Wikipedia talk:No original research

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by Blueboar (talk | contribs) at 21:24, 21 May 2010 (→‎Question about original research). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Template:NORtalk

WikiProject iconSpoken Wikipedia
WikiProject iconThis page is within the scope of WikiProject Spoken Wikipedia, a collaborative effort to improve the coverage of articles that are spoken on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.

Beginning of lead

One of my intentions when I originally added the first sentence,[1] was to have something that simply stated the idea and was easy to understand by everyone, without having to look up any terms etc . Unfortunately, it has morphed into something else, and its content has changed.[2] I think that this is partly because editors have grown use to writing for the other editors here instead of for all the silent readers who come to this page to learn about NOR. --Bob K31416 (talk) 14:35, 7 May 2010 (UTC)[reply]

That was mostly my edit you reverted (with a little bit by Crum)... My intent (and I suspect Crum's intent as well) was to simplify and clarify the wording to make it even easier for people to understand, but not to change the meaning. In what way did our edits change the meaning? (here is the dif as of your last change.) Blueboar (talk) 17:40, 7 May 2010 (UTC)[reply]
I'm not keen on the new last sentence of the first para; I think we're over-egging the pudding. SlimVirgin talk contribs 20:56, 7 May 2010 (UTC)[reply]
I agree about that last sentence being superfluous. It repeats a list of "things" which need to be sourced, different from the previous list, which may raise a question of what is the significance of those differences. It also repeats for the third time or so in the same paragraph the same warning not to publish anything new or unsourced, which seems shrill and unprofessional. Crum375 (talk) 21:27, 7 May 2010 (UTC)[reply]
OK... but I do think something should be said somewhere in the policy about Wikipedia not being the first place of publication for new ideas or information... that goes to the heart of what No Original Research is all about. Blueboar (talk) 00:33, 8 May 2010 (UTC)[reply]
BB, can you say what new information you see the phrase "Wikipedia is not meant to be the first place of publication for new material" adding? Consider that the first paragraph already says:
  • "Wikipedia is an encyclopedia, which as a tertiary source summarizes material already published elsewhere"
  • "Wikipedia does not publish original research...'original research' refers to material...not already published by reliable sources."
Crum375 (talk) 00:58, 8 May 2010 (UTC)[reply]
Well, I tried to weave the "not first" verbiage into the lead anyway. Thoughts? Crum375 (talk) 01:09, 8 May 2010 (UTC)[reply]

I thought the simple-english version that I originally put at the beginning of the lead was pretty good. Oh well.

Wikipedia is an encyclopedia which summarizes existing published information and is not meant to be a source of information that has not been published.

--Bob K31416 (talk) 01:47, 8 May 2010 (UTC)[reply]

I don't think anyone here disagrees with your words. The question is how to best integrate the idea behind them into the policy. Crum375 (talk) 02:36, 8 May 2010 (UTC)[reply]
Exactly right... it isn't that we disagree with Bob's words, or with the concept behind them... I think we simply have differing ideas on how to best phrase it in a clear and concise way. As for my initial edit... I simply found his language a bit long winded for my taste... but if others prefer it his way, I certainly will not object. I was more concerned by the fact that Bob thought my language changed the meaning in some way... I don't understand how it did, so I figured I would ask. Blueboar (talk) 02:48, 8 May 2010 (UTC)[reply]

Re "The question is how to best integrate the idea" and "how to best phrase it in a clear and concise way" -

The answer is what I originally proposed for the beginning of the lead.

Wikipedia is an encyclopedia which summarizes existing published information and is not meant to be a source of information that has not been published.

The above would replace the following recent additions which were an attempt "to best phrase it in a clear and concise way".

"Wikipedia is an encyclopedia, which as a tertiary source summarizes material already published elsewhere." and "Since Wikipedia should not be the first to publish new material..."

--Bob K31416 (talk) 12:37, 9 May 2010 (UTC)[reply]

Problems with your version: (1) "Information" carries an implication that it's correct, so "material" is better; (2) "existing published information" implies that there's such a thing as "non-existing" published information; (3) the second clause repeats the first.
I do wonder why any of this is necessary though (any version, not just Bob's). It makes the lead unnecessarily clunky and repetitive. SlimVirgin talk contribs 22:40, 9 May 2010 (UTC)[reply]
Your analysis doesn't use standard interpretations of english, so it isn't useful. Sorry, but thanks for trying. --Bob K31416 (talk) 06:32, 10 May 2010 (UTC)[reply]

RfC: Lead sentence of WP:NOR

What should the first few sentences say?

A) Wikipedia is a tertiary source summarizing material already published elsewhere; it is not a first publisher of new material. As such it does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources, and to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by any of the sources.
B) Wikipedia is an encyclopedia which summarizes existing published information, and is not meant to be a source of information that has not been published. Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources, and to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by any of the sources.
C) (current version) Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources, and to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by any of the sources.

Comments

(no threaded replies here, please)

  • Please note that version B consists of only adding a first sentence in front of the current version C. Version A also adds a first sentence but with a slight modification of the first sentence of the current version C. --Bob K31416 (talk) 15:26, 11 May 2010 (UTC)[reply]
  • I proposed Version B (i.e. the first sentence) because it is simple, informative, and easily understood by everyone. It clearly describes the need for editors to put into Wikpedia only published information. The first sentence that is version B was previously added to WP:NOR but it was modified into the first sentence of version A by Blueboar, Crum375, and SlimVirgin, who seemed to think that the basic idea was worthwhile but wanted to put it into different words. (Some of them may have changed their opinions since then.)
To avoid any misunderstanding, here is what I proposed, "Wikipedia is an encyclopedia which summarizes existing published information and is not meant to be a source of information that has not been published." --Bob K31416 (talk) 13:05, 12 May 2010 (UTC)[reply]
  • I added the original and simpler alternative, to simply state that we don't publish OR. Crum375 (talk) 15:51, 11 May 2010 (UTC)[reply]
  • I agree with the proposed change (version B), it's more direct and to the point. ThemFromSpace 18:46, 11 May 2010 (UTC) (refactored to specify version B, also per Mike Cline below) ThemFromSpace 12:51, 13 May 2010 (UTC)[reply]
  • I prefer version A... I think it is more direct and to the point. Blueboar (talk) 18:47, 11 May 2010 (UTC)[reply]
  • Version C. I don't think we should labour the point about what WP is at the beginning of this policy. But if we're going to say anything about it in the lead, then I prefer Version A because it's more succinct and better written. Policies need to be as succinct as possible or people won't read them, and that has to apply to every sentence and every paragraph. SlimVirgin talk contribs 22:55, 11 May 2010 (UTC)[reply]
  • Version C (the original longstanding version before the recent changes) as my preferred version, but I can live with A. Crum375 (talk) 23:01, 11 May 2010 (UTC)[reply]
  • Version B because it describes what WP actually is--an encyclopedia, instead of Version A which describes how WP is used (compared to primary and secondary sources). An encyclopedia is a tertiary source, tertiary sources are not necessarily encyclopedias. Version C is flawed because it does not say what WP is--an encyclopedia, but instead, falsely, says what WP does--publish (WP doesn't publish anything).--Mike Cline (talk) 00:27, 12 May 2010 (UTC)[reply]
  • Version C. Version A requires an understanding of the term tertiary source. A link is provided but it would be much easier to not use the term so early on. Version B is just silly. You could start every policy with "Wikipedia is an encyclopedia" but would it get us anywhere? Perhaps the person who wrote it thought that encyclopedias never publish original research. This is not true. For example, the Stanford Encyclopedia of Philosophy is a collection of essays by experts in the field. The NOR policy is required of Wikipedia because we are not credentialist. Yaris678 (talk) 13:49, 12 May 2010 (UTC)[reply]

Threaded discussion

  • I am moving this from above since we are supposed to put threaded discussion here... as I said above, I think version A is more direct and to the point. Blueboar (talk) 18:47, 11 May 2010 (UTC)[reply]
    • It's not direct because the term "tertiary source" needs to be looked up to be understood. Sorry, since I think it was your idea to put in "tertiary source". Regards, --Bob K31416 (talk) 19:16, 11 May 2010 (UTC)[reply]
      • Actually, no... I didn't add that term. But I still find version A more direct and to the point, even with the link. I suppose we will have to simply disagree on that. Blueboar (talk) 19:23, 11 May 2010 (UTC)[reply]
  • If C is wrong, so are A and B, because they both include it verbatim (it's the text that follows them). And WP does publish information, but only if it has been first published elsewhere. Crum375 (talk) 00:32, 12 May 2010 (UTC)[reply]
    • A and B precede C on the policy page. They don't replace C. I think Mike Cline said that C was flawed was because it didn't say what B said. Mike points out that B says that Wikipedia is an encyclopedia and C does not. Thus B completes the info in C, which Mike believes is flawed without it. Seems reasonable to me. --Bob K31416 (talk) 01:12, 12 May 2010 (UTC)[reply]
      • Mike said above "Version C is flawed because it does not say what WP is--an encyclopedia, but instead, falsely, says what WP does--publish (WP doesn't publish anything)." He says that it is wrong because it incorrectly says that Wikpedia "publishes", but that's plainly wrong, and even if he is right, it is included in both A and B, as I noted above. Crum375 (talk) 01:24, 12 May 2010 (UTC)[reply]
        • I think Mike Cline meant that C was flawed because it didn't say what B said. Mike seemed to point out that B says that Wikipedia is an encyclopedia etc and C does not. Thus B would add to the info in C. I think you may have inadvertently confused the issue since your remarks might be interpreted as meaning that C has been substantially changed, when in fact it is substantially the same as it was, with some minor cosmetic changes by SlimVirgin recently. Cheers, --Bob K31416 (talk) 01:12, 12 May 2010 (UTC)[reply]
          • I am confused: how can my note that "C is the original longstanding version" be interpreted as "C has been substantially changed"? Crum375 (talk) 01:36, 12 May 2010 (UTC)[reply]
            • My apologies for the confusion. I was focusing on the lead sentence, not the entire paragraph as indeed each reflects: D. All of the above. As a lead sentence however, Version B is the most concise and instructive. Additionally, if the phrase WP does not publishe original research was modified slightly to read WP should not contain original research it would be signfiicantly more instructive.--Mike Cline (talk) 01:48, 12 May 2010 (UTC)[reply]
  • The problem is that B is oddly written as it stands, and the policy needs to be well-written as far as possible. Holding a content RfC is fine, but writing-by-numbers doesn't work. SlimVirgin talk contribs 02:04, 12 May 2010 (UTC)[reply]
    • Could you clarify what you wrote, "The problem is that B is oddly written as it stands..." and "Holding a content RfC is fine, but writing-by-numbers doesn't work." Thanks. --Bob K31416 (talk) 03:42, 12 May 2010 (UTC)[reply]
As I said before, I feel the writing could be improved in a number of ways, and as Yaris said encyclopedias do publish OR, including the very good ones such as the EB, so starting the policy by saying WP is an encyclopedia implies that there's a link between that and no OR, which is wrong—but if it doesn't imply a link there's no point in mentioning it. SlimVirgin talk contribs 16:18, 12 May 2010 (UTC)[reply]
Seems like you are taking the term "encyclopedia" out of context. The full phrase was "Wikipedia is an encyclopedia which summarizes existing published information..." Regards, --Bob K31416 (talk) 16:25, 12 May 2010 (UTC)[reply]
It's a question of trying to make every word count. In this sentence "is an encyclopedia" isn't doing any work. SlimVirgin talk contribs 16:28, 12 May 2010 (UTC)[reply]
Actually, it's needed because without it the sentence wouldn't sound right. Read it over yourself without "is an encyclopedia which" and perhaps you will see what I mean. --Bob K31416 (talk) 16:37, 12 May 2010 (UTC)[reply]
I agree, but there are other things about the sentence that aren't quite right either, which is why I wasn't keen on it. SlimVirgin talk contribs 16:39, 12 May 2010 (UTC)[reply]
Please, let's discuss them. --Bob K31416 (talk) 16:43, 12 May 2010 (UTC)[reply]
I mentioned them above. [3] The key thing is that the policy should try to stay tight. It's already a bit repetitive, so it would be good not to add any more. We say WP does not publish OR, and we explain what that means. Adding that we're an encyclopedia says nothing relevant. Adding that we're a tertiary source or not a first publisher—these are just different ways of saying we don't publish material not already published by reliable sources. But we already say that, so I don't see the point of repeating it using different words. SlimVirgin talk contribs 16:50, 12 May 2010 (UTC)[reply]
Sigh... I think we are all agreeing on concept here... but are quibbling over the exact wording and language of how we say it. I agree with SV that Bob's proposed language seems clunky and somehow "not quite right"... but I am having difficulty expressing exactly what it is that I find wrong. I also understand her point about version A (although I still prefer it). This may simply be a question of what sounds best to each of us. Blueboar (talk) 17:03, 12 May 2010 (UTC)[reply]
It's also a question of word count. We have so many policies and guidelines now that people are overwhelmed, and the longer each of them is, the less likely it is to be read. So we shouldn't say in 20 words what we can say in 10. I feel we need to try to stick to that very closely. The policy's already a bit wordy in places. SlimVirgin talk contribs 17:24, 12 May 2010 (UTC)[reply]

