Wikipedia talk:No original research

Do not make analytic...claims about material found in a primary source

The policy contained the sentence "Do not make analytic, synthetic, interpretive, or evaluative claims about material found in a primary source." It seems obvious to me that this is about original/novel/unpublished claims (it doesn't include all claims - I interpret "claims" here as meaning statements - since we'e quite happy to make claims that we can cite to a good secondary source), but when I added the word "original" or "novel" to the sentence, it kept getting reverted. Can anyone explain this?--Kotniski (talk) 11:13, 27 January 2011 (UTC)

Please review my comment above of 21:17, 25 January 2011, which specifically addresses this issue. Please also review my comment above of 15:21, 26 January 2011, where I write "It's hard for me to have a discussion with people who apparently aren't reading my posts." Jayjg ^(talk) 17:57, 27 January 2011 (UTC)

Can you explain again, then? Your comment above doesn't make much sense. This whole page is about "no original research", now it seems you have some objection to un-original research as well?--Kotniski (talk) 18:01, 27 January 2011 (UTC)

Can you explain which specific parts you didn't understand? Jayjg ^(talk) 18:43, 27 January 2011 (UTC)

Your words: "Why on earth would we want "un-original" analytic, synthetic, interpretive, or evaluative claims about material found in a primary source?" "Un-original" claims are exactly the kind we want, surely? that's what this whole page (including its title) is essentially about.--Kotniski (talk) 18:57, 27 January 2011 (UTC)

Why do we want editors to make "analytic, synthetic, interpretive, or evaluative claims" of any sort? Shouldn't that be the prerogative of reliable secondary sources? Jayjg ^(talk) 20:49, 27 January 2011 (UTC)

And if reliable secondary sources make such a claim, are we not permitted to make it also? Paul August ☎ 21:22, 27 January 2011 (UTC)

No. We're allowed to reproduce the claims, attributed to the reliable secondary sources. We're not permitted to make them ourselves. Jayjg ^(talk) 00:48, 28 January 2011 (UTC)

What's the difference between "reproducing" and "making" them? If an article says "A is true", citing a reliable source, isn't that "making" a claim? Paul August ☎ 02:46, 28 January 2011

So this is just about semantics, clearly. Some people read "making" a claim as implying the originality of that claim, others don't. But in any case, this seems to justify the addition of a word like "original" or "novel", to clarify to both groups which reading is intended. --Kotniski (talk) 05:11, 28 January 2011 (UTC)

Also, the policy explicitly allows reproducing claims that are already in primary sources, because that is just a descriptive use of the primary source. In other words, if a primary source X say Y, we can reproduce this as "X says Y" and there is no original research issue. — Carl (CBM · talk)

OK, the sentence is back in place, with "unsourced" this time (though it could equally well be "original", "novel", or other word to that effect, per the above discussion).--Kotniski (talk) 11:27, 28 January 2011 (UTC)

It should be novel; WPV requires that hings can be sourced, not that they literally are sourced. There is no blanket policy against inserting unsourced material provided the material can be sourced in principle. — Carl (CBM · talk) 13:41, 28 January 2011 (UTC)

If every person takes it upon himself to make one small substantive change, the policy will soon be derailed. Four or five people have made a plea here for consensus to be obtained before anything of substance is changed. Please abide by that. If you can't gain consensus here, open an RfC. But however you do it, once multiple objections have been lodged, it has to be via discussion, compromise and consensus, because the policy has to remain stable. "Stable" doesn't mean unchanging, but it does mean not bouncing all over the place. SlimVirgin ^{TALK|CONTRIBS} 13:49, 28 January 2011 (UTC)

Oh come on, this is a tiny matter and it's been resolved in the discussion above. Matters of importance need discussion; insisting that every little change go through some sort of constitutional approval process is going to take everyone's time away from those matters of importance. --Kotniski (talk) 14:44, 28 January 2011 (UTC)

Well, you started this section, so I don't think you should complain about it taking your time away, and I don't really see that the issue has been resolved. I really wish you had waited to add your choice of word to the policy. -- Donald Albury 14:55, 28 January 2011 (UTC)

(to Kotniski) You say these are tiny matters, but others disagree. This change, for example, would cause confusion, and it alters wording that has been in the policy for years.

I've been in the same position—where I've assumed something is completely uncontentious and I've added it, only to be told I had changed the meaning in the view of some people. When it's experienced editors who say that, we have to take it seriously. I realized it's frustrating, and I also realize it can mean less-than-ideal writing might remain in the policy longer than it should. But the default position has to be stability for the sake of all the editors who rely on this policy, but who never join in discussions about it.

When you change a policy, you have to think of all the situations in which your words will be used and misused—not only situations you personally have in mind. SlimVirgin ^{TALK|CONTRIBS} 14:59, 28 January 2011 (UTC)

But look at the discussion above - this one has been resolved (unless you're seriously contesting my conclusion about different semantic interpretations - no-one else is). At least, that some word is needed here is hopefully clear to everyone now - we can try some different word, I don't care which one, but the sentence as you restored it to the policy is just wrong (if it's read in the way that many people would read it). This continual arguing about stupid and obvious things is a real waste of time, and I can only assume it's intended to drive people away from the task of editing this page.--Kotniski (talk) 15:08, 28 January 2011 (UTC)

I can't follow what you're talking about anymore, because it's too much at once. Which are the stupid and obvious things? SlimVirgin ^{TALK|CONTRIBS} 15:11, 28 January 2011 (UTC)

OK, it's obvious from the discussion above that there has to be a word like "novel" or whatever in that place so that everyone will read it right; and it's stupid to drag out discussions unnecessarily when the issue has to all intents and purposes been decided. Don't you see how incnosistent you're being - on one hand you're making us discuss every detail, and on the other you're complaining that the discussion is growing out of control. THIS IS NOT THE WIKI WAY.--Kotniski (talk) 15:23, 28 January 2011 (UTC)

Kotniski, not only did you re-add the word novel, which is currently contested, you made 2 other changes that, while I honestly believe you think are better, have subtle ramifications that absolutely need to be discussed first. You say above that this arguing is "intended to drive people away from the task of editing this page." In fact, I think that's exactly what we're trying to do. Policy pages are policy for a good reason--because they've survived a lot of long term work. Policy pages should be harder to change than articles. I could make points as to why each of the small changes you just introduced are problematic, but the bigger message that I think needs to come across here is that you shouldn't be making anything other than trivial changes (spelling, actual grammar errors, etc.) without consensus. Even if you think the change is right, and you do boldly make it, as soon as someone reverts you, it's got to come here until consensus is reached.Qwyrxian (talk) 15:18, 28 January 2011 (UTC)

"Novel" is not contested, what are you talking about. And once again, you haven't provided any arguments for reverting my changes, even though they were explained in edit summaries, you've just made a blind revert on the grounds that all change must be bad. Well, since you admit that you're trying to drive people away from editing this page (did you really mean that?), I'm going. It's clear improvements are not wanted and that the owners of this page want to bask in the false illusion that it's somehow close to perfection. Have it to yourselves, well done, another dangerous progressive successfully got rid of. Removing from watchlist and GOODBYE. Extremely disgusted,--Kotniski (talk) 15:29, 28 January 2011 (UTC)

Is everyone agreed on what "novel" means? What we want to do here is make sure it is clear that any interpretation presented is (dare I say "completely") supported by reliable published sources. In any case, I think you should have proposed "novel" here and waited for consensus before adding it. -- Donald Albury 16:47, 28 January 2011 (UTC)

Any interpretation has to be supported by secondary sources, as that passage already makes clear. So it's not clear what's meant by adding "novel" in that context. SlimVirgin ^{TALK|CONTRIBS} 17:19, 28 January 2011 (UTC)

Interpretation that is already in the primary source can just be sourced to the primary source, because all that we are doing in that case is making a descriptive claim (e.g. if we write "X says Y"). On the other hand, interpretation that is not already in a secondary source can't be cited to the source that doesn't contain it. So on the one hand, we can cite "non-novel" facts to the sources that contain them, and on the other hand we can't cite "novel" facts anywhere, by definition. — Carl (CBM · talk) 17:37, 28 January 2011 (UTC)

