doesn't educate teh reader or advance a position, it only takes up space. What would be appropriate is an exact description of what it is that X has been charged with. The fact that this is different than in the US is not illuminating to the reader. Parallel example: "I got married in India. Marriage law in India is different than in the U.S." CMummert · talk 23:15, 14 January 2007 (UTC)
My opinion: the above example is technically synthesis, but such synthesis of reliably sourced, non-controversial information should be allowed/encouraged (although I see the slippery slope argument that this can lead to "Earth is round [1]. Round is a 2 dimensional shape.[2] Earth is 2 dimensional"). As another example: is the following synthesis ?
"According to the U.S. Forest Service, Shoshone National Forest has the greatest number individual glaciers of any National Forest in the Rocky Mountains. The forest recreation guide lists 16 named and 140 unnamed glaciers within the forest, all in the Wind River Range. Forty-four of these glaciers are in the Fitzpatrick Wilderness, centered around the highest mountain peaks.[23] However, the state water board for Wyoming lists only 63 glaciers for the entire Wind River Range, and this includes areas outside the forest boundaries.[24]"
since one (wikilawyer?) could argue, based on the current policy language, that this quote synthesizes information from two different sources ([23] and [24]) to push the case that there is no agreement on the number of glaciers in the forest; and since it does not cite another source independently contrasting these figures. This sentence is from yesterday's FA Shoshone National Forest.
I don't have a proposal for how to modify the policy language, in order to demarcate the legitimate/disallowed "synthesis". But I do believe that the current language and example are not very clear. Abecedare 00:39, 15 January 2007 (UTC)
This is exactly the sort of situation I'm talking about, the reason why I'm so intent on some form of clarification or rewording being done on this policy. Clearly (at least in my opinion) that sentence should not be considered Original Research, as it is merely a statement of verifiable fact. I do not believe that it is the type of synthesis that WP:OR (or WP:SYNT, for that matter) set out to prohibit when it was conceived, and the text of the policy should clearly and in no uncertain terms reflect that. -- Y|yukichigai (ramble argue check) 01:06, 15 January 2007 (UTC)
I think everybody engaged in this discussion agree on these points! --Merzul 03:31, 15 January 2007 (UTC)
Regarding the FA example, while technically it is synthesis, I would not object to it, and I wouldn't object to the above example either; if I saw them in a real article. I'm still thinking in analogous lines to WP:V that "any synthesis that is challenged or likely to be challenged needs a reliable source." I'm not saying that they all must disappear, but that I support a more strict interpretation of synthesis, relying on editors to be reasonable when demanding sources. This is of course based on personal experience, I haven't yet seen anybody being unreasonable in demanding sources, but I've seen plenty of accuracy problems. Maybe I should take a look at Yukichigai's underlying case and see how serious this SYNT-abuse really is. (I have personally abused SYNT once, but I don't think the discussion that ensued is something that I'm embarrassed about.) --Merzul 03:31, 15 January 2007 (UTC)
The FA example about glaciers is not OR. Stating neutrally that two reliable sources disagree isn't synthesis, it's accurate writing. So long as both sources are reliable and neither represents a minority view, presenting both is appropriate. In fact, it is encouraged by the neutral POV guideline WP:NPOV. The synthesis example in WP:OR is just an explanation of one common source of "original research". The actual definition of OR is just: material that cannot be attributed to a reliable source. So long as everything in an article can be attributed, it isn't OR - end of story. (The question of NPOV enters when trying to decide how strongly or weakly to cover a minority opinion, but that is not related to OR.) CMummert · talk 03:35, 15 January 2007 (UTC)
I think everyone here agrees that the FA example is should not be considered OR; however I fail to see how it does not violate WP:SYNT (which means that it is technically OR)? Note: I think that is a problem with WP:SYNT, not with the quoted text. Also the policy specifically contradicts the reasonable expectation that "So long as everything in an article can be attributed, it isn't OR". Abecedare 03:50, 15 January 2007 (UTC)
Yes, and how can one formulate the policy, such that the glacier example is allowed, but other harmful cases are not. The glacier example shows the contradiction between two primary sources. What if we point out a contradiction of a similar nature between the Quran and the Bible in a neutral and matter of fact manner? What exactly is the difference, other than common sense? --Merzul 04:03, 15 January 2007 (UTC)
The definition of OR is at the top of the policy: "Original research (OR) is a term used in Wikipedia to refer to material that has not been published by a reliable source". The synthesis section describes one common way that unsourceable material can find its way into articles, but it isn't another definition of OR. It's an example. But this is all getting too lawyerish; the right way to interpret policies is with their intentions and common sense in mind. CMummert · talk 04:12, 15 January 2007 (UTC)
The synthesis section is extremely important, because the material is not about unsourceable material. It is about avoiding how well-sourced material can be cause of major embarrassment. The kind of example I'm talking about is what we have in the philosophy articles. For example, the theory of Marx is first simplified into "belief in God is the result of powerlessness under oppression" and then one editor wants to cite statistical figures that people in powerful positions also believe in God in order to refute Marx's theory. Now, all of this is sourced and good, except it is a synthesis of sourced material, and in a comparison study similar to the one Nature did, this would be one of the big blunders that would make philosophers laugh out laud. It is precisely because synthesis of sourced material is in some sense the core activity that we engage in that we need better guidance on this issue. My own rule of thumb is that I don't do the kind of synthesis that is usually done by people with PhDs, this means I don't mention a fact (now matter how well sourced) as even implicitly implying that it has anything to do with the refutation of a philosophical theory, unless the refutation is also attributed. In the case of the examples above, you need a PhD in comparative religion to contrast the Bible and the Quran, but you don't need a PhD to contrast the number of glaciers, so in a sense it comes down to respect for experts. In particular, it comes down to realizing that History, Philosophy, Religious Studies, Psychology etc., are also academic disciplines. So, yes, we are back to the intention of the policy. But I don't think I am the one who has misunderstood the intentions of the policy, rather the official example seems rather distant from the sort of crackpot articles and pseudo-philosophical speculations that I prefer were not mentioned on wikipedia. --Merzul 19:56, 15 January 2007 (UTC)

Back to current policy

Still, let's get legalistic again... My question is if you all agree that the following example is equivalent to the current example used in the policy?

Smith says that Jones committed plagiarism by copying references from another book. The Chicago Manual of Style does not call the copying of references plagiarism.

So let's forget about the issue of this being OR or not. The question is only if this is equivalent? Meaning has this less precise reformulation also removed the violation of SYNT? --Merzul 19:56, 15 January 2007 (UTC)

Yes, I think it has. A line is definitely crossed when you add a third explicit statement that itself isn't supported by anyone, but is someone's synthesis of the other two. So that is the easiest place to draw the line. Take away the third novel statement, and all you have are two supported statements from two different sources, that could easily both be relevant. Let the reader make their own inferences by putting the two statements together, but you can't nail it down as being a "synthesis", unless the synthesis (conclusion) is explicitly stated. ፈቃደ (ውይይት) 20:06, 15 January 2007 (UTC)
The two sentences aren't in the WP definition of OR because they can be attributed to reliable sources. But the second sentence is irrelevant, so it ought to be removed. It doesn't matter what the CMS calls plagiarism, Smith has still been charged with it. Adding a third, unattributed sentence saying "Therefore Smith didn't commit plagiarism" would certainly qualify as OR unless a source could be provided for that opinion. Putting well-sourced but leading sentences into articles to attempt to influence the reader falls into the POV policy but not into the OR policy.CMummert · talk 20:16, 15 January 2007 (UTC)

One last try

What about this one?

Smith says that Jones committed plagiarism by copying references from another book. Jones denies this, and says it's acceptable scholarly practice to use other people's books to find new references. The Chicago Manual of Style requires citation of the source actually consulted, but does not call violating this rule "plagiarism." Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

I don't see how this can violate any other policy, it is both relevant and neutral, unless any other style manual has a different definition, and let's assume they don't. If this is not synthesis, I will add this to the policy page as an example of removed synthesis, and if it is synthesis I will replace the current confusing examples. In either case, I will put this example into the policy, but YOU will decide, if this is or is not SYNT. --Merzul 21:12, 15 January 2007 (UTC)

  • I personally think this is synthesis, and if challenged by another editor would require a source. However, the above discussion shows this is a minority view, and so I'm putting it in as a counter-example. I know this is very hasty, but I'm going to sleep now, and well... it's not too difficult to revert the page! --Merzul 23:31, 15 January 2007 (UTC)
  • Please do not change the policy again. That none of the regular editors of this page are commenting is a sign that you have no consensus to make the change. SlimVirgin (talk) 23:36, 15 January 2007 (UTC)
Here is a slightly modified one:
"Smith says that Jones committed plagiarism by copying references from another book. Jones denies this, and says it's acceptable scholarly practice to use other people's books to find new references. The Chicago Manual of Style defines plagiarism as using a source's information, ideas, words, or structure without citing them. It also requires citation of the source actually consulted, but although it does not call violating this rule "plagiarism."
(for simple referencing, you can refer to Merzul's version os OLT 1, and mine as OLT 2)
Additionally I would disagree that the second and third sentence are irrelevant. For instance, if instead of plagiarism we had, "X was accused of voluntary culpable homicide"; shouldn't the next sentence explain how that term is defined by the relevant country's law ? I think the same principle applies to plagiarism, whose meaning will not be necessarily clear to an average reader.
And to clarify my position: I consider OLT 2, to be compliant with OR, NPOV etc. However any change to the policy should be vetted by a much larger quorum of editors. Perhaps this discussion can be moved to Wikipedia:Village pump (policy). Abecedare 23:39, 15 January 2007 (UTC)
Look, the whole Chicago Manual of Style reference is OR unless a reliable source has mentioned it in relation to the topic of the article. That's the bottom line, and to be honest, if you don't understand this part of the policy, you really ought not to be editing the policy page, because this is actually a very important part of the NOR policy. SlimVirgin (talk) 23:45, 15 January 2007 (UTC)
(1) With all due respect, I don't think your interpretation is universally accepted; for example CMummert 's reading is that the example is clearly not OR (although he has other objections to the paragraph); and I have been clear to mark my personal interpretations as such.
(2) I have not edited the policy, nor even expressed any intent to do so. In fact, in my last comment I clearly state, that we should not edit the policy without establishing consensus among a large group of editors. Given that, I do find your comment "if you don't understand this part of the policy, you really ought not to be editing the policy page," to be patronizing. But I'll assume good faith on your part and attribute the comment to a burst of frustration. Thanks. Abecedare 23:57, 15 January 2007 (UTC)
This happens a lot. Merzul is a new editor who has made 522 edits overall, and 288 to articles. He was thwarted at some point regarding material he wanted to add by someone quoting the NOR policy. So he has arrived at this policy, with almost no editing experience, and wants to change it to allow the kind of edits he feels he should be able to make. Merzul, you're welcome to discuss it here, but in truth, you'd be better going out there to edit the encyclopedia so that you gain more experience, and then you may start to see how this policy, as it's currently worded, makes sense. SlimVirgin (talk) 23:50, 15 January 2007 (UTC)
Sounds about right. FeloniousMonk 00:02, 16 January 2007 (UTC)
Except, I only wish I had been thwarted by somebody citing this policy. Is it the intention of "the regular editors" of this page to just ignore the fact that a large number of us are completely confused about this policy?? If the above example that I added is indeed original research, then why isn't it used as an example of synthesis in the first place? Should we just keep the policy deliberately ambiguous, so that new users are confused and only those with >40k edits can understand them? --Merzul 02:12, 16 January 2007 (UTC)
Merzul is synthesizing material to make a contentious claim; it is classic original research. Who says the Chicago Manual of Style is relevant to this specific incident? Who says that the Wikipedia editor has applied it correctly? Allowing this will, in fact, allow any and all original research. Jayjg (talk) 01:23, 16 January 2007 (UTC)
The POV guideline/policy covers this situation just fine. If every sentence is attributed to a reliable source, then there is no "original research". There may still be problems with other policies, or with common sense. CMummert · talk 01:46, 16 January 2007 (UTC)
Nonsense! The individual elements of original research can often be well-sourced. It's how you put them together that makes it original! Jayjg (talk) 02:01, 16 January 2007 (UTC)
"Original research (OR) is a term used in Wikipedia to refer to material that has not been published by a reliable source. " CMummert · talk 02:50, 16 January 2007 (UTC)
"Material" can also be conclusions, implications or juxtapositions, even if the individual elements are well sourced. Crum375 02:55, 16 January 2007 (UTC)
Juxtapositions are not even mentioned by the current phrasing. Two well-sourced direct quotes from reliable sources, one after another, cannot be original research as it is currently defined. But we are in agreement about the overall interpretation: even implied implications are inappropriate if they cannot be sourced. I disagree that it is this policy that covers them, but I don't think that lawyerlike interpretation is necessary to remove inappropriate unverifiable arguments. CMummert · talk 03:01, 16 January 2007 (UTC)
It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or synthesis to a reputable source - that's in the current policy, and is pretty clear. Jayjg (talk) 03:42, 16 January 2007 (UTC)
I'm not arguing for any change to the current policy, and I don't think you are, either. If we disagree about how we interpret the policy in theory, we still probably agree in practice about specific examples. CMummert · talk 03:56, 16 January 2007 (UTC)
  • Comment - I don't know if anyone else feels this way, but to me it would seem a lot less like a "synthesis" if the two items were simply split into two paragraphs. Putting two different cites in the same paragraph is a nuanced synthesis that they are supposed to be related or build a conclusion. Anyone else get that basic feeling, that this is part of it or has something to do with the definition? ፈቃደ (ውይይት) 01:41, 16 January 2007 (UTC)

The issue, IMO, is simply that NOR also includes 'NOJ' - No Original Juxtaposition. That means that we may not juxtapose items together and thereby create a new implication. If the juxtaposition and the implication can be properly attributed to a reliable source, we are covered. Otherwise, we are not. Crum375 01:55, 16 January 2007 (UTC)

If that is so, then why on earth does the policy example contain "If Jones's claim that he consulted the original sources is false, this..." It is the worst and most obnoxious red herring that I have ever seen. --Merzul 02:17, 16 January 2007 (UTC)
The second paragraph which begins with "If Jones's" is adding juxtaposed material plus novel unsourced conclusions. IMO the example would have been valid even if the second paragraph simply quoted the Chicago Manual of Style's definition of plagiarism at that point, without having a reliable source that quoted it in that specific context with Jones. Crum375 02:28, 16 January 2007 (UTC)
Again, the policy would be much clearer, that is much stronger, if it did not begin with that sentence. So my question is why is the policy is deliberately weakened, when all "regular editors" hold the stronger view. Is it because you are afraid that the policy might be abused, if the meaning was more obvious? --Merzul 02:35, 16 January 2007 (UTC)
It's clear enough as it is. You can source A, and you can source B, but putting them together might still be original research unless you find a source who has put them together in the context of the topic at hand. SlimVirgin (talk) 03:49, 16 January 2007 (UTC)
And therin lies what I was trying to get at in the first place. Read what you wrote again; someone could read that and come to the conclusion that unless an outside source had used the specific combination of sources A and B, WP:NOR would in every circumstance prohibit creating a section that used sources A and B in conjunction. This is the kind of wording that is present in the current text of the policy. Despite what it actually means, it is quite possible to misinterpret. -- Y|yukichigai (ramble argue check) 04:23, 16 January 2007 (UTC)
They would only misinterpret it if they missed the crucial phrase: if it serves to advance a position. SlimVirgin (talk) 04:43, 16 January 2007 (UTC)
Yes, exactly. And yet, it happens. Constantly. Repeatedly. Annoyingly. And it's often times impossible to persuade them that their interpretation is wrong, no matter how simple one makes the argument. -- Y|yukichigai (ramble argue

check) 08:48, 16 January 2007 (UTC)

I agree. The current section on synthesis - and especially the example on plagiarism - is hard to follow and interpret. To page is failing to set out the policy clearly, resulting in all these misunderstandings. Enchanter 23:52, 16 January 2007 (UTC)

Synthesis proposal: without the red herring

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 in order 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.[1] A and B, implies C is acceptable only if a reliable source has published this argument in relation to the topic of the article.