<outdent>Well, it looks like it's boiling down to a matter of judgement. I think the editors who are active here have lost touch with what it is like for less experienced editors to come to the policy page to learn about NOR, or even some experienced ones that have trouble understanding what is written in this policy. When you discuss these things among yourselves on this talk page, I think you develop understandings of what you put into this policy because of all the time that you spend discussing it here. It's a far different case for people who don't spend so much time thinking about this policy page, in my opinion. I guess that's the way it's going to be here. But Wikipedia seems to survive in spite of that. Anyhow, thanks for trying to do the best that you can. Regards, --Bob K31416 (talk) 19:10, 12 May 2010 (UTC)[reply]

Bob... don't give up yet... can you say what you want to say in fewer words? That may help. Blueboar (talk) 19:22, 12 May 2010 (UTC)[reply]
I wish I could find some way to satisfy you and still put in what I think is best, but that sentence is about as good as I can make it. Cheers, --Bob K31416 (talk) 19:34, 12 May 2010 (UTC)[reply]

Although I'm never sure about what others think here, I think we agree that the idea of the first sentence in A and B is to say what Wikipedia is, and do it in a way that leads into the subject of NOR. My preference is to do it in a way where none of the terms used need to be clarified for the reader. Perhaps we could move towards a meeting of the minds by first modifying version A so that it doesn't contain "tertiary source"? --Bob K31416 (talk) 01:33, 13 May 2010 (UTC)[reply]

As someone else noted above, it doesn't make sense to start every policy by a general "Wikipedia is..." statement. But ironically, if we were to do so, the most direct description of what "Wikipedia is" in relation to WP:NOR is a tertiary source, because this exactly conveys the fact that it should be a summary of secondary sources, and not include original research. Crum375 (talk) 01:53, 13 May 2010 (UTC)[reply]

If there's consensus that we ought to add something, how about this?

Wikipedia is not a first publisher of new material and does not publish what we call original research. The term "original research" refers to any material—such as facts, allegations, ideas, and stories—not already published by reliable sources. This includes any analysis or synthesis by Wikipedians of published material in which the analysis or synthesis advances a position not advanced by any of the sources.

SlimVirgin talk contribs 06:06, 13 May 2010 (UTC)[reply]

Here's SlimVirgin's version with even few words:
Wikipedia does not publish original research, a term referring to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. Original research includes any analysis or synthesis by Wikipedians of published material that advances a point of view (POV) not advanced by any of the sources.

--Mike Cline (talk) 10:25, 13 May 2010 (UTC)[reply]

I like Mike Cline's version although I wouldn't link to the guideline WP:IRS from this policy - it could imply the guideline has been upgraded.
Also, it seems to me that the main point that versions A and B are trying to get across is that Wikipedia comprises summaries of information published elsewhere. I agree that this is important... it deserves to be mentioned early on... but perhaps not so important that it deserves to be in the very first sentence. Perhaps the best way to do this is to revive a sentence that used to be in the SYNTH section, but isn’t there any more. i.e.
Wikipedia does not publish original research, a term referring to material - such as facts, allegations, ideas, and stories - not already published by reliable sources. Original research includes any analysis or synthesis by Wikipedians of published material that advances a point of view (POV) not advanced by any of the sources. Carefully summarizing or rephrasing a source without changing its meaning or implication does not violate this policy: it is good editing.
Or perhaps some other sentence about summarising... just a thought.
Yaris678 (talk) 13:00, 13 May 2010 (UTC)[reply]
Seems OK to me. Crum375 (talk) 13:31, 13 May 2010 (UTC)[reply]

I don't know how or why this got started, but what was wrong with the lead as it was?

Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by any of the sources. [4]

SlimVirgin talk contribs 15:47, 13 May 2010 (UTC)[reply]

I've restored the version of the lead we had before all the recent changes. If we want to add more words, let's gain consensus first rather than tweaking back and forth and making it more repetitive. SlimVirgin talk contribs 15:52, 13 May 2010 (UTC)[reply]
It is a falsehood to claim that a person can copy text or concepts from any published source, reliable or otherwise, and still call it original research. To do so is both false and plagiarism. I have revised the policy accordingly. Jc3s5h (talk) 17:14, 13 May 2010 (UTC)[reply]
SlimVirgin, Wow. That's an audacious move, considering it's one of the three options above that we are discussing. Well, I know from previous experience that if I tried to revert it back you would just edit war so.... Regards, --Bob K31416 (talk) 18:08, 13 May 2010 (UTC)[reply]
It's clear from this discussion that there was no consensus for the change in the first place, and some of the recent suggestions for improving the change took us closer to the wording before the change, so it would have been better had it been agreed before making it. SlimVirgin talk contribs 18:15, 13 May 2010 (UTC)[reply]
You and three other editors contributed to the version A on the policy page that you just replaced, by yourself, with version C. --Bob K31416 (talk) 00:58, 14 May 2010 (UTC)[reply]
It's equally clear from this discussion that there is no consensus for the old version any longer, either. WP:Consensus doesn't authorize reverting to an old version simply because it had support last month. I think you (and everyone) should have left it alone until the discussion ended (with optional addition of suitable tags to point other editors to this disucssion). WhatamIdoing (talk) 21:59, 13 May 2010 (UTC)[reply]
If no version has consensus, it really does not matter which version is currently on the page... what matters is that we work together to reach a new consensus version. In other words... don't argue about what doesn't have consensus... figure out what does. Blueboar (talk) 22:06, 13 May 2010 (UTC)[reply]
Re "If no version has consensus, it really does not matter which version is currently on the page..." - It does matter because the version that is on the page will be the version that remains if a sufficiently large enough consensus is not reached. SlimVirgin supports version C and SlimVirgin just put that version on the policy page at will. So far no one has supported SlimVirgin's action and two editors have opposed it. I just reverted SlimVirgin's action. I think it would be edit warring if SlimVirgin reverted it back without gaining consensus here. SlimVirgin, Try waiting for the resolution of this RFC like the rest of us, or are you special? --Bob K31416 (talk) 01:04, 14 May 2010 (UTC)[reply]
sigh... OK... I'll support her action... it is the version that was on the page before we started this discussion and thus, presumably, the last version to actually have a consensus. It might well be the case that it no longer enjoys that status, but neither has any of the proposed replacements. I see no harm in at least temporarily retaining it while we hammer out a consensus (if we can) on replacement language. More to the point, it is accepted that being BOLD on policy pages is not a good idea. Policy pages can change, but they should change very slowly and deliberately... and that means keeping old language until new language is finalized. If you need to warn people that the section is under discussion, tag it.
Personally, I think the "long standing version" (a term I am reluctant to use) is quite acceptable... Yes, I still prefer version A, but that is not the same thing as objecting to the other versions (which I don't). Version C is my second choice (for the above reasons)... and I could certainly live with your version if that gained consensus. Ultimately they all say the same thing in different words. Blueboar (talk) 01:25, 14 May 2010 (UTC)[reply]
Threaded discussion - Break 1

My motivation in all this is not to change policy but to make it clearer. I guess we have different ideas about how to do that. WP:NOR will be useful in any case. I just hope that it can improve in clarity and editors understand it better, so that there is more productive editing in Wikipedia by more editors, with fewer disputes.

I first came up with version B because I saw all information as being divided between that which is published and that which isn't. I tried to say, without any jargon, that Wikipedia is not concerned with unpublished information, since this is the basis of NOR. And I thought that would be a good beginning of WP:NOR.

Version B:

Wikipedia is an encyclopedia which summarizes existing published information and is not meant to be a source of information that has not been published.

--Bob K31416 (talk) 03:10, 14 May 2010 (UTC)[reply]

Bob's description of version B would be better if it showed the version he is changing from, with strkeout text, as well as what he wants to replace it with. Also, for logical consistency, version B should say something like "a source of information that has not been published elsewhere", since Wikipedia is a publication. Jc3s5h (talk) 04:40, 14 May 2010 (UTC)[reply]
Please note that version B consists of only adding a first sentence in front of the current version C. See RfC: Lead sentence of WP:NOR. Thanks for your suggestions. --Bob K31416 (talk) 14:24, 14 May 2010 (UTC)[reply]

All material in unreliable sources is original research?

Should the "No original research" policy use a definition that indicates all material in unreliable publications is original research? Jc3s5h (talk) 17:49, 13 May 2010 (UTC) revised 18:48 UTC.[reply]

I believe the definition of original research should be changed as follows:

The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources.

The current version confounds the concept of an idea being original, and the concept of being verifiable in a reliable source. This NOR policy should confine itself to originality, and the Verifiability policy should confine itself to verifiability. To do otherwise reduces the comprehension of the concepts because readers won't be able to tell which is which. It also encourages regular readers of the policies to use non-standard definitions of the phrase "original research" and confuses new editors who receive incomprehensible edit summaries when their contributions are deleted.

An example of why this definition is nonstandard, consider that corporations often practice defensive publication in non-peer-reviewed publications to prevent a later would-be inventor claiming the ideas as the would-be inventor's original idea. This illustrates that as far as the rest of the world is concerned, an idea is no longer original once it is published, whether the publication is good, bad, or indifferent. Jc3s5h (talk) 18:04, 13 May 2010 (UTC)[reply]

  • Oppose. I don't understand what point is being made, but removing that reliable sources are required would obviously be a major change to NOR, as well as to V, NPOV and BLP. SlimVirgin talk contribs 18:07, 13 May 2010 (UTC)[reply]
  • There would be no change whatsoever, only a more clear explanation about why certain contributions should be rejected. If I try to include material from the fake autobiography of Howard Hughes as if it were a true autobiography, my contribution should be deleted with a edit summary that mentions WP:V. If I add an article about my own proposal to replace the Gregorian calendar, the article should be deleted and the main theme of the deletion discussion should be this policy.
  • SlimVirgin's approach, if taken to it's logical conclusion, would be to have one policy, and every policy-releted deletion could just have an edit summary that says "Violates the WP:Master Policy. Jc3s5h (talk) 18:19, 13 May 2010 (UTC)[reply]
  • I'm afraid I don't understand your point, in that case. It's true that I don't think V and NOR should be separate, but I don't think that's the point you're making here. SlimVirgin talk contribs 18:21, 13 May 2010 (UTC)[reply]
  • My point is that there is more than one reason for rejecting a contribution. One is it is original (or appears to be original because there is no source cited and the challenger can't easily find a source). A different reason is that although there is a source, and therefore it isn't original research, the source is not reliable and therefore the information is not verifiable. The current definition claims these to different reasons are actually one reason, and it would be just fine to delete contributions that cite rotten sources with an edit summary that says "Undo per WP:No original research". Jc3s5h (talk) 19:06, 13 May 2010 (UTC)[reply]
  • I'm sorry for not getting this, Jc, but I'm still not seeing your point. We define OR on WP as material that has not been published by a reliable source. We define reliable source loosely here. Within that context, I can't follow what you mean by "The current definition claims these to different reasons are actually one reason, and it would be just fine to delete contributions that cite rotten sources with an edit summary that says 'Undo per WP:No original research'." Are you saying that providing a poor source shows it's not OR? Not really, because we define OR as material for which no reliable source exists. These are all just working definitions. SlimVirgin talk contribs 01:24, 15 May 2010 (UTC)[reply]
SlimVirgin, It seems like the term "original research" has been used in Wikipedia for information that is originated by the editor who tries to put it in Wikipedia. If the editor didn't originate the information but instead cites an unreliable source written by someone else, then the material is excluded because the source is unreliable, not because the editor originated it.
For instance, suppose a professor gives a seminar on his unpublished results and hands out the notes of the seminar. Then suppose a Wikpedia editor uses those notes as a source for a Wikipedia article. Isn't it more natural to exclude that information by saying that it is from an unreliable source, rather than it is original research by the editor? Regards, --Bob K31416 (talk) 02:54, 15 May 2010 (UTC)[reply]
No, that's a misunderstanding. It may originate with the editor who adds it, but it may not. "Original research" = material for which no reliable source exists," as the policy says. This is why the ArbCom, for example, has several times ruled that using certain non-reliable sources amounts to OR. It's just a definition. OR = "material that is not attributable to a reliable published source." SlimVirgin talk contribs 02:58, 15 May 2010 (UTC)[reply]
Re "This is why the ArbCom, for example, has several times ruled that using certain non-reliable sources amounts to OR. " - I would have to see the case that you are referring to, since it may be that the editor may have originated information by using unreliable sources to form conclusions that weren't in the unreliable sources.
However, if ArbCom called the editor's contributions original research, even though the editor didn't originate the information himself, that would still be an unnaturally defined term. The information that the editor did not originate, could have been excluded from Wikipedia without calling it "original research", because it was from an unreliable source. Regards, --Bob K31416 (talk) 03:34, 15 May 2010 (UTC)[reply]
Bob, I think you are missing the point. On WP we define "OR" as "material not attributable to a reliable source". It may not be what first comes to your mind, but that's the way it's defined in WP:OR. And I think we all agree that both OR and V, as defined on WP, are misnomers. See my post below addressing the terminology issue. Crum375 (talk) 03:49, 15 May 2010 (UTC