The examples we've seen here of interpretation from a primary source aren't really interpretation (Wellington saying Waterloo was a close thing); as you say, covered by "description." I can't understand anything after your first sentence, I'm afraid. SlimVirgin ^{TALK|CONTRIBS} 17:43, 28 January 2011 (UTC)

My point was that if the interpretation is already in a source then we can use it regardless of the type of source, and if it is not in any source then we can't use it in the first place. So there is nothing special about secondary sources making interpretations vs. primary sources making them. — Carl (CBM · talk) 17:47, 28 January 2011 (UTC)

As the policy says in several places, secondary sources are needed for interpretation/analysis. If you want to change that, you'd need very strong consensus, and not only on this page, because this is an idea that has strong support, one that Wikipedians rely on. SlimVirgin ^{TALK|CONTRIBS} 17:51, 28 January 2011 (UTC)

You just said that we can include Wellington's interpretation. Change your mind often? What exact sort of interpretation could we use from a secondary source that we could not use from a primary source? — Carl (CBM · talk) 17:54, 28 January 2011 (UTC)

We can report what Wellington said about Waterloo as a fact. But whether he was right about it being a near thing, or how near of a thing, is something we want from secondary sources. -- Donald Albury 18:04, 28 January 2011 (UTC)

If we had a primary source that said Wellington was right, we could cite that, too. We can make descriptive claims about the contents of primary sources without violating the NOR policy, even if the contents we are describing are analysis or interpretation. — Carl (CBM · talk) 18:08, 28 January 2011 (UTC)

Wellington can't analyse himself. What he says about himself and things he did is just a description. SlimVirgin ^{TALK|CONTRIBS} 18:12, 28 January 2011 (UTC)

Wellington is not analyzing himself, he is analyzing a battle. So can we cite Wellington's primary source for the interpretive claim he is making about the battle, or can't we? I put a different example below of an interpretive claim that we would routinely cite to primary literature. — Carl (CBM · talk) 18:16, 28 January 2011 (UTC)

That has already been explained: secondary sources offer an overview of primary sources. That's where we go for analysis. The IDIDNTHEARTHAT has made this page unreadable, so I'm not going to continue posting unless someone raises a new concern. SlimVirgin ^{TALK|CONTRIBS} 18:02, 28 January 2011 (UTC)

(ec) Aaargh... I was going to stay away but it irks me to see intelligent Wikipedians wasting their valuable editing time on such an idiotic conversation about such a trifling matter. The sentence in question says (simplifying the language, so no-one can pretend they don't get it): DO NOT INCLUDE ANALYSIS OF PRIMARY SOURCES. This is WRONG (and we all know it's wrong), because we CAN include analysis of primary sources, if that analysis is well-sourced. So to make the sentence a true statement of our practice, the word "analysis" must be qualified with something. With a word like "novel", "original", "unsourced" or something - take your pick - but the version with no qualifier at all is just wrong, however long it may have lurked unnoticed in the policy. Someone missed a word out when they wrote it, that's all.--Kotniski (talk) 18:03, 28 January 2011 (UTC)

Oh, and IDIDN'THEARTHAT as an accusation by Slim against others? That takes the cake.--Kotniski (talk) 18:03, 28 January 2011 (UTC)

Please stop the personal attacks. You're going too far with it. SlimVirgin ^{TALK|CONTRIBS} 18:12, 28 January 2011 (UTC)

Wow, now that takes even more cake. You can accuse others of something, but no-one else can accuse you of it?--Kotniski (talk) 18:14, 28 January 2011 (UTC)

Let's just be clear about something here; the insertion of the word "novel" adds confusion to the policy, and encourages editors to include their own analysis etc. It is not a "tiny matter", it is not just a "different semantic interpretation", and it has not "been resolved in the discussion above". There is no consensus that it is a useful addition; on the contrary, it is obviously an unhelpful change. Jayjg ^(talk) 18:21, 28 January 2011 (UTC)

I think the issue Kotniski is raising is that the NOR policy permits us to use primary sources for claims that are actually in the source (and thus not "novel"). For example, Wellington's analysis of the battle of Waterloo can be cited to Wellington, and so is not original research. I have written a more concrete example below. — Carl (CBM · talk) 18:24, 28 January 2011 (UTC)

No, it's far more trivial than that. It's that the sentence as written can reasonably be read as meaning "no analysis whatever", even if found in a secondary source, which is clearly absurd. I'm just flabbergasted that there are people who are so defensive about this policy that they turn a simple correction like this into a huge, unpleasant and time-wasting argument.--Kotniski (talk) 18:37, 28 January 2011 (UTC)

There's a blind spot of some kind here. "Hitler said: 'I have black hair'" is the same kind of statement as: "Hitler said: 'I nearly won the war'". Both are fine in articles about Hitler and what he said. But we don't add his analysis (as a primary source, an involved source) to articles about the Second World War, alongside the analysis of modern historians (secondary sources, uninvolved sources). This is accepted everywhere on Wikipedia. I can't understand why a small number of editors on this page suddenly object to it. SlimVirgin ^{TALK|CONTRIBS} 18:39, 28 January 2011 (UTC)

There's a blind spot, certainly, probably several, since we all seem to be talking past each other and making absolutely no sense. But I really don't see how I can be expected to explain any more clearly about this missing word (and I don't see what either your or Carl's examples have remotely to do with it).--Kotniski (talk) 18:43, 28 January 2011 (UTC)

We could certainly cite statements by Richard Nixon in the article on Watergate, and in fact we do at the moment, even statements that make interpretive claims. If Hitler thought he had almost won the war, we could quote him as saying that without violating the OR policy. — Carl (CBM · talk) 18:51, 28 January 2011 (UTC)

I don't see how it's on topic for this thread, but for the record, I agree with you. I don't think a quote from Hitler (analytic or otherwise) would necessarily be out of place in a WWII article. It might or might not be an effective way of conveying relevant information to the reader, but I can't see how it offends against our basic principles, as long as it's made clear that we are just quoting him and not agreeing with him.--Kotniski (talk) 18:59, 28 January 2011 (UTC)

Oh, possible light bulb flashing... Some people are interpreting the sentence to mean not "(claims about) material found in primary sources)", but "claims (about material) found in primary sources". I somehow doubt it, since no-one objected to the word "novel" by saying "how can it be novel if it's in a primary source?", but that interpretation would at least explain the direction this discussion has taken. Is anyone in fact reading it like that?--Kotniski (talk) 18:55, 28 January 2011 (UTC)

Kotinski, I find your remarks consistently difficult to understand because you insist on using phrases such as "the sentence". With the policy in constant flux and the volume of this discussion, it is impossible to determine the antecedent of "the sentence". Jc3s5h (talk) 19:31, 28 January 2011 (UTC)

The one quoted at the very start of this thread. Sorry for being the only one to be on topic.--Kotniski (talk) 19:46, 28 January 2011 (UTC)

I agree with Jc3s5h here. Regarding the issue, the problem is that the proposed change encourages editors to include their own interpretation/analysis, because it suddenly opens the loophole of the "un-original" interpretation/analysis being acceptable. Editors should not be adding their own interpretation/analysis, of any kind, "original", "un-original", etc. Instead, they should cite interpretation/analysis found in reliable secondary sources. If it's a notable "un-original" or "non-novel" interpretation/analysis, then it will be found in reliable secondary sources. Jayjg ^(talk) 02:06, 30 January 2011 (UTC)

We are also permitted to cite interpretation or analysis contained in primary sources, without violating NOR. There is nothing special about secondary sources in that regard, all that matters is that the interpretation or analysis is in some source. — Carl (CBM · talk) 02:18, 30 January 2011 (UTC)

I agree the sentence at the beginning of the section is not acceptable, in part because if I place a claim (with supporting footnote referring to source A) in an article using Wikipedia's voice, then Wikipedia makes the claim.