Here is an example from a Wikipedia article, with the names changed. The article was about Jones:

Smith says that Jones committed plagiarism by copying references from another book. Jones denies this, and says it's acceptable scholarly practice to use other people's books to find new references.

That much is fine. Now comes the unpublished synthesis of published material:

The Chicago Manual of Style requires citation of the source actually consulted, but does not call violating this rule "plagiarism." Instead, plagiarism is defined as using a source's information, ideas, words, or structure without citing them.

This 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. To make the paragraph consistent with this policy, a reliable source is needed that specifically comments on the Smith and Jones dispute and makes the same point about the Chicago Manual of Style and plagiarism. In other words, that precise analysis must have been published by a reliable source in relation to the topic before it can be published in Wikipedia.


So is this still SYNT? If the regulars agree with this, then I see no reason why you prefer the official version. (Except the fear of abuse). The reason I hate the official version is that everybody knows the first sentence is original research, so it really distracts from understanding what synthesis is! --Merzul 03:12, 16 January 2007 (UTC)

Which first sentence is OR? SlimVirgin (talk) 03:50, 16 January 2007 (UTC)
This is actually less clear, because the intervening text makes it look like a new paragraph, the purpose of the sentence you removed is to make it screamingly obvious that someone is trying to tie these thoughts together into a new synthesis. But while we're at it, would it be okay to let editors use the above section on this talkpage as a sort of sandbox to tweak the wording of the policy for various proposed improvements? ፈቃደ (ውይይት) 04:18, 16 January 2007 (UTC)
Please don't tweak it just yet. I want to know, if SlimVirgin would call this synthesis or not. The point is that, if they consider this synthesis, then this should be spelled out, because then we would not be confused about it. Clearly, if the above is synthesis, then the official example is OBVIOUSLY synthesis. --Merzul 12:05, 16 January 2007 (UTC)
Anyway, I'm talking about the first sentence of the official version that has been removed in this example to make it clear that the entire paragraph is OR. In the proposed example above, I don't have to even write "entire paragraph", because there is no confusion. You should know what I'm talking about! I'm amazed that in spite of all the evidence that the policy is not understood, you insist that it is clear as it stands. Please, just consider this more carefully, neither the above revert or the revert to my edit would have even been needed, if the policy was clearer. --Merzul 12:05, 16 January 2007 (UTC)

I just want to make it clear that it is me, Merzul, who is behind this conspiracy to clarify the synthesis example. Now, maybe I can state what I'm asking more clearly as two questions that can be more or less answered one after the other:

  1. Is the above example original research?
  2. If yes, then isn't it a more pedagogical illustration of synthesis?
    If no, then what made this OR and the above example NOR?

Thank you for your attention, I will now take the advice above and edit the encyclopedia for a change :) Still, I believe that I have a point, so it is hard to just leave this, but I'll try, so good bye! --Merzul 16:58, 16 January 2007 (UTC)

Interpreting text on an image: OR or not?

Map by MercatorHondius from 1634, carrying the label Sinus Arabicus

On Talk:Persian Gulf naming dispute there is a drawn-out on-going dispute. The article Persian Gulf naming dispute contains a cartographic gallery displaying sixteen historical maps in which the Persian Sea is labelled "Persian Gulf" in various variations and languages. This bolsters the claim, implied by the text of the article, that "Persian Gulf" is the time-honoured name for this body of water, and that the name "Arabian Gulf" has not been used for it before the rise of Arab nationalism in the sixties.

Some editors want to include a few historical maps like the one shown here, which – at least in their view – show that occasionally the label "Arabian Gulf" has been applied earlier to this body of water. Other editors oppose this. They maintain that the interpretation that the label "Arabian Gulf" on the map is applied there to the body of water generally known as the Persian Gulf – and not, for example, to the (nearby) Shatt al-Arab river – without support by "textual evidence" (a published text stating that the label on the map refers to this body of water) is original research and therefore not allowed.

The issue is of course much more general than just these maps; it may apply to all text on an image, and also to the interpretation of what is further depicted on an image. For example, for the map here: does the body of water on the map in which we see the text Sinus Arabicus show the Persian Gulf, and not the Arabian Sea? Is this assumption warranted without support from a citable source? How should the requirement of "no original research" be construed in cases like this? Is it correct to consider inclusion of these maps without support by further "textual evidence" a form of OR? What about the maps that are included in the gallery of Persian Gulf naming dispute?  --LambiamTalk 06:33, 17 January 2007 (UTC)

I don't think the fact that these sources are images rather than text should make any difference - the normal criteria to interpreting sources should apply. That is, just as with a text source, it would seem reasonable that a map could be used as a source for a statement that the map clearly supports (e.g. map X shows country Y to cover a certain area), but should not be interpreted to give statements that go beyond what the map says. If the map is not clear, due to unclear labelling, it shouldn't be used to back up a statement one way or the other, just as an unclear paragraph of text shouldn't be used as a source and interpreted as one thing or another. It would seem rather arbitrary to make a distinction to say that you could support a statement with a source that was written in text, but not one that states exactly the same thing in a map or diagram.
Enchanter 22:32, 17 January 2007 (UTC)
Th fact that the inscription is not in English is a minor problem. But the contributors appear to willing to say anything to maintain their PoV, which is the major problem. Septentrionalis PMAnderson 23:48, 17 January 2007 (UTC)
Yes, engaging in cartographic interpretation and publishing the results on Wikipedia in order to bolster a claim about historic word usage falls under even fairly loose definitions of the NOR policy. Jkelly 23:55, 17 January 2007 (UTC)
Why not just say that "according to a map by Jodocus Hondius, the body of water is named Sinus Arabicus. Other maps by x, y, and z describe it as ABC"? ≈ jossi ≈ (talk) 00:00, 18 January 2007 (UTC)
Because such edits[3] are instantly reverted[4]. According to a majority (consensus?) of editors on the talk page this is "original research" if not backed up by "textual evidence"[5]. My question is: is such separate textual evidence needed, something like "Dr So N. So writes in his book Historical Maps on page 123 that on the 1634 Mercator–Hondius map the Persian Gulf is labelled Sinus Arabicus"? One could say (and some editors do say) that inferring this fact from the map itself is a form of cartographic interpretation, and publishing the results on Wikipedia violates NOR policy, as Jkelly might seem to suggest also.  --LambiamTalk 01:09, 18 January 2007 (UTC)
At the very least, this map should be added to the "Cartographical gallery". A description of what each of the other maps says is provided, so presumably this one gets to play too. Reading what the map plainly labels the Persian Gulf is not OR, since any other reader can verify it by looking at the image provided. WP:RS allows "descriptive" claims about primary sources, and this map is certainly a primary source for the question of what names have been attached to the Persian Gulf by maps. The bigger problem with the overall article is WP:SYNT, which is a common problem in "dispute" articles. CMummert · talk 01:33, 18 January 2007 (UTC)
I have to agree with Jkelly here. When there is no textual evidence or there is textual evidence to the contrary of what the map is suppose to represent [6] (which seems to be the case here), it falls under NOR. Khoikhoi 03:16, 18 January 2007 (UTC)
Thanks to khoikhoi for pointing to the link above. If people read the arguments in the talk page, I have brought three maps from the same Atlas of Hondius fromt the same book[7] where in the maps three different designations are used for the same body of water. Thus one is forced to see what is the designation used in the text. And throughout the text, only Persian Gulf is used. Note there are very distorted maps from 500 years ago. Just in the area of Persia I found a map that calls central Persia as Assyria, and put the [red sea] on the bottom of the horn of Africa. Another put the red sea in the Indian Ocean. Another put the Gulf of Oman as the Sea of Persia. Thus when there is contradictory maps within one atlas, one needs to ascertain the text of the altas for clarification and they certainly go hand in hand. --alidoostzadeh 05:24, 18 January 2007 (UTC)
The section Khoikhoi cites is an incoherent rant, ending with a paragraph of personal attacks in FULL CAPS.
  • The chief intelligible argument in it is that the Renaissance mapmaker cannot have called the body of water the "Sinus Arabicus" because the U.S> Geographic Board decided some centuries later to call it the Persian Gulf.
  • The other argument seems to be that Hondius, or possibly Mercator, called the Red Sea the Sinus Arabicus on other maps in the same book (what book, btw? I will see if I can find an offline copy.) I've seen enough pre-modern cartography to find this unsurprising, even expected.
  • Related to this is some bizarre argument, which I do not follow, about the Indian Ocean being labelled the Mare Rubrum. Since the name Mare Rubrum/Erythraean Sea originally belonged to the Indian Ocean, and was only restricted to the present Red Sea in the last few centuries, I am underwhelmed.
As for the purposes of this policy: this is a primary source. It is evidence for what is plainly visible on it: the words Sinus Arabicus are indeed written on the Persian Gulf. Attacks on its meaning, authenticity, and reliability also belong in the article provided they are sourced. Without that, they are themselves Original Research. Septentrionalis PMAnderson 06:39, 18 January 2007 (UTC)
Actually the name Persian Gulf dates back to at least 2000 years ago (Strabo). In the text of Hondius which is based on Mercator it is exclusively called Persian Bay and there are maps with both Persian Gulf, Mare-e-el-Qatif and Sinus Arabicus designating the same area and note there are maps that designate Sinus Arabicus as the Arabian Sea and Red Sea in the same book. There are more than 400 ancient maps calling it Persian Gulf. The most important argument is that there is not single text that has called the Persian, Sinus Arabicus. Note I have a book with 400 ancient European maps calling it Persian Gulf. The US Geographic Board is following a 2000+ year old convention and more than 400+ European maps. Several lone maps (who in the same book have different designations for the same area) out of more than 400+ maps from 14th century to 19th century can not be seen as evidence over 400+ maps and Hondius himself which uses Persian Gulf in the text. As for the 400+ maps, see this book: [8]. --alidoostzadeh 13:06, 18 January 2007 (UTC)
It is not the role of WP to present arguments in favor of particular interpretations; you should familiarize yourself with WP:OR and particulalrly with WP:SYNT. If there are just a few old maps that use the "Arabian gulf" term, it is perfectly acceptable to say "A few maps, including ..., use the term "Arabian gulf", but almost all of the maps listed in ... use the term Persian Gulf." That phrasing is neutral and doesn't give undue weight to the small number of exceptions. CMummert · talk 13:14, 18 January 2007 (UTC)
Precisely; and especially on an article on a dispute, we should give both sides of the dispute, which is all that is desired here. We should not, and will not, decide one side is right (even if it is; I use "Persian Gulf" myself); and edit out the contentions of the other side. Septentrionalis PMAnderson 18:29, 18 January 2007 (UTC)

I just tried putting them back, since the one most in contention is clearly sourced, as above. I was reverted in ten minutes. Frankly, I think more severe action may be warranted. We have no business presenting an article which refuses to acknowledge sourced evidence. Septentrionalis PMAnderson 20:01, 18 January 2007 (UTC)

The problem is that those maps you mention are from the same book that has maps that call it Persian Gulf. And in the texts of those books it is called Persian Gulf. I have 400+ maps that call it Persian Gulf. (Overwhelming evidence). So several maps from the same books that use Persian Gulf in their text, and have maps of Persian alongside Sinus Arabicus is very dubious. And furthermore evidence was brought from neutral western sources that the name Arabian Gulf was recent political change in Arabic countries. That is why you will not find a single evidence in Arabic before the recent era of nationalism that has such a nam. And I am familiar with OR. Intrepretation of a map without proper textual evidence and without secondary sources is OR. So far there is not a single ancient Atlas and text that explicity in the text calls that body of water "Sinus Arabicus". Thus not only is the accusation of OR invalud but furthermore without providing even a single textual evidence, and not mentioning maps from the same book call it Persian Gulf and in their text have Persian Gulf, the otherside is committing OR. --alidoostzadeh 03:44, 19 January 2007 (UTC)
(edited indent) There is no "interpretation" of the maps in question being added to the article. Just stating what name the maps clearly put on the Persian Gulf is not interpretation when the maps are clearly labeled - it is a purelyt descriptive claim. WP:RS seems to support the idea tha descriptive claims of primary sources are permitted. And the fact that the book has maps that go both ways is an even stronger argument to mention both sides, for NPOV. CMummert · talk 03:50, 19 January 2007 (UTC)
One of the problem was that the other side did not acknowledge this. I think the fact that the book has two/three different maps and users the Persian Gulf solely in the text requires mentioning, but furthermore the fact that some maps placed Assyria in Central Persia and placed the red sea in various places (some below horn of Africa, some in the Indian Ocean, some in the red sea of today) and placed the sinus arabicus mainly in the read sea (all the in same text), these facts can be confusing for a entry called Persian Gulf. Thus descriptive claim is not sufficient without mentioning the textual evidence from the same book. I asked the other side to find one and only classic European text that describes the area as Sinus Arabicus. That is not too much to ask. (I know there is not a single source in Arabic for their claim either). The fact is that I brought sources that the name change was recent as well (during 1950s in Arab countries due to anti-Persian Hysteria). Note I only asked for one textual evidence. Yet they only brought a map from the same book that has Persian Gulf in another map and calls it Persian Bay throught the text of the Atlas. Also as I said I have 400+ maps from 14th century to 1925 that I can put in that article, which of course is not necessary, but the ratio will be more than 40+ to 1. But I think a map from the same book should be represenative of what the text in the book mentions. I am going to look at a major reference on saturday, but so far the other side has not brought a single evidence from one text. And if we are talking about representation (like the idea in wikipedia of how many goolge hits are received for a start of an entry), I have access to more than 400+ maps. --alidoostzadeh 04:15, 19 January 2007 (UTC)
Why does the article need 16 maps in the gallery if the only point of the gallery is to illustrate what a single text says? The article could just cite the text once and include a single map as a pretty picture. The only reasonable explanation for the inclusion of the 16 maps is that someone was trying to present evidence for the name "Persian Gulf". In that case, evidence for the other viewpoints should also be included, or all the maps should be removed. I don't see any support in WP:RS for your claim that a primary source must be described somewhere else before a descriptive claim a can be made about it. CMummert · talk 04:23, 19 January 2007 (UTC)
It is not only about a one text. There are numerous texts that refer to Persian Gulf while there is not a single text so far that has been found to call the Persian Gulf as Sinus Arabicus. I have dozens of them right now, and putting aside Greek, Persian, English texts, I have more than 30+ Arabic sources that call it Persian Gulf, whereas there is not a single Arabic source before the era of Nasser and pan-arabist nationalism (1950-1960) whose main proponent in recent time was Saddam that has called it Arabian Gulf. As I said there are more than 400+ maps from the 14th century to the 1924 and I'll be glad to provide texts for many of the maps. But as per evidence, if you think the textual evidence is sufficient, then I have no problem with that eiher. Since there are numerous texts from Strabo to Hondius and etc , while there is not a single text that has called it Sinus Arabicus. (This has been my main point and mentinoing a book with three different maps whose text only uses Persian Bay is not sufficient in my opinion). And if we are including evidence from the other POV, then how come they don't mention that in the same book they are referring to, the area is called Persian Gulf and the text calls it Persian Bay. --alidoostzadeh 04:33, 19 January 2007 (UTC)
I am all for a neutral description of the facts. As far a I can tell: the term "Persion gulf", in various langauges, is the most common, but a few maps use the term "Arabic Gulf", in various languages. CMummert · talk 04:39, 19 January 2007 (UTC)
Well "Arabic Gulf" is not in any text and the book that "Sinus Arabicus" is mentioned also has a map with Persian Gulf and in the text it is Persian Bay. Note I have also found European maps that call the Gulf of Oman as Sea of Basra and Persian Sea, again totallly mis-placed. I also brought the example of Assyria where the Assyrian empire clearly did not encompass central Iran. Thus by such a map, can we go and call central Persia as Assyria? Of course not. According to one good source: As recognized by the United States Board on Geographic names, the name of the body of water that lies between Iran and the Arab states of the Gulf Cooperation Council is the Persian Gulf. For political reasons, Arabs often refer to it as the Arab or Arabian Gulf. [9] --alidoostzadeh 04:46, 19 January 2007 (UTC)

This rule cannot work, does not work and will never work.