<outdent>Re "I think you are missing the point." - Then I don't think you were able to understand what I wrote. I considered the possibility that the Wikipedian term "original research" was intended to be unnaturally defined by those like yourself, SlimVirgin and others who write policy and desire that definition. All I did was point out that if that was the definition, where information that was not originated by the editor was called "original research", it was a needlessly unnatural definition because the subject material of the editor could have been excluded because it was from an unreliable source, and there was no need to call it "original research." --Bob K31416 (talk) 04:05, 15 May 2010 (UTC)[reply]

There are two separate issues. We use "OR" as shorthand for saying "this material seems to me to be unattributable," which can also mean "the sources you have provided are unreliable or otherwise unacceptable." But if your point about "unnatural" is that the term OR is a misnomer and can cause confusion for newcomers, we are in agreement, and that's what I tried to address in the section below. Crum375 (talk) 04:15, 15 May 2010 (UTC)[reply]
I've known for a long time that the term "original research" had a special meaning on Wikipedia, but not this special. I don't think many people on Wikipedia are aware of this very special meaning of "original research" that includes information that is not originated by an editor. If that actually is what WP:NOR currently states, it is an easy thing to clear up because I think very few Wikipedians make that interpretation. Then the present RFC is the type of change that is needed to fix WP:NOR. Your remarks have only made the change of the present RFC more valuable. --Bob K31416 (talk) 04:40, 15 May 2010 (UTC)[reply]
Bob, here is what the policy actually says, taken over the last five years (bold added):
  • 2005: "What is original research? Original research refers to material added to articles by Wikipedia editors that has not been published already by a reputable source."[5]
  • 2006: "Original research is a term used in Wikipedia to refer to material that has not been published by a reliable source."[6]
  • 2007: "Original research is a claim for which no reliable source can be found. Producing a reliable published source that advances the same claim taken in context is the only way to disprove an assertion that a claim constitutes original research."[7]
  • 2008: "to demonstrate that you are not presenting original research, you must cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented."[8]
  • 2009: "To demonstrate that you are not presenting original research, you must cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented."[9]
  • 2010: "The term original research refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources."[10]
As you can see, "OR" = "not published by RS". It's been this way for years, with no significant change, so this is clearly what we mean by "OR" on Wikipedia. Crum375 (talk) 12:16, 15 May 2010 (UTC)[reply]
I think we should go back to the use of "reputable". Reliable sources is what's required for verification of facts. WP:OR, which would be better framed as WP:Attribution, is about opinions, deduction, synthesis and analysis, not facts, and it is the reputability, not reliability, of the source that matters. That someone commented so, is not of interest because it was reliably sourced, but because it was sourced in a publication of repute. --SmokeyJoe (talk) 12:40, 15 May 2010 (UTC)[reply]
There are no "facts" on WP, only material which must be attributable to reliable sources to be included. WP:OR is simply WP shorthand for "material not attributable to reliable sources." The quality of reliable sources, which includes their "reputability", is addressed in WP:SOURCES. Crum375 (talk) 12:56, 15 May 2010 (UTC)[reply]
RE: facts. Have another read through Carbon. It is jam packed with facts, each verifiable to a primary source. NOR has no relevance to such articles. Wikipedia includes an abundance of facts, precise and incontestable. They are not spread uniformly across all articles. --SmokeyJoe (talk) 13:24, 15 May 2010 (UTC)[reply]
Today's fact can be tomorrow's fiction, and vice versa. It was a "fact" that Pluto was a planet up until four years ago. Dark matter and dark energy were considered science fiction by many until recently. We don't differentiate Wikipedia material based on what people call it, or whether it's considered "factual" or "true". The only relevant criterion we care about, for material to be added to this site (ignoring neutrality issues), is whether it is attributable to a reliable source. If it isn't, it's defined as "OR". Crum375 (talk) 13:45, 15 May 2010 (UTC)[reply]
OK, accept that facts vary perhaps continuously from incontestable to terminology, inferred and hypothetical. Don't agree that new editors understand why material sourced from less-than-reliable (how were they to know) sources is called original research. --SmokeyJoe (talk) 23:16, 15 May 2010 (UTC)[reply]
Crum375, According to what you've presented, WP:NOR says that information not originated by an editor, but appearing in an unreliable source, is "original research". Since such information from an unreliable source is simply excluded anyhow without that definition of "original research", would you like to see that unnatural aspect of the definition cleaned up with a simple change like the one in this RFC? Thanks. --Bob K31416 (talk) 14:32, 15 May 2010 (UTC)[reply]
I am sorry for being dense, but I am not still following. WP:NOR defines OR as material which is unattributable to a reliable source. This is the current definition, and has been consistently so for years, as I have shown above. So it's really very simple: either the material can be cited to a reliable source, in which case it is not OR, or it cannot, in which case it is OR. I don't see any "unnatural aspect" to the definition, or any in-between cases. The issue I referred to in my post below has to do with the choice of terminology and the artificial separation between NOR and V. Crum375 (talk) 15:04, 15 May 2010 (UTC)[reply]
Well, I tried and apparently I'm unable to get through to you, so have a nice day. Regards, --Bob K31416 (talk) 15:29, 15 May 2010 (UTC)[reply]
Thanks for trying. If you can think of a better way of explaining your concern, let me know. Crum375 (talk) 15:51, 15 May 2010 (UTC)[reply]

<outdent>Crum375, Hi again. Please note that the present RFC is for a change in the definition of original research. Are you aware of that? --Bob K31416 (talk) 15:58, 16 May 2010 (UTC)[reply]

  • Oppose I may be dense, but I don't understand why we should care about any material not published by a reliable source. We must emphasize that if we add any material not published by a reliable source, it is original research by our definition. Crum375 (talk) 18:25, 13 May 2010 (UTC)[reply]
  • I have revised the question to make it clearer. As it stands now, someone could add material from the fake Hughes autobiography as if it were actually written by Hughes. I could undo the edit and claim it is original research. But it isn't really original research, it's been published for decades. The current definition is FALSE. I also reject the notion of making up false definitions of words and then using the excuse that the definition only applies within this policy. Jc3s5h (talk) 18:53, 13 May 2010 (UTC)[reply]

For a view of how the outside world thinks about the meaning of original research, see this Nature policy. I believe I can boil their policy down: "if the main result, conclusion, or implications are not apparent from the other work" it may be original enough for Nature to consider publishing it. They do say that other factors might be considered, such as if the other work is not in English.

Of course, Nature wants to exclude work that is NOT original, while we want to exclude work that IS original. But still, Nature does not say previously published work might still be considered original, so long as the publication that published it is unreliable. Jc3s5h (talk) 19:19, 13 May 2010 (UTC)[reply]

  • Actually, I think Jc3 has a point here. It is an issue of the policy's scope... If I read a statement on some nutjob website, that material isn't original to me (the Wikipedia editor who wants to add it to an article). Thus, it isn't Original Research for me to use it (it would be piss-poor research on my part, but not original research). The scope of NOR is whether the material has been published... the "where" of publication falls within the scope of WP:V, WP:RS and other policies and guidelines. I would agree that we need to remind editors to look for reliable sources... but not in the opening few sentences... not in the definition of what OR is. Blueboar (talk) 20:13, 13 May 2010 (UTC)[reply]
    • I disagree. The current WP:NOR policy says (bold added): "If no source exists for something you want to add to Wikipedia, it is what we call original research. To demonstrate that you are not adding original research, you must be able to cite reliable published sources that are directly related to the topic of the article, and that directly support the material as presented." This is the essence of OR. If there are no reliable sources for some material, it is considered to be "original research" on WP. This does not mean the editor adding the material actually created or invented it out of thin air, or "researched" it, it's just that we call all unsourced or poorly sourced material on WP (for which an appropriate source cannot be found) "original research". That's been the policy and its definition since I came here, and it still makes good sense to me. Crum375 (talk) 20:38, 13 May 2010 (UTC)[reply]
Yeah... I know that is what the policy says, and that it has said it for a long time (do we know when that language was added?)... and, until now, I have agreed with everything in that passage. But I am beginning to think that it is wrong... and that we should remove the word "reliable", or mention it in a different way. The reality is, as long as you can show that your material has been published, you are demonstrating that you are not adding original research.
Remember, the fact that something passes WP:NOR does not automatically mean you can add it to an article. We do have other policies and guidelines. To be acceptable, material has to also pass WP:V, WP:NPOV and all our other policies and guidelines... However, as long as something is published, it simply isn't original research.
A few weeks ago, someone asked what the distinction between WP:NOR and WP:V was... I think this is a large part of it right here. When it comes to sourcing our material... NOR should tell our editors that our material must come from a published source. Leave it to WP:V to tell them that the source must be reliable. (of course, WP:ATT told them that the source had to be published and reliable at the same time... but...) Blueboar (talk) 21:13, 13 May 2010 (UTC)[reply]
I think the problem stems from the fact that WP:NOR sits on a shaky foundation as a separate policy. Logically, it should be an integral part of WP:V, and should focus, perhaps, on the issue of WP:SYN or other non-obvious unsourced material. In other words, we all understand that to say that "The Eiffel Tower is 1063 feet tall" we need a reliable source saying that. But when it comes to combining information from different sources and coming up with explicit or implicit conclusions, not everyone understands that we are not allowed to synthesize new conclusions to advance a position, while we are allowed, and even encouraged, to find reliable sources and use our own words to neutrally summarize what they write about an issue. Also, the distinction between the three different classes of reliable sources, WP:PSTS, and the need to avoid interpreting, analyzing or otherwise misusing primary sources, is not obvious to most editors. So at the moment, V focuses on the general need for reliable sources for everything we say, and NOR adds some fine tuning: no WP:SYN; descriptive, limited and careful use of primary sources; allowing no more than routine arithmetic calculations; how to handle translations, etc. So the problem is the basic title of NOR: it's really a subset of V. Ideally, it should just be subsumed by it, under WP:ATT, in which case we could just say, "this material is not attributed and I hereby challenge it". And if no source is provided within a reasonable time, "this material is not attributable, so I am removing it." No NOR terminology would be needed once we upgrade to this combined policy. Crum375 (talk) 22:07, 13 May 2010 (UTC)[reply]
  • Support such changes that remove statements that are not true in the normal use of English outside Wikipedia. "Original research" is material that is arguably reasonable, but that has not been previously published. "Original research" does not include nonsense, whether original nonsense, or published nonsense, or publish truth found only in unreliable sources. Original research is not the only reason to refuse material, and we should not expand definitions beyond readily accepted language. To do so is to make our most basic, prominent, introductory documents opaque to the uncultured, and this makes it difficult for new editors. It would be better for this policy to say "Material not attributable to a reliable source is not acceptable in Wikipedia." NOR explanations can go in the lower rationales. --SmokeyJoe (talk) 20:57, 13 May 2010 (UTC)[reply]
  • Comment The policy passage quoted above seems inconsistent, unclear. The 1st sentence simply says "no source", but the 2nd suddenly starts talking about reliable ones. Peter jackson (talk) 09:47, 14 May 2010 (UTC)[reply]
  • Support the removal of the mention of reliable sources from the lead of WP:NOR for the reasons given by Blueboar. However, I actually prefer the idea of merging WP:NOR into WP:V, as suggested by Slim[11] and Crum[12]. Yaris678 (talk) 12:08, 14 May 2010 (UTC)[reply]
    • Yaris, I took the liberty of fixing the link to my diff. Feel free to correct me if you meant something else. Crum375 (talk) 12:24, 14 May 2010 (UTC)[reply]
  • Support - "reliable sources" serves no useful purpose. Both "material not published" and "material not published by reliable sources" are original research. --Bob K31416 (talk) 16:16, 14 May 2010 (UTC)[reply]
  • Weak support to delete the expression "reliable sources" in the sentence under question. There's no point in us having two policies that mean the same thing. And we should normally use meanings that are the most commonly understood meaning. Doing so improves clarity. Maurreen (talk) 16:34, 14 May 2010 (UTC)[reply]
  • Comment - I think Jc3s5h makes a good point, though I don't see the harm of reiterating in three words ("by reliable sources") the need to use RSs. ... Kenosis (talk) 14:55, 15 May 2010 (UTC)[reply]
  • Support Original research is confusing, and it should not be further confused by making people think that content published in an unreliable source is original research. Such content is not removed because it is original research but because it is unreliable. Incidentally, unreliable should always be reasonably interpreted, and not interpreted in a bureaucratic way to mean simply mean published in a certain place. II | (t - c) 08:26, 20 May 2010 (UTC)[reply]