You might say that if the claim is interpreting a "primary source", then the supporting source A must inherently be a secondary source, at least for this particular purpose. Not so. If the supporting source A is close to the event, then it is a primary source even when it is interpreting primary sources. Jc3s5h (talk) 02:35, 30 January 2011 (UTC)

I still think you're all (especially Jayjg) looking for difficulties and complications that aren't there. The sentence in question doesn't even mention secondary sources. We can argue about whether interpretations found in primary sources can be included, but that's a different topic - the sentence as written (and as most of us are interpreting the words) would forbid us from reporting claims EVEN IN THE UNCONTROVERSIAL CASE where they come from secondary sources. Can you all take a step back and see if you see that now? Or if not, at least see how I'm reading the sentence to reach that conclusion, and explain how you're reading it differently? (I thought we'd already cleared that up when we identified the different interpretations of "make (claims)", above, but since at least one person is still arguing about it, perhaps there's something else.)--Kotniski (talk) 10:32, 30 January 2011 (UTC)

I see what you mean. The sentence is "Do not make analytic, synthetic, interpretive, or evaluative claims about material found in a primary source." If that is saying articles cannot make sourced interpretive claims, it's wrong. So it must be saying that articles can only make sourced interpretive claims. That makes sense to me. — Carl (CBM · talk) 13:20, 30 January 2011 (UTC)

Yeah... there is a difference between making an analytic, synthetic, interpretive, or evaluative claims about material found in a primary source (OR and thus Not Allowed)... and discussing an analytic, synthetic, interpretive, or evaluative claim that is contained in a primary source (Not OR if done right... so Allowed, but with strong caution and limitations).

The thing is... I think the same distinction applies to Secondary sources... It is just as much a violation of NOR to make such statements about secondary sources as it is to make such statements about primary sources. The difference between Primary and Secondary sources (as far as the concept of OR is concerned) is that there is more flexibility and there are fewer limitations and cautions that apply when describing such statements contained in secondary sources. Blueboar (talk) 14:37, 30 January 2011 (UTC)

Again, I think this is all just about semantics, as I said a long way above. You (and others) apparently interpret "making a claim" as meaning "stating something that hasn't been stated before", while I (and others) interpret it as meaning simply "stating something". But if the first interpretation holds, it would be enough for the sentence to read "Don't make claims (full stop)". And if the second interpretation holds, then (as I keep saying) the statement as quoted right at the start of this thread needs qualification with a word or phrase with the force of "original", "unsourced", "novel" or whatever, otherwise it would be expressly forbidding the inclusion of certain kinds of statements even if they are perfectly well sourced to multiple reliable secondary sources.--Kotniski (talk) 14:52, 30 January 2011 (UTC)

The sentence in question is giving direction to editors, not advice about article content. There's a world of difference between "do not make" and "articles should not contain". Jayjg ^(talk) 18:45, 30 January 2011 (UTC)

So Wikipedia's policy page is telling people not to make claims about primary sources as they go about their everyday lives?? That would be even more absurd - but it presumably isn't what you mean - what is it you mean?--Kotniski (talk) 18:59, 30 January 2011 (UTC)

The distinction here is between a claim made by an editor and a claim made by a reliable secondary source. Jayjg ^(talk) 20:19, 30 January 2011 (UTC)

So we're back to the same place again, I think. You think that an editor who repeats a claim from a reliable secondary source is not "making" that claim; others understand the word ("making") differently. Do you see now why the word "novel", "unsourced" or some such is needed? It's to make the sentence true for those who read it with the second interpretation of "making". (In fact the whole sentence is redundant and could happily be omitted without weakening the policy in any way, if you feel the wording is problematic.)--Kotniski (talk) 20:32, 30 January 2011 (UTC)

I do not see that novel or unsourced would help, seems to me it would cloud the issue. Jayjg is spot on, and an editor referencing a claim made in a primary source is not "making a claim". --Nuujinn (talk) 20:37, 30 January 2011 (UTC)

Exactly. When an editor inserts into an article a claim taken from a reliable secondary source, that editor is "referencing" or "citing" or "reproducing" the claim, not "making" it. Positing that policy readers will understand the word "making" here in an inexact and unusual way, and then trying to "solve" that problem by adding the word "novel", makes the policy less clear and more open to abuse. Jayjg ^(talk) 20:48, 30 January 2011 (UTC)

But, if that's the logic, we also cannot "make" claims about material in secondary sources, we can only "cite" claims about material in secondary sources. In other words, if "make" means stating claims that are not in sources, then we cannot "make" any claims whatsoever. But we are also free to "cite" claims from primary sources, as long as it's clear that we are accurately describing what's in those primary sources. — Carl (CBM · talk) 20:53, 30 January 2011 (UTC)

Except that primary sources should rarely be cited, because of the other issues with them (particularly around the relative ease with which they can be used for WP:NOR.) Jayjg ^(talk) 03:09, 8 February 2011 (UTC)

(ec) Well most people in this discussion read it that way, so I don't think it's unreasonable to assume that other people reading the policy will read it that way. Can we agree just to remove this sentence - whatever it's trying to say is already said countless times in the policy already. Or to rephrase it so that it doesn't use the apparently ambiguous expression "make...claims". --Kotniski (talk) 20:56, 30 January 2011 (UTC)

What to do with the sentence?

So, what are we doing with this sentence: removing or clarifying? (I note it could pretty much be combined with the sentence before it, which is perfectly OK as it does say "unless found in a secondary source").--Kotniski (talk) 14:35, 4 February 2011 (UTC)

The phrase "unless found in a secondary source" does not appear in the policy nor does it appear in this thread, save for this post and Kotinksi's pose of 14:35 UTC. Jc3s5h (talk) 16:20, 4 February 2011 (UTC)

Sorry, the actual words are "any interpretation needs a secondary source". We could expand that to say "any analysis, interpretation (blah blah) needs a secondary source, and then drop the sentence that has caused all this confusion. Or we could simply go back to the entirely unobjectionable solution of adding a word like "original" or "unsourced" to the "Do not make claims..." sentence.--Kotniski (talk) 17:08, 4 February 2011 (UTC)

I would think it would be infrequent, but there might be occasions to cite a primary source for interpretation of another primary source. For example, "President Jones stated that the bumbling character President Peters in Jack Smith's novel Some of the President's Men is a transparent allusion to himself and as such is totally false."<ref>Cite to White House web site </ref>. In this case the cited source would be secondary, since it is analyzing another work, except that it was created by a person close to the events. Of course someone will come along and say it would be better to cite a secondary source, and maybe it would, but it wouldn't be original research to cite the hypothetical president. Jc3s5h (talk) 17:23, 4 February 2011 (UTC)

Given the detailed and specific objections above, how on earth can you claim that there is an entirely unobjectionable solution of adding a word like "original" or "unsourced"? Your comment does not appear to be made in good faith. Jayjg ^(talk) 17:35, 6 February 2011 (UTC)

I think all the original objections have now been answered. Do you think there are any outstanding? Surely you see now what the problem is - how do you propose solving it? Would you be OK with the suggestion of combining the two sentences, as made above?--Kotniski (talk) 10:54, 7 February 2011 (UTC)

I don't think so. See my comments in the example section below. -- PBS (talk) 01:22, 8 February 2011 (UTC)

Kotniski, you continue to do this; people raise objections to your proposed changes, you explain to your own satisfaction why you think your proposals are valid anyway, and then insist the "objections have been answered". Responded to, perhaps, but not resolved in any way. Again to be clear, you have not demonstrated any problem with the current wording, but your proposals create problems. Jayjg ^(talk) 03:09, 8 February 2011 (UTC)

What??????????????? I've demonstrated the problem with the current wording over and over again - it would forbid the reporting even of sourced claims, which is absurd and not what anyone wants. And I really think that all objections have been answered by this simple, basic and uncontestable observation. Can we now end this absolutely absurdly long and contrived discussion, simply add the word "unsourced" which was undoubtedly supposed to be there all along, and move on to other matters? For the last time, is there any reasoned objection to this clarification of the wording?--Kotniski (talk) 15:10, 9 February 2011 (UTC)

You've stated there is a problem, but not "demonstrated" it. The current wording doesn't do what you say it does. And no matter what you "really think", merely declaring that "all objections have been answered by this simple, basic and uncontestable observation" doesn't make it so. Jayjg ^(talk) 03:32, 10 February 2011 (UTC)

Or if it makes it any clearer to those who still claim not to get it, I've added a more explicit parenthetical to the sentence in question: "(unless those claims are supported by secondary sources)". It would be neater to do it with just one word, but at least I hope with this version no-one can conceivably misunderstand or claim that the qualification is not needed.--Kotniski (talk) 15:26, 9 February 2011 (UTC)

I think the problem now is that you've created so many forest fires about this that no one really knows what you're saying anymore. I'd have agreed with you a couple of weeks ago about adding something about secondary sources (though not the way you want to word it), but now people are so confused that others are turning up to say that would be wrong. I therefore think the sentence is best left as it is for now. In all the years it's been there I've never seen it (here or on other pages) cause a problem.