All great writing, all good writing is without exception based on originality. If originality is removed from the equation, it is the same as just feeding the subject into some sort of computer search engine and have the computer spit out some bits and bytes and there is Wikipedia. But there are good articles in Wikipedia and without exception, these are created in spite of this rule not because of it. It is not possible to have a good article without originality, it just cannot be done. Throughout history in every case of good writing, the spark has been originality.

On controversial subjects this rule is even more of a disaster. There are all sorts of bad and rude idiotic ideas published and verified, and what is there to refute all this rubbish? It would be far better to let both sides present their strongest argument and let the reader decide the truth. This would insure that both sides of the argument get a fair hearing which is not now the case on Wikipedia.

It must also be remembered that this article is itself completely and totally originally research and its claims that this policy works are not shown by Wikipedia itself. Perhaps the strongest proof that this policy is a failure is that in so many articles the pictures are of far greater quality than the prose. This is very likely because the prose is far more restricted by this policy than the pictures.

I am surprised by the overall low quality of Wikipedia, and it certainly cannot be because of lack of effort. It must be bad policy.01001 03:47, 18 January 2007 (UTC)

WP:SYNT at Persian Gulf naming dispute ?

At the article Persian Gulf naming dispute, someone has gathered a collection of 16 maps that call the Persian Gulf the "Persian Gulf", apparently (?) in an effort to argue that this is the proper name. When maps with the name "Arabian Gulf" were added, they were removed multiple times. The issue was raised by Lambian above, but few people commented on it. It seems to me that the entire article has a WP:SYNT overtone, but especially the maps section. I am wrong? CMummert · talk 03:36, 19 January 2007 (UTC)

In terms of policy then of course you are right. The entire map section is a big SYNT! The maps in question are in no way different from the Chicago Manual of Style from the synthesis example: a source is being used to argue a case in a conflict that the source's authors were not addressing. However, since the analysis of historical maps seems quite relevant, surely there must be reliable secondary sources that have made such analysis, why not base our research on them? --Merzul 16:34, 19 January 2007 (UTC)

SYNT. Yes, it's me again.

I promise this is the last edit I make to a policy page, and I apologize for being so cheeky, but I simply couldn't leave this issue without at least having tried my best to clarify the issue. The reasons for making that change:

  1. Edit reflects consensus: "That would be OR" and the discussion that followed.
  2. Edit is needed: "surely only the first sentence is OR" and discussion.

If this too is reverted, I will gracefully accept it. I'm actually thankful that my stubbornness has been tolerated this far, and will not push it any further, but I am curious why this issue is not clarified on the policy page. --Merzul 17:29, 19 January 2007 (UTC)

That has to be the most childish revert in the history of wikipedia

When I said I would gracefully accept reverts, I didn't expect something as childish as "please don't change the original; it's a real example". Fine, I'm leaving. --Merzul 17:46, 19 January 2007 (UTC)

Before I leave, I just have to admit that my section title has to be the most childish one in the history of wikipedia. I overreacted, I always do! --Merzul 17:58, 19 January 2007 (UTC)

More reasonable question

I fully apologize for my above stunts, but I think I have at least understood the policy myself. Now, I consider the following as original research:

Mr. X, who wrote an article in a particular scholarly journal, was accused of plagiarism when it was discovered that he had copied a list of references from another journal article. Of plagiarism the Detroit Style Guide says that any reference cited must actually be used in the text, whilst the Cleveland Style Guide says that this is an acceptable way to create a working bibliography.

However, I am still unable to say why the following material from a featured article is not original research:

According to the U.S. Forest Service, Shoshone National Forest has the greatest number individual glaciers of any National Forest in the Rocky Mountains. The forest recreation guide lists 16 named and 140 unnamed glaciers within the forest, all in the Wind River Range. Forty-four of these glaciers are in the Fitzpatrick Wilderness, centered around the highest mountain peaks.[23] However, the state water board for Wyoming lists only 63 glaciers for the entire Wind River Range, and this includes areas outside the forest boundaries.[24]

This is synthesizing the sources to make the implicit claim that the forest recreational guide are slightly exaggerating the numbers, yet in some sense this is more acceptable. I can only think that this is because the editors found this more neutral/innocent. Maybe one can also draw the distinction between an analysis at the tertiary level (presenting different secondary sources neutrally) and the secondary level (presenting contradictions in primary sources), but I'm not sure how to define that properly... Maybe since the issue is so difficult, it is better to keep the policy ambiguous. --Merzul 19:16, 23 January 2007 (UTC)

Citing Jimmy Wales

The section on synthesis supports the statement:

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 in order 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.

with the footnote:

...Some who completely understand why Wikipedia ought not create novel theories of physics by citing the results of experiments and so on and synthesizing them into something new, may fail to see how the same thing applies to history.

I think this is misleading. This part of the policy is not directly based on this statement, (or any other statement by Jimmy, to my knowledge) - it was written by Wikipedians themselves, like most parts of policy. Where a statement is cited with a footnote, the reader has the right to assume that the statement is directly based upon the footnote, not a creative interpretation of it. Enchanter 22:41, 19 January 2007 (UTC)

There's nothing misleading about this. The footnote is a footnote, it's not meant to be a reference such as we would use in an article to source an assertion. Jkelly 22:45, 19 January 2007 (UTC)
Yes, a policy is not an external article, it is an internal document intended for WP editors engaged in encyclopedia building. We can use various types of rationale for our policies, including the founder's comments or views, combined with other language that is accepted by community consensus. Crum375 22:57, 19 January 2007 (UTC)
I'll concede that point - you're right that the footnote itself fairly reflects what Jimmy said. I agree the footnote can stay, provided that we don't interpret that as meaning that Jimmy endorses the current version of the policy in some way. This part of the policy goes quite a way beyond anything that he has said, to my knowledge. Enchanter 23:05, 19 January 2007 (UTC)
I don't think Jimbo's words are gospel per se, but OTOH they do serve as a general guideline. I assume that even he himself, like all of us, evolves his own views over time, and WP in general keeps evolving as it grows. I think what we have is a combination of the founder's original vision coupled with ongoing evolving consensus as well as common sense. Crum375 23:24, 19 January 2007 (UTC)
Agreed, although I think the synthesis section in particular needs a bit of work to reflect consensus. At the moment differemt people seem to be coming to different conclusions about what the existing wording means, and it's not altogether clear to me - in particular, I find the plagiarism example very hard to follow. Nor does the section clearly set out the rationale for why the policy is as it is. Enchanter 23:40, 19 January 2007 (UTC)
Personally I agree that the current version can be improved. I would like to see 'synthesis by juxtaposition' listed as a type of OR. I think that if John Doe is sourced as doing act X, and act X is sourced as being illegal in state Y, we can't say "John Doe did act X[1]. Act X is illegal in state Y[2]", unless we have a source that actually combines the 2 facts together; WP cannot be the original combiner of facts by juxtaposition in controversial cases or sensitive BLP issues. So if there is a way to include that, I'm all for it. The plagiarism example adds new verbiage, and by calling the entire paragraph OR we would be weakening the synthesis criterion, IMO. Crum375 00:09, 20 January 2007 (UTC)
Yes, this is what I understood is already the consensus, but even adding something like that was reverted. Now, unlike me, Crum has not behaved like a troll, so at least people should listen to him! ;) --Merzul 00:48, 20 January 2007 (UTC)

Translation into Dutch

I'm currently translating this project page into Dutch for NL-Wikipedia, but I'm sort of stuck... There's one word in Jimmy Wales' text that I just don't know. Google and loads of dictionaries weren't able to help me either, so here I am asking for your assistance.

See Wikipedia:No original research#Neutral point of view (NPOV): "If your viewpoint is held by an extremely small minority, then — whether it's true or not, whether you can prove it or not — it doesn't belong in Wikipedia, except perhaps in some ancilliary article. Wikipedia is not the place for original research."

Does he mean ancilliary or ancillary here? And what exactly does he mean? Don't worry, I do not expect Dutch translations ;) but what I would appreciate is some sort of paraphrase or alternative for this specific word so I know what to use in my Dutch version.


Kind regards, Erik1980 21:04, 23 January 2007 (UTC)
(nl:Gebruiker:Erik1980 -- Dutch talk page)

He means Ancillary, meaning supplementary, subordinate, accompanying. So, for example, we would not cover in detail the viewpoint that the Earth is flat in the artcle on the Earth, but instead include it in the ancillary article flat earth. Enchanter 19:16, 24 January 2007 (UTC)

Thanks for your explanation, I'll add this section to the Dutch version immediately! Erik1980 20:22, 24 January 2007 (UTC)

Attention needed of SYNT experts!

Could someone please take a look at what I believe is a perfect example of unacceptable synth here ?

A user called "Big Brother 1984" has recently come into the article Heaven with a very militant attitude and has concocted a new section on "The atheist view of heaven". The problem is that in order to concoct "the atheist view of heaven", he has synthesized a string of citations for support, that consist of quotes from various literary authors taken completely out of context, such as Edward Gibbon and George Orwell - neither of whom was writing on behalf of atheism in the quotes in question, and neither of whom was ever known to have been an atheist, nor any kind of spokesperson for atheism. As far as I can tell, this synthesis of disparate quotes from non-atheists, who were not talking about atheism or purporting to give an "atheist point of view", in order to create the "atheist view of Heaven" is a completely novel invention of his own. On the talk page, he refuses to acknowledge that it is a novel synthesis - even though he acknowledges that Gibbon and Orwell were each criticising governmental misuse of state-controlled religion, rather than the theological concept of Heaven per se: all secondary analyses agree that Gibbon was really criticising the pagan Roman Empire, while Orwell was criticising Stalinism, in the quotes in question - and certainly not formulating any "atheist doctrine", as he has done with these quotes. He has engaged in a revert war with me now for the past week, and at one point he peppered the rest of the article with gratuitous "fact" tags after I reverted him. Aside from that, I also have issues with undue, disproportionate weight now being given to the "atheist view of heaven" in an article about explaining the Jewish, Christian, Muslim, Hindu and Buddhist, and other concepts of Heaven - it just seems needlessly confrontational. ፈቃደ (ውይይት) 22:31, 23 January 2007 (UTC)

If your description is accurate, then this is just blatant misattribution. Abusing quotations to further a certain case is far worse than SYNT, and should not be done even with a single source. I looked through the policy, and false attribution related problems are not discussed, but I think it is just common sense that one should fairly represent the ideas of a source, not just pick what is suitable. --Merzul 00:00, 25 January 2007 (UTC)


This article's nutshell currently reads: "Articles may not contain any unpublished arguments, ideas, data, or theories; or any unpublished analysis or synthesis of published arguments, ideas, data, or theories that serves to advance a position." This is confusingly written, and I believe it's not a very good nutshell. I tried changing it to "Wikipedia articles may only contain information that's been published in other sources," but that was reverted. I'd like community help on coming up with a better nutshell for this policy. - Brian Kendig 23:49, 25 January 2007 (UTC)

I just posted my suggestion - let's see what others think. Crum375 00:00, 26 January 2007 (UTC)
I support this simplification of the nutshell - the nutshell is meant to be a concise summary, and the remaining detail can be in the rest of the policy page. Enchanter 23:45, 26 January 2007 (UTC)

There is no real substantive difference between stating and advancing a position

The section on synthesis is logically incoherent along with just plain being incoherent. Arguably the first sentence in any article is a synthesis, simply by choosing it as the first sentence. But any sentence after the first sentence must logically be a synthesis that advances a position. For any statement after the first statement can be any number of an infinite number of statements. And any change in any statement will change the meaning of the article to a greater or lesser extent,i.e, will serve to advance a position. Therefore, by this rule the whole of Wikipedia is breaking its own rules.

Of course there must be rules to enforce the veracity of this encyclopedia, but these rules here are written so strictly as to truly be meaningless. The rules as stated here must be broken for any article to be written.