Resolving the issue

I think we can resolve the issue with a minimum of change... mostly by simply moving some content around in the second paragraph. Here is what I suggest...

Blueboar's Suggested Language

Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published . It also refers to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by the sources.
If no source exists for something you want to add to Wikipedia, it is what we call original research. To demonstrate that you are not adding original research, you must be able to cite published sources that are directly related to the topic of the article, and that directly support the material as presented. Wikipedia's Verifiability policy requires that sources also be reliable. This means that all material added to articles must be attributable to a source that is both published and reliable, even if not actually attributed. A source must be provided for all quotations, and for anything challenged or likely to be challenged — and a source must exist even for material that is never challenged. For example: "Paris is the capital of France" needs no source because no one is likely to object to it, and we know that sources for that sentence exist.

Comments on Blueboar's suggestion

I think this language would make it clearer that while reliability isn't an OR issue... because of WP:V, you still need to be able to cite a reliable source when dealing with OR issues. Blueboar (talk) 16:56, 14 May 2010 (UTC)[reply]

I am sorry, but I fail to see why Wikipedia should be concerned with non-reliable sources. For us, if it's not reliable, it does not belong on this site, period. We call any material for which no reliable source can be found "original research", per our longstanding tradition. We can do away with the OR terminology by simply focusing on attribution, and clarifying that all material on Wikipedia must be attributable to a reliable source. Crum375 (talk) 17:28, 14 May 2010 (UTC)[reply]
I don't disagree... no where in my suggestion does it say say that you can use unreliable sources... all we are pointing out is that NOR requires published sources while WP:V says those sources must also be reliable. That's all. Blueboar (talk) 17:42, 14 May 2010 (UTC)[reply]
But if we are all in agreement that NOR is a subset of V, there is no special allowance for unreliable sources in NOR any more than in V, its parent. A source is either reliable, or not. Any material which does not have a reliable source is "original research", by our longstanding definition. And if material is original research, i.e. is not attributable to any reliable source, it does not belong on WP, period. If people don't like the terminology "original research", let's get rid of it by focusing on "attributability". Crum375 (talk) 17:50, 14 May 2010 (UTC)[reply]
See also my expanded points below. Crum375 (talk) 18:33, 14 May 2010 (UTC)[reply]
Certainly there are varying degrees of reliability, and also varying degrees of the extent that the source's reliability has been established in Wikipedia, including per the attributes defined in the two policies. North8000 (talk) 16:22, 16 May 2010 (UTC)[reply]
I don't think NOR is a subset of V... they are definitely related and overlap a lot... but they are distinct concepts. I think they are both sub-concepts of "Attribution" (which is why I was in favor of the WP:ATT merger way back when). Blueboar (talk) 20:01, 14 May 2010 (UTC)[reply]
As I see it, NOR is more subtle, since it focuses more on synthesized material which seemingly relies on reliable sources, while V focuses on the general need to supply sources to any added material. But I think we are in agreement that these distinctions in emphasis still belong under a single "attribution" policy. Crum375 (talk) 21:04, 14 May 2010 (UTC)[reply]
We should not misuse the phrase "original research" because Crum375's "longstanding tradition" is a tradition of the crowd that infests Wikipedia talk pages, not a longstanding tradition of the English language.
I actually support the idea of an attribution policy that merges verifiability, no original research, and the reliable sources guideline. But since Jimbo Wales didn't like it when that was attempted, that has little chance of happening. So I will settle for not torturing the English language.
Blueboar's language could be a way to go, but I think it is too long because it repeats (with attribution) too much material from other policies. Jc3s5h (talk) 20:59, 14 May 2010 (UTC)[reply]
I gather you don't think the material is worth repeating? Blueboar (talk) 00:30, 15 May 2010 (UTC)[reply]

I think repeating material in one policy in another makes both of them prone to becoming nonsense when one policy is edited and the other isn't. Also, both policies become so long that people overlook the provision that applies to the case they are interested in, because it's like looking for a needle in a haystack. Jc3s5h (talk) 00:41, 15 May 2010 (UTC)[reply]

I think that this proposal is an improvement. These changes will make this page more closely describe what the community does and what it means when it points someone to this page.
For Crum (and SV, and others): The real policy is what the community does and says. This page is currently (and has been for years, as proven above) an inaccurate description of the community's policy. The community does not—and never has—say that every single unattributable/unverifiable statement violates WP:NOR. Consider the case of fat-fingering a statistic while typing: The resulting claim ("01% of women will develop breast cancer" rather than "10% of women") is unattributable to any reliable source. But the community will revert it with an explanation that it violates WP:V, not WP:NOR.
The reason for this is that the community needs to be able to make a distinction between specific kinds of "no source says that", so that editors know exactly what needs to be corrected. Making NOR's scope cover every single sort of "no source says that" does not help the community.
As an analogy: Among older folks in parts of Europe, "scientific" is a synonym for "academic". So you get old men nattering on about their "scientific reading" of metaphysical poetry, which leaves all the physicists and chemists wondering how you run experiments on the meaning of an old poem. Conflating NOR with all of WP:V is just as silly as conflating "science" with "all of academia". WhatamIdoing (talk) 04:53, 17 May 2010 (UTC)[reply]

Terminology

I think most people agree with the core concepts of V and NOR. The problems seem to stem from poor terminology. There are two specific terminology issues which are perennial trouble makers, and they happen to be the very names of these two policies, both of which are unfortunate misnomers. "Verifiability" implies "verity", or "truth", yet its very first sentence tells us (correctly) that we don't care about truth, only sources. "Original research" implies Wikipedians working furtively in their secret basement lab and coming up with their own original inventions, while the policy tells us "to demonstrate that you are not adding original research, you must be able to cite reliable published sources." In other words, we don't care if you or someone else came up with this material, if it's not attributable to a published reliable source, it does not belong on WP. So both key policies start off with a bad title and use bad terminology. Does it surprise anyone that we keep running into confused editors and endless talk page threads?
The solution is simple: let's combine the two, and focus on attributability. The key: All material on Wikipedia must be attributable to a reliable source. And any unattributed material which is challenged, likely to be challenged, or quoted, must be cited inline to a reliable source. All other details are secondary. We can explain that synthesizing material from reliable sources to advance a position is not allowed, while searching for reliable sources and neutrally summarizing material from them, without creating (or implying) new conclusions is part of normal editing. We can explain that reliable sources fall into three classes, WP:PSTS, and explain that primary sources should be used sparingly and carefully. We can explain that simple arithmetic is allowed, but beyond that, reliable sources are required, and so on. In other words, all material currently in WP:V and WP:NOR can be included in WP:ATT, and the result will be simplicity, clarity, and harmony. Comments? Crum375 (talk) 18:16, 14 May 2010 (UTC)[reply]

I would definitely support any movement to revive WP:ATT. Unfortunately, I think you will find that there will be strong resistance to such a move... and as many people will oppose the idea as did the first time around (You might ask at VPP just to see if there is any favorable consensus.) Blueboar (talk) 20:11, 14 May 2010 (UTC)[reply]
I think at least part of the problem last time was that, despite our attempts to publicize it and get more inputs, some people still felt left out of the process. This time we would need to get very deep participation, with emphasis on addressing any counter-points raised by those who opposed in the last round. I think it would be very helpful to hear people's objections, even if nothing else comes out of it. Crum375 (talk) 20:36, 14 May 2010 (UTC)[reply]
In my opinion, there are 4 foundational points which govern article content (I'm presenting a simple view of each):
  1. Verifiability - all material must be attributable to a reliable source
  2. Non-synthesis - Do not combine sources to make a novel point
  3. Neutral language - Describe subjects objectively
  4. Due weight - Coverage in the article should be in proportion to that of published sources.
Right now, points 3 & 4 are both included in WP:NPOV, since they are closely related. A proposal to combine points 1 & 2, which are also related, makes sense to me.  --Joshua Scott (LiberalFascist) 20:28, 14 May 2010 (UTC)[reply]
I'm going to start trying in my userspace to work on a version of ATT that (a) combines NOR and V, (b) is as tight and as clear as possible, and (c) that takes into account some of the objections to both that I've seen over the years. If I can create a working first draft I'll then invite other people to take a look, and maybe we can produce something that would gain consensus. SlimVirgin talk contribs 00:58, 15 May 2010 (UTC)[reply]
  • Could we pretend to ignore the existence of WP:A and fix the terminology issues here. To help the new contributors, guidance should be simple and positive (not using "not"s). "All material on Wikipedia must be attributable to a reliable source" is good. "Wikipedia publishes material attributable to reliable sources" may be even better. OR is merely a common example of material not attributable to reliable sources, and is prominent only because there are people determined to promote (in good faith of course) their own whatever. --SmokeyJoe (talk) 03:13, 15 May 2010 (UTC)[reply]
    • Forget the existence of WP:ATT for a moment. WP:OR is simply an adjunct to WP:V. V says, effectively, that all material must be attributable to a reliable source. OR clarifies that "material" may also be an implication or synthesis created from reliable sources, and such implications or syntheses must also be attributable. OR also explains the concept of "primary source", and cautions editors that it must be used sparingly and carefully, and so on. The point is that OR is in reality an amplification of V, focusing on the more subtle aspects of "materials" which must be attributable. So it seems we are in agreement that to stop the confusion we have today, V and NOR should be combined. It also seems that WP:A ("everything on WP must be attributable") is the best way to do it, though other ideas are of course welcome. Crum375 (talk) 03:31, 15 May 2010 (UTC)[reply]
Speaking as just one WP user, one thing I think I can guarantee right now is that a combination of WP:NOR and WP:V into one page such as WP:ATT will definitely not stop or substantially reduce arguments over how policy affects practice in WP. Right now it's two policy pages covering related material but where WP:NOR is a much more nuanced and subtle extension of WP:V. Some folks seriously think that making logical sense of what it's called (specifically the argument has been to call it "attribution" rather than "verifiabilty" and "no original research") will reduce disputes and let everyone rest content that all the major logical flaws have been resolved and that the policy talk pages will now proceed to settle down from what some see as unnecessary arguments. I personally disagree with this implicit assumption, and think that a change of existing foundational policy (the three-part core content policy into two policies instead of three) will result in essentially no net improvement in how editors manage disagreements with one another and w.r.t. how this page and that at WT:V play out for the more experienced users who try to resolve issues on these talk pages.
....... Personally I like the idea of WP:ATT, conceptually at least, though the community roughly two years ago rejected it rather vociferously after it was labeled a policy page upon somewhat WP:CONLIMITED discussion among a limited number of users who happened to be regular WP "policy wonks", so to speak. A major issue is, of course, after nearly a decade of WP existence, countless WP users have come to rely on the basic existing format. What is proposed is a major change in policy format that has almost a decade of generally successful practical results for WP. Recall (or if you weren't here then and tracking this basic stuff, I suppose take somebody's word for it or go back and research it) that WP:Verifiability originally meant roughly that the expectation is for a contributor "to be able to to verify a particular piece of information" presented in an article. Since WP's origin, though, the expectation has increasingly become that users are expected to be prepared to cite the source of a contested piece of information in-line via a Harvard citation or equivalent.
....... As has been pointed out by Crum375 and others, there are very arguably more rationally acceptable ways of naming the policy in light of the growing expectation for in-line citations, and in light of the general similarity between V and NOR, that could readily be placed under one umbrella on one project policy page such as WP:ATT. I don't think it's necessary nor helpful on-balance to put all the same information on one page as opposed to two policy pages. IMO, the existing bifurcation of related policy provisions into two separate pages is not necessarily a net disadvantage to WP, nor need it necessarily be subject to WP:FORK standards, nor necessarily be subject to any other WP standard other than that they're the original three core content policies. Pretty much from the beginning (recall NOR was added after NPOV and V) they've been admitted to overlap and compliment one another. They are the original content policy pages, and they've been well consensused, and as I said countless WP users have come to rely upon this basic format. ... Kenosis (talk) 04:14, 15 May 2010 (UTC)[reply]