In future if you want to achieve policy change, the best approach is to propose one change and just let it sink in for a bit, rather than trying to force in multiple changes on several pages at once. That's the kind of thing the policies have to be protected against, so it causes people to move into defensive mode. SlimVirgin ^{TALK|CONTRIBS} 19:38, 9 February 2011 (UTC)

The other problem is that you keep proposing policy changes, then when people explain why they object to them, you explain to your own satisfaction why you think your proposals are valid anyway, and then insist the "objections have been answered". Responded to, perhaps, but not resolved in any way. Then, after declaring that you are correct, you make all sorts of policy changes that you've never even proposed here! Do you really wonder why people object to that? Jayjg ^(talk) 03:34, 10 February 2011 (UTC)

This is NOT POLICY CHANGE, it's just a trivial correction to a sentence that's worded wrongly. Of course it isn't policy that "you can't make interpretative [...] claims"; we all know that isn't what the page is supposed to say, there's simply a missing word or phrase in that sentence. That's all; and THAT MUCH no-one is disputing or objecting to. It's not me who's creating "forest fires", whatever that's supposed to mean - the excessive amount of talk-page discussion is caused by people like you who just obstinately refuse to get the point however clearly and however many times it's made.--Kotniski (talk) 09:19, 10 February 2011 (UTC)

Um - hate to say this but why not simplify the morass to Primary sources can only be used to make statements explicitly made in the primary source. for everything after "for that interpretation" up to the "Use extra caution ..."? Collect (talk) 11:57, 10 February 2011 (UTC)

I certainly agree with simplifying - though I don't think that wording is quite right - mostly we don't make statements that are explicitly made in the primary source. If the primary source contains a statement like "Jesus was laid in a manger", then the statement we make is not "Jesus was laid in a manger" but "XXX says that Jesus was laid in a manger". --Kotniski (talk) 12:07, 10 February 2011 (UTC)

Example

Here is a more concrete example. This paper [1] was published as Notre Dame J. Formal Logic Volume 51, Number 4 (2010), 475-484. Presumably it counts as a primary source in the classification of this policy.

In the first paragraph, the authors make this interpretive claim:

"The multiverse axioms express a certain degree of richness for the set-theoretic multiverse, flowing from a perspective that denies an absolute set-theoretic background."

If we wrote about this topic in our article on set theory, we certainly could use this source to cite that interpretation. So I'm not sure what the claim "secondary sources are needed for interpretation/analysis" is supposed to mean here. There is nothing in policy or practice that prevents us from using this source for that particular interpretation and analysis. — Carl (CBM · talk) 18:05, 28 January 2011 (UTC)

Hmm! That paper has a list of references about a page long. To the extent that the paper is summarizing, analyzing and interpreting the sources listed as references, it is a secondary source. Only the parts, if any, of the paper that present new work not based on the cited references should be treated as primary. Things aren't always clear cut. Best case would be if it is a review article. -- Donald Albury 18:25, 28 January 2011 (UTC)

It is not a review article, and those two authors are the ones who have developed the very idea of "multiverse axiom" that they are describing (this is one of the first papers about that topic). The sentence I quoted is not analyzing the references, it is giving the authors' original interpretation of the topic that they are studying. Nevertheless we can use the paper as a source for that claim. — Carl (CBM · talk) 18:27, 28 January 2011 (UTC)

Yes... that is because sources are allowed to contain original research... NOR is about what we say in our articles, not what our sources say. It all goes back to the original concept behind the NOR policy... "Wikipedia should not be the first place of publication for any fact... or for any analytic, interpretive or conclusionary statement about a fact". Blueboar (talk) 14:54, 30 January 2011 (UTC)

Take another example: "According to Wellington, the battle of Waterloo was 'the nearest-run thing you ever saw in your life'." That quoted analysis of the battle is acceptable, but to concatenate an an editorial analysis of what he wrote: "... but he of course did not know what he was talking about." is not unless it is backed up with a secondary source.

That of course is an extreme example which I doubt anyone would disagree with. But to take a more subtle example. The Russian Foreign Ministry has said "declaration of tragic events of that time as act of genocide of the Ukrainian nation is unilateral misinterpretation of history in favour of modern conformist political-and-ideological principles". That can be summarised in a number of ways here are two "The Russian Foreign Ministry has denied that the Holodomor was a genocide." and "The Holodomor genocide has been denied by the Russian Foreign Ministry.". The former is an acceptable summary but the latter draws on external evidence not presented in the quote, that the Holodomor was a genocide. The latter is not acceptable under this policy, because the sentence is not supported by the quote (the primary sources). However there is nothing novel in that analysis, I can find dozens of articles that draw that conclusion, which is why I do not think that novel should not be added to the sentence "Do not make analytic, synthetic, interpretive, or evaluative claims about material found in a primary source."

In Gumby Theatre there is the line spoken by a Python "Come in" followed by a crashing noise, to which the Python says "Open the door and come in". I had assumed, given the content of the rest of the paragraph, that the sentence obviously referred to analysis by editors of primary sources unsupported by secondary sources and given the context and so in my opinion there is no need to add anything to the sentence about secondary sources.

I think we need to step back and look at the raison d'être of this paragraph. The whole point of the paragraph is to stop someone doing original research--something that is very easy to do with primary sources (particularly if it involves a statistical analysis of the data in primary sources, or synthesising a position from the information contained in several of them). It is usually more difficult to do that sort of OR with secondary sources (unless they themselves are used as primary sources). So this is primarily a warning to editors editing in good faith to be careful because it is easy to do OR with primary sources. -- PBS (talk) 23:02, 6 February 2011 (UTC)

The first confusing thing in the sentence "Do not make analytic, synthetic, interpretive, or evaluative claims about material found in a primary source", for me, is whether it means:

1. Articles may not include interpretive claims about primary sources
2. Wikipedians should not make interpretive claims about primary sources in their daily lives

The first of those is false: our articles can, and do, make interpretive claims about material in primary sources. The second is irrelevant to Wikipedia. So it doesn't mean either of those.

At the risk of of putting a bean up you nose give an example. -- PBS (talk) 10:43, 9 February 2011 (UTC)

What the sentence is trying to keep out are interpretations for which a source cannot be given. But then the type of the source isn't important: if we can source the interpretation to a primary source, that's still a source we can use without having any original research. — Carl (CBM · talk) 02:21, 8 February 2011 (UTC)

It it is an interpretation from a published primary source and the interpretation of that primary source follows this policy then I do not see what the problem is that you are trying to show. -- PBS (talk) 10:43, 9 February 2011 (UTC)

Exactly. A much better example is provided by the 30-year scheduled declassification of the analysis and official policy decisions contained in the "Documents on foreign policy-series" published by the US, UK and other governments. In many instances, those documents reveal in-depth analyses and political agendas that are at odds with earlier accounts or assumptions contained in a host of published secondary sources. Wikipedia obviously cannot adopt a policy of only allowing those secondary sources to speak about the official policies of a government. So long as the analysis comes from an existing reliably published primary, secondary, or tertiary source, it does not originate from Wikipedia and cannot be considered original research. harlan (talk) 19:55, 8 February 2011 (UTC)

If it is an interpretation from a published primary source and the in-depth analyses and political agendas are used "to make straightforward, descriptive statements that any educated person, with access to the source but without specialist knowledge, will be able to verify are supported by the source." I don't see what the problem is. I can see however that one might want to use such as source as a secondary source in which case that is a matter for discussion on the talk page of an article. -- PBS (talk) 10:43, 9 February 2011 (UTC)