This article here is also breaking the rules for NPOV. Any set of rules must have an upside and a downside, and certainly a discussion as to why a set of rules is chosen over another set of rules must be included in this article. Or somewhere that is clearly linked to this article if this article is to truly have any validity.01001 20:45, 26 January 2007 (UTC)

(a) NPOV applies to articles and not policies' (b) here we are talking about new or unpublished synthesis of sources that are designed to advance a position, not about synthesis, or summaries of sources. ≈ jossi ≈ (talk) 20:57, 26 January 2007 (UTC)
(a)The policy is broken and there needs to be policy on the policy. (b)HUH? A synthesis or summary of sources is advancing a position. Writing an article is advancing a position. Changing the grammer of an article is advancing a position. This is a false distinction. And further this article does not even make a logical distinction.01001 00:22, 27 January 2007 (UTC)
All conclusions or analyses must be quoted from or citable to a reliable source. WP itself advances a position editorially when it effectively becomes a judge or an analyst, which is not acceptable per our neutrality and sourcing rules. Just writing a sentence or changing the grammar is not automatically 'advancing a position' - it has to sound as if WP has a clear bias. Crum375 00:42, 27 January 2007 (UTC)
You are wrong, just writing a sentence is advancing a position. It is beyond me how this very fundumental point is so misunderstood. I do not have the answer but a policy which is completely logically flawed cannot be it.01001 01:17, 27 January 2007 (UTC)
For us, 'advancing a position' means to express a bias. Can you give an example of expressing a bias by writing a sentence without mentioning any facts or leading to any conclusions? Crum375 01:31, 27 January 2007 (UTC)
Allow me to try an example, in this case, Hitler's views on the Jews. If I write, "Hitler viewed the Jews as an inferior people and the source of Germany's problems," I am merely stating a position: that Hitler hated the Jews. But if I were to write, "Jews are inferior," or even, "Hitler saw that the Jews were inferior and had caused Germany's problems," then I am advancing a position, as I am placing that position ahead of other, conflicting positions (e.g. the Jews were innocent scapegoats and Hitler was an asshole) and leading the reader to conclude that Hitler's view was the correct one. Make sense? -- Y|yukichigai (ramble argue check) 02:45, 27 January 2007 (UTC)
Clearly an excellent example for Godwin's Law ;^) Crum375 03:27, 27 January 2007 (UTC)
Ehh, I dunno. Technically it's a "comparison involving Nazis or Hitler", but it isn't comparing anything to Nazis or Hitler. Maybe it falls under Godwin's Addendum. :P -- Y|yukichigai (ramble argue check) 03:37, 27 January 2007 (UTC)
There is a difference between the article itself, which is not permitted to advance an unattributable position, and the activities of the editors, who of course edit the articles to reflect their personal opinions of what is appropriate. It only takes a little thought to see the difference between the two. CMummert · talk 01:50, 27 January 2007 (UTC)

Some friendly advice to 01001: Look, Wikipedia has a body of policies and guidelines that has been fine tuned by hundreds of editors over an extended period of time. Sure, we can always fine tune them a bit more, adjust here and there as necessary, etc., but make no mistake: the framework upon which Wikipedia stands works and has enabled us to create one of the most important encyclopedias in the world, based on that framework that you have chosen to criticize (I have seen your posts at WP:V and other policies.) So, my advice to you is: If you like the way Wikipedia works, great, your contributions are most welcome and while you are at it, your suggestions on how to improve our policies is welcome as well But if you don't, rather than attempt to change the system to your liking, you may want to try other wikis such as Wikiinfo or Citizendium. ≈ jossi ≈ (talk) 03:34, 27 January 2007 (UTC)

Yes, Wikipedia does work and most of it is miraculously wonderful. This has made it very popular and new editors will come and inevitably ask these kinds of questions. I have tried to explain something about synthesis at WP:ATTFAQ#Isn't all of Wikipedia unpublished synthesis?, but it would be nice if someone experienced considered rewriting it. --Merzul 02:13, 28 January 2007 (UTC)
Uh, Wikipedia is not exactly working. Any brief survey of the articles will see all kinds of problems.
Certainly, Wikipedia is needed and now that it is established it would be very hard to replace it with something else altogether even something else with better policy.01001 02:53, 2 February 2007 (UTC)

Picking patterns out of a timetable: OR?

The Eastern Suburbs & Illawarra railway line, Sydney has many scheduled trains per day, which go from stopping only at some more major stations to stopping at all stations. Is it original research to pick some of these as the primary ones and call the others variations, when the timetables don't do anything of the sort? (Eastern Suburbs & Illawarra railway line, Sydney#Stopping patterns) --NE2 07:42, 28 January 2007 (UTC)

Yes we can use English definitions in an normal way. Wjhonson 08:00, 28 January 2007 (UTC)
Can you explain what you mean by that? I did my own "research" with the timetable and was not able to pick out the patterns that are in the article. --NE2 08:05, 28 January 2007 (UTC)
If you using your normal senses cannot determine that the statements are apparent, then they probably are not. However you yourself said that the trains stop at "major" stations and then other stations. Your statement appears to distinguish between types of stop in some fashion. Wjhonson 08:10, 28 January 2007 (UTC)
Here's the schedule of the 182 weekday outbound trains: [10] There are many variants in which trains stop at or don't stop at some stations. All I can really see is which stations are served by the majority of trains. Now it's possible that some of it could be backed up by a footnote: "of the foo trains that run to bar, baz of them run in exactly this pattern", and that could be verified by the reader. But as it stands, I can't verify what's being said there about the stopping patterns. --NE2 08:17, 28 January 2007 (UTC)
Bingo. You said it yourself. You can't verify the statements. Take the above to the Talk page of the article, and remove the statements from the main page, or place a {{fact}} on the statement. Be sure to post to Talk why you're removing or tagging the statement and hopefully the other editors will join you in discussing the issue.Wjhonson 08:22, 28 January 2007 (UTC)
However, if every train stops at Main street and Grand Avenue, you can certainly say that in the article, even if your train schedule doesn't state it in those exact words. It's called simple analysis and we're allowed to do it. If you can verify by trivial inspection that the statement is true, then it's allowed. Wjhonson 08:24, 28 January 2007 (UTC)
I actually placed a [original research?] tag (for the second time; it had been replaced by a cite to the timetable about an hour ago). I also wrote Talk:Eastern Suburbs & Illawarra railway line, Sydney#Original research with stopping patterns. --NE2 08:32, 28 January 2007 (UTC)


I've resently have had a section I wrote deemed original research. I have pics to show the proof of what I wrote. My question is:

Can something that has pics (like of a comic book, TV show screen and along that line) to show proof still be deemed original research?

"THROUGH FIRE, JUSTICE IS SERVED!" 02:38, 29 January 2007 (UTC)

Proof of what? If you want to say that such-and-such picture proves thus-and-so, you need a published reliable source that says that. If you are drawing conclusions about what a picture shows, that is original research. -- Donald Albury 22:34, 30 January 2007 (UTC)
Could the original user post more details about what is being proved? CMummert · talk 22:41, 30 January 2007 (UTC)

its on Tokyo_Mew_Mew#Ribbon_translation_error, I really dont want to explain it, as I'm kinda tired of doing so, and I'm kinda sleepy. In a short, I used the pics to show the difference between the katakana to show there is a difference. I didnt draw any conclusions or whatever.

"THROUGH FIRE, JUSTICE IS SERVED!" 02:53, 1 February 2007 (UTC)

Thanks, that made all the difference. In this situation, the analysis in the article appears to be original research. You can correct this impression by adding additional sources.
Here is the problem: you are using primary sources and your personal knowledge in order to argue that a certain translation is incorrect. But whenever a WP article claims that a reliable source is incorrect, a second reliable source that explicitly says the first is incorrect is required. The purpose behind this policy is to avoid cranks in the sciences who argue that some well-established scientific theory is false because their flawed analysis shows that it is. Convincing them that their analysis is incorrect is impossible, so the analysis itself has been disallowed by policy.
In the article you mentioned, you could probably find a reliable source for the fact that the choon symbol is important, which would help, and you could rephrase the article so that it only says that another translation is possible rather than stating that a particular translation is wrong.
I appreciate that it is unpleasant to know something is wrong but be unable to say so explicitly. The problem is that in many other cases people erroneously believe things are wrong, and the policy to catch them also catches you. CMummert · talk 03:07, 1 February 2007 (UTC)

I'm not sure if I should say it is possibly correct as opposed to say its right is A.) The other mangas shows it "reborn", such as B.) asked my Japanese teacher just to be sure, as she a native speaker C.) Tokyopop is knowns for its typos, translation errors n such, so that kinda destroys their creditiblity D.) MANY vandals have tried to delete it, and I'm afraid once they see I said is "possibly the correct translation", ti will probaly lead them to vandalze, such as POV statements, blanking it, etc.

Getting sources is what I can do for sure. Thanks for the advice! "THROUGH FIRE, JUSTICE IS SERVED!" 23:17, 1 February 2007 (UTC)

The IAL population hydra

Hi - I'm curious what people here think of the discussion that has begun on talk:Ido. A user has decided to enforce policy there but since population estimates are so entrenched that simply enforcing WP:NOR won't do anything for the other pages on other IALs, both in English and pretty much every other language. Adding to the problem is a lot of mutual intelligibility between IALs - thanks to learning Ido I can read and write Esperanto, Novial, etc. and for all I know might count as a speaker of those as well depending on the definition. Making it even more difficult is the political nature of the movement - often 'number of speakers' means 'number of supporters' because people will choose to only use the language they feel to be the best even if they speak and understand another one. What to do? Mithridates 03:03, 29 January 2007 (UTC)

If the estimates are cited from published, reliable sources, they can be used. If they aren't they can't be used. If there are published, reliable sources analyzing why the estimates are questionable, they can also be used. Any unsourced analysis would be OR. -- Donald Albury 22:37, 30 January 2007 (UTC)

2nd Person

Some 2nd person "you" references were just cleaned-up from the WP:CCC text merged into WP:CON because they stuck out like a sore thumb after the merge. This policy article is written almost entirely in the 2nd person, almost like a how-to. Does anyone else think this is inappropriate? Dhaluza 03:06, 31 January 2007 (UTC)

One case of it stuck out at me just now as I was reviewing this policy... I rephrased it ever so gently. Personally I'm of two minds about it -- I think it's good to address the reader directly when we are providing specific instructions to them, but in cases where we're merely describing policies and their rationales, addressing "you" don't seem right. Using words like "we" and "our" are good, though; it promotes a sense of inclusiveness and community, which is a great thing to reinforce with our fellow editors, especially newer ones. -/- Warren 06:24, 4 February 2007 (UTC)
Using the 2nd person is considered informal, and I'm not sure this is appropriate for a policy. It is definitely inappropriate for other encyclopedic content like articles. I considered your point about community as well, but I think the bigger issue is that this sets a bad example, which people may then repeat in articles. Dhaluza 12:05, 4 February 2007 (UTC)

Reversed audio clips

Can I upload a short clip of backwards messages found in songs in the List of backmasked messages? The clips would, of course, be reversed versions of the original sections. And the existence of the backwards messages would be documented by reliable sources.

My tenative theory is that, since audio seems to generally fall under the same standards as images (for example, {{Music sample}} is in Category:Non-free image copyright tags), a short reversed sample would be similar to an image in being exempt from the Original Research policy. Of course, it would still have to meet the restrictions at Wikipedia:Music samples. Λυδαcιτγ 20:04, 31 January 2007 (UTC)

Barring any objections, User:Bearingbreaker92 and I are going to start uploading reversed clips. Λυδαcιτγ 02:16, 6 February 2007 (UTC)

Do not confuse the success of Wikipedia with the success of this policy

Certainly there is a place for an internet encyclopedia that we can all contribute. Wikipedia has filled that void and has now achieved a monopoly position there. The success of Wikipedia does not validate the original research policy. What would happen if the OR restriction was removed entirely?

Clearly, as pointed out previously the OR policy is logically invalid. It is hard to believe that Wikipedia needs to rely on a logically absurd policy. How could the policy be changed to be logically valid?

How is one to refute verified nonsense without OR?

Lastly, if we have a physics article where we have two sides of a right triangle can we infer the length of the third side?01001 04:56, 1 February 2007 (UTC)

First it's not clear that the OR policy is logically invalid. You yourself miscontrue the issue by then advancing the claim that it's logically "absurd", which makes no sense in logic. Second what exactly is "verified nonsense", I think you're trying to push a POV here. Lastly, what exactly does *physics* have to do with *geometry*? Nothing much. That's all. Wjhonson 07:52, 1 February 2007 (UTC)
I don't think it's clearly logically invalid. And how do we show that something is "verified nonsense"? Since when was it Wikipedia's job to refute it? If the OR restriction was removed entirely, Wikipedia would be used as a dumping ground for new theories, personal vendettas and all manner of rubbish, essentially becoming useless as an encyclopaedia. Trebor 07:55, 1 February 2007 (UTC)


The nutshell had an unfortunate example of a double negative, so I reworked it into a bulleted list to try to make it more clear. The IP edit source was an accident since I did not realize my login timed out. Dhaluza 11:54, 2 February 2007 (UTC)

Synthesis of published material -- request for clarification

I am trying to establish the limits of the rules about synthesis of published material. This is relevant to discussions going on at Talk:Controversy over criticism of Quebec society.

I'll use a fabricated example which may seem stupid, but definitive answers to which would help me understand the limits of the policy: There are articles on Wikipedia about each of the Canadian provinces, each reporting its (properly sourced) surface area. A user reads all these reports and concludes that Quebec is the largest province. If he adds a sentence to the article about Quebec stating that it is the largest province, but without providing a source, is that original research? To me, the plagiarism example clearly implies that it would be. John FitzGerald 18:35, 2 February 2007 (UTC)

If Quebec is the largest province, it would be easy to find a source to that effect. If you couldn't find a source who said exactly what you're saying, it would be OR, yes. OR is material for which no source can be found, not something for which no source has been offered. SlimVirgin (talk) 18:45, 2 February 2007 (UTC)
I find that a ludicrous idea. Think about it: if we could only state things that had specifically been stated by another source in the specific way we wanted to present the information then Wikipedia would be nothing but abject plagarism. All of Wikipedia is a compilation of information from multiple sources -- a "synthesis" of information, if you will. It is only OR when that synthesis promotes a certain position that cannot be sourced to a reliable source. In this case if a source has listed all of the province sizes in Canada and Quebec's listed size is the largest, then saying "Quebec is the largest province (according to source X)" isn't OR, it's a rewording of the information cited. It's basic math and a restatement of fact, not an advancement of position. -- Y|yukichigai (ramble argue check) 19:53, 9 February 2007 (UTC)
Another issue to consider here is controversy or 'fuzziness'. The surface area of provinces is a mathematical or geographic attribute which is clear, well-defined and non-controversial. So in this case, even without a source that directly includes the comparison, IMO if we are given reliable sourcing for each province's area, we may arrange them from largest to smallest as WP editors. As soon as the attribute we compare becomes fuzzy (e.g. most parks), or the issue becomes contentious (e.g. highest crime rate), we need to tread more carefully and rely on a source that will do the comparison for us. Other opinions welcome. Crum375 20:56, 2 February 2007 (UTC)

Thanks for the prompt answer. I'm afraid I find these issues rather thorny, so i have follow-ups. First of all, I'm not saying a source couldn't be found; I want to know if one need be found. If the criterion is being "clear, well defined, and non-controversial," I do not understand what is unclear, poorly defined, or controversial about the statement:

"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."

This statement is described in the article as original research. The statement seems abundantly clear and well defined to me, and any possible controversy can be resolved by consulting the Chicago Manual of Style. Furthermore, a secondary source is no more likely to be accurate about this issue than is a group of Wikipedia users who consult the manual.