(ec)Kenosis, "WP:ATT will definitely not stop or substantially reduce arguments over how policy affects practice in WP." Agreed, and that is not a proper objective here. Basic policy is not for dispute resolution, but for introducing basic concepts to new editors. POV pushers will find a way to wikilawyer even if we can sand down the terminology rough edges. --SmokeyJoe (talk) 05:05, 15 May 2010 (UTC)[reply]

Wouldn't it be nice if... Anyway I think I said my piece, and I apologize if my last piece was excessively lengthy. ... Kenosis (talk) 05:17, 15 May 2010 (UTC)[reply]
I get what you're saying, explanation unnecessary, but thanks, and no, your piece was not too lengthy, I've read much written by you and never have I seen you write anything not worth reading.
You might be missing my point. A few terminology rough points ("OR" includes sourced erroneous material; "secondary sources" include anything repeated) are problematic to new editors, and non-editors contemplating getting involved. They mean that our use of language is opaque. I don't disagree that the existing language was consensused by existing (encultured) editors; I disagree that this is a reason to not fix cases of misuse of standard use of the language. --SmokeyJoe (talk) 23:38, 15 May 2010 (UTC)[reply]
I definitely encourage "sanding down" the language of this page wherever possible. IMO, there's certainly some of that kind of cleanup to be done. When attempting this, it seems to me that collectively we've had problems in reconciling on the one hand a desire for concise language containing just the basics for new editors, with on the other hand a desire for language that also captures subtleties with which more experienced editors are familiar. Sometimes, as experienced editors know, the language comes into play in content disagreements among editors, where the subleties can be important. And what we often end up with is language that reads like "committee language", which is pretty much what it is-- generally coherent, but with bits and pieces of contributions stitched together here and there that still retain a bit of a Frankensteinian character.
.......W.r.t. the two illustrations you just gave, I think I agree. Secondary sources could perhaps be better explained-- a mere repetition in another source is not necessarily a secondary source for the primary source from which the repetition is drawn. And unsourced erroneous material is not OR, but instead is properly a WP:SOURCES analysis. . ... Kenosis (talk) 02:20, 16 May 2010 (UTC)[reply]
I think that an analysis of the "nuts and bolts" of this yields:
1. Structurally the subject of WP:OR is a subset of the subject of WP:VER.
2. I'm assuming / it sounds like WP:NOR was created to address the more specialized area of WP not being the place to present new ideas, new theories, I.E. not the place for Original Research (per the outside world's definition of that term which is different than Wikipedia's.) I'm guessing that somewhere along the line it left that track and mostly reverted to it's logical underpinnings which is basically wp:ver.
3. When one tries to understand what is happening regarding OR/NOR out in Wikipedialand, one must understand one huge thing. In common use, "OR" has become the noun to refer to a(n alleged) violation of wp:VER.
I would personally prefer something like wp:att which would "start over". A good chance to fix its fundamental problems which tweaking will never do. One is the over-emphasis on primary/secondary/tertiary to define source suitabilities. The other is that a strict interpretation of WP:NOR & WP:VER as written defines 90% of Wikipedia as being in violation. But, rather than starting by moving the mountain of combining the two policies, why not just start moving material that is obviously simply about sourcing out of WP:OR into WP:VER? A good place to start would be the most obviously misplaced one, the section on Primary/Secondary/Tertiary source types.
Sincerely, North8000 (talk) 13:29, 16 May 2010 (UTC)[reply]
Note that I'm still waiting for evidence to support your claim that "90% of Wikipedia" violates the policies. I have not seen the slightest evidence that even 10% violates these pages. WhatamIdoing (talk) 05:20, 17 May 2010 (UTC)[reply]
Tangential -- I tried a very quick check to see about the 90% figure. I pulled 10 random articles (not counting two without content). If that set is any indication, about 80 percent of our articles have less than a page of body text. The longest article in the set was Scissor-tailed Flycatcher.
The only thing I noticed that might be in violation was Mirza Dildar Baig. Maurreen (talk) 06:33, 17 May 2010 (UTC)[reply]
I had set my sights on dissecting that entire article which is a few hour job which I haven't gotten done. Also just was "unhooked" on a beach for 5 days. I think I'll just take a paragraph out of a good article and do this. But think about it. Every sentence contains on average 2-3 explicit and implicit statements. As interpreted by the regulars here (i.e. ignoring the "challenged or likely to be challenged" statement, which is ignored/negated by the mechanics of the policy) each statement needs a reference that literally supports it, and by a reference must meet several tests. Most of the time it must be secondary. But more importantly, each reference must meet the wp:ver criteria for reliability, including "reliable, third-party (independent), published sources with a reputation for fact-checking and accuracy;...Sources should directly support the material as it is presented in an article". And the person putting / keeping the material in must prove all of the above.
Now that I have written this, I think that my "90%" figure might be low.  :-) North8000 (talk) 11:00, 18 May 2010 (UTC)[reply]
See new section I'm adding below North8000 (talk) 12:47, 18 May 2010 (UTC)[reply]

Small question

I was wondering: should we put somewhere in there that this policy isn't a requirement to blindly acknowledge every reliable-looking source as such? For example, I've seen videos of newscasts from normally reputable sources that are obviously skewed and/or fake. That way, if, say, someone uses a video from a Chinese news network to "prove" the massacres at Tiananmen Squre didn't happen (just an example), it can be removed despite being from a "reliable source". Basically, a statement like, "This policy does not require editors to stop using their heads when looking at sources. If a document, photograph, or video purports to be from a reliable source, but is an obvious fake, or the source's coverage is blatantly skewed so as to constitute a fringe opinion despite claiming otherwise, its removal on those grounds would not constitute OR". Or would this lead to excessive wiki-lawyering, and/or is this covered by IAR already? I bring this up because I've seen a few blatant fakes (mostly videos) in I/P topics get removed with that rationale, and the removal is hotly contested as being OR, (things like "saying it looks edited and fake is OR" when the video is clearly fake to amateur eyes). The Blade of the Northern Lights (talk) 05:48, 15 May 2010 (UTC)[reply]

This is a question about the reliability of sources. I think it would be better placed at WT:Identifying reliable sources. That said, this post does make the case quite well for the suggestion made at #All material in unreliable sources is original research? Yaris678 (talk) 08:31, 15 May 2010 (UTC)[reply]
I only brought it here because the argument against removing these videos was, "Well, your saying it looks fake is OR", I wasn't exactly sure where to take it. I'll move it as necessary a little later on- I've got some other business to attend to. But it's at least good to hear that my idea might just have something behind it. The Blade of the Northern Lights (talk) 08:44, 15 May 2010 (UTC)[reply]
Ha ha. That was a big misunderstanding by whoever said "Well, your saying it looks fake is OR". NOR applies to the content of articles not the selection of sources. If we demanded published sources that told us which sources we could use we would go round in circles. It sounds like different people have different ideas about what actually happened. I think the best guide in that context is the policy WP:NPOV. If the editors of your article can't agree on how to apply NPOV, you could post a message at Wikipedia:Neutral point of view/Noticeboard. Yaris678 (talk) 09:43, 15 May 2010 (UTC)[reply]
I can sort of see where someone could get confused about OR in that way, but I'll take your suggestion and move this elsewhere. In this particular case, it was a debate (mercifully over now) about a video of a celebration supposedly glorifying a mass murderer. The video was so obviously edited and fake that even I noticed, but someone didn't want it removed due to the rationale I mentioned above. Eventually it was, because it was also tangientially related at best. But I'll move it over to WT:Identifying reliable sources. Thanks!! The Blade of the Northern Lights (talk) 19:32, 15 May 2010 (UTC)[reply]

Argument that today's main page FA violates a thorough application of WP:OR+WP:VER

The title was much later significantly renamed from "Sampler that 90% of Wikipedia violates a thorough application of WP:OR+WP:VER". North8000 (talk) 20:31, 18 May 2010 (UTC)[reply]

I'll take today's featured article. "David Alexander Johnson" The first sentence is "David Alexander Johnson (December 18, 1949 – May 18, 1980) was an American volcanologist with the United States Geological Survey (USGS) who was killed by the 1980 eruption of Mount St. Helens in Washington." WP:OR+WP:VER would define this EXCELLENT opening sentence as containing 6 violations.

This sentence contains the following statements:

  • Born 12/18/49
  • Died 5/18/80
  • Was an American
  • was a volcanologist
  • worked for USGS
  • killed by the volcano

None of these are referenced (and the lead is not exempt). So, per WP:OR+VER there are 6 violations in just the first sentence alone of today's featured article. This is a reflection on the policies, not on the article. If I were a deviant personality, or if it would serve my ulterior motives to do so, I could (while referencing WP policies and claiming moral superiority)delete the first sentence or hang 6 tags on it, and it would be a lot of work for the author to get it back in, and it would be footnoted mess when they did.North8000 (talk) 13:59, 18 May 2010 (UTC)[reply]

Ledes are except from being sourced - it's an optional style, and sourcing is only required if a direct quote is used there or is a contentious fact (which none of the above are). As long as the salient points are referenced in the body, there's no OR. --MASEM (t) 14:30, 18 May 2010 (UTC)[reply]
WP:VER and WP:NOR make no exception for the lead, so this violates them. As a sidebar, wp:lead says leads: "should be carefully sourced as appropriate"
There is no exception to the lead in the sense that all material there must be attributable, and challenged material attributed, as in any part of the article. But as a style issue, the location of the citations, esp. those which are less contentious (such as the ones in the example), may also be in the later text. If something is very contentious, or quoted, the citation should be included inline, close the material it supports. Crum375 (talk) 14:54, 18 May 2010 (UTC)[reply]
WP:VER & WP:OR make no distinction whatsoever for the lead.
So, let's say that I, right now, challenge the lead sentence as being unsourced and hang 6 tags on it. Per WP:VER & WP:NOR, can the author just tell me to go pound sand and revert me, or do they have provide in-line citations for those 6 facts in order to put it back in? — Preceding unsigned comment added by North8000 (talkcontribs)
Most likely you would be reverted, and that revert would likely be supported because 1) the material is referenced in the body in line with WP:LEAD and 2) none of the statements are the type that would be considered extraordinary or contentious. If it was something more like "John Smith was considered the greatest volcanologist ever", then you may have a case. --MASEM (t) 15:18, 18 May 2010 (UTC)[reply]
When a claim is repeated in an article in order to make the article more readable, it is not necessary to provide a citation for every instance of the claim. Jc3s5h (talk) 15:20, 18 May 2010 (UTC)[reply]
(ec)If you feel some material in the lead or elsewhere requires a source, you can ask the editors where the source is cited. If they say it's cited lower down in the text and you would like it moved higher, then it boils down to a style issue, which should be settled by consensus. But there is no question that all material must be attributable, with challenged and quoted material attributed inline. The only question, which is a style issue, is where exactly, and in how many places, to place that citation. Crum375 (talk) 15:22, 18 May 2010 (UTC)[reply]
Respectfully, you didn't answer my question. What happens next (per WP:VER & WP:NOR) if I hang those 6 tags on it? And don't forget that WP:VER says that the burden of proof is on the person who wants to retain/restore the material. North8000 (talk) 15:28, 18 May 2010 (UTC)[reply]
Since you are on notice, you would be reported at WP:ANI. Jc3s5h (talk) 15:30, 18 May 2010 (UTC)[reply]
I don't understand what you said. North8000 (talk) 15:50, 18 May 2010 (UTC)[reply]
Masem is correct... every single one of those statements is not only verifiable, but verified by one of the cited reliable sources. The fact that the citations are not included where the facts first appear (in the lede paragraph) does not mean that the facts are not referenced. All of the facts are referenced and cited later in the article. Thus, there is no WP:V or WP:NOR violation. If you were to tag them, I would politely point this out to you and remove the tags. If you pressed the point, I would eventually report you at ANI. Blueboar (talk) 15:34, 18 May 2010 (UTC)[reply]
Would you have to prove that all 6 are sourced in order to do so, or just make a blanket statement as you did? If you reported me without wasting your time proving that all 6 are sourced, you would be reporting me for enforcing WP policies......the burden of proof is on the person who wants to restore / keep the material. North8000 (talk) 16:08, 18 May 2010 (UTC)[reply]
North, if I made a blanket statement "all FA's are unsourced!", and added {{fact}} tags to every sentence in every FA, it won't take long for me to get blocked, and for everything to get reverted. The reason is that to improve the encyclopedia, we need to focus on individual points, one at a time. Any campaign of mass addition or removal of material to well reviewed articles (including adding templates), even if "armed" with quotes from policies, however well intentioned, is considered WP:POINTy and disruptive. Crum375 (talk) 16:33, 18 May 2010 (UTC)[reply]