Yes, that use of a source would be unproblematic. The question is how this interacts with the sentence ""Do not make analytic, synthetic, interpretive, or evaluative claims about material found in a primary source". In the example Harlan gave, we would be making an interpretive claim about a primary source, using information found in a primary source, but we would not be violating the OR policy. — Carl (CBM · talk) 20:32, 9 February 2011 (UTC)

Actually, we would be doing both. Jayjg ^(talk) 03:32, 10 February 2011 (UTC)

Are you claiming that we would violate the NOR policy by doing that? If so, you need to explain. The NOR policy directly allows using primary sources in that way; we may need to add "According to", but we are still free to use the source without violating NOR. For example, we could say, "According to Cable X [1], Cuba was of key strategic interest to the United States in the 1960s." That would not violate NOR, provided that the cable did say Cuba was of key strategic interest. NOR is not NPOV. — Carl (CBM · talk) 12:33, 10 February 2011 (UTC)

I think you have to take the sentence in the context of the paragraph it is located, (and read our polices with the foreknowledge that they are camels (horses designed by committees)). To misunderstand it (and to give it the twist that is given here) is to read it out of context. Providing the interpretation of the primary source is sourced with a cited source. (see my example above about the Holodomor). But as soon as a cited source is used the focus changes from the initial primary source to the source used to give the "analytic, synthetic ...". If I were writing a parser to check if the logic was being followed correctly then it seems to me obvious that recursion would handle it. -- PBS (talk) 03:41, 10 February 2011 (UTC)

I don't understand what you're saying here. — Carl (CBM · talk) 12:33, 10 February 2011 (UTC)

I don't think anyone does. But I don't understand why you just removed the qualification "unless supported by secondary sources" from the sentence in question. Your edit summary "a source is a source" doesn't make much sense - can you elucidate?--Kotniski (talk) 12:50, 10 February 2011 (UTC)

Under the NOR policy, there is no requirement that we can never use primary sources for interpretive claims. As long as anyone can verify that the primary source literally makes the interpretive claim that is sourced to it, there is no NOR issue with using a primary source in that way. So it is not true that we have to have a secondary source for any interpretive claim; we can use a primary, secondary, or tertiary source. That's what I meant by "a source is a source". — Carl (CBM · talk) 12:57, 10 February 2011 (UTC)

For example, here is an interpretive claim sourced to a primary source:

According to [2], geometric frustration was "particularly well studied" in the 1990s and 2000s, but not for ferroelectric materials.

There is no NOR issue with using the source in that way: anyone can verify that the claim being made is literally made by the source. — Carl (CBM · talk) 13:01, 10 February 2011 (UTC)

Yes, but we are not making that claim, only reporting it (because of the words "According to..."). So this doesn't falsify the statement that you objected to. And even if it did, it would then necessarily still falsify the form of the statement that you restored, since the statement you restored is stronger. Do you see this now?--Kotniski (talk) 14:28, 10 February 2011 (UTC)

No, I don't. Are you saying that if we just reworded the sentence to

Geometric frustration was "particularly well studied" in the 1990s and 2000s, but not for ferroelectric materials.[3]

Then it would violate the original research policy? I would also argue that the rephrased sentence is a perfectly fine example of using a primary source to make a claim that anyone can verify from the primary source, and is not a violation of NOR either. I wholeheartedly agree that the sentence I restored is worded poorly, but I thought your change was not an improvement. What we should do it just remove the sentence entirely, because it's wrong. If we source an interpretation to a primary source that literally gives that interpretation, that's not original research, it's source-based research. — Carl (CBM · talk) 14:34, 10 February 2011 (UTC)

I suspect your view on this would be contested. If the quotes you put round "particularly well studied" are not intended as scare quotes with meaning equivalent to "it has been stated that..." or some such phrase, and if the source you give is indeed considered to be a primary source in this context, then I think people would indeed say that it violated the policy. (Though I suspect it wouldn't actually be regarded as a primary source.)--Kotniski (talk) 14:43, 10 February 2011 (UTC)

So, about the original sentence, can we just put in the word "unsourced", as proposed a long time ago? The sentence is certainly wrong without such an addition; and exactly what "unsourced" means is discussed elsewhere in the policies.--Kotniski (talk) 08:44, 14 February 2011 (UTC)

Yes? OK? No objections?--Kotniski (talk) 08:11, 21 February 2011 (UTC)

We'd wan't "unsourceable" to match WP:V. We don't want to add another category to the list of things that are supposed to always have a source literally provided: "direct quotations, challenged, or likely to be challenged". — Carl (CBM · talk) 12:22, 21 February 2011 (UTC)

OK, agreed; "unsourceable" better than "unsourced".--Kotniski (talk) 12:51, 21 February 2011 (UTC)

RfC on citing oneself

{{rfctag|policy}}

In its section about citing oneself, should the No original research policy incorporate the following language from the Conflict of interest guideline?

Citing material you have written yourself and had independently published is allowed within reason, so long as it is relevant and complies with our neutrality and conflict of interest policies. Excessive self-citation is strongly discouraged. You should not place undue emphasis on your own work, and should include the work of others as appropriate. When in doubt, defer to the community's opinion.

Addition on February 6: Alternatively, as several below have suggested, we could remove this section entirely from NOR and leave it to the COI guideline.

SlimVirgin ^{TALK|CONTRIBS} 07:42, 28 January 2011 (UTC)

Comments

(no threaded replies here, please)

• Support. The section on citing oneself in NOR has lagged behind best practice. People citing their own material is not viewed favourably as a rule, unless the work is clearly notable and would have been added by someone else at some stage. I think the policy should adopt the language from COI of strong discouragement regarding excessive self-citation and undue emphasis. SlimVirgin ^{TALK|CONTRIBS} 07:45, 28 January 2011 (UTC)

• Or remove entirely. As several below are suggesting removing this section entirely from NOR, I'd be in favour of that too. SlimVirgin ^{TALK|CONTRIBS} 16:57, 6 February 2011 (UTC)

• Oppose, let's keep stuff in one place (in this case, the COI guideline). We don't even need a separate section on this page about "citing yourself" - just say that it is allowed, but with a few caveats, which appear to be: write in third person (obvious), be neutral (link to NPOV), and avoid conflicts of interest (link to relevant section of COI). That needs just one or two sentences, and will avoid the inevitable inconsistencies that arise when we try to document the same practices in two places.--Kotniski (talk) 10:39, 28 January 2011 (UTC)
• Oppose. Once the work has been published by a reliable publisher, it no longer meets our idea of original research. If there is some other reason to object to its presence in Wikipedia that should be ruled upon in some other guideline or policy. In addition, creating a clear explanation of a topic isn't easy and our technical articles are often criticized for being inaccessible to people who are not already expert in the topic; we shouldn't discourage textbook authors from drawing on their experience of what approaches to introducing the topic to newcomers work, and which approaches fail. Jc3s5h (talk) 11:43, 28 January 2011 (UTC)
• Support In fact I would add, both here & at COI, something like: "Where you have more standard academic sources than your own works available, these should be used in preference to your own works to reference matter covered as fully by them." I see Kotniski's point, but at the least there should be links to the COI section. Johnbod (talk) 12:36, 28 January 2011 (UTC)
• Oppose Citing published research is not OR. If might be a conflict of interest, but this is not the guideline on conflict of interest. Better explanations were already provided in the lengthy discussion above, which I include here by mention. — Carl (CBM · talk) 13:39, 28 January 2011 (UTC)
• Oppose As long as the research is published, there's no need for this page to object to it. Consider the citation at Edgewater Park Site — it's based on an article published in a major archaeological journal by several authors, including Dr. William Whittaker. Are we going to object to the fact that this source was added and the article written by User:Billwhittaker, a professional archaeologist? I agree that COI can be a problem (although I don't think it is with the Edgewater Park article), but that's a matter that should be discussed at the COI page, not here. Nyttend (talk) 16:30, 28 January 2011 (UTC)
• Support. This is clearly best practise. Jayjg ^(talk) 18:10, 28 January 2011 (UTC)
• Oppose. What part of this is Original Research? And why should we suggest that excessive citation of a single work and POV is acceptable if the author is not the editor - or doesn't admit to being the editor? That's a penalty for editing under your own name. Septentrionalis PMAnderson 20:38, 28 January 2011 (UTC)
• Oppose. I support the text (of course), and Johnbod's suggested expansion (although "academic" is probably not the best choice of words; perhaps "authoritative"?), but IMO this is not the right page to put it on. Spamming your own (properly published) publications into Wikipedia is not a NOR problem. It's a COI/bookspam problem. WhatamIdoing (talk) 02:07, 30 January 2011 (UTC)
• Support what it says... Oppose saying it in this policy. This has nothing to do with the concept of NOR, and thus belongs elsewhere. Blueboar (talk) 14:57, 30 January 2011 (UTC)
• Oppose putting it in wp:nor. Support what it says. This has nothing to do with wp:NOR. North8000 (talk) 16:53, 6 February 2011 (UTC)
• Oppose because it is only tangentially related to OR. Other stuff in this policy is only tangentially related to OR, but lets not make it worse. Yaris678 (talk) 23:40, 6 February 2011 (UTC)
• Support removing entirely. WP:COI is clear enough, and easy enough to find at WP:COI. Paraphrasing here complicates the guidance for first time readers, and so it counter-productive. --SmokeyJoe (talk) 06:05, 10 February 2011 (UTC)