I'm not trying to be difficult about this. I already am grateful for your statement of these criteria. I also understand perfectly your examples of other issues that would involve people in original research, and agree with you about them. John FitzGerald 21:19, 2 February 2007 (UTC)

I disagree with the assertion that you would need a source which said exactly that "Quebec is the largest province." If a person of average intellect can surmise that, among the 13 provinces of Canada, that Quebec has the largest land area, then we can state that. That is, simple assertions of fact do not require a source. This is the "We do not need to source that the Sky is Blue" rule. If we only parroted what each source said, we wouldn't be synthecizing, and our articles would be just a series of disjointed quotes. That doesn't make for smooth writing. What we actually do, in my opinion, is mostly indirect quoting, not direct quoting. That is, we paraphrase what our sources say. What we shouldn't do is make them say something quite skewed from what they do say. Controversial points need direct quotes with specific citations, but mundane points do not. Wjhonson 07:17, 3 February 2007 (UTC)

Since Canada has 10 provinces, not 13, the dangers of original research may have been made apparent by your post; I'll assume it's a typo, although people sometimes think Nunavut, NWT, and YT are provinces. In general, though, I agree with you. However, as I read it, the policy doesn't. Comparing sizes is inference. My concern is that the policy is too restrictive as written, and it is being interpreted too restrictively by some people (the discussion page i have already linked to provides examples). It also seems to me that your nightmare of articles consisting of "series of disjointed quotes" is becoming real. It seems to me the limits of this policy need to be specified more clearly. John FitzGerald 16:37, 4 February 2007 (UTC)

To put this another way, if comparing sizes is not considered inference, I don't see why comparing specific claim about behaviour to a definition in the Chicago Manual of Style is considered inference. This is the crux of much of the debate at Talk:Controversy over criticism of Quebec society. John FitzGerald 00:36, 6 February 2007 (UTC)
There are two differences. The first is what sort of judgments are implicit in the statement. If a single source gives areas for all ten provinces, identification of the largest province depends only upon fundamental facts about sets and positive rational numbers that are not subject to reasonable dispute. On the other hand, the statement about the Chicago Manual of Style's definition of plagiarism depends upon a judgment that the particular definition is authoritative in the circumstances. That is far from clear in the example, so we need someone with authority to make the connection. The second difference is the degree of expertise needed to make the comparison. One could reasonably ask fourth-graders, "Which province has the largest area?" On the other hand, whether a particular fact pattern conforms to a detailed set of rules is a complicated judgment best left to experts, or at least specially-empaneled laypersons. Robert A.West (Talk) 01:17, 6 February 2007 (UTC)
  • Note that I give the stricture that a single source gives areas for all ten provinces. Areas as found in distinct sources may be computed using incompatible methodologies, so we should avoid making inferences based on assumptions that may not be true. Robert A.West (Talk) 01:23, 6 February 2007 (UTC)

Nut Wikipedia is not written by or for fourth-graders. The only judgment required in the example – it seems to me – is whether a citation was provided or not. Obviously it doesn't seem like that to you. As I've said, the problem is that the policy needs to be clearer. As it stands, i would say the great majority of Wikipedia's articles constitute original research according to this policy.

One idea I have had – which well might not work, which is why I'm mentioning it here – is to let inferences like this stand until either disproven or demonstrated by discussion on the talk page to be questionable. My experience here suggests to me that people will agree to withdraw a questionable inference. John FitzGerald 13:41, 7 February 2007 (UTC)

An inference or conjecture vs. a simple summary of quoted text is a matter of degree. We also have the issue of juxtaposition, which IMO can be an even more subtle OR and/or expression of bias by WP. In non-controversial cases, we can afford to use a little more elaboration and analysis of the sources, whereas when we touch controversial topics, and even more so when a living person is involved, we try to cut the OR and juxtaposition to near-zero. If any reasonable editor spots a possible OR or bias, s/he may remove it, and it becomes the burden of the others to find a language that properly reflects the available sources (with due weight) and nothing more, while still being legible. Ideally we'd like to have secondary sources as much as possible, where the primary sources are already analyzed and compared, leaving us a much easier job as a tertiary source. Unfortunately we often just have the lower level primary sources and we have to make the best of the situation by following our WP:NPOV, WP:V and WP:NOR policies. Crum375 14:10, 7 February 2007 (UTC)
Thanks. That clarifies all the points made here. It also allows me to pst a qyery further down about an actual dispute going on right now. John FitzGerald 18:04, 9 February 2007 (UTC)


Guys I need your opinions urgently about NOR on an article Languages of Iran; as you will see from the history of the article and its talk page I have been told that we cannot take a calculator and add up the figures on this article and demonstrate the discrepancies because that is counted as Original research! Is this true? Is adding up the figures on a list and showing their discrepancies counted as original Research? Kiumars

First of all, some people will speak several languages, so one might expect the total to exceed the population. In the UK, for example, this would certainly be the case. But adding up figures itself is not too difficult to qualify as OR - and probably less work than finding a refernce in a library for most readers. So I suspect the problem is the spin you are putting on the answer you get. Stephen B Streater 09:37, 3 February 2007 (UTC)

I think it's inappropriate to label this OR, and especially confusing for new users to Wikipedia.

As far as I can see, the background to this is that someone noticed that the number of people in a table of speakers of different languages in Iran added up to greater than the population of Iran. This kind of research is to be encouraged - cross checking facts and figures with other sources to check that they make sense.

In my view, an appropriate response to this editor would be to look into this comment, and say something like "Thank you for pointing out this discrepancy. However, looking at the source, there's a good reason for it - the figures cover all speakers of each language, so people speaking more than one language are covered more than once." It might also be appropriate to clarify the article.

However, the actual response to this editor was to label his observation as OR. I don't think this was justified; he was doing straightforward research and fact checking, drawing no novel or contentious conclusions, and yet he was accused of violating the NOR policy, which has always been primarily aimed at keeping novel unpublished ideas out of Wikipedia.

As well as strongly discouraging original research, if we are to create a reliable encyclopedia we need to strongly encourage desirable, non-novel, unoriginal research, such as cross checking articles against multiple sources, carefully checking for reasonableness and consistency, and raising observations when facts and figures don't appear to stack up. For this reason, I think we should be very careful not to apply the NOR policy in a way that might discourage this sort of desirable research, and not to accuse editors of violating the NOR policy without good reason. Enchanter 12:33, 3 February 2007 (UTC)

A very well reasoned approach. Unfortunately, the OR rules as written preclude any such rational methods.01001 02:50, 16 February 2007 (UTC)
WP:ATT differs on this point. Septentrionalis PMAnderson 05:59, 16 February 2007 (UTC)

Where to put it

The second paragraph of the article says: "Wikipedia is not the place for original research." I think, you should, after that, include a link to one of the Wiki projects that is the place for OR. Mausy5043 12:35, 3 February 2007 (UTC)

Wikinfo springs to mind. Stephen B Streater 16:47, 4 February 2007 (UTC)
I just read up a bit on Wikinfo but it seems their fork was in order to circumvent WP:NPOV not WP:NOR. Wjhonson 00:44, 5 February 2007 (UTC)
I thought, more generally, the fork was to avoid heated arguments about what should be in articles. Stephen B Streater 10:47, 5 February 2007 (UTC)
I think that's more a factor of how many editors are present :) Wjhonson 17:07, 5 February 2007 (UTC)
I've checked on the Wikinfo website and they say they welcome original research. Stephen B Streater 17:45, 6 February 2007 (UTC)
Then I propose to add something in the article like "Wikipedia is not the place for original research. Other projects exist (like Wikinfo) that are more suitable for that." Pity though there is no Wikimedia project that covers that. It would be easier to link to and from it and the Commons. Mausy5043 17:07, 7 February 2007 (UTC)
But there is indeed such a Wikimedia project where research is welcomed, as I've mentioned before: v: ፈቃደ (ውይይት) 17:23, 7 February 2007 (UTC)
Wikiversity is for writing *textbooks*. What if I want to write a cookbook, or a novel, or an tour guide? Wjhonson 17:52, 7 February 2007 (UTC)
Wikibooks cookbook? [11] --Milo H Minderbinder 18:54, 7 February 2007 (UTC)

Additional opinions wanted - Angie's List

At Angie's List there's some info in the article that is unsourced speculation about the company, specifically numbers that are assumed from other numbers, numbers based on things like minimum wage and postal rates, and similar things. Also removes citation needed tags, insisting they aren't needed. Seems like blatant OR to me, but I don't want to get into a one on one situation with this editor (so far I've left most of it with citation tags, but I suspect sources don't exist for most and the claims will be deleted). Outside input would be helpful - the article is locked right now but explanation from someone else on the talk page would be helpful. Thanks. --Milo H Minderbinder 22:48, 6 February 2007 (UTC)

Photo manipulation

Just a note that I've edited the bit on images to reflect that manipulated images are not per se required to be listed on Images for Deletion. Only when the manipulation materially affects the encyclopedic value of the image should we be concerned. If there's a photo of a flower for example, it's normally considered acceptable (according to WP community consensus) to do things like alter saturation levels, exposure timing, and perhaps even artificially blur out the background to simulate shallow depth of field. On the other hand, it would be unacceptable to do things like fundamentally change the color of the flower (red -> yellow, etc.), remove or add important parts of the flower, and other things that would improperly represent the "reality" of the flower. There have been many lengthy discussions about the appropriateness of various and specific manipulations at WP:FPC, and anyone with questions may wish to look there. I trust that my addition to the WP:OR page is uncontroversial, but please let me know if anyone has a problem with it. --Dante Alighieri | Talk 22:23, 8 February 2007 (UTC)

How does Image:Xenu space plane.jpg fit in? --NE2 09:36, 9 February 2007 (UTC)
Perfectly well. The magic words, in this case, are "artist's impression". —Ilmari Karonen (talk) 15:24, 9 February 2007 (UTC)

Question about an audio sample

Hello, I have found an audio sample at the Electronic voice phenomenon article. This sample, which is supposed to be of paranormal electronic voices, was recorded by two people who then self-published it on their website. According to the Original Research policy page, I think they are not a reliable source, and it is original research?-MsHyde 00:53, 9 February 2007 (UTC)

The Original Research prohibitions apply to *you* as an editor, not to your sources. That is, *you* should not do original research. All primary sources, by their nature, do original research. Otherwise they'd be secondary sources. Wjhonson 07:14, 9 February 2007 (UTC)
"Original research (OR) is a term used in Wikipedia to refer to material that has not been published by a reliable source. It includes unpublished facts, arguments, concepts, statements, or theories, or any unpublished analysis or synthesis of published material that appears to advance a position — or which, in the words of Wikipedia's co-founder Jimmy Wales, would amount to a "novel narrative or historical interpretation."
So, if somebody self-publishes noise on their website, and says it is voices of the dead or extraterrestrials, they are a reliable source and it is not original research?-MsHyde 07:26, 9 February 2007 (UTC)
When their organization has 970,000 Google hits, then yes. Your opinion of the work is original research. The work itself is not. – Someguy0830 (T | C) 07:46, 9 February 2007 (UTC)
The three core policies of wikipedia are verifiability, no original research and neutral point of view. Being a reliable source is a *part* of verifiability, but not it's whole. No Original Research is a completely seperate policy. They are not related in any way, except that you must follow them all. I think you are confusing them with each other. You can be a reliable source and not do original research. You can do original research and not be a reliable source. You can do original research AND be a reliable source. I.E. you can have every combination. Having said that, you have to test if the website can pass the requirements of WP:V. Wjhonson 08:02, 9 February 2007 (UTC)
Wjohnson, I think you should take a look at this website. Also, this is web self-publishing--not a reliable third party publication. Again, they have self-published noise which they claim is the voices of the dead or extraterrestrials. It is absurd.-MsHyde 08:27, 9 February 2007 (UTC)
Again, your opinion, and you're misrepresenting how it's being used. The audio clip in question is used as an audible example of the EVP phenomenon. It is not being used as a source of any type, just as an example. It comes from a primary source, which is verfied by a secondary source within its text. – Someguy0830 (T | C) 08:34, 9 February 2007 (UTC)
I would think an article about EVP would need to have an example. I would think that without an example the article would be less useful. Wjhonson 08:38, 9 February 2007 (UTC)
The question is who is providing the example, and are they reliable, or is it original research. Also, isn't it a synthetic argument that the noise is noise of anything in particular. And it is being used to advance a position. The authors make money from EVP. How do we know it is not a recording of their dishwasher or vacuum?-MsHyde 08:44, 9 February 2007 (UTC)
Note also this: 08:47, 9 February 2007 (UTC)
The policy against original research does not apply to the sources we use. It applies to us only. To wikipedians. Us as editors. It does not apply to our sources. Wjhonson 08:50, 9 February 2007 (UTC)
That doesn't make any sense--you have just said the three policies work together. Also, the first line of the policy states that original research is published by unreliable sources. A self-published website with a conflict of interest is not a reliable source.-MsHyde 09:00, 9 February 2007 (UTC)
Note that the one concern in the link MsHyde provided is no longer a concern. I dealt with it. – Someguy0830 (T | C) 08:51, 9 February 2007 (UTC)
I do not think so. Also, do not remove the accuracy dispute and NPOV tags again.-MsHyde 09:00, 9 February 2007 (UTC)
What you think and what is are two very different things. Also, how about you stop adding them? I don't intend to sit by and watch you misuse tags to fit your POV. You've made your position on this article clear, and I for one do not intend to let you enforce it. – Someguy0830 (T | C) 09:03, 9 February 2007 (UTC)

MsHyde I think you're confusing editors with sources. You and I are editors. The Weekly World News, and that website you're referring to are sources. The policy against original research does not apply to sources. Zero, nunca, zilch, nada, nothing. Not applicable. The policy that prohibits original research *only* applies to MsHyde and Wjhonson and others like us here in wikipedia land. WE cannot do original research. MsHyde cannot go do an EVP and post it. Some other people, let's say John Brown, who has his own website can do original research all day long and we can cite to it under certain circumstances. That is why you need to read WP:V. The first line of what policy? If it says "original research is published by unreliable sources" then I need to fix it, because that makes no sense at all. Original research can be published by reliable sources as well. When the Wall Street Journal writes an article that uses new ideas, thats original research and yet they are a reliable source. I hope the distinction is a little more clear. Wjhonson 09:08, 9 February 2007 (UTC)

As I quoted above, the first line is: "Original research (OR) is a term used in Wikipedia to refer to material that has not been published by a reliable source."-MsHyde 09:20, 9 February 2007 (UTC)
External sources can do OR, so you can't exclude it on an OR basis. But you can certainly argue that the website isn't a reliable source and exclude it on those grounds. However, in an article about such a fringe topic, it might be acceptable to just leave it and make sure the text isn't "this is a recording of EVP" but "this is a recording that X claims is EVP". --Milo H Minderbinder 14:12, 9 February 2007 (UTC)
But see MsHyde that's quite a different statement. At first you said "original research is published by unreliable sources", secondly you said "Original research (OR) ...refer[s] to material that has not been published by a reliable source". Those two statements are quite different. OR has nothing to do with whether something is reliable or not. The two concepts are not related. Original Research only refers to *new* ideas (etc) whether published or not, whether by a reliable source or not. Reliable source only refers to the source whether doing original research or not. The two concepts are quite disjoint, so it seems we're going to have to clear up that quote to make that more clear. Wjhonson 17:26, 9 February 2007 (UTC)

Synthesis yet again

The "synthesis" section seems to be becoming contentious in relation to scientific/medical articles.

Suppose there's a situation where someone has published a medical/scientific claim that makes no sense in mainstream scientific terms (for instance, it's wildly in breach of thermodynamics or is just a meaningless usage of scientific terminology) - but no mainstream critic has yet published comment/refutal.

To what extent, if any, can one place context on the claim?

As things stand, it seems the "no synthesis" guideline is allowing non-mainstream views to be presented non-neutrally and with undue weight, since some editors are arguing that any kind of broader context counts as OR unless a specific counter-citation can be found.