North, original research is material for which no reliable source exists; it's not material that doesn't have a footnote next to it. About leads, there is no exception for leads, but a convention has arisen, particularly in featured articles, that non-contentious material is not referenced. I don't agree with that myself, but it's quite standard. It arose, I believe, because of Wikipedians with a background in academia who were used to the idea of article abstracts. SlimVirgin talk contribs 15:37, 18 May 2010 (UTC)[reply]

I picked the first sentence of the lead, but I could have done the same with material that is not in the lead. North8000 (talk) 16:08, 18 May 2010 (UTC)[reply]
Can you give us an example of material in that article for which you're reasonably sure no source exists? SlimVirgin talk contribs 16:14, 18 May 2010 (UTC)[reply]
I think that we just did a role reversal. What you just described is what I think that the criteria/procedure SHOULD be in order to tag or delete it. Under the current policies (and the interpretation of the regulars here) I don't have to say I think that no source exists, I can tag or delete it for merely being unsourced and then the burden of proof is on the person who wants to keep it to prove otherwise. North8000 (talk) 16:40, 18 May 2010 (UTC)[reply]
North, you can point to something in an article and challenge it, and editors will have to show you where the source for that material is, and defend it if you claim it doesn't directly support the material. But you need to be very specific: you can't just wave your arms and say "it's all unsourced!" Crum375 (talk) 16:45, 18 May 2010 (UTC)[reply]
North, the reason I've been pushing you to go through this exercise is because I have strongly suspected (and now been proven right) that you have completely misunderstood the policies.
As an example: Please show me the sentence in WP:V or WP:NOR that requires inline citations next to every single instance of a non-contentious fact (not including direct quotations).
You couldn't find any such sentence? Well, I let you in on the secret: You couldn't find it because it's not there. None of our sourcing policies have ever required this. The policy is that material must be "verifiable", not "already cited". If a suitable source exists -- anywhere in the world, in any language, in any medium -- then the material complies with WP:V. (The point of BURDEN is to help us identify situations in which a suitable source probably does not exist.)
Also, you might benefit from reading WP:LEAD#Citations, which addresses this point explicitly. WhatamIdoing (talk) 17:03, 18 May 2010 (UTC)[reply]
Just to tweak what you said: if a source exists, whether or not it's supplied, the material complies with NOR. If it's a quotation, or is material that's been challenged or is likely to be challenged, and it has an inline citation, then it complies with V. SlimVirgin talk contribs 17:15, 18 May 2010 (UTC)[reply]

Answering several folks:

In this discussion section I am making a point. Which is that, the policies as written make 90% of Wikipedia material vulnerable to the abuses that I describe. The common abuses in real life typically aren't to make such points. The two most common are to selectively knock out material to POV an article, or the social misfits who just go around doing such things to articles (and never creating anything) just because they enjoy doing so, being social misfits.

As before, we are debating two different things. Y'all are talking about the principles and intentions of the policies, and how high-minded people would resolve issues regarding them. an area where I am in total agreement with you. I am talking about the unintended consequences of the policies as written, and that they make 90% of Wikipedia content vulnerable to such abuse, and that such abuse is widespread.

And I think I can answer the question that nobody has been willing to answer directly. If I were to sanctimonously hang those 6 tags on that sentence, the editor that created this excellent material is going to have to waste their time proving that all 6 are sourced in order to remove the tags. North8000 (talk) 17:23, 18 May 2010 (UTC)[reply]

North, if your point is "can an editor intent on disruption create disruption and waste people's time?" The answer is "yes", any free and open project like Wikipedia is "vulnerable" to people intent on mayhem. This means that no matter what the policies said, the disruptive users will find a way to waste our time, so we might as well focus on writing reasonable policies to help reasonable people who want to help, not those bent on disruption. Crum375 (talk) 17:41, 18 May 2010 (UTC)[reply]
Yup, these things happen: People misapply policies and guidelines all the time. Editors who know better fix the resulting problems whenever they encounter them. Depending on the perceived motivation, this particular mistake is either called a bad-faith challenge or a simple misunderstanding of the policy.
North, you have repeatedly asserted on this page that the actual policies require a silly level of inline citations. You have been repeatedly told that the policies only require that material be verifiable rather than already supported by a specific inline citation. Do you now agree that the policies mean what they say, and thus that the sentence you quote above is fully compliant with Wikipedia's actual content policies? WhatamIdoing (talk) 17:49, 18 May 2010 (UTC)[reply]
Per WP:ver anything challenged requires an in-line citation. Per both policies and per the regulars here, the challenge need be only that it is unscourced. The challenger does not even need to say that the don't believe a source exists, they just need to say that it is unsourced. So, once it is challenged for being unsourced, it is in violation until it gets in-line citations. (not that I agree that it should be so) North8000 (talk) 18:18, 18 May 2010 (UTC)[reply]
We don't try to read people's minds; if they challenge material which is not attributed, it needs to be attributed. That's all there is to it. It's simple, and it works. You have yet to demonstrate any problem with it. Crum375 (talk) 18:28, 18 May 2010 (UTC)[reply]
Those are two new topics / discussions and I'm getting writer's cramp for the moment.  :-) Peace. North8000 (talk) 19:06, 18 May 2010 (UTC)[reply]
North, the six facts you identified in this sentence have not been challenged and are not likely to be challenged. Do we agree that these six facts therefore comply with the policies, as written, in their current form? WhatamIdoing (talk) 19:09, 18 May 2010 (UTC)[reply]
YES, until some misfit challenges them simply for being unsourced at which time they become in violation until they get in-line citations. North8000 (talk) 19:19, 18 May 2010 (UTC)[reply]
So? When they are challenged, they will be attributed (if they are not already). Whether on the same sentence or on another one is a question of style, to be decided by consensus. But there is no "violation", and no problem. Crum375 (talk) 19:23, 18 May 2010 (UTC)[reply]
I was just answering WhatamIdoing's question. My "so" was already said.
But my main point is that there is a problem, and that the following change (paraphrased) would go a long way towards fixing it:
  • To take action (tagging, deletion etc.) against an unsourced statement, you should also either state disagreement with the statement or that you believe that it is unsourcable.
  • (handle the more stringent BLP and quotation sourcing requirements separately)
Sincerely, North8000 (talk) 19:48, 18 May 2010 (UTC)[reply]
North, I am sure you are familiar with "if it ain't broke, don't fix it." In this case, you have yet to demonstrate that anything is "broke". There is no need to insist on a reason for a challenge: all we need is a challenge. And since all material on Wikipedia must be attributable, it means that the person who added the material should have a source for it, and it should be cited. That's easy, simple, and it works. Crum375 (talk) 19:56, 18 May 2010 (UTC)[reply]
Well, I guess we'll have to agree to disagree on several of those points. Sincerely, North8000 (talk) 20:09, 18 May 2010 (UTC)[reply]
North, if some "misfit" editor were to challenge your six facts for being "unsourced"... I would do exactly as I am doing with you... I would point out to them that these facts ARE, in fact, sourced... every single one of them. The facts are repeated (with citations) in the "Life and carrer" section of the article (the first section after the lede). I would then point the misfit editor to WP:LEDE to explain why they are cited in this section and not cited in the lede. I would explain that WP:BURDEN has in fact been complied with and that there is no WP:V or WP:NOR violation. If they insisted that the facts be cited in the lede... no problem... it takes but a second to cut and paste these existing citations. In short, what you think is a problem... isn't a problem. Blueboar (talk) 20:27, 18 May 2010 (UTC)[reply]
WP:VER says that once challenged they need in-line citations. But we're going in circles. Though somebody changed the title, this started out as a sampler (example), not an all-inclusive example of the things covered by my assertion. There are many other components. Any high level of summarization is wp:synthesis, most sources do not meet the criteria of wp:ver, etc. I think that yourself and Crum are committed to defending and keeping 100% the status quo regarding this (apologies if I misunderstand) I think we have to agree to disagree. I came here emboldened to pursue this issue because it is the policy most often questioned by experienced editors and most often abused by abusers. Probably I'll get worn down and give up and leave this discussion like all of the others. North8000 (talk) 21:05, 18 May 2010 (UTC)[reply]
I am definitely not committed to status quo ... In fact I have several issues with the current language of the policy. However, I also accept that current consensus does not agree with my view point, so I don't belabor the point. Blueboar (talk) 02:08, 19 May 2010 (UTC)[reply]
I'd also take Blueboar's basic approach in that instance. That is, I'd first see whether I could convince the concerned editor to withdraw the challenge.
North, one of the reasons that we don't require any bureaucratic overhead when a fact is challenged is that the reason for the challenge has no practical effect. Here's the system:
I challenge because... Your response is...
I think it's factually wrong Provide an inline source
I think it's unsourceable Provide an inline source
I'm disrupting Wikipedia to prove a point Provide an inline source
I'm wikilawyering against a POV I hate Provide an inline source
My teacher won't let me cite Wikipedia directly Provide an inline source
I'm feeling grumpy Provide an inline source
I don't understand the content policies Provide an inline source
The moon is full Provide an inline source
The sky is blue Provide an inline source
I'm a jerk Provide an inline source
Given that your response is always the same, it actually doesn't matter what my excuse (nominal or real) is for challenging material. If I challenge it, then you provide an inline source, full stop. Why I (claim to) have challenged it is irrelevant.
If you want to discuss your (erroneous) assertions that most sources used on Wikipedia fail WP:V and that summarizing sources violates WP:SYNTH, then I suggest starting new sections. WhatamIdoing (talk) 21:43, 18 May 2010 (UTC)[reply]
Excellent! That should be in the policy. :) SlimVirgin talk contribs 01:11, 19 May 2010 (UTC)[reply]

I don't understand your logic. You seem to be seeing that sameness of the response means that the cause is irrelevant. That does not follow. But either way, it does not refute my assertion that my proposal would reduce abuses. North8000 (talk) 23:59, 18 May 2010 (UTC)[reply]