Threaded discussion

• Can we have just a little on the pros and cons of this? Presumably there is a history to this proposal. Looks sensible, but could be creep? I think I read somewhere long ago that to cite yourself, you should propose it on the talk page and leave it to someone else. I thought that was a very good idea. --SmokeyJoe (talk) 08:53, 28 January 2011 (UTC)

• There's no history that I'm aware of. I just think the current wording doesn't stress enough that self-citation is frowned upon. We have a big problem with people using Wikipedia to promote themselves, and this policy currently suggests that that's okay, without warning people that they may run into problems if they try it. I think the wording in the COI guideline makes clearer that, while it's sometimes okay, if it gets to be excessive or violates UNDUE, it's not allowed. SlimVirgin ^{TALK|CONTRIBS} 09:00, 28 January 2011 (UTC)

I don't object a cross-reference to COI provided we make it clear it is a cross-reference and does not describe what original research is. If we call anything and everything original research, the definition of original research becomes "anything that the regulars at WT:NOR and the NOR noticeboard catch and don't like". Jc3s5h (talk) 21:52, 28 January 2011 (UTC)

We should not go back to "propose on the talk page" except for major additions. Johnbod (talk) 12:54, 28 January 2011 (UTC)

Please, do not make rules that can be used to throw the baby out with the bathwater. No matter how many rules you make, people who want to avoid intent will do so. Rules that are too stringent will just drive out the authors who are reasonable. There are situations I have described where substantive material that supports mention of many other people will just not be written if self-citation has to go into talk page first. Michael P. Barnett (talk) 15:15, 28 January 2011 (UTC)

Hi Michael, the issues we have to balance are wanting good sources, on the one hand, against people using Wikipedia to promote their own work, on the other. This proposal won't stop people from contributing their own work, but it asks that they not violate UNDUE (e.g. by making their work appear more significant within the literature than their peers might say it is), and that they don't add their own material excessively. That seems to be a reasonable compromise. SlimVirgin ^{TALK|CONTRIBS} 15:54, 28 January 2011 (UTC)

Very reasonable. Thanks! Michael P. Barnett (talk) 21:40, 28 January 2011 (UTC)

Slim... WP:UNDUE is a NPOV provision and not an NOR provision. My objection to your proposal isn't that we should not say this... my objection is that we should say it here. Blueboar (talk) 15:12, 30 January 2011 (UTC)

• This AfD discussion from within the past month involved probable self-citation. (Its circumstances may be addressed through the proposal qualifiers anyhow.) Throwing this out there in case I don't get time to make more substantive comments, and in case it might be useful/illustrative. –Whitehorse1 20:04, 28 January 2011 (UTC)

• Several people above imply that the "citing oneself" section ought not to be in NOR at all. I agree. I may put up a second option in the RfC about removal so people can choose that if they want. SlimVirgin ^{TALK|CONTRIBS} 02:10, 30 January 2011 (UTC)

I would support that. Blueboar (talk) 15:12, 30 January 2011 (UTC)

I agree / support that. North8000 (talk) 01:02, 7 February 2011 (UTC)

This seems to be the general view, then, so I've removed the section from this page (and changed the shortcuts to go to the COI section). I suggest that further proposals as to what we write about self-citing be made at WT:COI. I've started a thread on that talk page, where I've noted the text that was here (some of it might be incorporated into that guideline). --Kotniski (talk) 09:52, 6 February 2011 (UTC)

Slim Virgin put it back. 75.47.132.215 (talk) 00:50, 7 February 2011 (UTC)

Even though she apparently agrees with taking it out. Page ownership can go to extreme lengths at times... --Kotniski (talk) 10:55, 7 February 2011 (UTC)

So can we finish this...

It seems that "Wikipedia is not a bureaucracy" doesn't apply to this page, so here goes... Does anyone disagree with the assessment of the above discussion, that there is consensus to remove the section (on citing oneself) from this policy, and replace it with a link to the relevant section of the conflict of interest guideline?--Kotniski (talk) 11:20, 7 February 2011 (UTC)

I agree with that assessment, this edit, this edit and changing the shortcuts to redirect to the COI page, as you did. I am generally against bureaucracy but waiting to the end of and RfC is quite an important convention. That said, if no one objects to it fairly soon I think you can call WP:SNOW. Yaris678 (talk) 13:01, 7 February 2011 (UTC)

I don't see any reason to wait longer... the comments were very one sided in support of removal. Blueboar (talk) 13:07, 7 February 2011 (UTC)

RfCs are usually left open for 30 days. I don't mind closing this one after 14 days if comments have dried up, but there's no reason to close it before that, and people are still commenting. This was in the policy for years, so it's worth waiting a bit longer. Why the rush? SlimVirgin ^{TALK|CONTRIBS} 14:13, 7 February 2011 (UTC)

No-one was commenting until you reignited the debate by reverting the implementation of the decision. There's no rush, but on the other hand, why keep dead discussions open? It distracts people's attention from many other useful things they could be discussing or doing. (And there's no rule that just because someone puts an RfC template on a discussion, consensus can't be implemented for some number of days - think how that would be gamed. It's just that the bot needs a number of days after which it automatically removes the template. Since page deletion discussions are generally closed after 7 days, it would be perverse to insist that the deletion of just a small and redundant section of a page be discussed for several times longer than that.) --Kotniski (talk) 15:28, 7 February 2011 (UTC)

This passage has been here for years, and I know it was important to some people, so the least we can do is allow two weeks for comment. SlimVirgin ^{TALK|CONTRIBS} 15:41, 7 February 2011 (UTC)

Why two weeks? Articles are deleted in 7 days, and this is a far lesser step. --Kotniski (talk) 15:49, 7 February 2011 (UTC)

By further delaying, there is the real risk that editors who formed the consensus will get tired and leave the discussion, if they haven't already. There is potential for abuse of the consensus process. 75.47.155.210 (talk) 17:00, 7 February 2011 (UTC)

Looking over the number and quality of comments above, it is highly unlikely that the outcome will change, no matter how long the RFC stays open. If anything, the responses are likely to become slightly more lopsided.

RFCs should normally be kept open until the problem is solved, which could be ten minutes or ten months. The 30-day timer on the bot is purely arbitrary, designed primarily for making sure that the central lists aren't cluttered with year-old forgotten discussions.

However, those who perceive themselves to be on the "winning" side should avoid rushing to close the discussion when the minority view hasn't yet despaired of support. There is no deadline, and there is no pressing need to discourage people from commenting, especially when you have every reason to think that future comments are more likely to support your view than not. Ideally, the "winners" will adopt a posture of deep interest in ascertaining the community's true views, regardless of how long it takes for that view to appear, and behave with both confidence in the outcome and generosity towards the other side's hopes.