Any thoughts? Tearlach 13:17, 9 February 2007 (UTC)

Uhm... NPOV and Undue Weight should of course apply, but I'm afraid we shouldn't engage in original debunking. If you have doubts about the validity of a statement published in a reliable journal then you can use prose attributions like "Mr. Crackpot claims in an article published in Journal of Lax Standards that ..."; but again Undue weight still applies and clearly we should not include any information from dubious sources that would contradict information from peer-reviewed publications. --Merzul 13:52, 9 February 2007 (UTC)
We are dealing with exactly this situation in BDORT, which has just undergone Arbitration. There we have a situation that the main topic gets an article because notability was established because of a Medical Tribunal malpractice case, that punished a doctor who used an unproven technique. One issue that we had to deal with is that although the Tribunal reviewed one technique, it didn't review all of the several related ones. In our case they reviewed 'BDORT' but not 'Solar Papers' (for example). So we have no mainstream review for Solar Papers, yet we need to present it, since its notability is established indirectly via its association with BDORT via the same inventor and many of the same writings or seminars. The way we handled it is by making a statement that explains that there is no known mainstream review of this or other claims. This way, if any editor (or presumably proponent of the technique) can ever find an acceptable source showing mainstream review, it can be easily added. Conversely, if no such source is found, the statement remains valid. This helps us strike a balance between the somewhat conflicting requirements (in this case) of WP:NOR, WP:V and undue_weight. Crum375 14:16, 9 February 2007 (UTC)
Thanks. Specific examples I have in mind are Gillian McKeith and Mucoid plaque.
The former: GMcK has made statements such as "food contains vibrational charges", and some editors are saying we can't indicate this term to be scientifically meaningless unless some published statement to this effect can be found.
Ditto with the latter: the idea has acquired notability through a deal of publicity, enabling a lot of justifying material to be cited from the originator's website. Yet it has little specific published criticism, and it appears editors can't cite the basic digestive biology, and properties of products used, necessary to put the claims in context. Tearlach 15:31, 9 February 2007 (UTC)
I would naively prefer articles on crackpots to just cover the attention they have in popular media, and completely refrain from even including any scientific details that are not published in academic journals. Since that obviously can't be done, I just have to admire your diligent fight against pseudo-scientific nonsense with whatever solutions you have found. I admire the patience that is required to improve quality and remain civil, so as to think up things like Crum has suggested. Just keep up the good work, both of you! --Merzul 16:44, 9 February 2007 (UTC)
Thanks for the kind words, Merzul, they are very much appreciated :) Crum375 20:58, 9 February 2007 (UTC)

Reliable Source

As MsHyde correctly pointed out (see abpve), the intro confuses the concepts of reliable source (part of verifiability) with original research. The two concepts are quite disjoint. A reliable source may do original research, they may also not be doing original research. As well an unreliable source may do original research, and also may not do it. All four cases are possibilities. In a truth table, they are independent variables, free to move without restraint from each other. We should not, right here on the main page, be hopelessly confusing them with each other. I have reworded the offending section to remove the issue of reliability. All questions of reliability should be place on the appropriate policy page, which isn't this one. I'm open to discussion about the matter should others disagree. Wjhonson 17:34, 9 February 2007 (UTC)

Controversy over criticism of Quebec society

We're having a disagreement at Talk:Controversy over criticism of Quebec society about original research. I think to a large extent it's an argument at cross-purposes, so I will be encouraging the other people involved to post their views of the dispute here. Anyway, I raised a question about a section of the article dealing with the city of Westmount's opposition to a policy of the Quebec provincial government. This opposition was characterized by some politicians as Quebec-bashing and colonialism. My point is that it is not clear that Westmount's action constituted criticism of Quebec society, and that therefore any unsourced assumption that it does constitutes original research. Others seem to think (and I'll leave it to them to clarify whether I'm stating their argument correctly) that as long as a source can be provided showing that the accusation was made that Westmount's actions constitute criticism of Quebec society. To me that sounds like letting the accusers serve as their own verification. I think Stephen Harper wears a hairpiece, but I can't add that assertion to the article about him and then say it's verified by my claiming it's true. John FitzGerald 18:14, 9 February 2007 (UTC)

I think that it has to do with notability. Those who characterized the opposition as Quebec-bashing were high ranking officials in the Quebec government. Their opinions carry more weight than your average person. Peregrine981 04:41, 15 February 2007 (UTC)

They're also interested people, though. I don't believe they constitute reliable sources. However, no one who is not associated with the article seems interested in commenting, so we'll probably never know. John FitzGerald 13:47, 19 February 2007 (UTC)

Can we say that two conflicting sources conflict?

Can we point out that two sources are in conflict with each other?

Let's suppose I'm in the year 2050 writing an article on the life of one Avraham Ben Yosef. Having found a reliable Source A, I claim:

"On Feb. 10, 2007, Ben Yosef made a query on the WP:NOR Talk page (Source A)."

Some time later, however, I find an equally reliable Source B that states: "Ben Yosef died on February 9, 2007, from injuries sustained in the Cuban-Indonesian war."

Can I cite both sources and point out that they conflict? For instance:

On Feb. 10, 2007, according to Source A, Ben Yosef made a query on the WP:NOR Talk page.
(In apparent conflict with this account, however, Source B reported that "Ben Yosef died on February 9, 2007, from injuries sustained in the Cuban-Indonesian war.")

Is that possible? Or would it be OR to point out the contradiction? --Abenyosef 18:59, 10 February 2007 (UTC)

Although as a rule we should always seek out the best possible sources, in terms of both basic reliability and interpretation of facts, sometimes we are left, as you say, with conflicting sources that we have a duty (IMO) to identify and explain to the reader. I would always approach such a situation with utmost care - our goal as neutral observers is to minimize any commentary or interpretation, to absolute zero wherever possible. Two important considerations are controvery and BLP. In both cases we need to be even more careful than usual, trying our very best to both minimize editorial verbiage as well as to be as neutral in tone as possible. In the example you bring, it sounds to me as if there is not that much controversy, and the person is apparently deceased, hence it is not as critical as others, and your suggested wording sounds reasonable to me. I am ignoring the issue that in your example, making an edit on WP only indicates that someone knows your password, hardly a proof of being alive, and that WP itself can never be used as a source, so I'm just assuming a generic conflict of otherwise acceptable sources. Crum375 19:22, 10 February 2007 (UTC)
OK, here's the actual case:
Another position is that criticism of Israel or Zionism is not in itself antisemitic, but that anti-Zionism can be used to hide antisemitism. Dr. Martin Luther King Jr., in response to a question from the audience after a lecture at Harvard University shortly before his death in 1968, is reported to have said:
"When people criticize Zionists they mean Jews; you are talking anti-Semitism."[2][3]
(In apparent conflict with this account, however, the Harvard Crimson reported, on the day of King's death, that he "was last in Cambridge" on "April 23, 1967". [4]
My contribution would be the text in boldface. Does it amount to original research?--Abenyosef 19:56, 10 February 2007 (UTC)
Doesn't look like original research to me. You are faithfully reporting what the underlying sources say, and pointing out the apparent contradiction in what you report to the reader. NOR is a policy intended to stop you drawing new, novel conclusions from the sources, and is not intended to prevent you from making straightforward, unoriginal observations in line with the source material. Enchanter 23:45, 10 February 2007 (UTC)
Am I missing something or does the source not actually conflict? It states that MLK said this shortly before his death in 1968. He didn't say it in 1968, his death was in 1968. He said it "shortly before his death" so 1967 might qualify as "shortly". Wjhonson 00:33, 11 February 2007 (UTC)
IMO this does smack of NOR or POV. Because the dates are not really in conflict, 'shortly' is not a well defined time period, as Wjhonson notes. This is certainly not a good example for clearly conflicting sources. And someone reading this can easily form an impression that WP is trying to convey to the reader that maybe MLK didn't actually make that statement. Crum375 00:42, 11 February 2007 (UTC)
Well, that would be covered by the word "apparently," but still is a reasonable point. No precise date is given for the historical event analyzed, so one can't stick to one interpretation of the word "shortly," and the sources are not unquestionably conflicting. That would raise the point of whether ambiguous sources are reliable, but that's another story.
However, I don't agree with your point that another objection would be the impression a reader can form. So long as you don't write the conclusion into the article, it's not O.R.
If the sources actually were conflicting, some readers would think the Crimson is inaccurate, and some others would think MLK never made the statement, and it would be OK so long as Wikipedia itself does not draw a conclusion from the contradiction. --Abenyosef 12:50, 11 February 2007 (UTC)
I think our NOR policy is actually a little tighter. In general, if we are making a controversial statement, even indirectly (e.g. that the MLK speech did not occur), we as WP should not be the first to introduce it. IOW, we must rely on a reliable published source to make that interpretation or conjecture for us first. In this case, assuming we had a clear conflict, we would have to cite some other reliable source saying, for example, "MLK said X in speech in Cambridge according to Y, but according to Z he was not there on that date[1]" (where [1] is an existing reliable source). Because this is controversial and sensitive, we as WP cannot be the party that actually brings together the two sources Y and Z, which would infer possible doubt on the existence of that speech. If we did, it would be an example of synthesis, which is a subset of OR. Using the word 'apparently' does not change the fact that we are presenting a novel conclusion (MLK possibly did not make that speech) which was not published elsewhere by a reliable source. This also relates to WP being a tertiary source, which means it is a 'summarizer' of existing published information, not a creator of new information, interpretations or conclusions. Thanks, Crum375 13:34, 11 February 2007 (UTC)
Wikipedia is not a creator of new information -- but my point is that there is a difference between "saying something new" and "saying something for the first time." If you point out for the first time the logical relationship between two propositions, that's not new information (the realtionship is inherent to the two propositions). What you can't do in Wikipedia is state those two propositions without proper sources.
Now a point could be made that stating an evident logical relationship is itself a proposition and, thus, new information. Enter another Wikipedia requirement: writing should not be awkward. If I said in a Wikipedia article:
"Ben Yosef made a query on the WP:NOR Talk page on Feb. 10, 2007 (Source A). Ben Yosef was whipped to death with a string of sausages in Hamburg on Feb. 9, 2007 (Source B)."
That wouldn't violate NOR, whatever the previous conventional wisdom may have been. I'm giving existing information from two reliable sources. However, it would sound quite awkward. The reader would be left asking himself, "don't these two statements conflict with each other"? And in this case, I believe it would be reasonable to point out the conflict, even if it may sound like O.R. to those sticking to the most rigid interpretation of the concept. --Abenyosef 21:26, 11 February 2007 (UTC)
I agree that if you have two reliable sources, both directly related to the subject, that are in clear conflict, we may need some minimal 'glue logic' to explain the conflict. But there are caveats: for example, if you said: "Ben Yoseph claims to have received his Law degree at school X in 1994[1], but school X only started its Law program in 1996[2]", that IMO would be WP:OR, unless we can find some other reliable source to make the connection between sources 1 and 2. In my example, only one source directly applies to Ben Yoseph, the other is an indirect connection, and hence our statement would be synthesis, i.e. unacceptable for WP itself to make. Crum375 21:39, 11 February 2007 (UTC)
But then again, where's the synthesis in pointing out a conflict? It would be a synthesis to claim: "Therefore, either Ben Yosef or School X are lying about themselves." But it's not synthetical to claim "According to his autobiography, Ben Yoseph received his Law degree at school X in 1994. (In apparent conflict with that account, however, X's web site claims that the school started its Law program in 1996.)" A synthesis of concepts, in my view, necessarily implies drawing a conclusion about any other thing that the concepts themselves.--Abenyosef 00:22, 12 February 2007 (UTC)

(outdent) The OR or synthesis in this example is in creating a novel inference in a contentious issue, that was not previously published, whereas WP is supposed to simply summarize reviews by existing sources. In this case, the novel inference is that perhaps Ben Yoseph is lying regarding his education. Nobody has previously noticed that point (and published it), it is solely WP's diligent editors who are putting 2 and 2 together, and in contentious/controversial and especially BLP related issues, this type of inference would be considered OR by synthesis and is hence not acceptable. We would still be able to include source #1, but including source #2 at that point would constitute OR. Crum375 00:34, 12 February 2007 (UTC)

As we say in Spanish, you're threading thin -- but I think I can understand your point, and, more importantly, agree with it. What you're saying is that even joining together two pieces of information that come from RS's is O.R. so long as the joining has not been done by some other RS before. Which makes sense.
You've given me a very useful insight into this highly complicated issue.--Abenyosef 04:49, 12 February 2007 (UTC)
Your understanding of 'synthesis by juxtaposition' is essentially correct, or at least matches mine. But to refine it slightly, or thread your needle even thinner ;^) - this is what the current nutshell says: "Content should not be synthesized to advance a position". So if the synthesis is a mere convenience to the reader, and/or the issue is not contentious, it would probably be OK. "X had 40 homeruns in 2006.[1] This is the most for a rookie since Y in 1947[2]" - most likely OK, assuming valid sources. But as soon as it seems that WP itself is conveying a message: "X says he got his degree from school Y in 1989[1], but Y only started awarding degrees in 1992[2]" then we need the published source advancing that position (that X may be lying) for us. Crum375 17:19, 12 February 2007 (UTC)
A perfect explanation. :-) SlimVirgin (talk) 18:23, 12 February 2007 (UTC)
I disagree. The second example: "X says he got his degree from school Y in 1989[1], but Y only started awarding degrees in 1992[2]" is certainly valid, though it would be better to leave out the "but". Saying "X says he got his degree from school Y in 1989[1]; Y's website state they started awarding degrees in 1992[2]". This leaves it up to the reader to decide which source is more accurate.
The first example is actually more likely to be problematic, depending on exactly what the sources say: "X had 40 homeruns in 2006.[1] This is the most for a rookie since Y in 1947[2]" If source [1] speaks only to 2006 and source [2] speaks only to 1947, then it is possible that someone hit more home runs in any year in between. If source [1] says no one hit more home runs in the interim, then source [2] would probably not even be need. (I guess source [2] could be needed if [1] gave only the year of the record and [2] gave also the name Y.) Johntex\talk 18:52, 14 February 2007 (UTC)

How about if it's a simple case of two authoritative sports statistics web sites displaying conflicting information? For example, we are trying to figure out over on the Brett Favre article if the statistic for number of passing attempts is correct, or if the and statistics are (they differ by 1). Would it be OR to make a note identifying the discrepancy? PSUMark2006 talk | contribs 19:58, 12 February 2007 (UTC)

In this case, I think reporting both is a good idea; but sending e-mails to the sources in questions was an excellent course of action. Well done! --Merzul 20:09, 12 February 2007 (UTC)

Can we use O.R. to argue for or against a source?