He is saying that requiring an "explanation" for a challenge is useless, and would only increase discord. It's better to just say "I challenge" and then to provide the source. Crum375 (talk) 00:04, 19 May 2010 (UTC)[reply]
What abuse? None of the causes I named above are prohibited. Several are silly and several are stupid, but none are actually prohibited. WhatamIdoing (talk) 00:46, 19 May 2010 (UTC)[reply]
  • I think North8000 has some point, but it is so subtly complex that I think he should write an extended essay on it for him to be better understood. I suggest that he illustrate this essay with examples of good and bad practice within and outside wikipedia, and suggest solutions. Does he suggest referencing and cross-referencing multiple times per sentence, or does he suggest something else? --SmokeyJoe (talk) 00:17, 19 May 2010 (UTC)[reply]
Maybe you are right. But I'm not sure of the Wikipedia "syntax" on creating an essay. Do you just create an article except with a Wikipedia:xxxxx name and then put an essay template on it?. BTW I think that the article is fine. North8000 (talk) 02:54, 19 May 2010 (UTC)[reply]
Yes, you just write it and tag it. The official rules are documented here, but the short version is: Any editor can write whatever essays s/he wants. If there are problems (i.e., it is a strongly anti-consensus viewpoint), then you'll hear about either a proposal to move it into your userspace (where editors are given even broader discretion), or to have it deleted at WP:MFD (usual rules apply; most editors argue from common sense for essays). WhatamIdoing (talk) 05:08, 19 May 2010 (UTC)[reply]
Please remember these are editorial policies which imply--indeed presume--a very broad extent of discretion at every stage of the writing and editing process. The three core content policies describe methods by which editors are essentially a conduit for summarizing facts and ideas w.r.t. any given topic in the "entire world", so to speak. W.r.t. requiring sourcing, the WP:V policy says content must be cited to reliable sources if it's "contested or likely to be contested", as opposed to e.g., arguing endlessly over what's a fact or what's the truth. This has pretty much from the beginning been a standard WP editorial principle. Which has in turn, for example, lead some to argue on this and other policy talk pages that we don't deal in facts, but only in verification of what others say. But that argument is very incomplete because there are editorial decisions to be made in conveying the material to readers. Which leads me to three interrelated assertions at the moment.
..... First, the core content policies prescribe the basic methods and proscribe the approximate limits beyond which editorial discretion is not to be exercised.
..... Second, the content policies describe a basic editorial method across an extremely wide range of topic areas, many of which are hardly controversial at all, but a very significant number of which are definitely controversial in one respect or another, with all shades of controversiality-grey in between.
..... Third, there is a very big difference between article writing that's more than just citing "Advocate A says x about y and advocate B says z about it" and on the other hand "original research and original synthesis. Here too there is a broad range of discretion, shades of grey and permissible disagreement over both the facts according to reliable sources and how to best express them encyclopedically.

IMO, a broad accusation of multiple "violation[s]" of policy in a featured article as asserted at the top of this talk section, or in "90%" of articles as also alleged by North8000, is very misguided, mistaking editorial policy for some kind of draconian mandate about how contributors are to do each and every thing in the never-fully-complete editorial process in this project. Which is why WP:IAR is also a treasured part of the ever-evolving WP project--however imperfect it may be. W.r.t. the notion that the featured article on David Alexander Johnston has "violated" WP:NOR and/or WP:V as alleged at the beginning of this talk section, I have yet to see anyone contesting the notion that Mr. Johnston was, e.g., born on 18 December 1949, or any other basic fact about Johnston. If someone has reason to contest or be likely to contest this or any or any other fact about Johnston, I have yet to see it demonstrated here in a way that implicates fault in WP policy or in the mentioned article about Johnston. .... Kenosis (talk) 04:41, 19 May 2010 (UTC)[reply]

Thank you to WhatamIdoing for the info and Kenosis for the overview. I think that SmokeyJoe a few lines above that was right. I tend to think structurally, and how elements can interact including creating unintended consequences. In the outside world I often get called on to write by-laws, policies, constitutions of organizations, articles of incorporation, and occasionally laws. (also in running organizations and a company) My strengths in that area are more as a logician and understanding cause / effect and the related people dynamics rather than writing skills. I see that I have not communicated my central theme in this section. Basically, it's that when there is a large disparity between the "letter of the law" and reality, i.e. where most things that get done get done by bending or breaking the rules, you have a problem, including making the situation ripe for abuse. (And I was using one sentence of one FA as a quick attempt to show that such a disparity exists.) And when one wanders Wikipedia, one can see that there ARE some problems which are big and getting bigger. Wikipedia is a huge success story. The core policies are fundamentally good, and are just what Wikipedia needs. They also have some granularity, structural, organizational and unintended-consequences issues where fixing them would help Wikipedia, including in those areas where there are problems which are getting bigger. I might start that "essay" section....if so I'll mention it here. Who knows.... a long shot....if done a certain way, it could become a collaborate area for analysis and work on things which which are too long term / complex for successful handling a talk page+tweaks-on-a-major-policy format. A big thank you everyone, even to those I disagree with. Sincerely, North8000 (talk) 11:28, 19 May 2010 (UTC)[reply]

  • No offense, but why is this thread continuing? At most, this is not an original research question but a verifiability one. That should have been stated and then the thread left alone so we don't continue a long, unproductive discussion and distract from other threads with actual issues. I realize it's ironic for me to be saying that, but I think it needs to be said - and I'm ignoring an oft-used acronym which comes to mind out of civility. Incidentally, North8000, I try to reference leads if there's something that's likely to be questioned. If you really want to see them cited, look through the article and add a cite to an obituary to the lead, which would likely cover all the facts. I also sometimes cite every single fact, which can lead to the same citation being used multiple times in the same paragraph. Some people don't like the style but I think the convenience of verifiability and the prevention of my addition being deleted makes it worth it. North8000, I also see you're saying that you're trying to make a larger point about Wikipedia's ignore all rules but I don't see any constructive proposals here. II | (t - c) 08:23, 20 May 2010 (UTC)[reply]
I was done with this and signing off. My point was summed up in that last paragraph (and goes to the root of lots of actual issues) and quite different than you described. Sincerely, North8000 (talk) 11:27, 20 May 2010 (UTC)[reply]

POLL: Constant misunderstandings about NOR and my proposed solution

I note the the 'Terminology' section above [13]. Indeed huge amounts of WP time are constantly wasted by eds misunderstanding the NOR policy and OR. This is because the title is misleading, and focuses on 'research' rather than 'opinion'. Editors do not undertake research when they add their personal opinion to articles. So they don't think they have done anything wrong when they add their opinion, and don't understand that the only opinion WP wants is restricted to what can be found in secondary sources. Over and over again I have seen editors told their edits are OR, and they appear to be baffled, saying 'but I haven't done any research, this is a known fact'. It would be helpful to them, I think, if they aren't told that they have conducted 'Original Research' but that they are 'Opinion Restricted' to what they can find in secondary sources. Consequently I propose NOR should also stand for 'New Opinion Restricted' and OR should also stand for 'Opinion Restricted'. These would simply be additional terms that we add to the lede of this article. Perhaps the description of 'Opinion Restricted' could also state the case that WP is not advocacy journalism, nor is it a Third-person omniscient narrative with intrusive narrators and omniscient narrators. The result, I hope, is that editors will understand faster that their personal opinions are irrelevant, and will steer them toward sourcing. Please vote below. Obviously I strongly support this addition to the NOR article. -Chumchum7 (talk) 08:28, 20 May 2010 (UTC)[reply]

In practice "OR" has become the noun for any violation of wp:ver. I think that its use has become much broader than just "Research" or "Opinions". But it might help. North8000 (talk) 12:14, 20 May 2010 (UTC)[reply]
Perhaps this policy could have been named differently when it was created, but I think we are stuck with it. Unfortunately, there are some things that have become so ingrained in Wikipedia that it is all but impossible to change them. The term "Original research" and the abbreviation "OR" may have evolved into "Wikipedia Jargon"... but that jargon is so much a part of Wikipedia that trying to change it would be impossible.
Besides... NOR isn't just about new opinions (ie conclusions). It is also about not adding original interpretation and analysis and (most importantly) original synthesis of sources. Far more common than people who say "but I didn't do any research" are people who say "but this isn't my opinion... it's fact. I did exhaustive research on this, and my research proves what I say" (to which the appropriate response is: "you may well be right... but the NOR policy says that Wikipedia isn't the place to debut your proof. Go publish it somewhere else first, and then we can consider adding it to Wikipedia.) Blueboar (talk) 12:50, 20 May 2010 (UTC)[reply]
A better name for the policy would be no tenuous attribution. If someone adds something to Wikipedia based on analysis, interpretation or synthesis of sources and then cites those sources, it should be labbelled tenuous attribution. If someone adds something and cites no sources at all then that is an issue for WP:V. Yaris678 (talk) 13:17, 20 May 2010 (UTC)[reply]
A snappy alternative would be WP:STS... which could stand for both "Summarize The Sources" and "Stick to The Sources"... which are both what this policy is really all about. But, as we learned with WP:ATT... people like what they are used to, and dislike changing the names of established policies. Blueboar (talk) 13:28, 20 May 2010 (UTC)[reply]

I should emphasise I'm not suggesting we change the meaning or NOR nor am I trying to limit it. My point is that keep WP:NOR as it is, but add another definition, in order to deal with editors who add their opinion in good faith, and don't understand what 'original research' has to do with their edits, no matter how many times people try to explain it to them. Most time-wasting conflicts seem to start because editors wrongly assume they are entitled to include their personal opinions in articles. My proposed lede would be something like this: Wikipedia does not publish original research. The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources. It also refers to any analysis or synthesis by Wikipedians of published material, where the analysis or synthesis advances a position not advanced by the sources. A second, and equally important meaning of this policy is "new opinion restricted": Wikipedians may not add their personal opinion to articles. Opinion is restricted to what can be found in reliable secondary sources, and does not extend to what is in Wikipedians' hearts and minds. -Chumchum7 (talk) 14:44, 20 May 2010 (UTC)[reply]

If people add their own opinions and don't cite sources that happen to say the same thing then the policy they need to refer to is WP:V. Some people will say "I am removing your original research" but such statements are misleading. As with the earlier discussion, this confusion could be avoided if WP:NOR and WP:V were merged. Yaris678 (talk) 15:13, 20 May 2010 (UTC)[reply]
I agree... but how do we convince all the people who prefer to have two separate policies? Blueboar (talk) 15:46, 20 May 2010 (UTC)[reply]
I dunno. It looks like a really good idea from where I am standing and no one has really come out against the idea while it has been discussed recently. On the other hand, I was quite taken aback by the opposition to the seemingly obvious WP:PSTSPROP so I guess we can't just assume that common sense will prevail. Is it worth bringing up at the village pump? Perhaps we should wait till Slim comes up with a draft... or would that make people think it had been stitched up in advance?
What were the complaints last time? That the resultant policy was too long? That they liked the old terminology?
Yaris678 (talk) 21:50, 20 May 2010 (UTC)[reply]
You can read both the positive and the negative comments at Wikipedia:Attribution/Poll Blueboar (talk) 23:11, 20 May 2010 (UTC)[reply]
To really understand that poll, you have to remember the shock that ran throught the community when a prominent and powerful policy editor made this edit
https://secure.wikimedia.org/wikipedia/en/w/index.php?title=Wikipedia:No_original_research&oldid=111798785
(This is superseded by WP:ATT, as discussed on Wikipedia:Village pump (policy) and Wikipedia talk:Attribution.)
appear on watchlists. Note the bold red cross. Think what "superceded" means. Note that few were watchlisting WP:ATT and few keep up with editing at WP:VPP. Note the uniformative preceding history of edits to WP:NOR and WT:NOR. Out of the blue, a core policy was, fait accompli, overturned per an established concensus that the community didn't know about. In hindsight, no one should be surprised at the kneejerk reaction. It doesn't mean that what was done at WP:ATT was wrong, but it proved that can't move the community without keeping them abreast. Many of the comments at that poll were in kneejerk reaction, and as such, the meaning of the poll should be nullified. --SmokeyJoe (talk) 05:07, 21 May 2010 (UTC)[reply]
  • This is another thread which appears to conflate original research and verifiability. Whether something is an opinion, or whether the person who adds the content has an opinion on the content, is irrelevant; if something is unsourced it violates WP:V, and if something is sourced to a source which does not directly say what the source is implied to say, it violates WP:NOR. These are very distinct concepts and that's why separate pages is an absolute must. In response to Chumchum7: if a Wikipedia adds their new personal opinion to the article, but uses a reliable source (someone who holds their opinion), there's nothing wrong (although they should perhaps WP:ATTRIBUTEPOV). Chumchum7, can we mark this resolved and collapse? II | (t - c) 02:39, 21 May 2010 (UTC)[reply]
Blueboar, thanks for pointing me towards Wikipedia:Attribution/Poll. It made interesting reading. One thing that stood out to me was that quite a lot of people liked the idea of merging WP:NOR with WP:V, but they didn't like the idea of also merging in WP:RS. I completely agree with that position and note that we are not advocating merging in WP:RS/WP:IRS this time. My thoughts on the three pages are summarised by Rusty Cashman in point 97 of Wikipedia:Attribution/Poll #Neutral/qualified/compromise/other.
II, I agree with how you distinguish between WP:NOR and WP:V. However, most people seem not to be able to get that distinction into their head, which leads to people labelling everything OR. A merged policy would be simpler because people wouldn’t have to make the distinction. If we don’t go for a merge, an alternative suggestion would be to rephrase everything in WP:NOR that makes it look like WP:V. For example, it currently says “If no source exists for something you want to add to Wikipedia, it is what we call original research.” It should say “If no source exists for something you want to add to Wikipedia, it is what we call unverifiable.”
Yaris678 (talk) 07:51, 21 May 2010 (UTC)[reply]
Actually, I think the distinction can be made even easier...
  • If no source exists for a fact that you want to add to Wikipedia, it is what we call Unverifiable (cite WP:V).
  • If no source exists for an interpretation, analysis or conclusion that you want to add to Wikipedia, it is what we call original research (cite WP:NOR)
In both cases, the problem is lack of sources... the difference is what type of statement is not sourced. Blueboar (talk) 11:32, 21 May 2010 (UTC)[reply]
Having a large overlap between subjects and some duplicated content is not good. Something should be done. My advice would be to move the WP:VER type content that's in WP:NOR out of WP:NOR and into WP:VER, and leave WP:NOR smaller, focusing on the (the outside world meaning of) it's title. This would be much easier to do than to get a strong consensus on a merge and accomplish a merge. But a merge would also take care of the issue. North8000 (talk) 10:51, 21 May 2010 (UTC)[reply]