(The policy could be changed now, to reflect the current perception of consensus, but that need not stop the RFC, and any changes could be reversed later, if the RFC unexpectedly swings the other direction.) WhatamIdoing (talk) 17:29, 7 February 2011 (UTC)

You can even make the argument that since this is not really a policy change that an RFC is overkill, doubly so for a lengthy RFC. I've had some concern about ownership issues on core policy pages that are played out via. a double standard where the degree of review / consensus "required" varies depending on who is making the change. Not saying that such is the case here, but removal of material which is a duplicate of material in another policy, (and where the other policy is obviously it's proper home) which has had zero dissent after a substantial discussion IMHO makes this clearly time to roll on this. Sincerely, North8000 (talk) 17:45, 7 February 2011 (UTC)

Personally I support the removal, but I'm puzzled as to why people can't wait for the RfC to close. There's been a manic quality to some of the talk on this page recently. The only reason I opened the RfC in the first place was that edits I made to bring this passage in line with COI were reverted. Therefore, the way to move forward is with an RfC. Now there's a rush to close the RfC quickly, accompanied by abuse on my talk page. Why the pressing need for such speed? SlimVirgin ^{TALK|CONTRIBS} 17:51, 7 February 2011 (UTC)

There isn't any need, just as there was no need (or even point) in your reverting the closing action and causing yet another round of pointless arguing. As I've pointed out before, we could make much easier and pleasanter progress on this page if people (well specifically you) didn't keep undoing anything they see anyone else do, almost as a knee-jerk reaction, without having any real reason to disagree with it. That way (I mean if people making reverts had substantial reasons for disagreeing with the edits and explained those reasons clearly) we could avoid wasting time discussing matters needlessly, and focus on the matters which genuinely are in need of discussion.--Kotniski (talk) 18:44, 7 February 2011 (UTC)

If you feel the forest fire of discussion is a waste of time—and I agree—the solution is not to start it. The community has processes in place for a reason, and while it's not good to follow them blindly, it's also not good to thwart them pointlessly. In this case, I made a bold edit. I was reverted. I therefore opened an RfC. These are left open up to 30 days, and closed earlier—usually by the person who opened them or by someone uninvolved—if comments have dried up and no one objects. You are involved, you closed it early while people were still commenting, and there was an objection. It's not the end of the world. SlimVirgin ^{TALK|CONTRIBS} 19:05, 7 February 2011 (UTC)

But there was no objection, that's the point - at least, not a substantial one, just a procedural one, even though the procedure is not fixed, and seems to exist largely in your imagination (and even where we do have procedure, we don't have to follow it pointlessly as if we were some kind of bureaucracy).--Kotniski (talk) 19:30, 7 February 2011 (UTC)

I disagree with the assessment that there is consensus to remove the section on citing oneself from this policy, and replace it with a link to COI. That wasn't really proposed until Feb. 6. Most of the discussion was about incorporating language from COI in this policy. For someone who doesn't follow discussion here closely, it's far from clear that this RfC is about removing the citing oneself section.--SaskatchewanSenator (talk) 19:41, 7 February 2011 (UTC)

Everyone else seems to have got it - I proposed it on Jan 28, the same day as the RfC. Do you actually object to removing the section from this policy? Can you say why? --Kotniski (talk) 19:44, 7 February 2011 (UTC)

I agree that replacing the first sentence of this section with a link to the citing oneself section of COI is an improvement. I think the "If you are able to discover something new..." sentence is about OR and should be kept and either moved somewhere else or in this section with a new title.--SaskatchewanSenator (talk) 02:29, 8 February 2011 (UTC)

The Feb 6 proposal for removal was put in by Slim Virgin after restoring the section when consensus was previously reached for removal. There is definitely a problem with editing WP:NOR. 75.47.131.51 (talk) 20:06, 7 February 2011 (UTC)

Well, it's well over 14 days now, discussion has stopped and consensus is clear, so it's time to put this into effect (I'm about to do so). If someone can find a place to put the "If you are able to discover something new..." sentence, then feel free to re-add it, though for me it just repeats what's already said over and over again in this policy.--Kotniski (talk) 08:39, 14 February 2011 (UTC)

Slim, you seem to have missed the end of this discussion. I don't think that sentence is repetitive. Mostly the policy discusses what OR is not (ie already published). This sentence addresses what OR is.--SaskatchewanSenator (talk) 19:01, 19 February 2011 (UTC)

Medieval annals are to be treated as secondary sources

It is claimed here [4][5][6] that medieval Irish annals are secondary sources. The user adding them appears to be a genealogy enthusiast or something, but he is not competent at judging the value and reliability of difficult sources such as annals. As a result Wikipedia has a bit of a problem in Viking Age Irish Sea history as the user in question habitually uses such primary sources. It is pointless for me to talk to him, as he hates me for blocking him months ago. So can someone else explain how Wikipedia works when it comes to such sources? Be very cautious as he has a bit of a temper (I'm not kidding). Regards, Deacon of Pndapetzim (Talk) 21:46, 13 February 2011 (UTC)

Annals are routinely classified as Primary sources by historians... Wikipedia agrees. Annals are valuable historical records (indeed sometimes the only historical records) but, as primary sources, they have recognized limitations. When writing articles on historical events or people, it is often informative to discuss what the annals have to say about the event or person... but, when doing so, it is important to present what the annals say in a purely descriptive fashion, and avoid making any interpretive, analytical or conclusionary statements based upon what the Annals say. When it doubt, attribute and quote. Blueboar (talk) 23:12, 13 February 2011 (UTC)

Great, that's what I thought. Trouble will be convincing User:DinDraithous, who thinks tagging such references as potential original research to be 'vandalism' and 'hounding' (per links above). Deacon of Pndapetzim (Talk) 23:20, 13 February 2011 (UTC)

(I've posted the identical to the NOR noticedboard) This editor, Deacon, has been following me for months and is now officially hounding me, by targeting articles I have created and leaving them full of OR tags. He has been around Wikipedia for some years, even being an administrator, and has been exposed to countless similar articles full of the identical practice of citing the Irish annals and has done nothing. Examples of articles full of this practice: Sigtrygg Silkbeard (good article), Battle of Glenmama (good article). Wait for many, many more. They're coming. DinDraithou (talk) 05:18, 14 February 2011 (UTC)

I don't think tagging these references is such a great idea - there's nothing wrong with using primary sources per se, it's just a question of making it clear to the reader what source the information comes from (not stating the "facts" from those sources in Wikipedia's voice). A more constructive approach would be to try to reword the statements in question to make their status clearer.--Kotniski (talk) 07:31, 14 February 2011 (UTC)

• No, there are massive and fundamental problems with the use of these primary sources. The editors involved are encyclopaedists when they are editing wikipedia. They are not historians. There is a profession who's occupation is the generation of factive material from primary sources (such as medieval annals), the name of that profession is historian. Fifelfoo (talk) 07:48, 14 February 2011 (UTC)

Which is what I mean - don't generate new "facts" from these sources; but it's quite within Wikipedia's remit to report the fact that a particular source contains particular words.--Kotniski (talk) 08:14, 14 February 2011 (UTC)

• Not really, because reporting that a source contains text is a claim that this is
  1. Meaningful
  2. Implicitly containing facticity
  3. A genuine transmission of the original text
• All three of which require professional verification. Historical articles should be exclusively sourced to scholarly secondary sources, or sources which have gone through equivalent review. Primary sources ought to be used like pictures, only for illustrative content where the factual elements have been verified against scholarly secondary sources already. Fifelfoo (talk) 08:32, 14 February 2011 (UTC)

Annals are not secondary sources, but that doesn't mean they can't be used at all. I generally agree with Kotniski here. WP:PRIMARY says that "A primary source may only be used on Wikipedia to make straightforward, descriptive statements that any educated person, with access to the source but without specialist knowledge, will be able to verify are supported by the source." For a specific example, have a look at talk:Ottir Iarla, where I have questioned OR-tagging of sentences on the formula "[Primary source] describes him as...", referenced to a edited/translated academic (even if dated) publication of the source in question. Finn Rindahl (talk) 11:06, 14 February 2011 (UTC)