If, in a Talk page, I want to argue for or against a source other editors want to include, can I do and show my O.R.? Can I argue, in the [Talk:Vector Calculus] page, against an author who writes about calculus by doing some research on his life and showing that he has never studied mathematics, but has rather spent his whole life growing rhubarbs?--Abenyosef 19:16, 10 February 2007 (UTC)

You cannot use OR to argue, but if you mention a person rendering a professional opinion, it makes sense to include his/her education, position and expertise also (based on reliable sources), letting the readers judge his/her qualifications. Of course in general you also want to include the best secondary sources that are already interpreting the situation for us, so all we need to do is quote them. Crum375 19:22, 10 February 2007 (UTC)
No, you're not understanding me. I don't mean that I'll argue against some author in the Wikipedia article. I'll argue against him in the Talk page of that article, debating other guys who want to include that author as a source.
If, let's say, in the [Talk:Vector Calculus] page somone wanted to use some claim by one I. L. Literate, am I enabled to investigate his life and prove that he was not knowledgeable about calculus?--Abenyosef 19:51, 10 February 2007 (UTC)
You can investigate anyone's life, as we live in a free society, but none of that is relevant in WP and none of it belongs here. Our goal at WP is to present previously published information in an informative and interesting fashion, based on good quality sources. This means that even in the most technical article, you may have some individuals whose main skill is in presentation and/or graphics, others with expertise in sourcing, others acting as 'test readers', or others more qualified in the abstract technical details. Nevertheless, all must collaborate to produce a high quality article. A good article is a collaborative effort that reflects contributions from many individuals with many different skill sets - this is WP's beauty and strength, and we should not detract from it by restricting the potential contributors. Crum375 20:02, 10 February 2007 (UTC)
In general, NOR is a policy which it makes sense to apply to article pages, but not to talk pages. For example, suppose you are an expert on a certain branch of mathematics, and you spot a formula in an article that looks to you like it has an error in it, even though it is backed up by a reputable source. It would be wrong to just change the formula in the article, without finding another source, but absolutely appropriate to raise your concerns on the talk page.
The NOR policy is there to stop new, novel ideas that we have no way of checking being published in Wikipedia. On the other hand, writing an encyclopedia involves lots of unoriginal, non-novel research, including checking articles for consistency, checking against multiple sources, and using editors own expertise and experience to evaluate articles. Fact checking and wide background research to find the most appropriate sources are to be encouraged - they are vital to producing a reliable encyclopedia. Provided that your background research is in line with what reputable published sources say and not arriving at new, novel conclusions, it is to be encouraged. Enchanter 23:32, 10 February 2007 (UTC)
Well, but what if reputable publications endorse a source and I, through my original research, prove that source to be a piece of crap? Can't I use my research, which goes against conventional wisdom, to try and kill the use of the source in question at the Talk pages? And can't I remove the source from the Wikipedia article on my research?
I believe I can. I believe NOR applies exclusively to what you add to a page, and does not forbid you from removing a source or a statement from an article based on your research. NOR applies to your positive contributions, because Wikipedia is not a publisher of original research. If you use your O.R. to remove something, your O.R. will not be published and no "novel historical narrative" will be introduced.--Abenyosef 00:47, 12 February 2007 (UTC)
If you can prove that a particular source is of lower quality than previously thought, and convince the other editors of your position by consensus, then you can either remove it or assign less weight to it, as appropriate. If it's a BLP case, where we are hyper-sensitive, and you can show that a given source of controversial information is dubious, you may remove it as 'poorly sourced'. Of course if other editors disagree, you'll have to convince them, in a civil and collaborative fashion, or otherwise reach agreement. Crum375 00:59, 12 February 2007 (UTC)

Proposed addition: "Original research is allowed on Talk pages"

I'd like to suggest that the following paragraph be added to the policy:

Original research is allowed on Talk pages

Most Wikipedia activity takes place on the Talk pages, and users occasionally tend to believe that the prohibition to use original research applies to those pages as well, which is not the case. Actually, original research is not only allowed, but also welcome in Talk pages, especially in the following situations:
  • When an editor argues for or against the use of a certain source. In those cases, the editor may do some research about the source, find that it is reliable or unreliable, and argue for using it or not based on his research. Other users participating in the Talk page may accept or reject his arguments, but may not refuse to consider them on the grounds that they constitute original research.
  • When an editor finds a factual error in an article; for instance, a mistake in a mathematical equation. In that case, he may communicate his finding at the Talk page and argue for the removal of the flawed equation, but he may not write the correct one into the article without citing a source for it.
The first question an editor must ask when considering doing original research is: will this be written into the article itself? If the answer is a clear no, there is no problem. An editor may do some research to remove Source A from an article -- he will be deleting, not adding, something. What he cannot do is write "Source A is unreliable" into the article itself: that would be in violation of WP:NOR.

The reason I suggest this to be added is that actually an awful lot of editors believe you can't do O.R. to check a source.--Abenyosef 04:40, 12 February 2007 (UTC)

I personally don't think we need all of that. I just inserted 'included in an article' into the policy page - we'll see if it'll fly. Clearly the entire WP:OR policy applies only to articles. We actually do encourage editors to do research as you say, and not to blindly follow existing textbooks or other published materials. But at the end of the day, the article itself must be strictly based on existing sources. Crum375 05:00, 12 February 2007 (UTC)
Well, but you're an intelligent and sensible person. Less bright and more stubborn editors --and in my few weeks here I've stumbled into quite a few-- may need for these evident points to be explictly stated in the rules, so that the other editors will not have to explain to them over and over again that, like you say, the entire policy applies only to articles. --Abenyosef 06:10, 12 February 2007 (UTC)
Well, if you run into one of those editors now, point him/her to the words included in an article. If s/he still finds it insufficiently clear, then we can fine tune it more. In principle, I believe that simpler is better, especially in policies. Crum375 06:30, 12 February 2007 (UTC)
What can you do when someone says "All of Abenyosef's statement abouve is Original Research. Therefore, they [sic] cannot form the basis of a deletion"? Or "It is only your original research that asserts she misrepresents the source"? Or "Please stop filling the page with repetitive and irrelevant original research, let's stick to policy"? Or "Actually, WP:NOR applies quite explicitly to the way sources are chosen and used. It baffles me that you would imagine you understand Wikipedia's content policies better than I do"? See for yourself here and here. As you'll be able to see, they tried, and tried hard, to actually prove that I couldn't object to a source based on my O.R. When I pointed them to the words included in an article they ignored me. Maybe the fully explicit wording I'm proposing would be necessary to deal with such stubborn people.--Abenyosef 07:09, 12 February 2007 (UTC)
I've reviewed the article and it's associated Talk page. I did my own source-based research and found a scholarly book discussing this very controversy. I quoted and cited it on the article page including the controversy and misquoting hoax letter that stirred the pot in the first place. put their foot in their mouth here by not fact-checking, as did the editor who cited the 1969 article without actually reading it. Wjhonson 08:31, 12 February 2007 (UTC)

Allow me to insist that the paragraph at the beginning of this posting be written into the Project page. The problem is that the rules are not explicit enough and there are also not enough examples. I think it would be of great help to editors to clarify that, for instance, deletion of materials based on O.R. is not wrong per se.

There is a lot of confusion with regard to O.R. on Talk pages. Some editors, either not knowing that it's allowed or feigning not to know, block other editors from suggesting or questioning sources based on their O.R. I think it's time for the policy to be clear and unequivocal about it: O.R. is allowed on Talk pages.--Abenyosef 23:27, 12 February 2007 (UTC)

Don't try to change policy based on your misrepresentation of content conflicts. Your research on the source was not whether or not it was reliable, but rather, that you disagreed with the analysis made by the source, and therefore insisted it was "unreliable" because of that, and wanted to exclude it on that basis. Jayjg (talk) 04:00, 21 February 2007 (UTC)
Agreed. We can and do give people wide latitude on talk pages - people express their own POV's all the time for example. But while I hesitate to ban any kind of discussion from talk pages, the fact remains that OR on talk pages is utterly useless. Talk pages are meant for discussion that will lead to improvements in an article. If people go on and on with what is essentially OR, none of it will be allowed in the article - it ends up wasting a lot of time. Slrubenstein | Talk 13:03, 21 February 2007 (UTC)

Does adding statistics count as synthesizing data?

On the Brett Favre article, there is debate as to including a statistic concerning career number of points scored that can only be derived (as far as we can tell from available sources) by combining statistics available on a football reference site - specifically, the site does not provide the number of points outright, but rather the number of rushing and passing touchdowns scored, both of which would need to be multiplied by 6 and then summed to get the career number of points. Is this an acceptable use of verifiable data or would this be breaching WP:NOR? Thanks, PSUMark2006 talk | contribs 05:17, 12 February 2007 (UTC)

IMO if the statistics are very straight forward and can be verified by an average person, then there should no problem to provide it. For example, if we provide election results in parliamentary seats per party, we may calculate a percentage for each, given a sourced total number of seats. When it gets more complex the calculation may get challenged and it may then be needed to provide an outside source for it. Crum375 15:51, 12 February 2007 (UTC)
This is one of the items mentioned in the essay Wikipedia:These are not Original Research, which I wrote several months ago. I'm a little disappointed that no one has referenced it here & that it has not attracted much attention since then. -- llywrch 18:33, 21 February 2007 (UTC)

a case in point: synthesis

I'd like to describe a case possibly exemplifying original research. Does it indeed constitute original research?

Suppose there's a Wikipedia category listing some languages. A statement has been made in some Wikipedia article, that all languages on the list show a certain linguistic trait. However, the article cites no source for the claim. For each of the listed languages, the Wikipedia article about it states that the language it deals with shows the trait, and cites a reliable source for this claim. Does the unsourced statement constitute original research for the purposes of Wikipedia?

The actual incident that's triggered my interest in this question is described next. I've read a sentence in the Hebrew Wikibook for teaching Arabic, that states that Arabic is nowadays the only Semitic language with a case system. Looking for proof of this claim, i came to the English Wikipedia article Semitic languages, where it states: "The proto-Semitic three-case system [...] has disappeared everywhere in the [...] Semitic languages, although Modern Standard Arabic maintains such case endings [...]. An accusative ending -n is preserved in Ethiopian Semitic." This statement is not sourced. I've sent a message to the authors of this statement asking them to state their source, but they haven't responded yet, and it appears they've not been active in Wikipedia for several months. I've been thinking what i could do to substantiate this claim. One possibility would be to peruse the articles about each of the Semitic languages, looking for a reliably sourced claim that the language under discussion has no case system.

Itayb 09:00, 12 February 2007 (UTC)

Certainly not OR, it simply lacks attribution; a simple google search on "Semitic case endings" yields an article on JSTOR that confirms this. Should not be difficult for any editor with some knowledge in this field to add sources, otherwise I could add that source myself. Original research is only when no such source can possibly be found, and we rely on our own knowledge. --Merzul 11:24, 12 February 2007 (UTC)
Thanks, Merzul, for your reply. You wrote: "I could add that source myself". Please do. I'd appreciate it.
My question is more general, though. Is it considered original research to state that every member of some list has some property, when it is possible to cite a source confirming this claim for each member of the list separately, but there is no source that explicitly makes the generalization? Itayb 11:48, 12 February 2007 (UTC)
My opinion is that in the general case, if you have a list of items, each pointing to its own valid source(s), or ideally its own wiki article, and each has a given common well defined property P, we may say: "This is a list of items with property P", even though we cannot cite an existing source that contains exactly that list and makes that same connection. If the 'items' are living people, there may be BLP issues (as was recently discussed on the BLP Talk regarding individuals with a certain disease), and the source quality and quantity would have to be highest (e.g. more than a just a single news article in one paper, highly respected publications, etc.). This is just an 'inclination' - I'd have to see the specific list to decide, but I am sure many, or even most, existing WP list articles follow this model. Crum375 14:42, 12 February 2007 (UTC)
I'd say it's "trivial synthesis" to say that if "each member of a class has property X" then "ALL members of a class have property X". So that defeats the OR claim in my book. The next issue rises if the claim is challenged by a {{fact}} tag. The conflicting editor may find themselves creating a List of Semitic languages and then assigning a property "has no case system" and citing each seperately. We do have precedent for that sort of *list* on wikipedia. That is, a list of members of Class Y, each cited with seperate sources that state it. Wjhonson 15:32, 12 February 2007 (UTC)
Agreed. IMO the easiest and cleanest way to provide verifiable sources in a list is to point to existing wiki articles, where the validity of the sources and information for each list item can be individually vetted at length by (potentially) a separate group of editors. Crum375 15:47, 12 February 2007 (UTC)
Thanks for your interesting comments. I agree with you. One of the footnotes in the policy page quotes Jimmy Wales regarding the origin of the NOR rule: "It can be quite difficult for us to make any valid judgment as to whether a particular thing is true or not. It isn't appropriate for us to try to determine whether someone's novel theory of physics is valid; we aren't really equipped to do that. But what we can do is check whether or not it actually has been published in reputable journals or by reputable publishers." (Wales, Jimmy. "Original research", December 3, 2004) Therefore, sythesizing a generalization based on a small list of claims, which are easily verifiable, without any need of special skill or knowledge, should not be considered original research. Itayb 16:00, 12 February 2007 (UTC)
One caveat: if the resulting synthesis 'advances a position', i.e. seems to imply some conclusion, then it must be backed up by a source, especially in contentious cases. See for example the discussion thread above with Abenyosef regarding the use of synthesis in such cases. Crum375 16:56, 12 February 2007 (UTC)

Statistically proveable original research

Let's say I wanted to do a bit of research into which fighting game has the most selectable characters. Now in my opinion, this is original research that could easily be proven with a little work. It also does not seem like the kind of information one could found somewhere else to cite, at least not from a reliable source (even if reliable, it may be out of date). So if I, perhaps with the help of other Wikipedians, checked every fighting game to see which has the most, we should be able to cite ourselves, correct? I got to wondering about this after I realized that The King of Fighters Neowave has 47 selectable fighters, and was considering doing some original research into if this was the highest.

So would the following example perhaps work?

Super Fighter 13 has 82 selectable fighters, the highest of any fighting game.1
1. (Pretend this is a cite) As of February 14, 2007. See findings at List of fighting games by number of selectable characters (article would contain list of games that were checked followed by number of characters so that readers could check for themselves to see if "they checked x game"; if such an article is not considered to be encyclopedic, the information may be hosted somewhere else like User:SeizureDog/NumberOfFighters).

It seems to me that certain statistic type information such as this (highest, lowest, first, last, etc.) should be an exception to OR as it is information that can be undeniably proven by anyone who cares to check well enough. However, the results and process of the person's research should still be made available so that others can verify. Comments?--SeizureDog 13:32, 14 February 2007 (UTC)

It is my understanding that we do this type of thing all the time. We list all sorts of things by ranking, since it is just a synthesis of known information and trivial. This includes ranking of cities and states by population and other stats. I don't know if this applies, but there is at least some similarities. -- RM 14:10, 14 February 2007 (UTC)
Yes, my own opinion is that if the criteria are clear, stable, non-controversial, and well defined (in contrast to 'fuzzy' criteria), the information is well sourced and the statistics or comparison can be readily verified by an average reader, then we can report them as a service or convenience to our readers. So 'tallest buildings list' would be OK (i.e. sorted by officially reported height), but 'largest employers' would not be (fluctuates, fuzzy criteria, unreliable and/or conflicting data, etc.). Crum375 14:23, 14 February 2007 (UTC)

Perhaps I should make more clear what the policy problem is here. WP:CITE says "Note: Wikipedia articles and categories cannot be used as sources in and of themselves. Sources must be independent from Wikipedia." The problem isn't the List of fighting games by number of selectable characters article, it's the Super Fighter 13 article that uses it as a source. Can I just ignore this rule?--SeizureDog 15:22, 14 February 2007 (UTC)

Looks like too many elements of original research to me. How could anyone verify you've checked every fighting game? What's defined as a fighting game? There's a reason we need reliable sources for material and it's to avoid this type of thing. If you have to resort to citing a Wikipedia article, then the material shouldn't be included per WP:NOR. - Taxman Talk 17:02, 14 February 2007 (UTC)
  • I suggest linking to the list as you describe but to tone down the description to "a large number of selectable fighters compared to other games". That allows for the possibility that some game has gone unchecked in the analysis. Johntex\talk 18:39, 14 February 2007 (UTC)

Can a distributor's product manual be used as a source?