North, I like the idea but I don't think it is any more likely to gain consensus. If you remove the prohibition of X from WP:NOR, people will say "We can't allow people to do X! This is a fundamental principle of Wikipedia!" They will ignore the fact that X is also prohibited by WP:V. I think my idea of just rephrasing some sections of WP:V is more likely to gain acceptance... but even that could be problematic. The second sentence of WP:NOR is currently

The term "original research" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources.'

We could change it to either

The term "unverifiable" refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources.

or, since we are supposed to be writing about original research,

The term "original research" refers to material—such as facts, allegations, ideas, and stories—only tenuously supported by reliable sources.

Yaris678 (talk) 11:57, 21 May 2010 (UTC)[reply]

"original research refers to material...only tenuously supported by reliable sources": That's incorrect. On Wikipedia OR refers to any material not supported by reliable sources, period. Crum375 (talk) 12:33, 21 May 2010 (UTC)[reply]
Crum, you are certainly right, but by what you just said, wp:NOR is 100% the same as / a duplication of wp:ver and shouldn't exist. North8000 (talk) 12:49, 21 May 2010 (UTC)[reply]
I was thinking along the lines of moving 80% of WP:NOR to WP:VER, the remaining 20% would basically just "explain" / expand upon one aspect of / set of scenarios within wp:ver, that wp is not the place for putting forth new ideas, new research, new theories etc. In reality, that is sort of a merger, but an easier / more pragmatic way to get there. North8000 (talk) 12:49, 21 May 2010 (UTC)[reply]
Just some advice at this point... don't do anything so significant to this policy (and WP:V) without giving a LOT of warning... in as many venues as possible... especially at the Village Pump. If there is one thing that I have learned in more than five years of editing policy pages ... you ensure a knee-jerk rejection of any change if those who do not actively follow this talk page are taken by surprise. Blueboar (talk) 13:58, 21 May 2010 (UTC)[reply]
Sorry I was unclear. I didn't mean to sound like I was thinking of ME starting to make those changes. I was just suggesting that overall. North8000 (talk) 14:12, 21 May 2010 (UTC)[reply]

Distinction between original research and unverifiable material

Over recent posts there seems to have been many different ideas about what original research actually is and how it is distinct from unverifiable material. I can see some merit in each but it would be nice if we actually decided which to go for so that we don’t keep talking at cross purposes. The distinctions I can see are:

  1. Unverifiable means not supported by the sources whereas OR means only supported through interpretation, analysis or synthesis of the sources.
  2. Unverifiable means a fact not supported by sources whereas OR means an interpretation, analysis, synthesis or conclusion not supported by the sources.
  3. Unverifiable means not supported by the sources and OR means not supported by the sources.

Part of the difference between 1 and 2 is that in 2 the OR writer is making the analysis etc. explicit in the text whereas in 1 the OR writer is just relying on the reader to make the jump from the source to the conclusion. 2 also has the issue that someone interpreting the policy must be able to distinguish between a fact and a conclusion.

Those of opinion 3 tend to think the two policies should be merged. Some of those of the other opinions think they should be merged to avoid confusion whereas others hold the distinction dear. Yaris678 (talk) 14:44, 21 May 2010 (UTC)[reply]

I think there are two separate issues here, the meaning of OR on Wikipedia, and the need to keep the OR policy separate from V. As for the first issue, OR has had the same definition on WP over the last 5 years, which is "OR means material not attributable to a reliable source." Here are some sample definitions from OR policy over the last few years:
  • 2005: "What is original research? Original research refers to material added to articles by Wikipedia editors that has not been published already by a reputable source."[14]
  • 2006: "Original research is a term used in Wikipedia to refer to material that has not been published by a reliable source."[15]
  • 2007: "Original research is a claim for which no reliable source can be found. Producing a reliable published source that advances the same claim taken in context is the only way to disprove an assertion that a claim constitutes original research."[16]
  • 2008: "to demonstrate that you are not presenting original research, you must cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented."[17]
  • 2009: "To demonstrate that you are not presenting original research, you must cite reliable sources that are directly related to the topic of the article, and that directly support the information as it is presented."[18]
  • 2010: "The term original research refers to material—such as facts, allegations, ideas, and stories—not already published by reliable sources."[19]
Clearly the definition has not strayed much from the basic meaning of "OR = 'not attributable to a reliable source'." Crum375 (talk) 15:40, 21 May 2010 (UTC)[reply]
Unlike Crum375, I don't care what the talk page regulars think the terms mean, or how that group have used the terms in the past. Unverifiable, in the context of source-based research, means the information cannot be found in a reliable publication. Original research, from the point of view of a publisher, means unpublished. Jc3s5h (talk) 17:10, 21 May 2010 (UTC)[reply]
My message above had nothing to do with "talk page regulars". I was referring to and quoted what the WP:OR policy actually says, and has said, over the last five years, as its own definition of what it is. That reflects what the entire Wikipedia community means by "OR", not some limited subset. And just like "verifiability", WP has always had its own definition of OR, which may be different than that of other groups. Crum375 (talk) 17:24, 21 May 2010 (UTC)[reply]
The WP def of OR isn't just contrary to the definition of it in one or another group, it's in conflict with the universal connotation of that term in the English language. As long as that conflict exists, we can expect to have to keep repeating over and over and over that once you enter Wikipedia, the real world meaning of the term no longer apples. If you write down a famous 1,000 year old quote and it is unsourced, in WP that is "Original Research" North8000 (talk) 17:49, 21 May 2010 (UTC)[reply]
There is no such concept as "universal definition" of anything. Every group has its own definition of things. Some are more common than others, but none are identical everywhere. In the WP case, as I noted above in the #Terminology section, both "verifiability" and "OR" have a somewhat peculiar definition, which can cause confusion. But the actual wiki definitions of both terms have not changed much over the years. Crum375 (talk) 17:58, 21 May 2010 (UTC)[reply]
My comment was that the WP definition of the term is very different than the general meaning of the term outside of WP. Ask a bunch of non-wikipedia-familiar people what "original research" means, and I think that the answers will be similar and nowhere near the WP definition. Implying that I was saying there was a universal definition (definition being more specific/precise) is sort changing what I said to a straw man version of it. But no biggee. North8000 (talk) 19:50, 21 May 2010 (UTC)[reply]
Crum375, the 04:34, 10 June 2006 version of the original research page says in the lead: "the only way to demonstrate that you are not doing original research is to cite reliable sources which provide information that is directly related to the topic of the article, and to adhere to what those sources say". This is what original research is about, and it is not the same as verifiability per se. Content can be plenty verifiable - it can be attributed to a reliable source which directly says what the content says, but if it is in the wrong article, it can still be original research (VERY COMMON). Some of the quotes you provide reflect this understanding (2008, 2009), but others don't. Something can be verifiable, but if the verification requires too much particular expertise and analysis (which may nevertheless be completely valid), it is original research. In other words, it can be "attributed" to reliable sources such as original scientific papers, but it can still be original research. It seems as if you're trying to pull out random sentences from versions of the article to try to make the point that verifiability and original research are broadly synonymous and mean that there aren't reliable sources. Since there are 3-4 threads which have this misunderstanding here, it's no surprise that certain versions of the policy have had an unsophisticated and basically incorrect understanding of the actual policy. It's also worth point out that if something is uncited, it can either be original research (no source around) or unverifiable (person forgot to add a source). II | (t - c) 21:07, 21 May 2010 (UTC)[reply]

North8000: you're right that the terms here don't match the terms in other areas. I agree with Crum that the definition of "original research" is essentially the same as "unverifiable". The only possible change over the years has been an increased focus on synthesis. The basic meaning has been pretty stable. — Carl (CBM · talk) 20:15, 21 May 2010 (UTC)[reply]

I 100% agree that you are both right on that. North8000 (talk) 20:57, 21 May 2010 (UTC)[reply]
The term "original research" in the academic world doesn't mean "not based on reliable sources". Students writing research papers and academics writing papers use reliable sources all the time in their original research--indeed they're expected to use reliable sources. What we do in WP is more comparable to, well, writing an encyclopedia, or writing a report on a topic, without synthesizing those sources in support of an original thesis, POV or other conclusion which puts forward an argument based on the cited reliable sources. This practice is well tolerated--even encouraged, in an original research paper. But it's not what we do in WP. And therein lies the important difference between WP:V and WP:NOR. ... Kenosis (talk) 20:46, 21 May 2010 (UTC)[reply]


Re Crum375's remark "both 'verifiability' and 'OR' have a somewhat peculiar definition, which can cause confusion." - We have the opportunity to make the "OR" definition less peculiar with the change proposed in the RFC above. As it is now, the definition of "OR" includes the case where editors' don't try to put their original thoughts into Wikipedia, but try to include material by citing unreliable sources. This prohibition belongs in WP:VER where it currently is, not in WP:NOR as part of the definition of "OR", where it is peculiar and a source of confusion. And this can be fixed easily with the RFC above.--Bob K31416 (talk) 20:49, 21 May 2010 (UTC)[reply]

Question about original research

Recent edits that I made to Tai chi chuan and Chen Wangting were good-faith-reverted by an editor indicating that original research cannot be accepted. He also directed me here.

The edit in question has to do with the birth year of Chen Wangting, a martial arts grandmaster who lived in the 1600s. I have found inconsistencies about his birth year between all articles about him in Wikipedia (English, German, French and Portuguese versions); they all have different birth years. So to resolve this issue, I asked Grandmaster Chen Zhenglei, a Chen family member who is also a world-recognized authority in Tai Chi Chuan and a direct descendant of Chen Wangting. He wrote me back indicating that the birth year of Chen Wangting is not well known, and what year the family uses. I can produce the email that he sent me in support of this.

I then proceeded to edit the articles with the official year of birth that the family uses, and indicated Chen Zhenglei as the source. I believe that grandmaster Chen Zhenglei must be recognized as a reliable source, as he is one of the few spokesmen for the family, so any direct communication from such an authority must be legit and reliable. Consequently, I find that the reversion of my edits was improper.

So if there are any doubts about propriety of my edits, can somebody indicate to me what is the proper way of documenting the reliability of responses of this kind? What recourse does one have to prove the validity of these types of edits?

Thanks for any help.

(I have also asked this in Wikipedia_talk:No_original_research/Noticeboard

 Bruno  20:38, 21 May 2010 (UTC)

What you did is the very definition of Original research. You are adding information that has not been published in a reliable source. Your information may be correct... but because it has not been published it constitutes OR (it also violates WP:Verifiability).
Even if published (say on a web-site), we would still question the material. However we would do so for a different reason... family members do not constitute a reliable source. Blueboar (talk) 21:24, 21 May 2010 (UTC)[reply]