Large sections of the text of these articles are clearly based on nothing but primary source readings. It isn't even debatable if this is acceptable, Finn & Kotniski. Even if it is permitted to report that a primary source says something, that does not means using the primary sources as the major sources for an article's narrative. As Finn knows, I am happy to overlook safe examples or to consider references individually, but the editor refuses to discuss anything when I bring messages to his talk page and revert wars to keep tags out. Deacon of Pndapetzim (Talk) 11:53, 14 February 2011 (UTC)

So it sounds as if the problem you're identifying isn't one of "original research" per se, but of overuse of primary sources in the article generally. If you find it useful in some way to tag this problem, then I would suggest one single tag at the top of the article (or maybe on the talk page), rather than lots of OR tags all through the article. --Kotniski (talk) 12:28, 14 February 2011 (UTC)

No, not at all.The problem is articles being written based on primary sources, as I've already said. I have been using the top tags for articles too, but using the text specific tags too to assist the editor(s) involved. Deacon of Pndapetzim (Talk) 12:38, 14 February 2011 (UTC)

Isn't that what I said? (And if the editors don't feel the tags assist them, isn't that a good reason to stop putting them back?)--Kotniski (talk) 12:46, 14 February 2011 (UTC)

I tend to agree with Kotniski ... the problem isn't mentioning what the Primary sources say (if done descriptively, this isn't OR) ... the problem is the overuse and over reliance of those primary sources. That problem is resolved doing by additional research - finding more secondary sources on the subject and discussing what they say. Blueboar (talk) 14:22, 14 February 2011 (UTC)

Sorry, was thrown off by Kotniski's false distinction between OR and overuse of primary sources. So yes, these articles are written based on [the user's interpretation of] primary sources, but consist of individual reference problems too. @Kotniski, I don't think your other point is very relevant and is a bit distracting. The user reverts everything and he is not the only one who should be reviewing. As I was saying, there is no mutual exclusivity between general article OR violations and ones that can be particularised to particular references. Regards, Deacon of Pndapetzim (Talk) 15:49, 14 February 2011 (UTC)

I note that we are also discussing this at WP:NORN (the NOR noticeboard)... I suggest that we end discussion here (which is really supposed to be for discussing edits to the policy itself, and not for article issues) and continue discussion at the noticeboard. Thanks Blueboar (talk) 17:42, 14 February 2011 (UTC)

Confused

It says that primary sources cannot be "interpreted." But a check with the dictionary says that "interpret" means "to explain or tell the meaning of," i.e. to create a summary. So how, exactly, are primary sources to be used in ways that do not involve explaining or telling the meaning of them. Straightforward description of what a source says is an interpretation. 96.39.62.90 (talk) 23:43, 13 February 2011 (UTC)

Source says: "John Smith was born 12 April 1826." That is precisely the extent of information which can be ascribed to the source. We cannot say "six months after his parents married" nor "in a log cabin" as that is not in the exact source used. Collect (talk) 23:55, 13 February 2011 (UTC)

OK, but in that case, it seems like the wording should be changed a bit. As it stands, the guideline currently basically says "primary sources can be used, but you can't actually make any claims about what they say." 71.88.35.24 (talk) 00:51, 14 February 2011 (UTC)

We can make claims about what they say, we try to avoid making claims about what they mean (unless we can properly source those claims). That seems quite consistent with my understanding of "cannot be interpreted".--Kotniski (talk) 07:18, 14 February 2011 (UTC)

Kotniski is correct... but "claims" is such a loaded and confusing word... to rephrase... we can create blunt descriptive statements as to what a primary source explicitly says... but we must avoid creating statements (either overtly or through implication) as to what the primary source means. Blueboar (talk) 14:30, 14 February 2011 (UTC)

Yes, I agree that talk of "making claims" is not the best way of phrasing this sort of thing (and the parts of the policies that use that phrase should probably be reworded). --Kotniski (talk) 14:51, 14 February 2011 (UTC)

I think the larger problem is that "what a source says" and "what a source means" are not really distinct concepts. In fact, I suspect that to many (perhaps even most) people, the two terms are interchangable. Certainly I don't intuitively understand a distinction, and I doubt it's because I'm thick. 71.88.35.24 (talk) 01:21, 15 February 2011 (UTC)

So how about "the words a source contains" versus "the meaning of those words"?--Kotniski (talk) 11:24, 15 February 2011 (UTC)

The only thing that a source "says" is a direct quote. Anything else is to some degree an interpretation, or a selective distillation which also changes it. So the situation isn't as "cut and dried" as this policy pretends that it is. When the article situation is amicable, it gets worked out by common sense and consensus and it works. When there is a battle going on, then such issues make this policy a failure; it's gets used as POV and warfare weapon, e.g. to selectively knock out material that one doesn't agree with via saying it's "OR". North8000 (talk) 12:30, 15 February 2011 (UTC)

To interpret does not mean to create a summary. Look it up in the OED and a good thesaurus. NOR states "summarize what they say in your own words, with each statement in the article attributable to a source that makes that statement explicitly. Source material should be carefully summarized or rephrased without changing its meaning or implication. Take care not to go beyond what is expressed in the sources, or to use them in ways inconsistent with the intention of the source, such as using material out of context" and not to "advance a position not directly and explicitly supported by the source"; "that article statements should not rely on unclear or inconsistent passages, or on passing comments. Passages open to multiple interpretations should be precisely cited or avoided." and not to "Draw[] conclusions not evident in the reference". Шизомби (Sz) (talk) 16:26, 15 February 2011 (UTC)

ATTENTION Admins, let's have a name change...

The name of this particular guideline should be changed to "No unresearched material." I say this because information not derived either from primary/secondary literature or from a controlled experiment one then has peer-reviewed is not research. Calling it "original research" implies otherwise even if it is meant to have a negative connotation. The Mysterious El Willstro (talk) 07:22, 16 February 2011 (UTC)

I disagree. I'm pretty sure what I do in the lab every day counts as research, but I can't write about it on Wikipedia because it's original. Someguy1221 (talk) 07:25, 16 February 2011 (UTC)

"No unpublished material" might be closer to the mark, but that will just result in people saying that the material most certainly was published—by themselves, right on Wikipedia. I don't think that a name change is a productive use of time. If you run into problems, try pointing your fellow disputants to the explanations at WP:MEDRS, which deals with this in the context of medicine ("But this herb killed cells in the petri dish, so it must be the cure for cancer!"). WhatamIdoing (talk) 17:11, 16 February 2011 (UTC)

Transcripts

I don't understand what this is saying that's new:

Similar considerations apply to transcriptions of audio and video sources (although obviously without the requirement to post the original – unless the transcription requires translation as well).

Also, similar considerations to what? SlimVirgin ^{TALK|CONTRIBS} 14:35, 22 February 2011 (UTC)

Similar considerations to those that apply to translations. (I.e. if we don't have a transcription available, then editors can supply their own, etc.) I'm sure it's redundant to something that's said somewhere else, but that applies to almost everything on this page.--Kotniski (talk) 14:55, 22 February 2011 (UTC)

It would be good not to add more redundancy in that case, and this is clearly redundant, especially in NOR. I'm not seeing how it's connected to OR at all in fact, so I'm going to remove it. If you want to argue for its addition, please do that without reverting again. SlimVirgin ^{TALK|CONTRIBS} 14:57, 22 February 2011 (UTC)

This was already discussed a long time ago, no-one objected. Surely you can see the connection though - some people might claim that transcribing spoken words into written words constitutes "original research", just as they might claim that changing French words into English words constitutes "original research" - while we're explaining that they're wrong in the second case, we might just as well explain that they're wrong in the first case. I don't see why you say this is "clearly redundant", or (even if it is) why you think that's a reason to delete anything - if we were to delete everything that's redundant on this page, then there would be nothing left. (I think it would be good to think seriously about doing that, in fact.) --Kotniski (talk) 15:04, 22 February 2011 (UTC)

Why we have two pages saying much the same things

Can someone explain why we should keep two pages, WP:V and WP:NOR, which make effectively the same points, and have material divided or duplicated between them in what seems to be fairly random fashion? Can we not combine them into one? Or if we are to keep two pages, can we try to define the scope of each in some logical way, so we know what should be said here, and what should be said there?--Kotniski (talk) 15:20, 22 February 2011 (UTC)