I would like some opinions on the Juice Plus article. It seems to be well-sourced, but some of the sources are (in my opinion) a bit dubious, such as letters to the editor, the product's distributor manual, and bottle labels. Especially since there have some studies which indicate that the bottle labels are not entirely accurate. It's my opinion that this article is straying well into the realm of original research, and I would appreciate some assistance in sorting through the claims. --Elonka 16:57, 14 February 2007 (UTC)

Letters to the editor can be used to source the claim that X wrote Y in a letter to Z Editor, but not for much else, such as the moon is made of green cheese. Similar for the bottle. The less reliable the source, the less it can be relied on to back general facts and must only be used to back the fact that the source stated X. Actually, come to think of it, the product label isn't good for much at all. If it has no date and the info on it is changed, how would that be able to be checked. Unless there are places that publish product label information and their changes, that's not a reliable source. - Taxman Talk 17:02, 14 February 2007 (UTC)
As for product labels, many acceptable sources do not include a publication date. This is especially true for web sites. If sources that lack publication dates were not acceptable sources, there would be no need for the Chicago Manual of Style (14th ed.) to have paragraph 15.175, which is titled "Date not indicated". (I must admit that that paragraph applies to books, and not all works.) --Gerry Ashton 18:05, 14 February 2007 (UTC)
What the manufacturer says about their own product should be viewed the same as a self-published article. An author may indeed state how she feels about her book or what it means, but we should quote and cite that statement, not present it as fact. And it may certainly be countered by another reliable source stating the opposite. Wjhonson 18:31, 14 February 2007 (UTC)
A further issue is that of verifiability. A manual that's online is one thing, but something that's only in the hands of distributors is not accessible to Wikipedia editors for verification. --Elonka 20:13, 14 February 2007 (UTC)
But sources don't have to be on-line. That is not a requirement. Some sources are easy to verify, others are not. Johntex\talk 20:18, 14 February 2007 (UTC)
In my opinion, a good source is something which should be accessible to the general public. Otherwise anyone could come in and say, "Oh yeah, I have papers which prove this. But I keep them in a safe, too bad you can't get at them." If something's not accessible via the web or a television broadcast or a library loan of some sort, that makes it a very chancy source. --Elonka 20:35, 14 February 2007 (UTC)
I'd agree with that. A source doesn't have to be online, but it has to be available to the general public. If it's only available to distributors then it would be an internal company document and not a RS. --Milo H Minderbinder 21:08, 14 February 2007 (UTC)
The company document in question (the distributor’s manual) was used is support of 2 claims, Neither claim could be considered contentious and there are no other sources that contradict or question either of them. The first claim is that “Allicin and other undisclosed phytochemicals” are added to the product. This is stated in the manual and there is no reason to doubt the claim.
The second claim for which the manual was cited was as follows: “In accordance with the Dietary Supplement Health and Education Act of 1994, NSAs distributor manual claims that Juice Plus is not intended to prevent, treat, cure, or mitigate any disease.” This statement is a legal requirement of the Dietary Supplement Health and Education Act. If the manufacturer did not make such a statement, the product could not legally be sold in the US. It is a self evident claim and there is no reason to doubt that it is true.
The manual is in fact available to the general public; it can be purchased by anyone for $7 at this URL[12]. It therefore is not an internal company document since it is publicly available. Lastly, there are copies of the manual available on the internet that can be viewed. I can post a URL to a PDF copy of the manual from a 3rd party website if any of the editors here would like to verify the statements in question? Rhode Island Red 15:33, 16 February 2007 (UTC)

(unindent) In the citation, I would mention the URL where it can be ordered, since it is not available through the same channels as most sources (libraries and bookstores). Linking to a 3rd party site is called a convenience link. This is allowed if the 3rd party site does not appear to be violating copyright, and is not an extremist site or pushing some point of view. --Gerry Ashton 15:48, 16 February 2007 (UTC)

It should be noted that in some areas the product manual is considered a reliable primary source. For example, with aircraft, the specifications and operational limits specified in the flight manual are the official numbers, and conflicting data would only be relevant under extraordinary circumstances (for example if they were contradicted in an accident board report). So product manuals should not be automatically discounted just because the manufacturer is the source. All sources need to be used with common sense. Dhaluza 00:02, 17 February 2007 (UTC)

Prohibiting Original Thinking Precludes Quality Work

Never in human history has a quality work been written without original thinking. Wikipedia is no exception. Where there is quality work in Wikipedia, we find original thinking and ultimately original research. It must be so. Such has it always been and such it always will be.01001 03:03, 16 February 2007 (UTC)

We do, in fact, create original work all the time. Except that our original work and thinking is in finding and vetting the appropriate verifiable sources, balancing them out so as to present a neutral picture, using interesting and engaging language to summarize those sources. As a result, we are producing a vast collection of original, informative, interesting, well sourced, balanced articles that are freely available to all.
This is most certainly original work - just look at all the other publications copying our articles - but we do not advance new positions, new science, or even new synthesis of existing information. Instead, we carefully summarize the existing published body of knowledge, per our mandate as an encyclopedia, which is a tertiary source. Crum375 04:36, 16 February 2007 (UTC)
These are pretty words and they all sound very nice. Unfortunately your logic is fatally flawed. As soon as you give an article a title and then write the first sentence, logically you are advancing a position. Further, every sentence after the first is further advancing a position. If a sentence has meaning it must advance a position. The very meaning of the sentence advances a position.
This is very important because this policy not only fails in theory, but it also fails in practice. There are many high quality articles in Wikipedia, but many articles suffer. In every case, the high quality articles are created despite this policy not because of it.01001 01:32, 17 February 2007 (UTC)
It sounds like you are possibly addressing WP:NPOV along with WP:NOR, and I would probably concede your point that it is nearly impossible to be totally devoid of POV, since the very presentation and arrangement of things can 'advance a position' in a subtle way. So we would be in agreement there. However, the beauty and power of Wikipedia is that it is a dynamic creature, that everyone can edit, any time. And by having different people constantly tweaking things and trying to aim for the magical 'sweet spot' of perfectly balanced neutrality, based on the available reliable and verifiable sources and their preponderance, we can always get closer to perfection, though perhaps never quite get there. Bottom line is that, as you say, we have a lot of high quality articles, and some not as good, but we keep improving all of them, all the time, always getting better, and that's what really counts. Crum375 02:34, 17 February 2007 (UTC)
In practice, that's not how Wikipedia is working. Wikipedia has two different categories of articles:
1) Articles that many people are interested in editing. These are subject to constant vetting and are for the most part O.R.-free and POV-free -- not because their editors are in pursuit of any "perfect balance," but because their competing biases cancel each other resulting in a reasonable degree of neutrality.
2) Articles that few people care for. These articles include enormous amounts of O.R., especially synthesis.
The original idea behind Wikipedia --summarizing existing information-- may thus be evolving towards something new and more complex, consisting of: a) good summarizations of those topics that attract many people's attention and for which abundant information is available, especially on the Web; b) useful, if not fully reliable, articles that include a large amount of original research about those topics that are not so popular and for which little information is to be found, especially on the Internet.
While the official policy continues to be "No original research," the actual philosophy of the encyclopedia may be changing to something like "some original research is better than nothing, when no verifiable sources are easily available." --Abenyosef 19:46, 17 February 2007 (UTC)
I agree with you that unfortunately there is a large number of poorly written articles. By that I mean both poor or no sourcing, as well poor writing style. However, I feel that these are gradually being addressed, although I have no hard statistics to prove it. In any case, your 'category b' articles are not somehow excluded from our rules - they are simply in the form of 'works-in-progress', and require fixing. If an editor discovers such an article, there are multiple templates to choose from to mark it as requiring work, and hopefully someone will eventually start fixing it. Is it a perfect system? - of course not - nothing ever is, but it is our working model, and if any specific article gets tagged as poorly written or sourced, it will be improved. Crum375 22:46, 17 February 2007 (UTC)

Wikipedia:Attribution: Merger of Wikipedia:Verifiability and Wikipedia:No original research

Wikipedia:Attribution, a proposal to subsume and replace Wikipedia:Verifiability and Wikipedia:No original research, is ready to be implemented. Please review the document and discuss any problems on the talk page. —Centrxtalk • 23:09, 16 February 2007 (UTC)

See the top of this page. SlimVirgin (talk) 02:42, 17 February 2007 (UTC)

Request for comments on Original research in Historical revisionism article

I have come accross a case in where an editor is resusing to accept that he has engaged in original research by drawing his own conclusions from sources. An editor of the Historical revisionism article includes two examples of historical revisionism - Serbia and the Yugoslav Wars and Macedonism - on what I see as original research.

In the case of Serbia he cites a number of Serbian language (I presume) nationalist websites plus a report from the UN about Serb war crimes to claim this as an example of historical revisionism. None of the sources state anything about Serbian historical revisionism. I see this as a clear cut case of Synthesis of published material serving to advance a position. The same situation applies to Macedonism, which is not supported by any sources claiming it as an example of historical revisionism. I have informed him of this and suggested that the examples be removed but he does not agree. Any comments? Anyone willing to take action (it would be better if an admin did it than if I acted as a party to the discussion)? Regards Osli73 14:05, 20 February 2007 (UTC)

Conflict between Policies

I gather that WP:ATT is now considered Policy... a merging of WP:V and WP:NOR. I have no problem with this, but I do have a concern with how this is being done. At the moment all three Policies are up and running... and this is going to lead to confusion and argument. The three policies do not completely agree with each other (or to put it more exactly, they agree with each other in surface substance, but they differ in greatly in tone and emphysis). This is especially noticable in the area of reliability of sources (especially Self-published sources - a frequent area of controversy) and how they are dealt with. Look at them side by side: WP:V#Sources, WP:NOR#Sources, WP:ATT#Reliable_sources. As a frequent contributer at WP:RS... a guideline that is supposed to help explain this particular aspect of Policy... I am seeing this conflict of tone writ large. Many of the questions we are asked involve parsing Policy statements and intent. Those of us at WP:RS agree that we need to conform what is said there to what is stated in Policy, but right now there is confusion as to which policy we should conform to. I would therefore request that, if WP:ATT is indeed confirmed as Policy, we redirect WP:V and WP:NOR to that page. If not, please move WP:ATT back to "proposed" status until the community can reach consensus. I don't care which, but we need clarity and not confusion. I have posted this request at the Village Pump Policy Page as well. Thank you. Blueboar 15:32, 23 February 2007 (UTC)

I personally am pretty hard-core about NOR, V, and of course RS. I say this with two caveats: first, I am an academic and know that the standards by which I would hold any source concerning an article concerning an academic topic are pretty strict ... and I acknowledge that people in other professions writing articles on other topics would find my own standards either twoo strict or wholley inappropriate. Before I explain the second caveat let me share my own sense of the trend at Wikipedia: when we started Jimbo's and Larry Sanger's vision for whatever number of reasons privileged NPOV - I mean, the the extent to which the question was, could the idea of NPOV be operationalized into a viable policy, and could a wiki-community exist and collaborate on an encyclopedia guided by NPOV? That is, Wikipedia to an extent was (I think) founded as an experiment in putting NPOV into practice. Early articles - including many to which I contributed and some that I wrote, had no sources at all and probably wouldn't have held up to our NOR standards of today. Perhaps this was because Wikipedia was founded based on a trust in the expertise of editors and that the wiki-process, constrained only by NPOV, would eliminate all that was false or inaccurate leaving only that which was, implicitly, reliable. I see the past several years as characterized by an increasing awareness that NPOV alone was inadequate to guarentee the encyclopedic nature of the project (or, the quality of the encyclopedia). And, thus, an increasing appreciation of the importance of the values expressed in the constellation of policies and guidelines NPOV, V, and RS. In fact, I think one reason there are these three is precisely because of the direction of the trend - as we become mor eand more committed to externally verifiable standards of research, we have developed more specific and rigourous policies. Thus, first came NOR telling people what not to do. Then V, providing a minimal standard of what people should do, along with two guidelines, first CS (again providing a minimal standard) and finally RS, providing a more rigoroous standard. As I said, I am ALL for this. But the point I am trying to make is this: I think that five years ago most Wikipedians were really committed to NPOV, but today there are more Wikipedians who are committed to NOR, maybe even more than NPOV. We have not abandoned the old values, but some of us - I include myself - are more fired up by some new values. So here is my second caveat: I think that there will be times when there is an unavoidable tension between NPOV and NOR/V/RS. Specifically, there will be times when all major views are not represented in equally rigorous sources, and people may feel that a view must be included to provide balance even if it comes from less reliable sources than other views. i do not want to argue specifics. Obviously I personally would be very careful in these cases and probably insist that with enough research one can find highly reliable sources for any major view i.e. there is no conflict. nevertheless, I do think that NPOV and NOR/V/RS do reflect fundamentally different principles and I believe both are equally important. I think when there is an apparent clash between the two principles editors should seek a solution on a case-by-case basis. So frankly while I agree that NOR, V, and RS ought to be made fully consistent with one another, I do not think this should and maybe not even could happen with NPOV. And this is why I think right now there are inconsistencies between NOR and V, which are policies, and RS, which is only a guideline - it is to allow enough room to play when there is a conflict with NPOV. If NOR, V, and also RS are consolidated (i.e. if RS is to be a policy and not just a guideline) I think we will need to be attentive about this, that guaranteeing NPOV may require a range of different types of sources that range between different levels of reliability or rigor, and that our policies need to be flexible enough for conflicts to be resolved not merely by appealing to some policy that covers every possibility but rather through the good judgement that comes out of an open discussion among informed editors who are acting in good faith - on a case-by-case basis. I am probably the last person here who would ask "what would Larry do?" but I do hope that there are still enough people who remember the original and motivating zeal of NPOV who are participating in this process. Slrubenstein | Talk 16:33, 23 February 2007 (UTC)
NPOV is still important, and still a Pokicy. WP:ATT is intended as a merger of WP:V and WP:NOR (and to some extent WP:RS), keeping the essence of both but unifying them into one Policy. The folks who wrote this actually did a good job, so I am not complaining about what they did... only how. The problem as I see it is that a unified Policy has been accepted without getting rid of the ones it was intended to replace. This is causing confusion. Blueboar 16:53, 23 February 2007 (UTC)
I saw comments somewhere explaining the idea was to run the new policy in parallel for a short period, then replace the old ones with redirects. Since ATT has been tagged as a policy since Feb. 15, I expect to see the redirects any day now. It has only been 8 days, and people will be shocked when the redirects appear. CMummert · talk 17:42, 23 February 2007 (UTC)

Now historical

Being bold, I have marked this as {{historical}}. That should end the confusion alluded to above. (Now let's see how long before I get reverted...) --EngineerScotty 18:11, 23 February 2007 (UTC)

Inappropriate editing

Even though the policy has been merged, to try to suppress discussion here is not appropriate. If we want to discuss things here, we should be free to do so, just like on any other talk page. Wjhonson 08:46, 4 March 2007 (UTC)

Why bother? This policy is no longer in effect, having been superseded by Wikipedia:Attribution, and there is no point to discussing the provisions of the former policy here. The policy is best discussed at Wikipedia talk:Attribution. -- Donald Albury 11:47, 5 March 2007 (UTC)
I'm just saying it should not be forbidden. Wjhonson 17:55, 5 March 2007 (UTC)
I agree. Who "banned" it, anyway? Ninetywazup?Review meMy ToDo 22:58, 6 March 2007 (UTC)
This former policy should not be edited, so what is there to discuss? If you want to address something that was part of this former policy, you would do better to bring it up at Wikipedia talk:Attribution. -- Donald Albury 01:52, 7 March 2007 (UTC)